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.
98 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
99 adjusted the size of a major service buffer; this failed because the
100 buffer was in use at the time. Change to a compile-time increase in the
101 buffer size, when this authenticator is compiled into exim.
103 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
104 previous fast-mode was untenable in the face of glibs using mmap to
105 support larger malloc requests.
107 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
108 New values supported, if defined on system where compiled:
109 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
110 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
112 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
113 running a limited number of proceses in parallel, a benefit is gained. The
114 amount varies with the platform hardware and load. The use of the option
115 queue_run_in_order means we cannot do this, as ordering becomes
122 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
123 side, to decode a TLS 1.3 packet capture.
125 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
126 Previously the default library behaviour applied, sending two, each in
129 JH/03 Debug output for ACL now gives the config file name and line number for
132 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
134 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
136 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
137 buffer overrun for (non-chunking) other transports.
139 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
140 TLS1.3, means that a server rejecting a client certificate is not visible
141 to the client until the first read of encrypted data (typically the
142 response to EHLO). Add detection for that case and treat it as a failed
143 TLS connection attempt, so that the normal retry-in-clear can work (if
144 suitably configured).
146 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
147 and/or domain. Found and fixed by Jason Betts.
149 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
150 configuration). If a CNAME target was not a wellformed name pattern, a
153 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
154 the OS reports them interleaved with other addresses.
156 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
157 used both for input and for a verify callout, both encrypted, SMTP
158 responses being sent by the server could be lost. This resulted in
159 dropped connections and sometimes bounces generated by a peer sending
162 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
163 string. Previously it was possible to cause undefined behaviour in a
164 library routine (usually a crash). Found by "zerons".
166 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
169 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
170 API was removed, so update to use the newer ones.
172 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
173 any timeout set, is taking a long time. Previously we would hang on to a
174 rotated logfile "forever" if the input was arriving with long gaps
175 (a previous attempt to fix addressed lack, for a long time, of initial
178 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
179 shared (NFS) environment. The length of the tempfile name is now
180 4 + 16 ("hdr.$message_exim_id") which might break on file
181 systems which restrict the file name length to lower values.
184 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
185 shared (NFS) environment.
187 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
188 did for all versions <4.90). Notably -M, -m, --invert, -I may be
191 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
192 on some platforms for bit 31.
194 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
195 to changes apparently associated with TLS1.3 handling some of the APIs
196 previously used were either nonfunctional or inappropriate. Strings
197 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
198 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
199 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
200 This affects log line X= elements, the $tls_{in,out}_cipher variables,
201 and the use of specific cipher names in the encrypted= ACL condition.
203 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
205 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
206 verification result was not updated unless hosts_require_ocsp applied.
208 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
209 queue_list_requires_admin set to false, non-admin users were denied the
212 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
213 directory-of-certs mode. Previously they were advertised despite the
216 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
217 A single TCP connection by a client will now hold a TLS connection open
218 for multiple message deliveries, by default. Previoud the default was to
221 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
222 default. If built with the facility, DANE will be used. The facility
223 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
225 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
226 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
227 must be defined and you must still, unless you define DISABLE_TLS, manage
228 the the include-dir and library-file requirements that go with that
229 choice. Non-TLS builds are still supported.
231 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
234 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
235 default. If the platform supports and has the facility enabled, it will
236 be requested on all coneections.
238 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
239 controlled by the build-time option SUPPORT_PIPE_CONNECT.
241 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
243 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
244 success-DSN messages. Previously the From: header was always the default
245 one for these; the option was ignored.
247 JH/28 Fix the timeout on smtp response to apply to the whole response.
248 Previously it was reset for every read, so a teergrubing peer sending
249 single bytes within the time limit could extend the connection for a
250 long time. Credit to Qualsys Security Advisory Team for the discovery.
252 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
253 delivery address, which leaked information of the results of local
254 forwarding. Change to the original envelope recipient address, per
257 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
258 requested. Previously not bounce was generated and a log entry of
259 error ignored was made.
261 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
263 JH/32 Introduce a general tainting mechanism for values read from the input
264 channel, and values derived from them. Refuse to expand any tainted
265 values, to catch one form of exploit.
267 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
268 was unused and the unexpanded text used for the test. Found and
269 fixed by Ruben Jenster.
271 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
272 an attempt to use a TLS library read routine dereffed a nul pointer,
275 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
276 them in our smtp response.
278 JH/36 Have the general router option retry_use_local_part default to true when
279 any of the restrictive preconditions are set (to anything). Previously it
280 was only for check_local user. The change removes one item of manual
281 configuration which is required for proper retries when a remote router
282 handles a subset of addresses for a domain.
284 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
285 link count into consideration.
287 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
288 caused the extension of big_buffer, the following lines were ignored.
290 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
291 accordance with RFC 2308. Previously there was no expiry, so a longlived
292 receive process (eg. due to ACL delays) versus a short SOA value could
295 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
297 JH/39 Promote DMARC support to mainline.
299 JH/40 Bug 2452: Add a References: header to DSNs.
301 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
302 parameters. The relevant library call is documented as "Deprecated: This
303 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
304 3.6.0, DH parameters are negotiated following RFC7919."
306 HS/06 Change the default of dnssec_request_domains to "*"
308 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
309 carried on and emitted a BDAT command, even when PIPELINING was not
312 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
313 buffer was used for the filename, resulting in a trap when tainted
314 arguments (eg. $domain) were used.
316 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
317 recommended to avoid a possible server-load attack. The feature can be
318 re-enabled via the openssl_options main cofiguration option.
320 JH/45 local_scan API: documented the current smtp_printf() call. This changed
321 for version 4.90 - adding a "more data" boolean to the arguments.
322 Bumped the ABI version number also, this having been missed previously;
323 release versions 4.90 to 4.92.3 inclusive were effectively broken in
324 respect of usage of smtp_printf() by either local_scan code or libraries
325 accessed via the ${dlfunc } expansion item. Both will need coding
326 adjustment for any calls to smtp_printf() to match the new function
327 signature; a FALSE value for the new argument is always safe.
329 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
330 the file-offset (which the Linux syscall does, and exim expects); this
331 resulted in an indefinite loop.
333 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
334 to do ARC verification. The Authentication-Results: header line added
335 by the configuration then had no ARC item.
341 JH/01 Remove code calling the customisable local_scan function, unless a new
342 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
344 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
345 non-signal-safe functions being used.
347 JH/03 Bug 2269: When presented with a received message having a stupidly large
348 number of DKIM-Signature headers, disable DKIM verification to avoid
349 a resource-consumption attack. The limit is set at twenty.
351 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
352 report of oldest_pass in ${authres } in consequence, and separate out
353 some descriptions of reasons for verification fail.
355 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
356 files in the spool were present and unlocked. A queue-runner could spot
357 them, resulting in a duplicate delivery. Fix that by doing the unlock
358 after the unlink. Investigation by Tim Stewart. Take the opportunity to
359 add more error-checking on spoolfile handling while that code is being
362 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
363 No known attacks, no CVE, this is defensive hardening.
365 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
366 a queue-runner could start a delivery while other operations were ongoing.
367 Cutthrough delivery was a common victim, resulting in duplicate delivery.
368 Found and investigated by Tim Stewart. Fix by using the open message data
369 file handle rather than opening another, and not locally closing it (which
370 releases a lock) for that case, while creating the temporary .eml format
371 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
373 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
374 $sender_verify_failure/$recipient_verify_failure to "random".
376 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
379 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
380 Previously this would segfault.
382 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
385 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
386 like zero, since the resolver should be doing this for us, But we need one
387 as a CNAME but no MX presence gets the CNAME returned; we need to check
388 that doesn't point to an MX to declare it "no MX returned" rather than
389 "error, loop". A new main option is added so the older capability of
390 following some limited number of chain links is maintained.
392 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
394 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
395 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
396 now-deprecated earlier definitions used only specified the range up to TLS
397 1.2 (in the older-version library docs).
399 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
401 JH/15 Rework TLS client-side context management. Stop using a global, and
402 explicitly pass a context around. This enables future use of TLS for
403 connections to service-daemons (eg. malware scanning) while a client smtp
404 connection is using TLS; with cutthrough connections this is quite likely.
406 JH/16 Fix ARC verification to do AS checks in reverse order.
408 JH/17 Support a "tls" option on the ${readsocket } expansion item.
410 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
411 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
412 Previously the "utf8" would be re-prepended for every additional message.
414 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
415 Previously thery were accepted, resulting in issues when attempting to
416 forward messages to a non-supporting MTA.
418 PP/02 Let -n work with printing macros too, not just options.
420 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
421 one parent address was copied, and bogus data was used at delivery-logging
422 time. Either a crash (after delivery) or bogus log data could result.
423 Discovery and analysis by Tim Stewart.
425 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
426 Previously if the string ended mid-character, we did not insert the
427 promised '?' replacement.
429 PP/04 Documentation: current string operators work on bytes, not codepoints.
431 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
432 should pack well giving a smaller memory footprint so better caching and
433 therefore performance. Group the declarations where this can't be done so
434 that the byte-sized flag variables are not interspersed among pointer
435 variables, giving a better chance of good packing by the compiler.
437 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
438 non-null, to avoid issues with sites running BATV. Previously reports were
439 sent with an empty envelope sender so looked like bounces.
441 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
442 The ignore_error flag wasn't being returned from the filter subprocess so
443 was not set for later routers. Investigation and fix by Matthias Kurz.
445 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
446 and a msg:complete for the whole, when a message is manually removed using
447 -Mrm. Developement by Matthias Kurz, hacked on by JH.
449 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
450 a "Gnu special" function, asprintf() in the DB utility binary builds; I
451 hope that is portable enough.
453 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
454 requiring a known-CA anchor certificate; make it now rely entirely on the
455 TLSA as an anchor. Checking the name on the leaf cert against the name
456 on the A-record for the host is still done for TA (but not for EE mode).
458 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
459 included in delivery lines for non-proxied connections, when compiled with
460 SUPPORT_SOCKS and running with proxy logging enabled.
462 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
463 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
464 move the existing event to fire before the normal logging of message
465 failure so that custom logging is bracketed by normal logging.
467 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
468 msg:fail:internal event. Developement by Matthias Kurz.
470 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
471 far too small for todays use of crypto signatures stored there. Go all
472 the way to the max DNS message size of 64kB, even though this might be
473 overmuch for IOT constrained device use.
475 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
476 copy a string over itself. The library routine is documented as not
477 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
479 JH/32 For main options check_spool_space and check_inode_space, where the
480 platform supports 64b integers, support more than the previous 2^31 kB
481 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
482 the previous G, M, k.
484 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
485 $authenticated_fail_id variable on authentication failure. Previously
488 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
489 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
490 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
491 GNUTLS_SEC_PARAM_MEDIUM.
493 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
494 side. Previously we would continue as if no SNI had been received.
496 JH/36 Harden the handling of string-lists. When a list consisted of a sole
497 "<" character, which should be a list-separator specification, we walked
498 off past the nul-terimation.
500 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
501 causes) even when the retry time is not yet met. Previously they were
502 not, meaning that when (say) an account was over-quota and temp-rejecting,
503 and multiple senders' messages were queued, only one sender would get
504 notified on each configured delay_warning cycle.
506 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
508 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
509 cipher-suites, an error can be left on the stack even for a succeeding
510 accept; this results in impossible error messages when a later operation
513 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
514 return error codes indicating retry. Under TLS1.3 this becomes required.
516 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
517 it only wrote the new authenticators, resulting in a lack of tracking of
518 peer changes of ESMTP extensions until the next cache flush.
520 JH/41 Fix the loop reading a message header line to check for integer overflow,
521 and more-often against header_maxsize. Previously a crafted message could
522 induce a crash of the recive process; now the message is cleanly rejected.
524 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
525 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
531 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
532 When redis_servers is set to a list of > 1 element, and the Redis servers
533 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
534 case of MOVED into a DEFER case instead, thus moving the query onto the
535 next server in the list. For a cluster of N elements, all N servers must
536 be defined in redis_servers.
538 GF/02 Catch and remove uninitialized value warning in exiqsumm
539 Check for existence of @ARGV before looking at $ARGV[0]
541 JH/01 Replace the store_release() internal interface with store_newblock(),
542 which internalises the check required to safely use the old one, plus
543 the allocate and data copy operations duplicated in both (!) of the
544 extant use locations.
546 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
547 modifier. This matches the restriction on the commandline.
549 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
550 Previously only the last row was returned.
552 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
553 we assumed that tags in the header were well-formed, and parsed the
554 element content after inspecting only the first char of the tag.
555 Assumptions at that stage could crash the receive process on malformed
558 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
559 While running the DKIM ACL we operate on the Permanent memory pool so that
560 variables created with "set" persist to the DATA ACL. Also (at any time)
561 DNS lookups that fail create cache records using the Permanent pool. But
562 expansions release any allocations made on the current pool - so a dnsdb
563 lookup expansion done in the DKIM ACL releases the memory used for the
564 DNS negative-cache, and bad things result. Solution is to switch to the
565 Main pool for expansions.
566 While we're in that code, add checks on the DNS cache during store_reset,
567 active in the testsuite.
568 Problem spotted, and debugging aided, by Wolfgang Breyha.
570 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
571 When none of the hosts presented to a transport match an already-open
572 connection, close it and proceed with the list. Previously we would
573 queue the message. Spotted by Lena with Yahoo, probably involving
576 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
577 Previously a spurious "250 OK id=" response was appended to the proper
580 JH/08 The "support for" informational output now, which built with Content
581 Scanning support, has a line for the malware scanner interfaces compiled
582 in. Interface can be individually included or not at build time.
584 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
585 by the template makefile "src/EDITME". The "STREAM" support for an older
586 ClamAV interface method is removed.
588 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
589 rows affected is given instead).
591 JH/11 The runtime Berkeley DB library version is now additionally output by
592 "exim -d -bV". Previously only the compile-time version was shown.
594 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
595 SMTP connection. Previously, when one had more receipients than the
596 first, an abortive onward connection was made. Move to full support for
597 multiple onward connections in sequence, handling cutthrough connection
598 for all multi-message initiating connections.
600 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
601 routers. Previously, a multi-recipient message would fail to match the
602 onward-connection opened for the first recipient, and cause its closure.
604 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
605 a timeout on read on a GnuTLS initiating connection, resulting in the
606 initiating connection being dropped. This mattered most when the callout
607 was marked defer_ok. Fix to keep the two timeout-detection methods
610 JH/15 Relax results from ACL control request to enable cutthrough, in
611 unsupported situations, from error to silently (except under debug)
612 ignoring. This covers use with PRDR, frozen messages, queue-only and
615 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
617 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
618 metadata, resulting in a crash in free().
620 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
621 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
622 Broken also in d185889f4, with init system revamp.
624 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
625 Previously we abruptly closed the connection after reading a malware-
626 found indication; now we go on to read the "scan ok" response line,
629 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
630 ACL. Previously, a crash would result.
632 JH/19 Speed up macro lookups during configuration file read, by skipping non-
633 macro text after a replacement (previously it was only once per line) and
634 by skipping builtin macros when searching for an uppercase lead character.
636 JH/20 DANE support moved from Experimental to mainline. The Makefile control
637 for the build is renamed.
639 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
640 was allocated for every new TLS startup, meaning one per message. Fix
641 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
643 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
644 reported the original. Fix to report (as far as possible) the ACL
645 result replacing the original.
647 JH/23 Fix memory leak during multi-message connections using STARTTLS under
648 OpenSSL. Certificate information is loaded for every new TLS startup,
649 and the resources needed to be freed.
651 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
653 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
656 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
657 DATA response info to the (existing) per-recipient response info for
658 the "C=" log element. It can have useful tracking info from the
659 destination system. Patch from Simon Arlott.
661 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
662 length value. Previously this would segfault.
664 HS/02 Support Avast multiline protoocol, this allows passing flags to
665 newer versions of the scanner.
667 JH/28 Ensure that variables possibly set during message acceptance are marked
668 dead before release of memory in the daemon loop. This stops complaints
669 about them when the debug_store option is enabled. Discovered specifically
670 for sender_rate_period, but applies to a whole set of variables.
671 Do the same for the queue-runner and queue-list loops, for variables set
672 from spool message files. Do the same for the SMTP per-message loop, for
673 certain variables indirectly set in ACL operations.
675 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
676 as a multi-recipient message from a mailinglist manager). The coding had
677 an arbitrary cutoff number of characters while checking for more input;
678 enforced by writing a NUL into the buffer. This corrupted long / fast
679 input. The problem was exposed more widely when more pipelineing of SMTP
680 responses was introduced, and one Exim system was feeding another.
681 The symptom is log complaints of SMTP syntax error (NUL chars) on the
682 receiving system, and refused recipients seen by the sending system
683 (propating to people being dropped from mailing lists).
684 Discovered and pinpointed by David Carter.
686 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
687 replaced by the ${authresults } expansion.
689 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
691 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
692 allows proper process termination in container environments.
694 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
695 Previously the "final dot" had a newline after it; ensure it is CR,LF.
697 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
698 and "err_perm", deprecated since 4.83 when the RFC-defined words
699 "temperror" and "permerror" were introduced.
701 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
702 transport-filters or DKIM-signing. The restriction was lost in the
703 consolidation of verify-callout and delivery SMTP handling.
704 Extend the restriction to also cover ARC-signing.
706 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
707 in defer=pass mode supply a 450 to the initiator. Previously the message
710 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
711 tls_require_ciphers is used as before.
713 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
714 "pass_unscanned". Only tmpfails from the scanner are written to
715 the paniclog, as they may require admin intervention (permission
716 denied, license issues). Other scanner errors (like decompression
717 bombs) do not cause a paniclog entry.
719 JH/36 Fix reinitialisation of DKIM logging variable between messages.
720 Previously it was possible to log spurious information in receive log
723 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
724 triggered odd behaviour from Outlook Express clients.
726 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
729 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
730 since the IETF WG has not yet settled on that versus the original
731 "bare" representation.
733 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
734 Previously the millisecond value corrupted the output.
735 Fix also for syslog_pid=no and log_selector +pid, for which the pid
736 corrupted the output.
742 JH/01 Rework error string handling in TLS interface so that the caller in
743 more cases is responsible for logging. This permits library-sourced
744 string to be attached to addresses during delivery, and collapses
745 pairs of long lines into single ones.
747 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
748 during configuration. Wildcards are allowed and expanded.
750 JH/02 Rework error string handling in DKIM to pass more info back to callers.
751 This permits better logging.
753 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
754 do not close it down and have the child transport start it up again on
755 the passed-on TCP connection. Instead, proxy the child (and any
756 subsequent ones) for TLS via a unix-domain socket channel. Logging is
757 affected: the continued delivery log lines do not have any DNSSEC, TLS
758 Certificate or OCSP information. TLS cipher information is still logged.
760 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
761 identical IP addresses on different listening ports. Will also affect
764 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
765 add noisy ifdef guards to special-case this sillyness.
766 Patch from Bernd Kuhls.
768 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
769 than 255 are no longer allowed.
771 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
772 Disable the session-cache too, which might reduce our load. Since we
773 currrectly use a new context for every connection, both as server and
774 client, there is no benefit for these.
775 GnuTLS appears to not support tickets server-side by default (we don't
776 call gnutls_session_ticket_enable_server()) but client side is enabled
777 by default on recent versions (3.1.3 +) unless the PFS priority string
780 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
781 <https://reproducible-builds.org/specs/source-date-epoch/>.
783 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
784 the check for any unsuccessful recipients did not notice the limit, and
785 erroneously found still-pending ones.
787 JH/08 Pipeline CHUNKING command and data together, on kernels that support
788 MSG_MORE. Only in-clear (not on TLS connections).
790 JH/09 Avoid using a temporary file during transport using dkim. Unless a
791 transport-filter is involved we can buffer the headers in memory for
792 creating the signature, and read the spool data file once for the
793 signature and again for transmission.
795 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
796 4.77 as the kernel support then wasn't solid, having issues in 64bit
797 mode. Now, it's been long enough. Add support for FreeBSD also.
799 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
800 case where the routing stage had gathered several addresses to send to
801 a host before calling the transport for the first, we previously failed
802 to close down TLS in the old transport process before passing the TCP
803 connection to the new process. The new one sent a STARTTLS command
804 which naturally failed, giving a failed delivery and bloating the retry
805 database. Investigation and fix prototype from Wolfgang Breyha.
807 JH/12 Fix check on SMTP command input synchronisation. Previously there were
808 false-negatives in the check that the sender had not preempted a response
809 or prompt from Exim (running as a server), due to that code's lack of
810 awareness of the SMTP input buffering.
812 PP/04 Add commandline_checks_require_admin option.
813 Exim drops privileges sanely, various checks such as -be aren't a
814 security problem, as long as you trust local users with access to their
815 own account. When invoked by services which pass untrusted data to
816 Exim, this might be an issue. Set this option in main configuration
817 AND make fixes to the calling application, such as using `--` to stop
820 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
821 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
822 responses to those, into a single TLS record each way (this usually means
823 a single packet). As a side issue, smtp_enforce_sync now works on TLS
826 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
827 affects you only if you're dancing at the edge of the param size limits.
828 If you are, and this message makes sense to you, then: raise the
829 configured limit or use OpenSSL 1.1. Nothing we can do for older
832 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
833 cmdline element to get the documented default one. Previously it was
836 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
837 are unsuable for non-message-specific reasons - eg. connection timeout,
838 banner-time rejection.
840 JH/16 Fix logging of delivery remote port, when specified by router, under
843 PP/06 Repair manualroute's ability to take options in any order, even if one
844 is the name of a transport.
847 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
849 JH/17 Change the list-building routines interface to use the expanding-string
850 triplet model, for better allocation and copying behaviour.
852 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
853 Previously it was constructed the first time a possibly-matching string
854 was met in the configuration file input during startup; now it is done
857 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
858 compatible one, to avoid the (poorly documented) possibility of a config
859 file in the working directory redirecting the DB files, possibly correpting
860 some existing file. CVE-2017-10140 assigned for BDB.
862 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
863 cache-hot. Previously, although the result was properly cached, the
864 initial verify call returned a defer.
866 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
867 the main verify for receipient in uncached-mode.
869 JH/22 Retire historical build files to an "unsupported" subdir. These are
870 defined as "ones for which we have no current evidence of testing".
872 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
873 if present. Previously it was ignored.
875 JH/24 Start using specified-initialisers in C structure init coding. This is
876 a C99 feature (it's 2017, so now considered safe).
878 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
879 if was a fixed-sized field and bitmask ops via macros; it is now more
882 PP/07 GitHub PR 56: Apply MariaDB build fix.
883 Patch provided by Jaroslav Škarvada.
885 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
886 during Coverity cleanups [4.87 JH/47]
887 Diagnosis and fix provided by Michael Fischer v. Mollard.
889 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
890 the right size to place the terminating semicolon on its own folded
891 line, the header hash was calculated to an incorrect value thanks to
892 the (relaxed) space the fold became.
894 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
895 and confused the parent.
897 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
898 which could crash as a result. This could lead to undeliverable messages.
900 JH/28 Logging: "next input sent too soon" now shows where input was truncated
903 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
904 matters on fast-turnover and PID-randomising systems, which were getting
905 out-of-order delivery.
907 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
908 a possibly-overlapping copy. The symptom was that "Remote host closed
909 connection in response to HELO" was logged instead of the actual 4xx
912 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
913 Previously only that bufferd was discarded, resulting in SYMTP command
916 JH/32 DKIM: when a message has multiple signatures matching an identity given
917 in dkim_verify_signers, run the dkim acl once for each. Previously only
918 one run was done. Bug 2189.
920 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
921 "panic the current process" to "deliberately defer". The panic log is
922 still written with the problem list name; the mail and reject logs now
923 get a temp-reject line for the message that was being handled, saying
924 something like "domains check lookup or other defer". The SMTP 451
925 message is still "Temporary local problem".
927 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
928 A crafted sequence of BDAT commands could result in in-use memory beeing
929 freed. CVE-2017-16943.
931 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
932 from SMTP input. Previously it was always done; now only done for DATA
933 and not BDAT commands. CVE-2017-16944.
935 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
936 to the message (such as an overlong header line). Previously this was
937 not done and we did not exit BDAT mode. Followon from the previous item
938 though a different problem.
944 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
945 than -2003 did; needs libidn2 in addition to libidn.
947 JH/02 The path option on a pipe transport is now expanded before use.
949 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
950 Patch provided by "Björn", documentation fix added too.
952 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
953 missing a wire-to-host endian conversion.
955 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
956 close after a BDAT command line could be taken as a following command,
957 giving a synch failure. Fix by only checking for synch immediately
958 before acknowledging the chunk.
960 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
961 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
962 macro. Patches provided by Josh Soref.
964 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
965 Previously we did not; the RFC seems ambiguous and VRFY is not listed
966 by IANA as a service extension. However, John Klensin suggests that we
969 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
970 the dkim code may be unix-mode line endings rather than smtp wire-format
971 CRLF, so prepend a CR to any bare LF.
973 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
974 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
976 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
977 the first were themselves being wrongly included in the feed into dkim
978 processing; with most chunk sizes in use this resulted in an incorrect
979 body hash calculated value.
981 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
982 DKIM signature block, for verification. Although advised against by
983 standards it is specifically not ruled illegal.
985 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
987 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
988 missing a body hash (the bh= tag).
990 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
991 It seems that HAProxy sends the Proxy Protocol information in clear and
992 only then does a TLS startup, so do the same.
994 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
995 TCP connections (such as for Spamd) unless the daemon successfully set
996 Fast Open mode on its listening sockets. This fixes breakage seen on
997 too-old kernels or those not configured for Fast Open, at the cost of
998 requiring both directions being enabled for TFO, and TFO never being used
999 by non-daemon-related Exim processes.
1001 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1002 endings, at least on the first header line. Try to canonify any that get
1003 past that check, despite the cost.
1005 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1006 now limited to an arbitrary five deep, while parsing addresses with the
1007 strip_excess_angle_brackets option enabled.
1009 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1010 instead leave the unprompted TLS handshake in socket buffer for the
1011 TLS library to consume.
1013 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1015 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1017 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1018 region in one operation, for speed, and this leaves assigned pointers
1019 dangling. Add checks run only under the testsuite which checks all
1020 variables at a store-reset and panics on a dangling pointer; add code
1021 explicitly nulling out all the variables discovered. Fixes one known
1022 bug: a transport crash, where a dangling pointer for $sending_ip_address
1023 originally assigned in a verify callout, is re-used.
1025 PP/06 Drop '.' from @INC in various Perl scripts.
1027 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1029 PP/08 Reduce a number of compilation warnings under clang; building with
1030 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1031 should be warning-free.
1033 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1035 HS/01 Fix portability problems introduced by PP/08 for platforms where
1036 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1038 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1039 chunk. This allows us to accept broken chunked messages. We need a more
1040 general solution here.
1042 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1043 already-broken messages in the queue.
1045 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1047 JH/19 Fix reference counting bug in routing-generated-address tracking.
1053 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1054 supports it and a size is available (ie. the sending peer gave us one).
1056 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1057 years of being deprecated). The replacements are the ACLs
1058 acl_smtp_mime and acl_not_smtp_mime.
1060 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1061 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1062 or even an in-clear connection were permitted. Now, if the host lookup
1063 was dnssec and dane was requested then the host is only used if the
1064 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1065 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1066 if one fails this test.
1067 This means that a poorly-configured remote DNS will make it incommunicado;
1068 but it protects against a DNS-interception attack on it.
1070 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1071 non-noisy when a race steals the message being considered.
1073 JH/05 If main configuration option tls_certificate is unset, generate a
1074 self-signed certificate for inbound TLS connections.
1076 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1077 in rewrites and routers.
1079 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1080 and logged a warning sing 4.83; now they are a configuration file error.
1082 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1083 (lacking @domain). Apply the same qualification processing as RCPT.
1085 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1087 JH/10 Support ${sha256:} applied to a string (as well as the previous
1090 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1091 a cutthrough deliver is pending, as we always want to make a connection.
1092 This also avoids re-routing the message when later placing the cutthrough
1093 connection after a verify cache hit.
1094 Do not update it with the verify result either.
1096 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1097 when routing results in more than one destination address.
1099 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1100 signing (which inhibits the cutthrough capability). Previously only
1101 the presence of an option was tested; now an expansion evaluating as
1102 empty is permissible (obviously it should depend only on data available
1103 when the cutthrough connection is made).
1105 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1106 the relevant preceding SMTP command did not note the pipelining mode.
1108 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1109 Previously they were not counted.
1111 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1112 as one having no matching records. Previously we deferred the message
1113 that needed the lookup.
1115 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1116 distinguished as "(=".
1118 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1119 for missing MX records. Previously it only worked for missing A records.
1121 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1123 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1124 after the data-go-ahead and data-ack. Patch from Jason Betts.
1126 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1127 even for a "none" policy. Patch from Tony Meyer.
1129 JH/22 Fix continued use of a connection for further deliveries. If a port was
1130 specified by a router, it must also match for the delivery to be
1133 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1134 When one of the recipients of a message was unsuitable for the connection
1135 (has no matching addresses), we lost track of needing to mark it
1136 deferred. As a result mail would be lost.
1138 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1140 JH/25 Decoding ACL controls is now done using a binary search; the source code
1141 takes up less space and should be simpler to maintain. Merge the ACL
1142 condition decode tables also, with similar effect.
1144 JH/26 Fix problem with one_time used on a redirect router which returned the
1145 parent address unchanged. A retry would see the parent address marked as
1146 delivered, so not attempt the (identical) child. As a result mail would
1149 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1150 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1151 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1154 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1155 Main config options check_{log,spool}_{inodes,space} are now
1156 100 inodes, 10MB unless set otherwise in the configuration.
1158 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1159 Previously it only applied to the main-section connection policy
1162 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1164 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1165 by me. Added RFC7919 DH primes as an alternative.
1167 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1168 are not in the system include path.
1170 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1171 GnuTLS, when a session startup failed (eg because the client disconnected)
1172 Exim did stdio operations after fclose. This was exposed by a recent
1173 change which nulled out the file handle after the fclose.
1175 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1176 signed directly by the cert-signing cert, rather than an intermediate
1177 OCSP-signing cert. This is the model used by LetsEncrypt.
1179 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1181 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1182 an incoming connection.
1184 HS/02 Bug 1802: Do not half-close the connection after sending a request
1187 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1188 fallback to "prime256v1".
1190 JH/34 SECURITY: Use proper copy of DATA command in error message.
1191 Could leak key material. Remotely exploitable. CVE-2016-9963.
1197 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1198 and 3.4.4 - once the server is enabled to respond to an OCSP request
1199 it does even when not requested, resulting in a stapling non-aware
1200 client dropping the TLS connection.
1202 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1203 support variable-length bit vectors. No functional change.
1205 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1206 The I= interface field on outgoing lines is now after the H= remote
1207 host field, same as incoming lines. There is a separate
1208 outgoing_interface log selector which allows you to disable the
1211 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1212 If not running log_selector +smtp_connection the mainlog would be held
1213 open indefinitely after a "too many connections" event, including to a
1214 deleted file after a log rotate. Leave the per net connection logging
1215 leaving it open for efficiency as that will be quickly detected by the
1216 check on the next write.
1218 HS/01 Bug 1671: Fix post transport crash.
1219 Processing the wait-<transport> messages could crash the delivery
1220 process if the message IDs didn't exist for some reason. When
1221 using 'split_spool_directory=yes' the construction of the spool
1222 file name failed already, exposing the same netto behaviour.
1224 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1225 mime_regex ACL conditions.
1227 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1228 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1229 to HELO, local diagnostic string.
1231 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1232 log line to debug. Even when configured with a tls authenticator many
1233 client connections are expected to not authenticate in this way, so
1234 an authenticate fail is not an error.
1236 HS/02 Add the Exim version string to the process info. This way exiwhat
1237 gives some more detail about the running daemon.
1239 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1240 matter for fast-change records such as DNSBLs.
1242 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1243 retry record, even if constant. There may be multiple transports with
1244 different interface settings and the retry behaviour needs to be kept
1247 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1249 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1251 JH/10 Bug 840: fix log_defer_output option of pipe transport
1253 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1254 hosts_randomize. This matters a lot when combined with mua_wrapper.
1256 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1257 ${quote_pgsql:<string>} operator.
1259 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1261 JH/14 Change header file name for freeradius-client. Relevant if compiling
1262 with Radius support; from the Gentoo tree and checked under Fedora.
1264 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1266 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1267 option transport_filter, meaning no filtering.
1269 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1271 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1272 defaults to "*" (all hosts). The variable is now available when not built
1273 with TLS, default unset, mainly to enable keeping the testsuite sane.
1274 If a server certificate is not supplied (via tls_certificate) an error is
1275 logged, and clients will find TLS connections fail on startup. Presumably
1276 they will retry in-clear.
1277 Packagers of Exim are strongly encouraged to create a server certificate
1278 at installation time.
1280 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1281 with the $config_file variable.
1283 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1284 in transport context, after the attempt, and per-recipient. The latter type
1285 is per host attempted. The event data is the error message, and the errno
1286 information encodes the lookup type (A vs. MX) used for the (first) host,
1287 and the trailing two digits of the smtp 4xx response.
1289 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1290 to write to mainlog (or rejectlog, paniclog) in the window between file
1291 creation and permissions/ownership being changed. Particularly affects
1292 installations where exicyclog is run as root, rather than exim user;
1293 result is that the running daemon panics and dies.
1295 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1297 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1298 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1299 "pri" and "weight". Note that the previous implicit priority given by the
1300 list order is no longer honoured.
1302 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1303 for DKIM processing.
1305 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1306 by defining SUPPORT_SOCKS.
1308 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1309 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1310 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1311 variables are renamed to proxy_{local,external}_{address,port}.
1313 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1314 for a build by defining SUPPORT_I18N
1316 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1317 of the query string, and make ${quote_redis:} do that quoting.
1319 JH/29 Move Events support from Experimental to mainline, enabled by default
1320 and removable for a build by defining DISABLE_EVENT.
1322 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1324 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1325 cached by the daemon.
1327 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1328 by defining LOOKUP_REDIS. The libhiredis library is required.
1330 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1331 keys are given for lookup.
1333 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1334 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1335 only supported when built with TLS support. The PolarSSL SHA routines
1336 are still used when the TLS library is too old for convenient support.
1338 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1339 openssl_options), for security. OpenSSL forces this from version 1.1.0
1340 server-side so match that on older versions.
1342 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1343 allocation for $value could be released as the expansion processing
1344 concluded, but leaving the global pointer active for it.
1346 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1347 and to use the domains and local_parts ACL conditions.
1349 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1350 incorrectly not doubled on cutthrough transmission, hence seen as a
1351 body-termination at the receiving system - resulting in truncated mails.
1352 Commonly the sender saw a TCP-level error, and retransmitted the message
1353 via the normal store-and-forward channel. This could result in duplicates
1354 received - but deduplicating mailstores were liable to retain only the
1355 initial truncated version.
1357 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1359 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1361 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1362 we're in there, support oversigning also; bug 1309.
1364 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1366 HS/04 Add support for keep_environment and add_environment options.
1368 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1369 either intentional arithmetic overflow during PRNG, or testing config-
1372 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1373 delivery resulted in actual delivery. Cancel cutthrough before DATA
1376 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1377 rejects a recipient: pass the reject to the originator.
1379 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1380 Many were false-positives and ignorable, but it's worth fixing the
1383 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1384 for the new environment-manipulation done at startup. Move the routines
1385 from being local to tls.c to being global via the os.c file.
1387 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1388 an extract embedded as result-arg for a map, the first arg for extract
1389 is unavailable so we cannot tell if this is a numbered or keyed
1390 extraction. Accept either.
1396 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1399 JH/02 The smtp transport option "multi_domain" is now expanded.
1401 JH/03 The smtp transport now requests PRDR by default, if the server offers
1404 JH/04 Certificate name checking on server certificates, when exim is a client,
1405 is now done by default. The transport option tls_verify_cert_hostnames
1406 can be used to disable this per-host. The build option
1407 EXPERIMENTAL_CERTNAMES is withdrawn.
1409 JH/05 The value of the tls_verify_certificates smtp transport and main options
1410 default to the word "system" to access the system default CA bundle.
1411 For GnuTLS, only version 3.0.20 or later.
1413 JH/06 Verification of the server certificate for a TLS connection is now tried
1414 (but not required) by default. The verification status is now logged by
1415 default, for both outbound TLS and client-certificate supplying inbound
1418 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1421 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1422 Status Notification (bounce) messages are now MIME format per RFC 3464.
1423 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1424 under the control of the dsn_advertise_hosts option, and routers may
1425 have a dsn_lasthop option.
1427 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1428 default, modifiable by a malware= option. The list separator for
1429 the options can now be changed in the usual way. Bug 68.
1431 JH/10 The smtp_receive_timeout main option is now expanded before use.
1433 JH/11 The incoming_interface log option now also enables logging of the
1434 local interface on delivery outgoing connections.
1436 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1437 if the interface and destination host and port all match.
1439 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1442 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1443 Patch from Andrew Lewis.
1445 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1446 now supports optional time-restrictions, weighting, and priority
1447 modifiers per server. Patch originally by <rommer@active.by>.
1449 JH/16 The spamd_address main option now supports a mixed list of local
1450 and remote servers. Remote servers can be IPv6 addresses, and
1451 specify a port-range.
1453 JH/17 Bug 68: The spamd_address main option now supports an optional
1454 timeout value per server.
1456 JH/18 Bug 1581: Router and transport options headers_add/remove can
1457 now have the list separator specified.
1459 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1462 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1465 JH/21 Support for the A6 type of dns record is withdrawn.
1467 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1468 rather than the verbs used.
1470 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1471 from 255 to 1024 chars.
1473 JH/24 Verification callouts now attempt to use TLS by default.
1475 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1476 are generic router options now. The defaults didn't change.
1478 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1479 Original patch from Alexander Shikoff, worked over by JH.
1481 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1482 files if ZCAT_COMMAND is not executable.
1484 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1486 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1488 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1489 Normally benign, it bites when the pair was led to by a CNAME;
1490 modern usage is to not canonicalize the domain to a CNAME target
1491 (and we were inconsistent anyway for A-only vs AAAA+A).
1493 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1495 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1496 when evaluating $sender_host_dnssec.
1498 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1499 $sender_helo_dnssec variable.
1501 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1503 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1505 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1507 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1508 documented as working, but never had. Support all but $spam_report.
1510 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1511 added for tls authenticator.
1513 HS/03 Add perl_taintmode main config option
1519 TL/01 When running the test suite, the README says that variables such as
1520 no_msglog_check are global and can be placed anywhere in a specific
1521 test's script, however it was observed that placement needed to be near
1522 the beginning for it to behave that way. Changed the runtest perl
1523 script to read through the entire script once to detect and set these
1524 variables, reset to the beginning of the script, and then run through
1525 the script parsing/test process like normal.
1527 TL/02 The BSD's have an arc4random API. One of the functions to induce
1528 adding randomness was arc4random_stir(), but it has been removed in
1529 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1530 function when detected.
1532 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1533 cause callback expansion.
1535 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1536 syntax errors in an expansion can be treated as a string instead of
1537 logging or causing an error, due to the internal use of bool_lax
1538 instead of bool when processing it.
1540 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1541 server certificates when making smtp deliveries.
1543 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1545 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1547 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1549 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1550 Merged patch from Sebastian Wiedenroth.
1552 JH/05 Fix results-pipe from transport process. Several recipients, combined
1553 with certificate use, exposed issues where response data items split
1554 over buffer boundaries were not parsed properly. This eventually
1555 resulted in duplicates being sent. This issue only became common enough
1556 to notice due to the introduction of connection certificate information,
1557 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1559 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1560 size buffer was used, resulting in syntax errors when an expansion
1563 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1564 version 3.3.6 or later.
1566 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1567 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1568 both become "event_action", the variables become $event_name, $event_data
1569 and $event_defer_errno. There is a new variable $verify_mode, usable in
1570 routers, transports and related events. The tls:cert event is now also
1571 raised for inbound connections, if the main configuration event_action
1574 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1575 early OCSP support, but no stapling (appears to be less than 1.0.0).
1577 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1578 server certificate names available under the smtp transport option
1579 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1582 JH/10 Time-related extraction expansions from certificates now use the main
1583 option "timezone" setting for output formatting, and are consistent
1584 between OpenSSL and GnuTLS compilations. Bug 1541.
1586 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1587 encoded parameter in the incoming message. Bug 1558.
1589 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1590 include certificate info, eximon was claiming there were spoolfile
1593 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1595 JH/14 Log delivery-related information more consistently, using the sequence
1596 "H=<name> [<ip>]" wherever possible.
1598 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1599 are problematic for Debian distribution, omit them from the release
1602 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1604 JH/16 Fix string representation of time values on 64bit time_t architectures.
1607 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1608 output list separator was used.
1613 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1614 checkers that were complaining about end of non-void function with no
1617 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1618 This was a regression introduced in 4.83 by another bugfix.
1620 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1622 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1623 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1629 TF/01 Correctly close the server side of TLS when forking for delivery.
1631 When a message was received over SMTP with TLS, Exim failed to clear up
1632 the incoming connection properly after forking off the child process to
1633 deliver the message. In some situations the subsequent outgoing
1634 delivery connection happened to have the same fd number as the incoming
1635 connection previously had. Exim would try to use TLS and fail, logging
1636 a "Bad file descriptor" error.
1638 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1639 utilities have not been installed.
1641 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1642 temporary space as the ACL may create new global variables.
1644 TL/01 LDAP support uses per connection or global context settings, depending
1645 upon the detected version of the libraries at build time.
1647 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1648 to extract and use the src ip:port in logging and expansions as if it
1649 were a direct connection from the outside internet. PPv2 support was
1650 updated based on HAProxy spec change in May 2014.
1652 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1654 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1655 Properly escape header and check for NULL return.
1657 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1660 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1662 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1663 characters in header names, implemented as a verify condition.
1664 Contributed by Michael Fischer v. Mollard.
1666 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1667 results permerror and temperror. Previous values are deprecated but
1668 still accepted. In a future release, err_perm and err_temp will be
1669 completely removed, which will be a backward incompatibility if the
1670 ACL tests for either of these two old results. Patch contributed by
1671 user bes-internal on the mailing list.
1673 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1675 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1676 selectors, in both main and reject logs.
1678 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1681 JH/07 Add malware type "sock" for talking to simple daemon.
1683 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1685 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1686 routers/transports under cutthrough routing.
1688 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1689 numbers. Touch up "bool" conditional to keep the same definition.
1691 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1693 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1695 JH/12 Expand items in router/transport headers_add or headers_remove lists
1696 individually rather than the list as a whole. Bug 1452.
1698 Required for reasonable handling of multiple headers_ options when
1699 they may be empty; requires that headers_remove items with embedded
1700 colons must have them doubled (or the list-separator changed).
1702 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1703 view the policy declared in the DMARC record. Currently, $dmarc_status
1704 is a combined value of both the record presence and the result of the
1707 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1709 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1710 dnslookup router and the smtp transport (applying to the forward
1713 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1714 of ldap servers used for a specific lookup. Patch provided by Heiko
1717 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1718 New variable $lookup_dnssec_authenticated for observability.
1720 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1721 Patch submitted by Lars Timman.
1723 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1725 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1726 Requires trusted mode and valid format message id, aborts otherwise.
1727 Patch contributed by Heiko Schlichting.
1729 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1730 certextract with support for various fields. Bug 1358.
1732 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1733 is requested by default, modifiable by smtp transport option
1736 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1737 operate on certificate variables to give certificate fingerprints
1738 Also new ${sha256:cert_variable}.
1740 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1742 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1743 Christian Aistleitner.
1745 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1747 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1748 file. Patch from Wolfgang Breyha.
1750 JH/25 Expand the coverage of the delivery $host and $host_address to
1751 client authenticators run in verify callout. Bug 1476.
1753 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1754 align with daemon_smtp_ports. Bug 72.
1756 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1757 support and error reporting did not work properly.
1759 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1760 and is readable. Patch from Andrew Colin Kissa.
1762 TL/14 Enhance documentation of ${run expansion and how it parses the
1763 commandline after expansion, particularly in the case when an
1764 unquoted variable expansion results in an empty value.
1766 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1768 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1769 by renaming away from it. Addresses GH issue 15, reported by
1772 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1774 TL/15 SECURITY: prevent double expansion in math comparison functions
1775 (can expand unsanitized data). Not remotely exploitable.
1782 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1784 PP/02 Make -n do something, by making it not do something.
1785 When combined with -bP, the name of an option is not output.
1787 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1790 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1791 $sender_host_name and config options to manage this, and basic check
1794 PP/05 DSCP support for outbound connections and control modifier for inbound.
1796 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1797 (Only plugin which currently uses this is kerberos4, which nobody should
1798 be using, but we should make it available and other future plugins might
1799 conceivably use it, even though it would break NAT; stuff *should* be
1800 using channel bindings instead).
1802 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1803 name; added for Sendmail compatibility; requires admin caller.
1804 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1805 just ignore it); requires trusted caller.
1806 Also parse but ignore: -Ac -Am -X<logfile>
1809 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1811 TL/02 Add +smtp_confirmation as a default logging option.
1813 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1814 Patch by Magnus Holmgren from 2007-02-20.
1816 TL/04 Bugzilla 1281 - Spec typo.
1817 Bugzilla 1283 - Spec typo.
1818 Bugzilla 1290 - Spec grammar fixes.
1820 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1822 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1824 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1825 mailing list by by Dmitry Isaikin.
1827 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1829 JH/02 Support "G" suffix to numbers in ${if comparisons.
1831 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1833 NM/01 Bugzilla 1197 - Spec typo
1834 Bugzilla 1196 - Spec examples corrections
1836 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1838 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1839 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1842 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1843 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1845 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1846 "acl {{name}{arg}...}", and optional args on acl condition
1849 JH/05 Permit multiple router/transport headers_add/remove lines.
1851 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1853 JH/07 Avoid using a waiting database for a single-message-only transport.
1854 Performance patch from Paul Fisher. Bugzilla 1262.
1856 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1859 JH/09 Add $headers_added variable, with content from use of ACL modifier
1860 add_header (but not yet added to the message). Bugzilla 199.
1862 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1863 Pulled from Bugzilla 817 by Wolfgang Breyha.
1865 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1867 (nb: this is the same fix as in Exim 4.80.1)
1869 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1872 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1875 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1877 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1878 advertise SMTP AUTH mechanism to us, instead of a generic
1879 protocol violation error. Also, make Exim more robust to bad
1880 data from the Dovecot auth socket.
1882 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1884 When a queue runner is handling a message, Exim first routes the
1885 recipient addresses, during which it prunes them based on the retry
1886 hints database. After that it attempts to deliver the message to
1887 any remaining recipients. It then updates the hints database using
1890 So if a recipient address works intermittently, it can get repeatedly
1891 deferred at routing time. The retry hints record remains fresh so the
1892 address never reaches the final cutoff time.
1894 This is a fairly common occurrence when a user is bumping up against
1895 their storage quota. Exim had some logic in its local delivery code
1896 to deal with this. However it did not apply to per-recipient defers
1897 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1899 This change adds a proper retry rule check during routing so that the
1900 final cutoff time is checked against the message's age. We only do
1901 this check if there is an address retry record and there is not a
1902 domain retry record; this implies that previous attempts to handle
1903 the address had the retry_use_local_parts option turned on. We use
1904 this as an approximation for the destination being like a local
1905 delivery, as in LMTP.
1907 I suspect this new check makes the old local delivery cutoff check
1908 redundant, but I have not verified this so I left the code in place.
1910 TF/02 Correct gecos expansion when From: is a prefix of the username.
1912 Test 0254 submits a message to Exim with the header
1916 When I ran the test suite under the user fanf2, Exim expanded
1917 the header to contain my full name, whereas it should have added
1918 a Resent-Sender: header. It erroneously treats any prefix of the
1919 username as equal to the username.
1921 This change corrects that bug.
1923 GF/01 DCC debug and logging tidyup
1924 Error conditions log to paniclog rather than rejectlog.
1925 Debug lines prefixed by "DCC: " to remove any ambiguity.
1927 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1929 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1930 Bug spotted by Jeremy Harris; was flawed since initial commit.
1931 Would have resulted in OCSP responses post-SNI triggering an Exim
1932 NULL dereference and crash.
1934 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1936 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1937 Bug detection, analysis and fix by Samuel Thibault.
1938 Bugzilla 1331, Debian bug #698092.
1940 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1942 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1943 Server implementation by Todd Lyons, client by JH.
1944 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1945 config variable "prdr_enable" controls whether the server
1946 advertises the facility. If the client requests PRDR a new
1947 acl_data_smtp_prdr ACL is called once for each recipient, after
1948 the body content is received and before the acl_smtp_data ACL.
1949 The client is controlled by both of: a hosts_try_prdr option
1950 on the smtp transport, and the server advertisement.
1951 Default client logging of deliveries and rejections involving
1952 PRDR are flagged with the string "PRDR".
1954 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1955 fclose(). Diagnosis by Todd Lyons.
1957 PP/17 Update configure.default to handle IPv6 localhost better.
1958 Patch by Alain Williams (plus minor tweaks).
1961 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1962 This is now consistent with GnuTLS, and is now documented: the
1963 previous undocumented portable approach to treating the option as
1964 unset was to force an expansion failure. That still works, and
1965 an empty string is now equivalent.
1967 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1968 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1969 not performing validation itself.
1971 PP/20 Added force_command boolean option to pipe transport.
1972 Patch from Nick Koston, of cPanel Inc.
1974 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1977 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1979 PP/21 Fix eximon continuous updating with timestamped log-files.
1980 Broken in a format-string cleanup in 4.80, missed when I repaired the
1981 other false fix of the same issue.
1982 Report and fix from Heiko Schlichting.
1985 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1986 Report from Prashanth Katuri.
1988 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1989 It's SecureTransport, so affects any MacOS clients which use the
1990 system-integrated TLS libraries, including email clients.
1992 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1993 using a MIME ACL for non-SMTP local injection.
1994 Report and assistance in diagnosis by Warren Baker.
1996 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1998 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2000 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2001 last id that failed so it may be referenced in subsequent ACL's.
2003 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2006 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2007 ldap library initialization, allowing self-signed CA's to be
2008 used. Also properly sets require_cert option later in code by
2009 using NULL (global ldap config) instead of ldap handle (per
2010 session). Bug diagnosis and testing by alxgomz.
2012 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2013 the src/util/ subdirectory.
2015 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2016 renamed to Transport Post Delivery Action by Jeremy Harris, as
2019 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2020 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2021 redis_servers = needs to be configured which will be used by the redis
2022 lookup. Patch from Warren Baker, of The Packet Hub.
2024 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2026 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2027 hostname or reverse DNS when processing a host list. Used suggestions
2028 from multiple comments on this bug.
2030 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2032 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2033 Missed a few lines, added it to make the runtest require no keyboard
2036 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2037 contains upper case chars. Make router use caseful_local_part.
2039 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2040 support when GnuTLS has been built with p11-kit.
2046 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2048 This, or similar/improved, will also be change PP/11 of 4.82.
2054 PP/01 Handle short writes when writing local log-files.
2055 In practice, only affects FreeBSD (8 onwards).
2056 Bugzilla 1053, with thanks to Dmitry Isaikin.
2058 NM/01 Bugzilla 949 - Documentation tweak
2060 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2063 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2065 PP/02 Implemented gsasl authenticator.
2067 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2069 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2070 `pkg-config foo` for cflags/libs.
2072 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2073 with rest of GSASL and with heimdal_gssapi.
2075 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2076 `pkg-config foo` for cflags/libs for the TLS implementation.
2078 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2079 properties get this fed in as external SSF. A number of robustness
2080 and debugging improvements to the cyrus_sasl authenticator.
2082 PP/08 cyrus_sasl server now expands the server_realm option.
2084 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2085 Patch by Jeremy Harris.
2087 PP/10 Added dbmjz lookup type.
2089 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2091 PP/12 MAIL args handles TAB as well as SP, for better interop with
2092 non-compliant senders.
2093 Analysis and variant patch by Todd Lyons.
2095 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2096 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2097 Patch from Dirk Mueller <dmueller@suse.com>
2099 PP/13 tls_peerdn now print-escaped for spool files.
2100 Observed some $tls_peerdn in wild which contained \n, which resulted
2101 in spool file corruption.
2103 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2104 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2105 or write after TLS renegotiation, which otherwise led to messages
2108 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2109 as a tracking header (ie: a signed header comes before the signature).
2110 Patch from Wolfgang Breyha.
2112 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2113 comma-sep list; embedded commas doubled.
2115 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2117 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2119 Report and patch from Dmitry Banschikov.
2121 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2122 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2123 failures. We appear to now support TLS1.1+ with Exim.
2125 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2126 lets Exim select keys and certificates based upon TLS SNI from client.
2127 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2128 before an outbound SMTP session. New log_selector, +tls_sni.
2130 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2131 NULL dereference. Report and patch from Alun Jones.
2133 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2134 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2135 Not seeing resolver debug output on NetBSD, but suspect this is a
2136 resolver implementation change.
2138 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2139 Left warnings. Added "eximon gdb" invocation mode.
2141 PP/21 Defaulting "accept_8bitmime" to true, not false.
2143 PP/22 Added -bw for inetd wait mode support.
2145 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2146 locate the relevant includes and libraries. Made this the default.
2148 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2149 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2151 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2152 This may cause build issues on older platforms.
2154 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2155 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2156 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2157 Added SNI support via GnuTLS too.
2158 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2160 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2162 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2163 Applied second patch from Janne, implementing suggestion to default
2164 multiple-strings-in-record handling to match SPF spec.
2166 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2168 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2169 read-only, out of scope).
2170 Patch from Wolfgang Breyha, report from Stuart Northfield.
2172 PP/29 Fix three issues highlighted by clang analyser static analysis.
2173 Only crash-plausible issue would require the Cambridge-specific
2174 iplookup router and a misconfiguration.
2175 Report from Marcin Mirosław.
2177 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2179 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2180 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2181 As part of this, removing so much warning spew let me fix some minor
2182 real issues in debug logging.
2184 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2185 assignment on my part. Fixed.
2187 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2188 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2189 Janne Snabb (who went above and beyond: thank you).
2191 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2192 string otherwise requires a connection and a bunch more work and it's
2193 relatively easy to get wrong. Should also expose TLS library linkage
2196 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2197 64-bit ${eval} (JH/03).
2199 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2200 GNU libc to support some of the 64-bit stuff, should not lead to
2201 conflicts. Defined before os.h is pulled in, so if a given platform
2202 needs to override this, it can.
2204 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2205 protection layer was required, which is not implemented.
2206 Bugzilla 1254, patch from Wolfgang Breyha.
2208 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2209 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2210 tls_dhparam take prime identifiers. Also unbreak combination of
2211 OpenSSL+DH_params+TLSSNI.
2213 PP/39 Disable SSLv2 by default in OpenSSL support.
2219 PP/01 Solaris build fix for Oracle's LDAP libraries.
2220 Bugzilla 1109, patch from Stephen Usher.
2222 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2224 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2227 TF/02 Fix a couple more cases where we did not log the error message
2228 when unlink() failed. See also change 4.74-TF/03.
2230 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2231 lock up or crash if it happened to be inside a call to libc when it
2232 got a SIGUSR1 from exiwhat.
2234 The SIGUSR1 handler appends the current process status to the process
2235 log which is later printed by exiwhat. It used to use the general
2236 purpose logging code to do this, but several functions it calls are
2237 not safe for signals.
2239 The new output code in the SIGUSR1 handler is specific to the process
2240 log, and simple enough that it's easy to inspect for signal safety.
2241 Removing some special cases also simplifies the general logging code.
2242 Removing the spurious timestamps from the process log simplifies
2245 TF/04 Improved ratelimit ACL condition.
2247 The /noupdate option has been deprecated in favour of /readonly which
2248 has clearer semantics. The /leaky, /strict, and /readonly update modes
2249 are mutually exclusive. The update mode is no longer included in the
2250 database key; it just determines when the database is updated. (This
2251 means that when you upgrade Exim will forget old rate measurements.)
2253 Exim now checks that the per_* options are used with an update mode that
2254 makes sense for the current ACL. For example, when Exim is processing a
2255 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2256 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2257 must specify per_mail/readonly. If you omit the update mode it defaults to
2258 /leaky where that makes sense (as before) or /readonly where required.
2260 The /noupdate option is now undocumented but still supported for
2261 backwards compatibility. It is equivalent to /readonly except that in
2262 ACLs where /readonly is required you may specify /leaky/noupdate or
2263 /strict/noupdate which are treated the same as /readonly.
2265 A useful new feature is the /count= option. This is a generalization
2266 of the per_byte option, so that you can measure the throughput of other
2267 aggregate values. For example, the per_byte option is now equivalent
2268 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2270 The per_rcpt option has been generalized using the /count= mechanism
2271 (though it's more complicated than the per_byte equivalence). When it is
2272 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2273 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2274 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2275 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2276 Note that using per_rcpt with a non-readonly update mode in more than
2277 one ACL will cause the recipients to be double-counted. (The per_mail
2278 and per_byte options don't have this problem.)
2280 The handling of very low rates has changed slightly. If the computed rate
2281 is less than the event's count (usually one) then this event is the first
2282 after a long gap. In this case the rate is set to the same as this event's
2283 count, so that the first message of a spam run is counted properly.
2285 The major new feature is a mechanism for counting the rate of unique
2286 events. The new per_addr option counts the number of different
2287 recipients that someone has sent messages to in the last time period. It
2288 behaves like per_rcpt if all the recipient addresses are different, but
2289 duplicate recipient addresses do not increase the measured rate. Like
2290 the /count= option this is a general mechanism, so the per_addr option
2291 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2292 example, measure the rate that a client uses different sender addresses
2293 with the options per_mail/unique=$sender_address. There are further
2294 details in the main documentation.
2296 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2298 TF/06 Removed a few PCRE remnants.
2300 TF/07 Automatically extract Exim's version number from tags in the git
2301 repository when doing development or release builds.
2303 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2304 Bugzilla 879. Patch from Paul Fisher.
2306 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2307 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2310 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2312 PP/05 Variable $av_failed, true if the AV scanner deferred.
2313 Bugzilla 1078. Patch from John Horne.
2315 PP/06 Stop make process more reliably on build failure.
2316 Bugzilla 1087. Patch from Heiko Schlittermann.
2318 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2319 Bugzilla 1089. Patch from Heiko Schlittermann.
2321 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2322 Bugzilla 1131. Patch from Holger Weiß.
2324 PP/09 Handle IPv6 addresses with SPF.
2325 Bugzilla 860. Patch from Wolfgang Breyha.
2327 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2329 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2332 PP/11 match_* no longer expand right-hand-side by default.
2333 New compile-time build option, EXPAND_LISTMATCH_RHS.
2334 New expansion conditions, "inlist", "inlisti".
2336 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2338 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2340 PP/14 fix log_write() format string regression from TF/03.
2341 Bugzilla 1152. Patch from Dmitry Isaikin.
2347 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2349 PP/02 Harmonised TLS library version reporting; only show if debugging.
2350 Layout now matches that introduced for other libraries in 4.74 PP/03.
2352 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2354 PP/04 New "dns_use_edns0" global option.
2356 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2359 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2360 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2362 TK/01 Updated PolarSSL code to 0.14.2.
2363 Bugzilla 1097. Patch from Andreas Metzler.
2365 PP/07 Catch divide-by-zero in ${eval:...}.
2366 Fixes bugzilla 1102.
2368 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2371 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2372 format-string attack -- SECURITY: remote arbitrary code execution.
2374 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2375 time unintentionally subject to list matching rules, letting the header
2376 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2377 arbitrary string expansion). This allowed for information disclosure.
2379 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2380 INT_MIN/-1 -- value coerced to INT_MAX.
2386 NM/01 Workaround for PCRE version dependency in version reporting
2389 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2390 This fixes portability to compilers other than gcc, notably
2391 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2393 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2394 makefiles for portability to HP-UX and POSIX correctness.
2396 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2397 Also via indented variable definition in the Makefile.
2398 (Debugging by Oliver Heesakkers).
2400 PP/02 Restore caching of spamd results with expanded spamd_address.
2401 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2403 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2404 Improves build reliability. Fix from: Frank Elsner
2406 NM/02 Fix wide character breakage in the rfc2047 coding
2407 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2409 NM/03 Allow underscore in dnslist lookups
2410 Fixes bug 1026. Patch from Graeme Fowler
2412 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2413 Code patches from Adam Ciarcinski of NetBSD.
2415 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2418 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2419 is logged, to avoid truncation. Patch from John Horne.
2421 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2422 Patch from Jakob Hirsch.
2424 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2425 SQL string expansion failure details.
2426 Patch from Andrey Oktyabrski.
2428 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2429 Patch from Simon Arlott.
2431 PP/09 New lookups functionality failed to compile on old gcc which rejects
2432 extern declarations in function scope.
2433 Patch from Oliver Fleischmann
2435 PP/10 Use sig_atomic_t for flags set from signal handlers.
2436 Check getgroups() return and improve debugging.
2437 Fixed developed for diagnosis in bug 927 (which turned out to be
2440 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2441 Patch from Mark Zealey.
2443 PP/12 Bugzilla 1056: Improved spamd server selection.
2444 Patch from Mark Zealey.
2446 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2447 Based on patch from Heiko Schlittermann.
2449 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2450 Patch from Uwe Doering, sign-off by Michael Haardt.
2452 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2453 variable declaration deep within a block. Bug and patch from
2456 PP/15 lookups-Makefile IRIX compatibility coercion.
2458 PP/16 Make DISABLE_DKIM build knob functional.
2460 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2461 Patch by Simon Arlott
2463 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2464 variable argument macros. Our copy now differs from upstream.
2470 TF/01 Failure to get a lock on a hints database can have serious
2471 consequences so log it to the panic log.
2473 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2474 controlled using the smtp_confirmation log selector.
2476 TF/03 Include the error message when we fail to unlink a spool file.
2478 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2479 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2480 for maintaining out-of-tree patches for some time.
2482 PP/01 Bugzilla 139: Documentation and portability issues.
2483 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2484 Handle per-OS dynamic-module compilation flags.
2486 PP/02 Let /dev/null have normal permissions.
2487 The 4.73 fixes were a little too stringent and complained about the
2488 permissions on /dev/null. Exempt it from some checks.
2489 Reported by Andreas M. Kirchwitz.
2491 PP/03 Report version information for many libraries, including
2492 Exim version information for dynamically loaded libraries. Created
2493 version.h, now support a version extension string for distributors
2494 who patch heavily. Dynamic module ABI change.
2496 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2497 privilege escalation vulnerability whereby the Exim run-time user
2498 can cause root to append content of the attacker's choosing to
2501 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2504 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2505 If dropping privileges for untrusted macros, we disabled normal logging
2506 on the basis that it would fail; for the Exim run-time user, this is not
2507 the case, and it resulted in successful deliveries going unlogged.
2508 Fixed. Reported by Andreas Metzler.
2514 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2515 only prepend for the Resent-* case. Fixes regression introduced in
2516 Exim 4.70 by NM/22 for Bugzilla 607.
2518 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2519 increasing numbers of administrators be bitten by this.
2521 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2523 PP/03 Bugzilla 994: added openssl_options main configuration option.
2525 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2527 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2529 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2531 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2532 without a peer certificate, leading to a segfault because of an
2533 assumption that peers always have certificates. Be a little more
2534 paranoid. Problem reported by Martin Tscholak.
2536 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2537 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2538 NB: ClamAV planning to remove STREAM in "middle of 2010".
2539 CL also introduces -bmalware, various -d+acl logging additions and
2540 more caution in buffer sizes.
2542 PP/09 Implemented reverse_ip expansion operator.
2544 PP/10 Bugzilla 937: provide a "debug" ACL control.
2546 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2548 PP/12 Bugzilla 973: Implement --version.
2550 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2552 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2554 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2556 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2557 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2558 ignore trailing whitespace.
2560 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2562 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2565 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2566 Notification from Dr Andrew Aitchison.
2568 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2569 ExtendedDetectionInfo response format.
2570 Notification from John Horne.
2572 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2575 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2576 XSL and documented dependency on system catalogs, with examples of how
2579 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2582 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2583 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2584 configuration file which is writeable by the Exim user or group.
2586 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2587 of configuration files to cover files specified with the -C option if
2588 they are going to be used with root privileges, not just the default
2591 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2592 option (effectively making it always true).
2594 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2595 files to be used while preserving root privileges.
2597 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2598 that rogue child processes cannot use them.
2600 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2601 run-time user, instead of root.
2603 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2604 Exim run-time user without dropping privileges.
2606 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2607 result string, instead of calling string_vformat() twice with the same
2610 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2611 for other users. Others should always drop root privileges if they use
2612 -C on the command line, even for a whitelisted configure file.
2614 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2616 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2622 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2623 $data_path, and $header_path variables; fixed documentation bugs and
2626 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2627 exipick to access non-standard spools, including the "frozen" queue
2630 NM/01 Bugzilla 965: Support mysql stored procedures.
2631 Patch from Alain Williams
2633 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2635 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2636 Patch from Andreas Metzler
2638 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2639 Patch from Kirill Miazine
2641 NM/05 Bugzilla 671: Added umask to procmail example.
2643 JJ/03 installed exipick 20100323.0, fixing doc bug
2645 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2646 directory. Notification and patch from Dan Rosenberg.
2648 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2650 TK/02 Improve log output when DKIM signing operation fails.
2652 MH/01 Treat the transport option dkim_domain as a colon separated
2653 list, not as a single string, and sign the message with each element,
2654 omitting multiple occurences of the same signer.
2656 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2657 Bugzilla 985, 986. Patch by Simon Arlott
2659 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2660 Patch by Simon Arlott
2662 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2663 MBX locking. Notification from Dan Rosenberg.
2669 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2671 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2673 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2675 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2677 NM/04 Bugzilla 915: Flag broken perl installation during build.
2683 TK/01 Added patch by Johannes Berg that expands the main option
2684 "spamd_address" if it starts with a dollar sign.
2686 TK/02 Write list of recipients to X-Envelope-Sender header when building
2687 the mbox-format spool file for content scanning (suggested by Jakob
2690 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2691 (http://www.dcc-servers.net/) support via dccifd. Activated by
2692 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2694 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2695 by Mark Daniel Reidel <mr@df.eu>.
2697 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2698 When building exim an external PCRE library is now needed -
2699 PCRE is a system library on the majority of modern systems.
2700 See entry on PCRE_LIBS in EDITME file.
2702 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2703 conversation. Added nologin parameter to request.
2704 Patch contributed by Kirill Miazine.
2706 TF/01 Do not log submission mode rewrites if they do not change the address.
2708 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2710 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2711 log files in place. Contributed by Roberto Lima.
2713 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2715 TF/03 Bugzilla 615: When checking the local_parts router precondition
2716 after a local_part_suffix or local_part_prefix option, Exim now
2717 does not use the address's named list lookup cache, since this
2718 contains cached lookups for the whole local part.
2720 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2721 Robert Millan. Documentation is in experimental-spec.txt.
2723 TF/04 Bugzilla 668: Fix parallel build (make -j).
2725 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2727 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2728 Patch provided by Jan Srzednicki.
2730 TF/05 Leading white space used to be stripped from $spam_report which
2731 wrecked the formatting. Now it is preserved.
2733 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2734 that they are available at delivery time.
2736 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2738 TF/08 TLS error reporting now respects the incoming_interface and
2739 incoming_port log selectors.
2741 TF/09 Produce a more useful error message if an SMTP transport's hosts
2742 setting expands to an empty string.
2744 NM/06 Bugzilla 744: EXPN did not work under TLS.
2745 Patch provided by Phil Pennock.
2747 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2748 Patch provided by Richard Godbee.
2750 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2751 acl_smtp_notquit, added index entry.
2753 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2754 Patch provided by Eugene Bujak.
2756 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2757 accept(). Patch provided by Maxim Dounin.
2759 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2760 Patch provided by Phil Pennock.
2762 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2764 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2765 Patch provided by Brad "anomie" Jorsch.
2767 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2768 Patch provided by Dean Brooks.
2770 TK/05 Add native DKIM support (does not depend on external libraries).
2772 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2773 Patch provided by Graeme Fowler.
2775 NM/16 Bugzilla 851: Documentation example syntax fix.
2777 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2779 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2782 NM/19 Bugzilla 745: TLS version reporting.
2783 Patch provided by Phil Pennock.
2785 NM/20 Bugzilla 167: bool: condition support.
2786 Patch provided by Phil Pennock.
2788 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2789 clients. Patch provided by Phil Pennock.
2791 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2792 Patch provided by Brad "anomie" Jorsch.
2794 NM/23 Bugzilla 687: Fix misparses in eximstats.
2795 Patch provided by Heiko Schlittermann.
2797 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2798 Patch provided by Heiko Schlittermann.
2800 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2801 plus update to original patch.
2803 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2805 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2806 Patch provided by David Brownlee.
2808 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2810 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2812 NM/30 Bugzilla 888: TLS documentation bugfixes.
2814 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2816 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2817 Unlike the original bugzilla I have changed all shell scripts in src tree.
2819 NM/33 Bugzilla 898: Transport filter timeout fix.
2820 Patch by Todd Rinaldo.
2822 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2823 Patch by Serge Demonchaux.
2825 NM/35 Bugzilla 39: Base64 decode bug fixes.
2826 Patch by Jakob Hirsch.
2828 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2830 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2832 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2834 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2840 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2841 ALT-N's libdkim that I have put here:
2842 http://duncanthrax.net/exim-experimental/
2844 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2845 were called 'true' and it seems that C99 defines that as a reserved
2846 keyword to be used with 'bool' variable types. That means you could
2847 not include C99-style headers which use bools without triggering
2848 build errors in sieve.c.
2850 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2851 as mailq or other aliases. Changed the --help handling significantly
2852 to do whats expected. exim_usage() emits usage/help information.
2854 SC/01 Added the -bylocaldomain option to eximstats.
2856 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2858 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2860 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2866 PH/01 Another patch from the Sieve maintainer.
2868 PH/02 When an IPv6 address is converted to a string for single-key lookup
2869 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2870 dots are used instead of colons so that keys in lsearch files need not
2871 contain colons. This was done some time before quoting was made available
2872 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2873 (notated using the quote facility) so as to distinguish them from IPv4
2874 keys. This meant that lookups for IP addresses in host lists did not work
2875 for iplsearch lookups.
2877 This has been fixed by arranging for IPv6 addresses to be expressed with
2878 colons if the lookup type is iplsearch. This is not incompatible, because
2879 previously such lookups could never work.
2881 The situation is now rather anomalous, since one *can* have colons in
2882 ordinary lsearch keys. However, making the change in all cases is
2883 incompatible and would probably break a number of configurations.
2885 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2888 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2889 conversion specification without a maximum field width, thereby enabling
2890 a rogue spamd server to cause a buffer overflow. While nobody in their
2891 right mind would setup Exim to query an untrusted spamd server, an
2892 attacker that gains access to a server running spamd could potentially
2893 exploit this vulnerability to run arbitrary code as the Exim user.
2895 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2896 $primary_hostname instead of what libspf2 thinks the hosts name is.
2898 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2899 a directory entry by the name of the lookup key. Previously, if a
2900 symlink pointed to a non-existing file or a file in a directory that
2901 Exim lacked permissions to read, a lookup for a key matching that
2902 symlink would fail. Now it is enough that a matching directory entry
2903 exists, symlink or not. (Bugzilla 503.)
2905 PH/03 The body_linecount and body_zerocount variables are now exported in the
2908 PH/04 Added the $dnslist_matched variable.
2910 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2911 This means they are set thereafter only if the connection becomes
2914 PH/06 Added the client_condition to authenticators so that some can be skipped
2915 by clients under certain conditions.
2917 PH/07 The error message for a badly-placed control=no_multiline_responses left
2918 "_responses" off the end of the name.
2920 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2922 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2923 (without spaces) instead of just copying the configuration text.
2925 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2927 PH/11 Added $max_received_linelength.
2929 PH/12 Added +ignore_defer and +include_defer to host lists.
2931 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2932 way in which PCRE > 7.0 is built.
2934 PH/14 Implemented queue_only_load_latch.
2936 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2937 MAIL command. The effect was to mangle the value on 64-bit systems.
2939 PH/16 Another patch from the Sieve maintainer.
2941 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2943 PH/18 If a system quota error occurred while trying to create the file for
2944 a maildir delivery, the message "Mailbox is full" was not appended to the
2945 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2946 only to a quota excession during the actual writing of the file.
2948 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2949 characters?) which causes problems in log lines. The DN values are now
2950 passed through string_printing() before being added to log lines.
2952 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2953 and InterBase are left for another time.)
2955 PH/21 Added message_body_newlines option.
2957 PH/22 Guard against possible overflow in moan_check_errorcopy().
2959 PH/23 POSIX allows open() to be a macro; guard against that.
2961 PH/24 If the recipient of an error message contained an @ in the local part
2962 (suitably quoted, of course), incorrect values were put in $domain and
2963 $local_part during the evaluation of errors_copy.
2969 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2970 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2973 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2974 issue a MAIL command.
2976 PH/02 In an ACL statement such as
2978 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2980 if a client was not listed at all, or was listed with a value other than
2981 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2982 the condition was not true (as it should be), so access was not denied.
2983 The bug was that the ! inversion was incorrectly passed on to the second
2984 item. This has been fixed.
2986 PH/03 Added additional dnslists conditions == and =& which are different from
2987 = and & when the dns lookup returns more than one IP address.
2989 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2990 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2992 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2993 FSYNC, which compiles an option called disable_fsync that allows for
2994 bypassing fsync(). The documentation is heavily laced with warnings.
2996 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2998 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2999 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3000 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3001 including adding "make clean"; (3) Added -fPIC when compiling the test
3002 dynamically loaded module, to get rid of a warning.
3004 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3005 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3006 = 0s. The bug is otherwise harmless.
3008 PH/07 There was a bug in the dovecot authenticator such that the value of
3009 $auth1 could be overwritten, and so not correctly preserved, after a
3010 successful authentication. This usually meant that the value preserved by
3011 the server_setid option was incorrect.
3013 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3015 PH/09 Installed PCRE release 7.0.
3017 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3018 run for batched SMTP input. It is now run at the start of every message
3019 in the batch. While fixing this I discovered that the process information
3020 (output by running exiwhat) was not always getting set for -bs and -bS
3021 input. This is fixed, and it now also says "batched" for BSMTP.
3023 PH/11 Added control=no_pipelining.
3025 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3026 patch, slightly modified), and move the expansion of helo_data till after
3027 the connection is made in the smtp transport (so it can use these
3030 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3032 PH/14 Added log_selector = +pid.
3034 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3036 PH/16 Add ${if forany and ${if forall.
3038 PH/17 Added dsn_from option to vary the From: line in DSNs.
3040 PH/18 Flush SMTP output before performing a callout, unless control =
3041 no_callout_flush is set.
3043 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3044 was true (the default) a successful delivery failed to delete the retry
3045 item, thus causing premature timeout of the address. The bug is now
3048 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3050 PH/21 Long custom messages for fakedefer and fakereject are now split up
3051 into multiline responses in the same way that messages for "deny" and
3052 other ACL rejections are.
3054 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3055 with slight modification.
3057 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3058 draft, changing the syntax and factoring some duplicate code".
3060 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3061 for deliveries of the second and subsequent messages over the same SMTP
3064 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3065 ${reduce, with only minor "tidies".
3067 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3069 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3070 expansion side effects.
3072 PH/27 When a message times out after an over-quota error from an Exim-imposed
3073 quota, the bounce message says "mailbox is full". This message was not
3074 being given when it was a system quota that was exceeded. It now should
3077 MH/03 Made $recipients available in local_scan(). local_scan() already has
3078 better access to the recipient list through recipients_list[], but
3079 $recipients can be useful in postmaster-provided expansion strings.
3081 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3082 in the case of a MAIL command with additional options following the
3083 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3084 were accidentally chopped off.
3086 PH/29 SMTP synchronization checks are implemented when a command is read -
3087 there is a check that no more input is waiting when there shouldn't be
3088 any. However, for some commands, a delay in an ACL can mean that it is
3089 some time before the response is written. In this time, more input might
3090 arrive, invalidly. So now there are extra checks after an ACL has run for
3091 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3092 pipelining has not been advertised.
3094 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3096 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3097 returned in the SMTP response when smtp_return_error_details was false.
3098 This has been fixed.
3100 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3101 instead of the C I/O that was originally supplied, because problems were
3102 reported on Solaris.
3104 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3105 Exim which did not show up earlier: it was assuming that a call to
3106 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3107 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3108 was a macro that became an assignment, so it seemed to work. This has
3109 changed to a proper function call with a void return, hence the compile
3110 error. Exim's code has been fixed.
3112 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3115 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3117 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3119 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3120 because it turns out to be too big or there is a timeout) there is no
3121 "Completed" line in the log. When some messages of this type were
3122 selected by exigrep, they were listed as "not completed". Others were
3123 picked up by some special patterns. I have improved the selection
3124 criteria to be more general.
3126 PH/38 The host_find_failed option in the manualroute router can now be set
3127 to "ignore", to completely ignore a host whose IP address cannot be
3128 found. If all hosts are ignored, the behaviour is controlled by the new
3129 host_all_ignored option.
3131 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3132 homing or because of multiple MX records with /mx) generated more than
3133 one IP address, and the following item turned out to be the local host,
3134 all the secondary addresses of the first item were incorrectly removed
3135 from the list, along with the local host and any following hosts (which
3136 is what is supposed to happen).
3138 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3139 whoever called Exim into the -H file. In the case of the daemon it was
3140 behaving confusingly. When first started, it used values for whoever
3141 started the daemon, but after a SIGHUP it used the Exim user (because it
3142 calls itself on a restart). I have changed the code so that it now always
3145 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3146 message are rejected with the same error (e.g. no authentication or bad
3147 sender address), and a DATA command is nevertheless sent (as can happen
3148 with PIPELINING or a stupid MUA), the error message that was given to the
3149 RCPT commands is included in the rejection of the DATA command. This is
3150 intended to be helpful for MUAs that show only the final error to their
3153 PH/42 Another patch from the Sieve maintainer.
3155 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3156 Eximstats - Fixed some broken HTML links and added missing column headers
3158 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3159 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3161 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3163 PH/43 Yet another patch from the Sieve maintainer.
3165 PH/44 I found a way to check for a TCP/IP connection going away before sending
3166 the response to the final '.' that terminates a message, but only in the
3167 case where the client has not sent further data following the '.'
3168 (unfortunately, this is allowed). However, in many cases there won't be
3169 any further data because there won't be any more messages to send. A call
3170 to select() can be used: if it shows that the input is "ready", there is
3171 either input waiting, or the socket has been closed. An attempt to read
3172 the next input character can distinguish the two cases. Previously, Exim
3173 would have sent an OK response which the client would never have see.
3174 This could lead to message repetition. This fix should cure that, at
3175 least in a lot of common cases.
3177 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3178 advertised in response to EHLO.
3184 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3185 fixed by 4.65/MH/01 (is this a record?) are fixed:
3187 (i) An empty string was always treated as zero by the numeric comparison
3188 operators. This behaviour has been restored.
3190 (ii) It is documented that the numeric comparison operators always treat
3191 their arguments as decimal numbers. This was broken in that numbers
3192 starting with 0 were being interpreted as octal.
3194 While fixing these problems I realized that there was another issue that
3195 hadn't been noticed. Values of message_size_limit (both the global option
3196 and the transport option) were treated as octal if they started with 0.
3197 The documentation was vague. These values are now always treated as
3198 decimal, and I will make that clear in the documentation.
3204 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3205 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3208 MH/01 Don't check that the operands of numeric comparison operators are
3209 integers when their expansion is in "skipping" mode (fixes bug
3210 introduced by 4.64-PH/07).
3212 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3213 child addresses, Exim now panics and dies. Previously, because the count
3214 is held in a short int, deliveries were likely to be lost. As such a
3215 large number of recipients for a single message is ridiculous
3216 (performance will be very, very poor), I have chosen to impose a limit
3217 rather than extend the field.
3223 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3224 leftover -K file (the existence of which was triggered by #402).
3225 While we were at it, introduced process PID as part of the -K
3226 filename. This should rule out race conditions when creating
3229 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3230 processing considerably. Previous code took too long for large mails,
3231 triggering a timeout which in turn triggers #401.
3233 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3234 in the DK code in transports.c. sendfile() is not really portable,
3235 hence the _LINUX specificness.
3237 TF/01 In the add_headers option to the mail command in an Exim filter,
3238 there was a bug that Exim would claim a syntax error in any
3239 header after the first one which had an odd number of characters
3242 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3243 callout verification, Exim cached a "reject" for the entire domain. This
3244 is correct for most verifications, but it is not correct for a recipient
3245 verification with use_sender or use_postmaster set, because in that case
3246 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3247 case of MAIL FROM:<> rejection from other early rejections (e.g.
3248 rejection of HELO). When verifying a recipient using a non-null MAIL
3249 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3250 Whatever the result of the callout, the value of the domain cache is
3251 left unchanged (for any other kind of callout, getting as far as trying
3252 RCPT means that the domain itself is ok).
3254 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3257 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3258 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3259 session, the daemon ignores ECONNECT errors and logs others; it now
3260 ignores EPIPE as well.
3262 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3263 (quoted-printable decoding).
3265 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3266 later the small subsequent patch to fix an introduced bug.
3268 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3270 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3272 PH/08 An error is now given if message_size_limit is specified negative.
3274 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3275 to be given (somewhat) arbitrary names.
3277 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3280 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3281 miscellaneous code fixes
3283 PH/10 Added the log_reject_target ACL modifier to specify where to log
3286 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3287 hostname. This is wrong, because it relates to the incoming message (and
3288 probably the interface on which it is arriving) and not to the outgoing
3289 callout (which could be using a different interface). This has been
3290 changed to use the value of the helo_data option from the smtp transport
3291 instead - this is what is used when a message is actually being sent. If
3292 there is no remote transport (possible with a router that sets up host
3293 addresses), $smtp_active_hostname is used.
3295 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3296 tweaks were necessary in order to get it to work (see also 21 below):
3297 (a) The code assumed that strncpy() returns a negative number on buffer
3298 overflow, which isn't the case. Replaced with Exim's string_format()
3300 (b) There were several signed/unsigned issues. I just did the minimum
3301 hacking in of casts. There is scope for a larger refactoring.
3302 (c) The code used strcasecmp() which is not a standard C function.
3303 Replaced with Exim's strcmpic() function.
3304 (d) The code set only $1; it now sets $auth1 as well.
3305 (e) A simple test gave the error "authentication client didn't specify
3306 service in request". It would seem that Dovecot has changed its
3307 interface. Fortunately there's a specification; I followed it and
3308 changed what the client sends and it appears to be working now.
3310 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3313 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3314 address A is aliased to B and C, where B exists and C does not. Without
3315 -v the output is "A verified" because verification stops after a
3316 successful redirection if more than one address is generated. However,
3317 with -v the child addresses are also verified. Exim was outputting "A
3318 failed to verify" and then showing the successful verification for C,
3319 with its parentage. It now outputs "B failed to verify", showing B's
3320 parentage before showing the successful verification of C.
3322 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3323 look up a TXT record in a specific list after matching in a combined
3326 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3327 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3328 they consult the DNS. I had assumed they would set it the way they
3329 wanted; and indeed my experiments on Linux seem to show that in some
3330 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3331 To be on the safe side, however, I have now made the interface to
3332 host_find_byname() similar to host_find_bydns(), with an argument
3333 containing the DNS resolver options. The host_find_byname() function now
3334 sets these options at its start, just as host_find_bydns() does. The smtp
3335 transport options dns_qualify_single and dns_search_parents are passed to
3336 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3337 of host_find_byname() use the default settings of RES_DEFNAMES
3338 (qualify_single) but not RES_DNSRCH (search_parents).
3340 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3341 spool_read_header() do less string testing, by means of a preliminary
3342 switch on the second character of optional "-foo" lines. (This is
3343 overdue, caused by the large number of possibilities that now exist.
3344 Originally there were few.) While I was there, I also converted the
3345 str(n)cmp tests so they don't re-test the leading "-" and the first
3346 character, in the hope this might squeeze out yet more improvement.
3348 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3349 flag allowing group syntax was set by the header_syntax check but not
3350 turned off, possible causing trouble later; (2) The flag was not being
3351 set at all for the header_verify test, causing "group"-style headers to
3352 be rejected. I have now set it in this case, and also caused header_
3353 verify to ignore an empty address taken from a group. While doing this, I
3354 came across some other cases where the code for allowing group syntax
3355 while scanning a header line wasn't quite right (mostly, not resetting
3356 the flag correctly in the right place). These bugs could have caused
3357 trouble for malformed header lines. I hope it is now all correct.
3359 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3360 with the "reply" argument non-NULL. The code, however (which originally
3361 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3362 but it didn't always do it. This confused somebody who was copying the
3363 code for some other use. I have removed all the tests.
3365 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3366 feature that was used to support insecure browsers during the U.S. crypto
3367 embargo. It requires special client support, and Exim is probably the
3368 only MTA that supported it -- and would never use it because real RSA is
3369 always available. This code has been removed, because it had the bad
3370 effect of slowing Exim down by computing (never used) parameters for the
3371 RSA_EXPORT functionality.
3373 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3374 authenticator to fail if there's a tab character in the incoming data
3375 (there should never be unless someone is messing about, as it's supposed
3376 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3377 if the connection is using TLS or if the remote IP is the same as the
3378 local IP, and the "valid-client-cert option" if a client certificate has
3381 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3382 authenticators. This can be used for authorization after authentication
3383 succeeds. (In the case of plaintext, it servers for both authentication
3386 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3387 if any retry times were supplied.
3389 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3390 connection, where there is no client IP address to check. In this
3391 situation, the verify now always succeeds.
3393 PH/25 Applied John Jetmore's -Mset patch.
3395 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3397 PH/27 In a string expansion for a processed (not raw) header when multiple
3398 headers of the same name were present, leading whitespace was being
3399 removed from all of them, but trailing whitespace was being removed only
3400 from the last one. Now trailing whitespace is removed from each header
3401 before concatenation. Completely empty headers in a concatenation (as
3402 before) are ignored.
3404 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3405 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3407 PH/29 [Removed. This was a change that I later backed out, and forgot to
3408 correct the ChangeLog entry (that I had efficiently created) before
3409 committing the later change.]
3411 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3412 address errors (4xx response to RCPT) over the same connection as other
3413 messages routed to the same hosts. Such deliveries are always "forced",
3414 so retry times are not inspected. This resulted in far too many retries
3415 for the affected addresses. The effect occurred only when there were more
3416 hosts than the hosts_max_try setting in the smtp transport when it had
3417 the 4xx errors. Those hosts that it had tried were not added to the list
3418 of hosts for which the message was waiting, so if all were tried, there
3419 was no problem. Two fixes have been applied:
3421 (i) If there are any address or message errors in an SMTP delivery, none
3422 of the hosts (tried or untried) are now added to the list of hosts
3423 for which the message is waiting, so the message should not be a
3424 candidate for sending over the same connection that was used for a
3425 successful delivery of some other message. This seems entirely
3426 reasonable: after all the message is NOT "waiting for some host".
3427 This is so "obvious" that I'm not sure why it wasn't done
3428 previously. Hope I haven't missed anything, but it can't do any
3429 harm, as the worst effect is to miss an optimization.
3431 (ii) If, despite (i), such a delivery is accidentally attempted, the
3432 routing retry time is respected, so at least it doesn't keep
3433 hammering the server.
3435 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3436 in ${readsocket because some servers need this prod.
3438 PH/32 Added some extra debug output when updating a wait-xxx database.
3440 PH/33 The hint "could be header name not terminated by colon", which has been
3441 given for certain expansion errors for a long time, was not being given
3442 for the ${if def:h_colon_omitted{... case.
3444 PH/34 The spec says: "With one important exception, whenever a domain list is
3445 being scanned, $domain contains the subject domain." There was at least
3446 one case where this was not true.
3448 PH/35 The error "getsockname() failed: connection reset by peer" was being
3449 written to the panic log as well as the main log, but it isn't really
3450 panic-worthy as it just means the connection died rather early on. I have
3451 removed the panic log writing for the ECONNRESET error when getsockname()
3454 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3455 runs only) independently of the message's sender address. This meant
3456 that, if the 4xx error was in fact related to the sender, a different
3457 message to the same recipient with a different sender could confuse
3458 things. In particular, this can happen when sending to a greylisting
3459 server, but other circumstances could also provoke similar problems.
3460 I have changed the default so that the retry time for these errors is now
3461 based a combination of the sender and recipient addresses. This change
3462 can be overridden by setting address_retry_include_sender=false in the
3465 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3466 remote server are returned as part of bounce messages. This was not
3467 happening for LMTP over a pipe (the lmtp transport), but now it is the
3468 same for both kinds of LMTP.
3470 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3471 in header lines that were in fact CNAMEs. This is no longer the case.
3473 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3474 and queue runs started by the daemon processed all messages. This has
3475 been fixed so that -R and -S can now usefully be given with -q<time>.
3477 PH/40 Import PCRE release 6.7 (fixes some bugs).
3479 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3481 PH/42 Give an error if -q is specified more than once.
3483 PH/43 Renamed the variables $interface_address and $interface_port as
3484 $received_ip_address and $received_port, to make it clear that these
3485 values apply to message reception, and not to the outgoing interface when
3486 a message is delivered. (The old names remain recognized, of course.)
3488 PH/44 There was no timeout on the connect() call when using a Unix domain
3489 socket in the ${readsocket expansion. There now is.
3491 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3492 be meaningful with "accept".
3494 SC/01 Eximstats V1.43
3495 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3497 SC/02 Eximstats V1.44
3498 Use a glob alias rather than an array ref in the generated
3499 parser. This improves both readability and performance.
3501 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3502 Collect SpamAssassin and rejection statistics.
3503 Don't display local sender or destination tables unless
3504 there is data to show.
3505 Added average volumes into the top table text output.
3507 SC/04 Eximstats V1.46
3508 Collect data on the number of addresses (recipients)
3509 as well as the number of messages.
3511 SC/05 Eximstats V1.47
3512 Added 'Message too big' to the list of mail rejection
3513 reasons (thanks to Marco Gaiarin).
3515 SC/06 Eximstats V1.48
3516 Mainlog lines which have GMT offsets and are too short to
3517 have a flag are now skipped.
3519 SC/07 Eximstats V1.49 (Alain Williams)
3520 Added the -emptyok flag.
3522 SC/08 Eximstats V1.50
3523 Fixes for obtaining the IP address from reject messages.
3525 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3526 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3527 whitespace changes from 4.64-PH/27
3529 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3532 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3533 are found, allow negative numbers in numeric criteria)
3535 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3537 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3538 to match changes made in 4.64-PH/43
3540 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3542 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3543 "domain" is actually an IP address, and force a failure. This locks out
3544 those revolvers/nameservers that support "A-for-A" lookups, in
3545 contravention of the specifications.
3547 PH/48 When a host name was looked up from an IP address, and the subsequent
3548 forward lookup of the name timed out, the host name was left in
3549 $sender_host_name, contrary to the specification.
3551 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3552 restricted to single-key lookups, Exim was not diagnosing an error if
3553 * or *@ was used with a query-style lookup.
3555 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3557 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3558 long ago, but noone interested enough thought of it. Let's just say that
3559 the "1.1" means that there are some new functions that weren't there at
3560 some point in the past.
3562 PH/51 Error processing for expansion failure of helo_data from an smtp
3563 transport during callout processing was broken.
3565 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3566 tested/used via the -bh/-bhc/-bs options.
3568 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3569 bug, fixed in subsequent PCRE releases).
3571 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3572 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3574 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3580 SC/01 Use a glob alias rather than an array ref in eximstats generated
3581 parser. This improves both readability and performance.
3583 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3584 Don't display local sender or destination tables in eximstats unless
3585 there is data to show.
3586 Added average volumes into the eximstats top table text output.
3588 SC/03 Collect data on the number of addresses (recipients) as well
3589 as the number of messages in eximstats.
3591 TF/01 Correct an error in the documentation for the redirect router. Exim
3592 does (usually) call initgroups() when daemonizing.
3594 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3595 with consistent privilege compared to when running as a daemon.
3597 TF/03 Note in the spec that $authenticated_id is not set for local
3598 submissions from trusted users.
3600 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3601 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3603 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3604 by adding some example configuration directives to the default
3605 configuration file. A little bit of work is required to uncomment the
3606 directives and define how usernames and passwords are checked, but
3607 there is now a framework to start from.
3609 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3610 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3611 without this. I don't know how relevant this is to other LDAP libraries.
3613 PH/02 Add the verb name to the "unknown ACL verb" error.
3615 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3617 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3619 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3620 directory not expanded when it should be if an expanded home directory
3621 was set for the address (which is overridden by the transport).
3623 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3626 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3627 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3628 because it is too late at that time, and has no effect.
3630 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3631 security issue with \' (bugzilla #107). I could not use the
3632 PQescapeStringConn() function, because it needs a PGconn value as one of
3635 PH/08 When testing addresses using -bt, indicate those final addresses that
3636 are duplicates that would not cause an additional delivery. At least one
3637 person was confused, thinking that -bt output corresponded to deliveries.
3638 (Suppressing duplicates isn't a good idea as you lose the information
3639 about possibly different redirections that led to the duplicates.)
3641 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3642 systems where poll() doesn't work, in particular OS X.
3644 PH/10 Added more information to debugging output for retry time not reached.
3646 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3647 operations in malware.c.
3649 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3652 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3653 syslog_duplication set false, log lines that would normally be written
3654 both the the main log and to the reject log were not written to syslog at
3657 PH/14 In the default configuration, change the use of "message" in ACL warn
3658 statements to "add_header".
3660 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3661 not followed by a command (e.g. "seen endif").
3663 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3664 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3667 PH/17 Added extra conditions to the default value of delay_warning_condition
3671 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3672 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3673 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3676 The Auto-Submitted: and various List- headers are standardised, whereas I
3677 don't think Precedence: ever was.
3679 PH/18 Refactored debugging code in route_finduser() to show more information,
3680 in particular, the error code if getpwnam() issues one.
3682 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3683 This is apparently needed in addition to the PH/07 change above to avoid
3684 any possible encoding problems.
3686 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3687 but not after initializing Perl.
3689 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3690 output them only if debugging. By default they are written stderr,
3691 apparently, which is not desirable.
3693 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3696 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3699 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3701 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3702 authenticated or an ident call has been made. Suppress the default
3703 values for $authenticated_id and $authenticated_sender (but permit -oMai
3704 and -oMas) when testing with -bh.
3706 PH/24 Re-jigged the order of the tests in the default configuration so that the
3707 tests for valid domains and recipients precede the DNS black list and CSA
3708 tests, on the grounds that those ones are more expensive.
3710 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3711 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3712 This bug exists in every version of Exim that I still have, right back to
3715 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3716 However, an attempt to turn on case-sensitivity in a regex key by
3717 including (?-i) didn't work because the subject string was already
3718 lowercased, and the effects were non-intuitive. It turns out that a
3719 one-line patch can be used to allow (?-i) to work as expected.
3725 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3726 other effects) broke the use of negated acl sub-conditions.
3728 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3731 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3732 "Deny" causes Exim to reject the incoming connection with a 554 error.
3733 Unfortunately, if there is a major crisis, such as a disk failure,
3734 tcp-wrappers gives "deny", whereas what one would like would be some
3735 kind of temporary error. A kludge has been added to help with this.
3736 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3737 554 error is used if errno is still zero or contains ENOENT (which occurs
3738 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3741 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3743 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3744 errors. Otherwise a message that provokes a temporary error (when other
3745 messages do not) can cause a whole host to time out.
3747 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3748 addresses were routed directly to files or pipes from a redirect router.
3749 File deliveries just didn't batch; pipe deliveries might have suffered
3752 PH/06 A failure to get a lock for a hints database would erroneously always say
3753 "Failed to get write lock", even when it was really a read lock.
3755 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3756 of 0600. This has been changed to use the value of the lockfile_mode
3757 option (which defaults to 0600).
3759 PH/08 Applied small patch from the Sieve maintainer.
3761 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3762 folder from quota calculations, a direct delivery into this folder messed
3763 up the contents of the maildirsize file. This was because the regex was
3764 used only to exclude .Trash (or whatever) when the size of the mailbox
3765 was calculated. There was no check that a delivery was happening into an
3766 excluded directory. This bug has been fixed by ignoring all quota
3767 processing for deliveries into excluded directories.
3769 PH/10 Added the maildirfolder_create_regex option to appendfile.
3775 PH/01 The code for finding all the local interface addresses on a FreeBSD
3776 system running IPv6 was broken. This may well have applied to all BSD
3777 systems, as well as to others that have similar system calls. The broken
3778 code found IPv4 interfaces correctly, but gave incorrect values for the
3779 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3780 that it would not match correctly against @[] and not recognize the IPv6
3783 PH/02 The ipliteral router was not recognizing addresses of the form user@
3784 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3786 PH/03 Added disable_ipv6.
3788 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3789 decoded form, because it is most often used to construct To: headers
3790 lines in autoreplies, and the decoded form may well be syntactically
3791 invalid. However, $reply_address has leading white space removed, and all
3792 newlines turned into spaces so that the autoreply transport does not
3795 PH/05 If group was specified without a user on a router, and no group or user
3796 was specified on a transport, the group from the router was ignored.
3798 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3799 for visible compile-time settings that can be used to change these
3800 numbers, for those that want even more. Backwards compatibility with old
3801 spool files has been maintained. However, going back to a previous Exim
3802 release will lost any variables that are in spool files.
3804 PH/07 Two small changes when running in the test harness: increase delay when
3805 passing a TCP/IP connection to a new process, in case the original
3806 process has to generate a bounce, and remove special handling of
3807 127.0.0.2 (sic), which is no longer necessary.
3809 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3810 be the same on different OS.
3812 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3815 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3816 whether --show-vars was specified or not
3818 JJ/02 exipick: Added support for new ACL variable spool format introduced
3821 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3822 syntactically invalid From: or Reply-to: line, and a filter used this to
3823 generate an autoreply, and therefore failed to obtain an address for the
3824 autoreply, Exim could try to deliver to a non-existent relative file
3825 name, causing unrelated and misleading errors. What now happens is that
3826 it logs this as a hard delivery error, but does not attempt to create a
3829 PH/11 The exinext utility has a -C option for testing purposes, but although
3830 the given file was scanned by exinext itself; it wasn't being passed on
3831 when Exim was called.
3833 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3834 an end-of-file indication when reading a command response.
3836 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3837 compiled. In many other places in Exim, IPv6 addresses are always
3838 recognized, so I have changed this. It also means that IPv4 domain
3839 literals of the form [IPV4:n.n.n.n] are now always recognized.
3841 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3842 used if the router is not running as root, for example, when verifying at
3843 ACL time, or when using -bh. The debugging output from this situation was
3844 non-existent - all you got was a failure to exec. I have made two
3847 (a) Failures to set uid/gid, the current directory, or a process leader
3848 in a subprocess such as that created by queryprogram now generate
3849 suitable debugging output when -d is set.
3851 (b) The queryprogram router detects when it is not running as root,
3852 outputs suitable debugging information if -d is set, and then runs
3853 the subprocess without attempting to change uid/gid.
3855 PH/15 Minor change to Makefile for building test_host (undocumented testing
3858 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3859 additional section of a DNS packet that returns MX or SRV records.
3860 Instead, it always explicitly searches for A/AAAA records. This avoids
3861 major problems that occur when a DNS server includes only records of one
3862 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3863 fixed another bug: if SRV records were looked up and the corresponding
3864 address records were *not* found in the additional section, the port
3865 values from the SRV records were lost.
3867 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3868 using the correct key (the original address) when searching the retry
3869 rules in order to find which one to use for generating the retry hint.
3871 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3872 adding the default one. Similarly, if it contains a Reply-To: header, the
3873 errors_reply_to option, if set, is not used.
3875 PH/19 When calculating a retry time, Exim used to measure the "time since
3876 failure" by looking at the "first failed" field in the retry record. Now
3877 it does not use this if it is later than than the arrival time of the
3878 message. Instead it uses the arrival time. This makes for better
3879 behaviour in cases where some deliveries succeed, thus re-setting the
3880 "first failed" field. An example is a quota failure for a huge message
3881 when small messages continue to be delivered. Without this change, the
3882 "time since failure" will always be short, possible causing more frequent
3883 delivery attempts for the huge message than are intended.
3884 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3886 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3887 $1, $2, $3) because the numerical variables can be reset during some
3888 expansion items (e.g. "match"), thereby losing the authentication data.
3890 PH/21 Make -bV show the size of off_t variables so that the test suite can
3891 decide whether to run tests for quotas > 2G.
3893 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3894 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3895 filecount value is greater than 2G or if a quota value is greater than 2G
3896 on a system where the size of off_t is not greater than 4, a panic error
3899 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3900 never match. The debug and -bh output now contains an explicit error
3901 message indicating a malformed IPv4 address or mask.
3903 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3904 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3905 PH/23 above applies.
3907 PH/25 Do not write to syslog when running in the test harness. The only
3908 occasion when this arises is a failure to open the main or panic logs
3909 (for which there is an explicit test).
3911 PH/26 Added the /no_tell option to "control=freeze".
3913 PH/27 If a host name lookup failed very early in a connection, for example, if
3914 the IP address matched host_lookup and the reverse lookup yielded a name
3915 that did not have a forward lookup, an error message of the form "no IP
3916 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3917 could be logged. Now it outputs the IP address instead of "NULL".
3919 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3920 allows the sender and the authenticated sender to be set when
3921 submitting a message from within Exim. Since child_open_exim() is
3922 documented for local_scan(), the new function should be too.
3924 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3925 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3926 results in an empty string is now treated as unset.
3928 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3930 PH/31 Added sender_verify_fail logging option.
3932 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3933 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3934 tidied the source and removed it altogether.
3936 PH/33 When a queue run was abandoned because the load average was too high, a
3937 log line was always written; now it is written only if the queue_run log
3938 selector is set. In addition, the log line for abandonment now contains
3939 information about the queue run such as the pid. This is always present
3940 in "start" and "stop" lines but was omitted from the "abandon" line.
3942 PH/34 Omit spaces between a header name and the colon in the error message that
3943 is given when verify = headers_syntax fails (if there are lots of them,
3944 the message gets confusing).
3946 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3947 names, as there are now some PTR records that contain slashes. This check
3948 is only to protect against broken name servers that fall over on strange
3949 characters, so the fact that it applies to all lookups doesn't matter.
3951 PH/36 Now that the new test suite is complete, we can remove some of the
3952 special code in Exim that was needed for the old test suite. For example,
3953 sorting DNS records because real resolvers return them in an arbitrary
3954 order. The new test suite's fake resolver always returns records in the
3957 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3958 bounces) except when queue_only is set, to avoid logging races between
3959 the different processes.
3961 PH/38 Panic-die if .include specifies a non-absolute path.
3963 PH/39 A tweak to the "H" retry rule from its user.
3965 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3966 a label. They prevented compilation on older perls.
3968 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3969 a warning to be raised on newish perls.
3971 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3972 on queue. Changes to match documented behaviour of showing count of
3973 messages matching specified criteria.
3975 PH/40 Changed the default ident timeout from 30s to 5s.
3977 PH/41 Added support for the use of login_cap features, on those BSD systems
3978 that have them, for controlling the resources used by pipe deliveries.
3980 PH/42 The content-scanning code uses fopen() to create files in which to put
3981 message data. Previously it was not paying any attention to the mode of
3982 the files. Exim runs with umask(0) because the rest of the code creates
3983 files with open(), and sets the required mode explicitly. Thus, these
3984 files were ending up world-writeable. This was not a big issue, because,
3985 being within the spool directory, they were not world-accessible. I have
3986 created a function called modefopen, which takes an additional mode
3987 argument. It sets umask(777), creates the file, chmods it to the required
3988 mode, then resets the umask. All the relevant calls to fopen() in the
3989 content scanning code have been changed to use this function.
3991 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3992 to 24 hours. This avoids potential overflow problems when processing G
3993 and H retry rules. I suspect nobody ever tinkers with this value.
3995 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3997 PH/45 When the plaintext authenticator is running as a client, the server's
3998 challenges are checked to ensure they are valid base64 strings. By
3999 default, the authentication attempt is cancelled if an invalid string is
4000 received. Setting client_ignore_invalid_base64 true ignores these errors.
4001 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4002 they are received. Thus, the responses can be made to depend on the
4003 challenges. If an invalid string is ignored, an empty string is placed in
4006 PH/46 Messages that are created by the autoreply transport now contains a
4007 References: header, in accordance with RFCs 2822 and 3834.
4009 PH/47 Added authenticated_sender_force to the smtp transport.
4011 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4013 PH/49 Installed latest patch from the Sieve maintainer.
4015 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4016 was also set, the appendfile transport was unnecessarily scanning a
4017 directory of message files (e.g. for maildir delivery) to find the count
4018 of files (along with the size), even though it did not need this
4019 information. It now does the scan only if it needs to find either the
4020 size of the count of files.
4022 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4024 PH/52 Two bugs concerned with error handling when the smtp transport is
4027 (i) Exim was not creating retry information for temporary errors given
4028 for individual recipients after the DATA command when the smtp transport
4029 was used in LMTP mode. This meant that they could be retried too
4030 frequently, and not timed out correctly.
4032 (ii) Exim was setting the flag that allows error details to be returned
4033 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4034 recipients that were returned after the DATA command.
4036 PH/53 This is related to PH/52, but is more general: for any failing address,
4037 when detailed error information was permitted to be returned to the
4038 sender, but the error was temporary, then after the final timeout, only
4039 "retry timeout exceeded" was returned. Now it returns the full error as
4040 well as "retry timeout exceeded".
4042 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4043 do this, and (what is worse) MTAs that accept it.
4045 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4046 will now be deprecated.
4048 PH/56 New os.c-cygwin from the Cygwin maintainer.
4050 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4051 formats (previously only available in exim formats via -bpr, -bpru,
4052 and -bpra. Now also available in native and exiqgrep formats)
4054 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4055 with very large, slow to parse queues
4057 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4059 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4061 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4062 responses to authentication challenges, though it was showing the
4063 challenges; (ii) I've removed the CR characters from the debug output for
4066 PH/58 Allow for the insertion of a newline as well as a space when a string
4067 is turned into more than one encoded-word during RFC 2047 encoding. The
4068 Sieve code now uses this.
4070 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4071 data_4xx, lost_connection, tls_required.
4073 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4074 message was being sent as an SMTP response.
4076 PH/61 Add -l and -k options to exicyclog.
4078 PH/62 When verifying, if an address was redirected to one new address, so that
4079 verification continued, and the new address failed or deferred after
4080 having set something in $address_data, the value of $address_data was not
4081 passed back to the ACL. This was different to the case when no
4082 redirection occurred. The value is now passed back in both cases.
4084 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4085 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4086 use login_cap.h, so on its own it isn't the distinguishing feature. The
4087 new name refers directly to the setclassresources() function.
4089 PH/65 Added configuration files for NetBSD3.
4091 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4093 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4094 is preferred over IPv4.
4096 PH/68 The bounce_return_message and bounce_return_body options were not being
4097 honoured for bounces generated during the reception of non-SMTP messages.
4098 In particular, this applied to messages rejected by the ACL. This bug has
4099 been fixed. However, if bounce_return_message is true and bounce_return_
4100 body is false, the headers that are returned for a non-SMTP message
4101 include only those that have been read before the error was detected.
4102 (In the case of an ACL rejection, they have all been read.)
4104 PH/69 The HTML version of the specification is now built in a directory called
4105 spec_html instead of spec.html, because the latter looks like a path with
4106 a MIME-type, and this confuses some software.
4108 PH/70 Catch two compiler warnings in sieve.c.
4110 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4111 function verify_get_ident() calls ip_connect() to connect a socket, but
4112 if the "connect()" function timed out, ip_connect() used to close the
4113 socket. However, verify_get_ident() also closes the socket later, and in
4114 between Exim writes to the log, which may get opened at this point. When
4115 the socket was closed in ip_connect(), the log could get the same file
4116 descriptor number as the socket. This naturally causes chaos. The fix is
4117 not to close the socket in ip_connect(); the socket should be closed by
4118 the function that creates it. There was only one place in the code where
4119 this was missing, in the iplookup router, which I don't think anybody now
4120 uses, but I've fixed it anyway.
4122 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4123 well as to direct DNS lookups. Otherwise the handling of names in host
4124 lists is inconsistent and therefore confusing.
4130 PH/01 Two changes to the default runtime configuration:
4132 (1) Move the checks for relay_from_hosts and authenticated clients from
4133 after to before the (commented out) DNS black list checks.
4135 (2) Add control=submission to the relay_from_hosts and authenticated
4136 clients checks, on the grounds that messages accepted by these
4137 statements are most likely to be submissions.
4139 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4141 (1) Generate an error if the third argument for the ${prvs expansion is
4144 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4147 (3) Reset the variables that are obtained from the first argument of
4148 ${prvscheck and used in the second argument before leaving the code,
4149 because their memory is reclaimed, so using them afterwards may do
4152 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4153 one (it's much easier than Tom thought :-).
4155 (5) Because of (4), we can now allow for the use of $prvscheck_result
4156 inside the third argument.
4158 PH/03 For some reason, the default setting of PATH when running a command from
4159 a pipe transport was just "/usr/bin". I have changed it to
4162 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4163 anything to be listed in the output from -bV.
4165 PH/05 When a filter generated an autoreply, the entire To: header line was
4166 quoted in the delivery log line, like this:
4168 => >A.N.Other <ano@some.domain> <original@ddress> ...
4170 This has been changed so that it extracts the operative address. There
4171 may be more than one such address. If so, they are comma-separated, like
4174 => >ano@some.domain,ona@other.domain <original@ddress> ...
4176 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4177 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4178 not being looked up in the rDNS to get a host name, Exim was showing the
4179 IP address twice in Received: lines, even though the IP addresses were
4180 identical. For example:
4182 Received: from [1.2.3.4] (helo=[1.2.3.4])
4184 However, if the real host name was known, it was omitting the HELO data
4185 if it matched the actual IP address. This has been tidied up so that it
4186 doesn't show the same IP address twice.
4188 PH/07 When both +timestamp and +memory debugging was on, the value given by
4189 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4190 called by the debug printing, thereby overwriting the timestamp buffer.
4191 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4193 PH/08 When the original message was included in an autoreply transport, it
4194 always said "this is a copy of the message, including all the headers",
4195 even if body_only or headers_only was set. It now gives an appropriate
4198 PH/09 Applied a patch from the Sieve maintainer which:
4200 o fixes some comments
4201 o adds the (disabled) notify extension core
4202 o adds some debug output for the result of if/elsif tests
4203 o points to the current vacation draft in the documentation
4204 and documents the missing references header update
4208 o fixes a bug in processing the envelope test (when testing
4209 multiple envelope elements, the last element determined the
4212 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4213 Electronic Mail") by including:
4215 Auto-submitted: auto-generated
4217 in the messages that it generates (bounce messages and others, such as
4218 warnings). In the case of bounce messages for non-SMTP messages, there was
4219 also a typo: it was using "Auto_submitted" (underscore instead of
4220 hyphen). Since every message generated by Exim is necessarily in response
4221 to another message, thes have all been changed to:
4223 Auto-Submitted: auto-replied
4225 in accordance with these statements in the RFC:
4227 The auto-replied keyword:
4229 - SHOULD be used on messages sent in direct response to another
4230 message by an automatic process,
4232 - MUST NOT be used on manually-generated messages,
4234 - MAY be used on Delivery Status Notifications (DSNs) and Message
4235 Disposition Notifications (MDNs),
4237 - MUST NOT be used on messages generated by automatic or periodic
4238 processes, except for messages which are automatic responses to
4241 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4242 to the default Received: header definition.
4244 PH/12 Added log selector acl_warn_skipped (default on).
4246 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4247 variables because (a) they are in the wrong storage pool and (b) even if
4248 they were copied, it wouldn't work properly because of the caching.
4250 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4251 checking when decoding. Apparently there are clients that generate
4252 overlong encoded strings. Why am I not surprised?
4254 PH/15 If the first argument of "${if match_address" was not empty, but did not
4255 contain an "@" character, Exim crashed. Now it writes a panic log message
4256 and treats the condition as false.
4258 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4260 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4261 extension "envelope-auth". The code is finished and in agreement with
4262 other implementations, but there is no documentation so far and in fact,
4263 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4264 not changing the active code.
4266 Print executed "if" and "elsif" statements when debugging is used. This
4267 helps a great deal to understand what a filter does.
4269 Document more things not specified clearly in RFC3028. I had all this
4270 sorted out, when out of a sudden new issues came to my mind. Oops."
4272 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4275 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4276 canonical form (as documented). However, after a host name lookup from
4277 the IP address, check_host() was doing a simple string comparison with
4278 addresses acquired from the DNS when checking that the found name did
4279 have the original IP as one of its addresses. Since any found IPv6
4280 addresses are likely to be in abbreviated form, the comparison could
4281 fail. Luckily, there already exists a function for doing the comparison
4282 by converting both addresses to binary, so now that is used instead of
4283 the text comparison.
4285 PH/20 There was another similar case to PH/19, when a complete host name was
4286 given in a host list; looking up its IP address could give an abbreviated
4287 form, whereas the current host's name might or might not be abbreviated.
4288 The same fix has been applied.
4294 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4295 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4298 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4299 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4301 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4303 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4304 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4305 Also, for "Auto-Submitted", treat anything other than "no" as
4306 non-personal, in accordance with RFC 3834. (Previously it treated
4307 anything starting "auto-" as non-personal.)
4309 TF/01 The control=submission/name=... option had a problem with syntax
4310 errors if the name included a slash character. The /name= option
4311 now slurps the rest of the string, so it can include any characters
4312 but it must come last in the list of options (after /sender_retain
4315 PH/05 Some modifications to the interface to the fake nameserver for the new
4323 TK/01 Added the "success_on_redirect" address verification option. See
4324 NewStuff for rationale and an example.
4326 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4328 PH/02 Patch to exigrep to allow it to work on syslog lines.
4330 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4331 fread() to skip over the body file's header line, because in Cygwin the
4332 header line is locked and is inaccessible.
4334 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4335 co-exist for some time) to make it clear that it is the Exim ID that is
4336 referenced, not the Message-ID: header line.
4338 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4339 string_format() function, because snprintf() does not exist on all
4342 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4343 ${stat: expansion item.
4345 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4346 protocol synchronization error", to keep the pedants happy.
4348 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4349 well as for IRIX systems, when gcc is being used. See the host.c source
4352 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4354 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4357 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4358 transport and to the smtp transport in LMTP mode.
4360 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4362 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4363 run a filter in a subprocess. This could lead to confusion in subsequent
4364 lookups in the parent process. There should also be a search_tidyup() at
4365 the end of the subprocess.
4367 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4368 only if the host matched helo_try_verify_hosts, which caused the
4369 verification to occur when the EHLO/HELO command was issued. The ACL just
4370 tested the remembered result. Now, if a previous verification attempt has
4371 not happened, "verify = helo" does it there and then.
4373 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4375 TK/03 Fix log output including CR from clamd.
4377 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4378 exist provoked a memory error which could cause a segfault.
4380 PH/15 Installed PCRE 6.2
4382 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4384 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4385 of the problem. Specifically, suggested +O2 rather than +O1 for the
4388 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4390 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4391 record, the port was not correctly logged when the outgoing_port log
4392 selector was set (it logged the transort's default port).
4394 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4395 fallback_hosts, and "hosts" in the smtp transport.
4397 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4398 host errors such as "Connection refused".
4400 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4401 authentication with radiusclient 0.4.9:
4403 - Error returned from rc_read_config was caught wrongly
4404 - Username/password not passed on to radius server due to wrong length.
4406 The presumption is that some radiusclient API changes for 4.51/PH/17
4407 were not taken care of correctly. The code is still untested by me (my
4408 Linux distribution still has 0.3.2 of radiusclient), but it was
4409 contributed by a Radius user.
4411 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4412 expanding the "port" option of the smtp transport.
4414 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4415 while reading a MIME header. Thanks to Tom Hughes for a patch.
4417 PH/24 Include config.h inside local_scan.h so that configuration settings are
4420 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4421 that in an ACL for RCPT (for example), you can examine exactly what was
4424 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4425 commands, but it was not correctly comparing the address with the actual
4426 client host address. Thus, it would show the EHLO address in Received:
4427 header lines when this was not necessary.
4429 PH/27 Added the % operator to ${eval:}.
4431 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4432 it should be ignoring failures (because with -C, for example, it has lost
4433 privilege). It wasn't ignoring creation failures other than "already
4436 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4439 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4440 because an input line was too long, either on its own, or by virtue of
4441 too many continuations, the temporary file was not being removed, and the
4442 return code was incorrect.
4444 PH/31 Missing "BOOL" in function definition in filtertest.c.
4446 PH/32 Applied Sieve patches from the maintainer.
4448 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4450 PH/33 Added "verify = not_blind".
4452 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4453 Local/Makefile (with some defaults set). These are used in built scripts
4454 such as exicyclog, but they have never been used in the exim_install
4455 script (though there are many overriding facilities there). I have
4456 arranged that the exim_install script now takes note of these two
4459 PH/35 Installed configuration files for Dragonfly.
4461 PH/36 When a locally submitted message by a trusted user did not contain a
4462 From: header, and the sender address was obtained from -f or from an SMTP
4463 MAIL command, and the trusted user did not use -F to supply a sender
4464 name, $originator_name was incorrectly used when constructing a From:
4465 header. Furthermore, $originator_name was used for submission mode
4466 messages from external hosts without From: headers in a similar way,
4467 which is clearly wrong.
4469 PH/37 Added control=suppress_local_fixups.
4471 PH/38 When log_selector = +received_sender was set, and the addition of the
4472 sender made the log line's construction buffer exactly full, or one byte
4473 less than full, an overflow happened when the terminating "\n" was
4476 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4477 when the result of a list match is failure because a DNS lookup failed.
4479 PH/40 RM_COMMAND is now used in the building process.
4481 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4482 the "build-* directories that it finds.
4484 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4485 address in a domain literal was a prefix of an interface address.
4487 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4488 when verifying a sender address, unless rewrite_headers is false.
4490 PH/44 Wrote a long comment about why errors_to addresses are verified as
4491 recipients, not senders.
4493 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4494 the ratelimit ACL was added.
4496 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4498 PH/46 Added extra information about PostgreSQL errors to the error string.
4500 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4501 avoiding the need to install special zones in a real server. This is
4502 backwards compatible; if it can't find the fake resolver, it drops back.
4503 Thus, both old and new test suites can be run.
4505 TF/02 Added util/ratelimit.pl
4507 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4508 clock is set back in time.
4510 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4511 Candler <B.Candler@pobox.com>.
4513 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4514 OK for addresses that are the result of redirections.
4516 PH/48 A number of further additions for the benefit of the new test suite,
4517 including a fake gethostbyname() that interfaces to the fake DNS resolver
4520 TF/06 The fix for widen_domains has also been applied to qualify_single and
4521 search_parents which are the other dnslookup options that can cause
4524 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4527 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4529 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4530 service" if it encountered an MX record with an empty target hostname.
4531 The message is now "an MX or SRV record indicated no SMTP service".
4533 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4534 if the DNS of the sending site is misconfigured. This is quite a
4535 common situation. This change restores the behaviour of treating a
4536 helo verification defer as a failure.
4538 PH/51 If self=fail was set on a router, the bounce message did not include the
4539 actual error message.
4545 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4547 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4548 command itself ended in error, the underlying message about the transport
4549 filter timeout was being overwritten with the pipe command error. Now the
4550 underlying error message should be appended to the second error message.
4552 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4554 PH/02 Reduce the amount of output that "make" produces by default. Full output
4555 can still be requested.
4557 PH/03 The warning log line about a condition test deferring for a "warn" verb
4558 was being output only once per connection, rather than after each
4559 occurrence (because it was using the same function as for successful
4560 "warn" verbs). This seems wrong, so I have changed it.
4562 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4563 it should not have, which might have caused a crash in the right
4564 circumstances, but probably never did.
4566 PH/04 Installed a modified version of Tony Finch's patch to make submission
4567 mode fix the return path as well as the Sender: header line, and to
4568 add a /name= option so that you can make the user's friendly name appear
4571 TF/03 Added the control = fakedefer ACL modifier.
4573 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4574 Mark Lowes for thorough testing.
4576 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4578 TK/03 Merged latest SRS patch from Miles Wilton.
4580 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4581 with the definition in sysexits.h (which is #included earlier).
4582 Fortunately, Exim does not actually use EX_OK. The code used to try to
4583 preserve the sysexits.h value, by assuming that macro definitions were
4584 scanned for macro replacements. I have been disabused of this notion,
4585 so now the code just undefines EX_OK before #including unistd.h.
4587 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4588 in the smtp transport. When a block could not be written in a single
4589 write() function, the timeout was being re-applied to each part-write.
4590 This seems wrong - if the receiver was accepting one byte at a time it
4591 would take for ever. The timeout is now adjusted when this happens. It
4592 doesn't have to be particularly precise.
4594 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4595 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4597 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4598 without a domain if the check to <postmaster@domain> fails.
4600 SC/01 Eximstats: added -xls and the ability to specify output files
4601 (patch written by Frank Heydlauf).
4603 SC/02 Eximstats: use FileHandles for outputting results.
4605 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4607 SC/04 Eximstats: fixed display of large numbers with -nvr option
4609 SC/05 Eximstats: fixed merging of reports with empty tables.
4611 SC/06 Eximstats: added the -include_original_destination flag
4613 SC/07 Eximstats: removed tabs and trailing whitespace.
4615 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4617 TK/06 MBOX spool code: Add real "From " MBOX separator line
4618 so the .eml file is really in mbox format (even though
4619 most programs do not really care). Patch from Alex Miller.
4621 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4622 The latter is generated from $received_to and is only set if the
4623 message has one envelope recipient. SA can use these headers,
4624 obviously out-of-the-box. Patch from Alex Miller.
4626 PH/08 The ${def test on a variable was returning false if the variable's
4627 value was "0", contrary to what the specification has always said!
4628 The result should be true unless the variable is empty.
4630 PH/09 The syntax error of a character other than { following "${if
4631 def:variable_name" (after optional whitespace) was not being diagnosed.
4632 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4633 accidental colon was present, for example, could give incorrect results.
4635 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4636 result is used (not including appendfile, where other changes are about
4639 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4640 This involved changing a lot of size variables from int to off_t. It
4641 should work with maildirs and everything.
4643 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4644 spamd dying while we are connected to it.
4646 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4649 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4650 function for BDB 4.3.
4652 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4654 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4655 the sources for PCRE have been re-arranged and more files are now
4658 PH/15 The code I had for printing potentially long long variables in PH/11
4659 above was not the best (it lost precision). The length of off_t variables
4660 is now inspected at build time, and an appropriate printing format (%ld
4661 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4662 to be "long long int" or "long int". This is needed for the internal
4663 formatting function string_vformat().
4665 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4666 the configuration file to be ":syslog", then the script "guesses" where
4667 the logs files are, rather than using the compiled in default. In our
4668 case the guess is not the same as the compiled default, so the script
4669 suddenly stopped working when I started to use syslog. The patch checks
4670 to see if log_file_path is "". If so, it attempts to read it from exim
4671 with no configuration file to get the compiled in version, before it
4672 falls back to the previous guessing code."
4674 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4675 implementing BATV in an Exim configuration. See NewStuff for the gory
4678 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4679 Makefile that are specific to HP-UX.
4681 PH/18 If the "use_postmaster" option was set for a recipient callout together
4682 with the "random" option, the postmaster address was used as the MAIL
4683 FROM address for the random test, but not for the subsequent recipient
4684 test. It is now used for both.
4686 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4687 patch removes a few documentation additions to RFC 3028, because the
4688 latest draft now contains them. It adds the new en;ascii-case comparator
4689 and a new error check for 8bit text in MIME parts. Comparator and
4690 require names are now matched exactly. I enabled the subaddress
4691 extension, but it is not well tested yet (read: it works for me)."
4693 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4694 rework some of the code of TK/09 above to avoid the hardwired use of
4695 "%lld" and "long long". Replaced the call to snprintf() with a call to
4698 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4699 records point to non-existent hosts", "retry timeout exceeded", and
4700 "retry time not reached for any host after a long failure period".
4702 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4703 experimental DomainKeys support:
4705 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4706 (2) On an error such as an illegally used "control", the wrong name for
4707 the control was given.
4709 These problems did NOT occur unless DomainKeys support was compiled.
4711 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4713 PH/24 Added ${if match_ip condition.
4715 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4716 databases so that it will be absolutely obvious if a crash occurs in the
4717 DB library. This is a regular occurrence (often caused by mis-matched
4720 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4721 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4722 picked up on a user's system that detects such things. There doesn't seem
4723 to be a gcc warning option for this - only an attribute that has to be
4724 put on the function's prototype. It seems that in Fedora Core 4 they have
4725 set this on a number of new functions. No doubt there will be more in due
4728 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4729 specify a transport. However, if an address that was verified by such a
4730 router was the subject of a callout, Exim crashed because it tried to
4731 read the rcpt_include_affixes from the non-existent transport. Now it
4732 just assumes that the setting of that option is false. This bug was
4733 introduced by 4.51/PH/31.
4735 PH/28 Changed -d+all to exclude +memory, because that information is very
4736 rarely of interest, but it makes the output a lot bigger. People tend to
4737 do -d+all out of habit.
4739 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4740 code in os-type was giving problems when libc.so lives in lib64, like on
4743 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4744 aren't the modern standard, and it seems that some systems' include files
4745 don't always have them. Exim was already checking for some of the newer
4746 ones like T_AAAA, and defining it itself. I've added checks for all the
4747 record types that Exim uses.
4749 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4750 not automatically generating a new one, as it is supposed to. This
4751 prevented TLS from working. If the file did exist, but contained invalid
4752 data, a new version was generated, as expected. It was only the case of a
4753 non-existent file that was broken.
4755 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4756 with a change in libdomainkeys > 0.64.
4758 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4759 from DNS. If the selector record carries the flag, it now has
4760 precedence over the domain-wide flag.
4762 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4764 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4765 the use of a port name that isn't defined in /etc/services) occurred, the
4766 message was deferred as in a normal delivery, and thus remained on the
4767 spool, instead of being failed because of the mua_wrapper setting. This
4768 is now fixed, and I tidied up some of the mua_wrapper messages at the
4771 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4772 the messages in a hash of arrays rather than using individual hashes.
4773 This is a bit cleaner and results in dramatic memory savings, albeit
4774 at a slight CPU cost.
4776 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4777 as requested by Marc Sherman.
4779 SC/10 Eximstats: added histograms for user specified patterns as requested
4782 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4784 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4785 fopen() in the content-scanning modules that did not already have it.
4791 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4792 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4794 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4796 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4798 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4799 to test Sieve filters that use "vacation".
4801 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4802 that changes the way the GnuTLS parameters are stored in the cache file.
4803 The new format can be generated externally. For backward compatibility,
4804 if the data in the cache doesn't make sense, Exim assumes it has read an
4805 old-format file, and it generates new data and writes a new file. This
4806 means that you can't go back to an older release without removing the
4809 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4810 work if there are any delivery delays because "one_time" forces the
4811 parent to be marked "delivered", so its unseen clone is never tried
4812 again. For this reason, Exim now forbids the simultaneous setting of
4815 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4816 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4817 entry if you want to know the details. The fix, however, neglected to
4818 consider the case when local delivery batching is involved. The test for
4819 "previously delivered" was not happening when checking to see if an
4820 address could be batched with a previous (undelivered) one; under
4821 certain circumstances this could lead to multiple deliveries to the same
4824 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4825 in its include files, and this causes problems building Exim.
4827 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4828 header_syntax) but Exim was just ignoring anything given after a slash.
4829 In particular, this caused confusion with an attempt to use "verify =
4830 reverse_host_lookup/defer_ok". An error is now given when options are
4831 supplied for verify items that do not have them. (Maybe reverse_host_
4832 lookup should have a defer_ok option, but that's a different point.)
4834 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4835 defined by RFC 821) to 2048, because there were problems with some AUTH
4836 commands, and RFC 1869 says the size should be increased for extended
4837 SMTP commands that take arguments.
4839 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4842 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4843 "unknown" error; now it says that the functionality isn't in the binary.
4845 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4846 an address' error message when a string expansion fails (syntax or
4847 whatever). Otherwise the password may appear in the log. Following change
4848 PH/42 below, there is no longer a chance of it appearing in a bounce
4851 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4853 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4854 of its addresses was ever tried. (Bugzilla bug #2.)
4856 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4857 the result incorrectly in the debug output. (It correctly added a newline
4858 to what was transported.)
4860 TF/01 Added $received_time.
4862 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4863 commented out examples of how to interface to a virus scanner and to
4864 SpamAssassin. Also added commented examples of av_scanner and
4865 spamd_address settings.
4867 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4868 and controls are allowed in which ACLs. There were a couple of minor
4869 errors. Some of the entries in the conditions table (which is a table of
4870 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4871 negation of where the condition IS allowed.
4873 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4875 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4876 header file does not have a version number, so I've had to invent a new
4877 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4878 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4879 radiusclient), but it was contributed by a Radius user.
4881 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4882 files or directories, for OpenSSL.
4884 PH/19 When an Exim process that is running as root has to create an Exim log
4885 file, it does so in a subprocess that runs as exim:exim so as to get the
4886 ownership right at creation (otherwise, other Exim processes might see
4887 the file with the wrong ownership). There was no test for failure of this
4888 fork() call, which would lead to the process getting stuck as it waited
4889 for a non-existent subprocess. Forks do occasionally fail when resources
4890 run out. I reviewed all the other calls to fork(); they all seem to check
4893 PH/20 When checking for unexpected SMTP input at connect time (before writing
4894 the banner), Exim was not dealing correctly with a non-positive return
4895 from the read() function. If the client had disconnected by this time,
4896 the result was a log entry for a synchronization error with an empty
4897 string after "input=" when read() returned zero. If read() returned -1
4898 (an event I could not check), uninitialized data bytes were printed.
4899 There were reports of junk text (parts of files, etc) appearing after
4902 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4904 PH/22 Added support for macro redefinition, and (re)definition in between
4905 driver and ACL definitions.
4907 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4908 forgetting to use the resulting value; it was using the unexpanded value.
4910 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4911 hadn't been configured. The fix is from Juergen Kreileder, who
4912 understands it better than I do:
4914 "Here's what I see happening with three configured cyrus_sasl
4915 authenticators configured (plain, login, cram-md5):
4917 On startup auth_cyrus_sasl_init() gets called for each of these.
4918 This means three calls to sasl_listmech() without a specified mech_list.
4919 => SASL tests which mechs of all available mechs actually work
4920 => three warnings about OTP not working
4921 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4923 With the patch, sasl_listmech() also gets called three times. But now
4924 SASL's mech_list option is set to the server_mech specified in the the
4925 authenticator. Or in other words, the answer from sasl_listmech()
4926 gets limited to just the mech you're testing for (which is different
4928 => the return list contains just 'plain' or 'login', 'cram-md5' or
4929 nothing depending on the value of ob->server_mech.
4931 I've just tested the patch: Authentication still works fine,
4932 unavailable mechs specified in the exim configuration are still
4933 caught, and the auth.log warnings about OTP are gone."
4935 PH/25 When debugging is enabled, the contents of the command line are added
4936 to the debugging output, even when log_selector=+arguments is not
4939 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4940 answer is "GNU", and only if the return is "GNU/something" is the answer
4943 PH/27 $acl_verify_message is now set immediately after the failure of a
4944 verification in an ACL, and so is available in subsequent modifiers. In
4945 particular, the message can be preserved by coding like this:
4947 warn !verify = sender
4948 set acl_m0 = $acl_verify_message
4950 Previously, $acl_verify_message was set only while expanding "message"
4951 and "log_message" when a very denied access.
4953 PH/28 Modified OS/os.c-Linux with
4955 -#ifndef OS_LOAD_AVERAGE
4956 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4958 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4959 nomenclature these days.)
4961 PH/29 Installed patch from the Sieve maintainer that adds the options
4962 sieve_useraddress and sieve_subaddress to the redirect router.
4964 PH/30 In these circumstances:
4965 . Two addresses routed to the same list of hosts;
4966 . First host does not offer TLS;
4967 . First host accepts first address;
4968 . First host gives temporary error to second address;
4969 . Second host offers TLS and a TLS session is established;
4970 . Second host accepts second address.
4971 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4972 and peerdn, if requested) that were in fact used only for the second
4975 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4976 attention to any local part prefix or suffix that was matched by the
4977 router that accepted the address. It now behaves in the same way as it
4978 does for delivery: the affixes are removed from the local part unless
4979 rcpt_include_affixes is set on the transport.
4981 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4982 timeout during the DATA phase of an incoming message.
4984 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4985 applied a patch sanctioned by the Sieve maintainer.
4987 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4988 the uid or gid is negative. A case of a negative gid caused this to be
4989 noticed. The fix allows for either to be negative.
4991 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4992 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4995 PH/36 The change PH/12 above was broken. Fixed it.
4997 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4998 the grounds that routing the same address twice would always produce the
4999 same answer. This might have been true once, but it is certainly no
5000 longer true now. Routing a child address may depend on the previous
5001 routing that produced that child. Some complicated redirection strategies
5002 went wrong when messages had multiple recipients, and made Exim's
5003 behaviour dependent on the order in which the addresses were given.
5005 I have moved the duplicate checking until after the routing is complete.
5006 Exim scans the addresses that are assigned to local and remote
5007 transports, and removes any duplicates. This means that more work will be
5008 done, as duplicates will always all be routed, but duplicates are
5009 presumably rare, so I don't expect this is of any significance.
5011 For deliveries to pipes, files, and autoreplies, the duplicate checking
5012 still happens during the routing process, since they are not going to be
5015 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5016 It corrects a timeout issue with spamd. This is Ian's comment: "The
5017 background is that sometimes spamd either never reads data from a
5018 connection it has accepted, or it never writes response data. The exiscan
5019 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5020 blindly assumes that writes won't block so it may never time out."
5022 PH/39 Allow G after quota size as well as K and M.
5024 PH/40 The value set for $authenticated_id in an authenticator may not contain
5025 binary zeroes or newlines because the value is written to log lines and
5026 to spool files. There was no check on this. Now the value is run through
5027 the string_printing() function so that such characters are converted to
5028 printable escape sequences.
5030 PH/41 $message_linecount is a new variable that contains the total number of
5031 lines in the message. Compare $body_linecount, which is the count for the
5034 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5035 bounce and delay warning messages, except in certain special cases, which
5038 (a) An SMTP error message from a remote host;
5039 (b) A message specified in a :fail: redirection;
5040 (c) A message specified in a "fail" command in a system filter;
5041 (d) A message specified in a FAIL return from the queryprogram router;
5042 (e) A message specified by the cannot_route_message router option.
5044 In these cases only, Exim does include the error details in bounce and
5045 warning messages. There are also a few cases where bland messages such
5046 as "unrouteable address" or "local delivery error" are given.
5048 PH/43 $value is now also set for the "else" part of a ${run expansion.
5050 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5051 being worked on, but at least Exim now implements the latest version to
5054 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5055 process to complete was treated as a delivery failure, a timeout while
5056 writing the message to the pipe was logged, but erroneously treated as a
5057 successful delivery. Such timeouts include transport filter timeouts. For
5058 consistency with the overall process timeout, these timeouts are now
5059 treated as errors, giving rise to delivery failures by default. However,
5060 there is now a new Boolean option for the pipe transport called
5061 timeout_defer, which, if set TRUE, converts the failures into defers for
5062 both kinds of timeout. A transport filter timeout is now identified in
5065 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5066 systems where "make" and "gmake" are different, calling "gmake" at top
5067 level broke things. I've arranged for the value of $(MAKE) to be passed
5068 from the Makefile to this script so that it can call the same version of
5072 A note about Exim versions 4.44 and 4.50
5073 ----------------------------------------
5075 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5076 changes of various kinds. As a consequence, a big documentation update was
5077 needed. This delayed the release for rather longer than seemed good, especially
5078 in the light of a couple of (minor) security issues. Therefore, the changes
5079 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5080 release. So 4.44 and 4.50 are in effect two different branches that both start
5083 I have left the 4.50 change log unchanged; it contains all the changes since
5084 4.43. The change log for 4.44 is below; many of its items are identical to
5085 those for 4.50. This seems to be the most sensible way to preserve the
5086 historical information.
5092 1. Minor wording change to the doc/README.SIEVE file.
5094 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5095 computation of the current number of files was incorrect.
5097 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5098 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5101 4. Give more explanation in the error message when the command for a transport
5102 filter fails to execute.
5104 5. There are several places where Exim runs a non-Exim command in a
5105 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5106 was being done only for the command run by the queryprogram router. It is
5107 now done for all such subprocesses. The other cases are: ${run, transport
5108 filters, and the commands run by the lmtp and pipe transports.
5110 6. Added CONFIGURE_GROUP build-time option.
5112 7. Some older OS have a limit of 256 on the maximum number of file
5113 descriptors. Exim was using setrlimit() to set 1000 as a large value
5114 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5115 systems. I've change it so that if it can't get 1000, it tries for 256.
5117 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5118 was an oversight, and furthermore, ever since the addition of extra
5119 controls (e.g. 4.43/32), the checks on when to allow different forms of
5120 "control" were broken. There should now be diagnostics for all cases when a
5121 control that does not make sense is encountered.
5123 9. Added the /retain_sender option to "control=submission".
5125 10. $recipients is now available in the predata ACL (oversight).
5127 11. Tidy the search cache before the fork to do a delivery from a message
5128 received from the command line. Otherwise the child will trigger a lookup
5129 failure and thereby defer the delivery if it tries to use (for example) a
5130 cached ldap connection that the parent has called unbind on.
5132 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5133 of $address_data from the recipient verification was clobbered by the
5134 sender verification.
5136 13. The value of address_data from a sender verification is now available in
5137 $sender_address_data in subsequent conditions in the ACL statement.
5139 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5141 15. Added a new option "connect=<time>" to callout options, to set a different
5144 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5145 was its contents. (It was OK if the option was not defined at all.)
5147 17. A "Completed" log line is now written for messages that are removed from
5148 the spool by the -Mrm option.
5150 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5151 information about exactly what failed.
5153 19. Added -dd to debug only the daemon process.
5155 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5156 handles timeouts, both on the server side and network timeouts. Renamed the
5157 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5159 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5160 It is now set to "smtps".
5162 22. $host_address is now set to the target address during the checking of
5163 ignore_target_hosts.
5165 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5166 being passed; this would have caused $sender_host_name to have been used if
5167 matching the list had actually called for a host name (not very likely,
5168 since this list is usually IP addresses). A host name is now passed as
5171 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5172 code that specifies a non-restarting handler (typically sigaction() in
5173 modern systems) in an attempt to fix a rare and obscure crash bug.
5175 25. Narrowed the window for a race in the daemon that could cause it to ignore
5176 SIGCHLD signals. This is not a major problem, because they are used only to
5177 wake it up if nothing else does.
5179 26. A malformed maildirsize file could cause Exim to calculate negative values
5180 for the mailbox size or file count. Odd effects could occur as a result.
5181 The maildirsize information is now recalculated if the size or filecount
5184 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5185 support for a long time. Removed HAVE_SYS_VFS_H.
5187 28. Installed the latest version of exipick from John Jetmore.
5189 29. In an address list, if the pattern was not a regular expression, an empty
5190 subject address (from a bounce message) matched only if the pattern was an
5191 empty string. Non-empty patterns were not even tested. This was the wrong
5192 because it is perfectly reasonable to use an empty address as part of a
5193 database query. An empty address is now tested by patterns that are
5194 lookups. However, all the other forms of pattern expect the subject to
5195 contain a local part and a domain, and therefore, for them, an empty
5196 address still always fails if the pattern is not itself empty.
5198 30. Exim went into a mad DNS loop when attempting to do a callout where the
5199 host was specified on an smtp transport, and looking it up yielded more
5200 than one IP address.
5202 31. Re-factored the code for checking spool and log partition space into a
5203 function that finds that data and another that does the check. The former
5204 is then used to implement four new variables: $spool_space, $log_space,
5205 $spool_inodes, and $log_inodes.
5207 32. The RFC2047 encoding function was originally intended for short strings
5208 such as real names; it was not keeping to the 75-character limit for
5209 encoded words that the RFC imposes. It now respects the limit, and
5210 generates multiple encoded words if necessary. To be on the safe side, I
5211 have increased the buffer size for the ${rfc2047: expansion operator from
5214 33. It is now permitted to omit both strings after an "if" condition; if the
5215 condition is true, the result is "true". As before, when the second string
5216 is omitted, a false condition yields an empty string. This makes it less
5217 cumbersome to write custom ACL and router conditions.
5219 34. Failure to deliver a bounce message always caused it to be frozen, even if
5220 there was an errors_to setting on the router. The errors_to setting is now
5223 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5224 canonical form (fully expanded) before being placed in
5225 $sender_host_address.
5227 36. The table in the code that translates DNS record types into text (T_A to
5228 "A" for instance) was missing entries for NS and CNAME. It is just possible
5229 that this could have caused confusion if both these types were looked up
5230 for the same domain, because the text type is used as part of Exim's
5231 per-process caching. But the chance of anyone hitting this buglet seems
5234 37. The dnsdb lookup has been extended in a number of ways.
5236 (1) There is a new type, "zns", which walks up the domain tree until it
5237 finds some nameserver records. It should be used with care.
5239 (2) There is a new type, "mxh", which is like "mx" except that it returns
5240 just the host names, not the priorities.
5242 (3) It is now possible to give a list of domains (or IP addresses) to be
5243 looked up. The behaviour when one of the lookups defers can be
5244 controlled by a keyword.
5246 (4) It is now possible to specify the separator character for use when
5247 multiple records are returned.
5249 38. The dnslists ACL condition has been extended: it is now possible to supply
5250 a list of IP addresses and/or domains to be looked up in a particular DNS
5253 39. Added log_selector=+queue_time_overall.
5255 40. When running the queue in the test harness, wait just a tad after forking a
5256 delivery process, to get repeatability of debugging output.
5258 41. Include certificate and key file names in error message when GnuTLS fails
5259 to set them up, because the GnuTLS error message doesn't include the name
5260 of the failing file when there is a problem reading it.
5262 42. Allow both -bf and -bF in the same test run.
5264 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5266 44. The "Exiscan patch" is now merged into the mainline Exim source.
5268 45. Sometimes the final signoff response after QUIT could fail to get
5269 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5270 < 0 before doing a fflush(). This bug looks as though it goes back to the
5271 introduction of TLS in release 3.20, but "sometimes" must have been rare
5272 because the tests only now provoked it.
5274 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5275 (this can affect the format of dates).
5277 47. exim_tidydb, when checking for the continued existence of a message for
5278 which it has found a message-specific retry record, was not finding
5279 messages that were in split spool directories. Consequently, it was
5280 deleting retry records that should have stayed in existence.
5282 48. Steve fixed some bugs in eximstats.
5284 49. The SPA authentication driver was not abandoning authentication and moving
5285 on to the next authenticator when an expansion was forced to fail,
5286 contradicting the general specification for all authenticators. Instead it
5287 was generating a temporary error. It now behaves as specified.
5289 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5290 (the order specifies the preference for clients). The order is now AES256,
5291 AES128, 3DES, ARCFOUR128.
5293 51. Small patch to Sieve code - explicitly set From: when generating an
5296 52. Exim crashed if a remote delivery caused a very long error message to be
5297 recorded - for instance if somebody sent an entire SpamAssassin report back
5298 as a large number of 550 error lines. This bug was coincidentally fixed by
5299 increasing the size of one of Exim's internal buffers (big_buffer) that
5300 happened as part of the Exiscan merge. However, to be on the safe side, I
5301 have made the code more robust (and fixed the comments that describe what
5304 53. Now that there can be additional text after "Completed" in log lines (if
5305 the queue_time_overall log selector is set), a one-byte patch to exigrep
5306 was needed to allow it to recognize "Completed" as not the last thing in
5309 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5310 patch that reportedly fixes this has been added. I am not expert enough to
5311 create a test for it. This is what the patch creator wrote:
5313 "I found a little strange behaviour of ldap code when working with
5314 Windows 2003 AD Domain, where users was placed in more than one
5315 Organization Units. When I tried to give exim partial DN, the exit code
5316 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5317 But simultaneously result of request was absolutely normal ldap result,
5318 so I produce this patch..."
5320 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5321 REFERENCE, so I have modified the code to exclude the patch when that macro
5324 55. Some experimental protocols are using DNS PTR records for new purposes. The
5325 keys for these records are domain names, not reversed IP addresses. The
5326 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5327 leaves it alone. Component reversal etc. now happens only for IP addresses.
5330 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5332 57. Double the size of the debug message buffer (to 2048) so that more of very
5333 long debug lines gets shown.
5335 58. The exicyclog utility now does better if the number of log files to keep
5336 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5338 59. Two changes related to the smtp_active_hostname option:
5340 (1) $smtp_active_hostname is now available as a variable.
5341 (2) The default for smtp_banner uses $smtp_active_hostname instead
5342 of $primary_hostname.
5344 60. The host_aton() function is supposed to be passed a string that is known
5345 to be a valid IP address. However, in the case of IPv6 addresses, it was
5346 not checking this. This is a hostage to fortune. Exim now panics and dies
5347 if the condition is not met. A case was found where this could be provoked
5348 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5349 components; fortuitously, this particular loophole had already been fixed
5350 by change 4.50/55 above.
5352 If there are any other similar loopholes, the new check in host_aton()
5353 itself should stop them being exploited. The report I received stated that
5354 data on the command line could provoke the exploit when Exim was running as
5355 exim, but did not say which command line option was involved. All I could
5356 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5357 running as the user.
5360 61. There was a buffer overflow vulnerability in the SPA authentication code
5361 (which came originally from the Samba project). I have added a test to the
5362 spa_base64_to_bits() function which I hope fixes it.
5365 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5366 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5368 63. The daemon start-up calls getloadavg() while still root for those OS that
5369 need the first call to be done as root, but it missed one case: when
5370 deliver_queue_load_max is set with deliver_drop_privilege. This is
5371 necessary for the benefit of the queue runner, because there is no re-exec
5372 when deliver_drop_privilege is set.
5374 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5375 This has been fixed.
5377 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5378 was in use, was not putting the data itself into the right store pool;
5379 consequently, it could be overwritten for a subsequent message in the same
5380 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5383 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5385 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5386 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5387 the function were treating the return as a boolean value, which happened to
5388 work because 0=false and not-0=true, but is not correct code.
5390 68. The host_aton() function was not handling scoped IPv6 addresses (those
5391 with, for example, "%eth0" on the end) correctly.
5393 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5394 negated items (that is, ~something) in unsigned ints. Some compilers
5395 apparently mutter when there is no cast.
5397 70. If an address verification called from an ACL failed, and did not produce a
5398 user-specific message (i.e. there was only a "system" message), nothing was
5399 put in $acl_verify_message. In this situation, it now puts the system
5402 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5403 session; change 4.31/43 added the unwanted input to the log line - except
5404 that it did not do this in the start of session case. It now does.
5406 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5407 This is wrong and can cause the other end to generate a synchronization
5408 error if it is another Exim or anything else that does the synchronization
5409 check. A QUIT command is no longer sent after a timeout.
5411 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5412 during host lookups.
5414 74. The defer_ok option of callout verification was not working if it was used
5415 when verifying addresses in header lines, that is, for this case:
5417 verify = header_sender/callout=defer_ok
5419 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5420 those file descriptors could be used for SMTP connections. If anything
5421 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5422 could be sent to the SMTP client, causing chaos. The daemon now opens
5423 stdin, stdout, and stderr to /dev/null when it puts itself into the
5426 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5427 log by default. The user can override this with suitable Perl magic.
5429 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5430 the log message when discard triggers, was not working for the DATA ACL or
5431 for the non-SMTP ACL.
5433 78. Error message wording change in sieve.c.
5435 79. If smtp_accept_max_per_host was set, the number of connections could be
5436 restricted to fewer than expected, because the daemon was trying to set up
5437 a new connection before checking whether the processes handling previous
5438 connections had finished. The check for completed processes is now done
5439 earlier. On busy systems, this bug wouldn't be noticed because something
5440 else would have woken the daemon, and it would have reaped the completed
5443 80. If a message was submitted locally by a user whose login name contained one
5444 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5445 It caused a spool format error. I have fixed the spool reading code. A
5446 related problem was that the "from" clause in the Received: line became
5447 illegal because of the space(s). It is now covered by ${quote_local_part.
5449 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5452 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5454 83. Patch from Sieve maintainer to fix unterminated string problem in
5455 "vacation" handling.
5457 84. Some minor changes to the Linux configuration files to help with other
5458 OS variants using glibc.
5460 85. One more patch for Sieve to update vacation handling to latest spec.
5463 ----------------------------------------------------
5464 See the note above about the 4.44 and 4.50 releases.
5465 ----------------------------------------------------
5471 1. Change 4.43/35 introduced a bug that caused file counts to be
5472 incorrectly computed when quota_filecount was set in an appendfile
5475 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5476 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5479 3. Give more explanation in the error message when the command for a transport
5480 filter fails to execute.
5482 4. There are several places where Exim runs a non-Exim command in a
5483 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5484 was being done only for the command run by the queryprogram router. It is
5485 now done for all such subprocesses. The other cases are: ${run, transport
5486 filters, and the commands run by the lmtp and pipe transports.
5488 5. Some older OS have a limit of 256 on the maximum number of file
5489 descriptors. Exim was using setrlimit() to set 1000 as a large value
5490 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5491 systems. I've change it so that if it can't get 1000, it tries for 256.
5493 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5494 was an oversight, and furthermore, ever since the addition of extra
5495 controls (e.g. 4.43/32), the checks on when to allow different forms of
5496 "control" were broken. There should now be diagnostics for all cases when a
5497 control that does not make sense is encountered.
5499 7. $recipients is now available in the predata ACL (oversight).
5501 8. Tidy the search cache before the fork to do a delivery from a message
5502 received from the command line. Otherwise the child will trigger a lookup
5503 failure and thereby defer the delivery if it tries to use (for example) a
5504 cached ldap connection that the parent has called unbind on.
5506 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5507 of $address_data from the recipient verification was clobbered by the
5508 sender verification.
5510 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5511 was its contents. (It was OK if the option was not defined at all.)
5513 11. A "Completed" log line is now written for messages that are removed from
5514 the spool by the -Mrm option.
5516 12. $host_address is now set to the target address during the checking of
5517 ignore_target_hosts.
5519 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5520 being passed; this would have caused $sender_host_name to have been used if
5521 matching the list had actually called for a host name (not very likely,
5522 since this list is usually IP addresses). A host name is now passed as
5525 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5526 code that specifies a non-restarting handler (typically sigaction() in
5527 modern systems) in an attempt to fix a rare and obscure crash bug.
5529 15. Narrowed the window for a race in the daemon that could cause it to ignore
5530 SIGCHLD signals. This is not a major problem, because they are used only to
5531 wake it up if nothing else does.
5533 16. A malformed maildirsize file could cause Exim to calculate negative values
5534 for the mailbox size or file count. Odd effects could occur as a result.
5535 The maildirsize information is now recalculated if the size or filecount
5538 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5539 support for a long time. Removed HAVE_SYS_VFS_H.
5541 18. Updated exipick to current release from John Jetmore.
5543 19. Allow an empty sender to be matched against a lookup in an address list.
5544 Previously the only cases considered were a regular expression, or an
5547 20. Exim went into a mad DNS lookup loop when doing a callout where the
5548 host was specified on the transport, if the DNS lookup yielded more than
5551 21. The RFC2047 encoding function was originally intended for short strings
5552 such as real names; it was not keeping to the 75-character limit for
5553 encoded words that the RFC imposes. It now respects the limit, and
5554 generates multiple encoded words if necessary. To be on the safe side, I
5555 have increased the buffer size for the ${rfc2047: expansion operator from
5558 22. Failure to deliver a bounce message always caused it to be frozen, even if
5559 there was an errors_to setting on the router. The errors_to setting is now
5562 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5563 canonical form (fully expanded) before being placed in
5564 $sender_host_address.
5566 24. Updated eximstats to version 1.33
5568 25. Include certificate and key file names in error message when GnuTLS fails
5569 to set them up, because the GnuTLS error message doesn't include the name
5570 of the failing file when there is a problem reading it.
5572 26. Expand error message when OpenSSL has problems setting up cert/key files.
5575 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5576 (this can affect the format of dates).
5578 28. exim_tidydb, when checking for the continued existence of a message for
5579 which it has found a message-specific retry record, was not finding
5580 messages that were in split spool directories. Consequently, it was
5581 deleting retry records that should have stayed in existence.
5583 29. eximstats updated to version 1.35
5584 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5585 1.35 - bugfix such that pie charts by volume are generated correctly
5587 30. The SPA authentication driver was not abandoning authentication and moving
5588 on to the next authenticator when an expansion was forced to fail,
5589 contradicting the general specification for all authenticators. Instead it
5590 was generating a temporary error. It now behaves as specified.
5592 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5593 (the order specifies the preference for clients). The order is now AES256,
5594 AES128, 3DES, ARCFOUR128.
5596 31. Small patch to Sieve code - explicitly set From: when generating an
5599 32. Exim crashed if a remote delivery caused a very long error message to be
5600 recorded - for instance if somebody sent an entire SpamAssassin report back
5601 as a large number of 550 error lines. This bug was coincidentally fixed by
5602 increasing the size of one of Exim's internal buffers (big_buffer) that
5603 happened as part of the Exiscan merge. However, to be on the safe side, I
5604 have made the code more robust (and fixed the comments that describe what
5607 33. Some experimental protocols are using DNS PTR records for new purposes. The
5608 keys for these records are domain names, not reversed IP addresses. The
5609 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5610 leaves it alone. Component reversal etc. now happens only for IP addresses.
5613 34. The host_aton() function is supposed to be passed a string that is known
5614 to be a valid IP address. However, in the case of IPv6 addresses, it was
5615 not checking this. This is a hostage to fortune. Exim now panics and dies
5616 if the condition is not met. A case was found where this could be provoked
5617 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5618 components; fortuitously, this particular loophole had already been fixed
5619 by change 4.50/55 or 4.44/33 above.
5621 If there are any other similar loopholes, the new check in host_aton()
5622 itself should stop them being exploited. The report I received stated that
5623 data on the command line could provoke the exploit when Exim was running as
5624 exim, but did not say which command line option was involved. All I could
5625 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5626 running as the user.
5629 35. There was a buffer overflow vulnerability in the SPA authentication code
5630 (which came originally from the Samba project). I have added a test to the
5631 spa_base64_to_bits() function which I hope fixes it.
5634 36. The daemon start-up calls getloadavg() while still root for those OS that
5635 need the first call to be done as root, but it missed one case: when
5636 deliver_queue_load_max is set with deliver_drop_privilege. This is
5637 necessary for the benefit of the queue runner, because there is no re-exec
5638 when deliver_drop_privilege is set.
5640 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5641 was in use, was not putting the data itself into the right store pool;
5642 consequently, it could be overwritten for a subsequent message in the same
5643 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5646 38. Sometimes the final signoff response after QUIT could fail to get
5647 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5648 < 0 before doing a fflush(). This bug looks as though it goes back to the
5649 introduction of TLS in release 3.20, but "sometimes" must have been rare
5650 because the tests only now provoked it.
5656 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5657 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5658 Now it returns an integer. A number of calls were still expecting a T/F
5659 return. Fortuitously, in all cases, the tests worked in OK situations,
5660 which is the norm. However, things would have gone wrong on any write
5661 failures on the smtp file descriptor. This function is used when sending
5662 messages over SMTP and also when doing verify callouts.
5664 2. When Exim is called to do synchronous delivery of a locally submitted
5665 message (the -odf or -odi options), it no longer closes stderr before doing
5668 3. Implemented the mua_wrapper option.
5670 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5672 5. Implemented the functions header_remove(), header_testname(),
5673 header_add_at_position(), and receive_remove_recipient(), and exported them
5676 6. If an ACL "warn" statement specified the addition of headers, Exim already
5677 inserted X-ACL-Warn: at the start if there was no header name. However, it
5678 was not making this test for the second and subsequent header lines if
5679 there were newlines in the string. This meant that an invalid header could
5680 be inserted if Exim was badly configured.
5682 7. Allow an ACL "warn" statement to add header lines at the start or after all
5683 the Received: headers, as well as at the end.
5685 8. Added the rcpt_4xx retry error code.
5687 9. Added postmaster_mailfrom=xxx to callout verification option.
5689 10. Added mailfrom=xxxx to the callout verification option, for verify=
5692 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5693 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5695 12. Inserted some casts to stop certain compilers warning when using pointer
5696 differences as field lengths or precisions in printf-type calls (mostly
5697 affecting debugging statements).
5699 13. Added optional readline() support for -be (dynamically loaded).
5701 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5702 same clock tick as a message's arrival, so that its received time was the
5703 same as the "first fail" time on the retry record, and that message
5704 remained on the queue past the ultimate address timeout, every queue runner
5705 would try a delivery (because it was past the ultimate address timeout) but
5706 after another failure, the ultimate address timeout, which should have then
5707 bounced the address, did not kick in. This was a "< instead of <=" error;
5708 in most cases the first failure would have been in the next clock tick
5709 after the received time, and all would be well.
5711 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5712 being recognized when the domain list was tested by the match_domain
5713 condition in an expansion string.
5715 16. Added the ${str2b64: operator.
5717 17. Exim was always calling setrlimit() to set a large limit for the number of
5718 processes, without checking whether the existing limit was already
5719 adequate. (It did check for the limit on file descriptors.) Furthermore,
5720 errors from getrlimit() and setrlimit() were being ignored. Now they are
5721 logged to the main and panic logs, but Exim does carry on, to try to do its
5722 job under whatever limits there are.
5724 18. Imported PCRE 5.0.
5726 19. Trivial typo in log message " temporarily refused connection" (the leading
5729 20. If the log selector return_path_on_delivery was set and an address was
5730 redirected to /dev/null, the delivery process crashed because it assumed
5731 that a return path would always be set for a "successful" delivery. In this
5732 case, the whole delivery is bypassed as an optimization, and therefore no
5735 21. Internal re-arrangement: the function for sending a challenge and reading
5736 a response while authentication was assuming a zero-terminated challenge
5737 string. It's now changed to take a pointer and a length, to allow for
5738 binary data in such strings.
5740 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5742 23. Exim was not respecting finduser_retries when seeking the login of the
5743 uid under which it was called; it was always trying 10 times. (The default
5744 setting of finduser_retries is zero.) Also, it was sleeping after the final
5745 failure, which is pointless.
5747 24. Implemented tls_on_connect_ports.
5749 25. Implemented acl_smtp_predata.
5751 26. If the domain in control=submission is set empty, Exim assumes that the
5752 authenticated id is a complete email address when it generates From: or
5753 Sender: header lines.
5755 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5756 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5757 chown and chgrp in /bin and hostname in /usr/bin.
5759 28. Exim was keeping the "process log" file open after each use, just as it
5760 does for the main log. This opens the possibility of it remaining open for
5761 long periods when the USR1 signal hits a daemon. Occasional processlog
5762 errors were reported, that could have been caused by this. Anyway, it seems
5763 much more sensible not to leave this file open at all, so that is what now
5766 29. The long-running daemon process does not normally write to the log once it
5767 has entered its main loop, and it closes the log before doing so. This is
5768 so that log files can straightforwardly be renamed and moved. However,
5769 there are a couple of unusual error situations where the daemon does write
5770 log entries, and I had neglected to close the log afterwards.
5772 30. The text of an SMTP error response that was received during a remote
5773 delivery was being truncated at 512 bytes. This is too short for some of
5774 the long messages that one sometimes sees. I've increased the limit to
5777 31. It is now possible to make retry rules that apply only when a message has a
5778 specific sender, in particular, an empty sender.
5780 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5781 it possible to be selective about when SMTP synchronization is enforced.
5783 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5785 32. Implemented hosts_connection_nolog.
5787 33. Added an ACL for QUIT.
5789 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5792 35. Added mailbox_size and mailbox_filecount to appendfile.
5794 36. Added control = no_multiline_responses to ACLs.
5796 37. There was a bug in the logic of the code that waits for the clock to tick
5797 in the case where the clock went backwards by a substantial amount such
5798 that the microsecond fraction of "now" was more than the microsecond
5799 fraction of "then" (but the whole seconds number was less).
5801 38. Added support for the libradius Radius client library this is found on
5802 FreeBSD (previously only the radiusclient library was supported).
5808 1. When certain lookups returned multiple values in the form name=value, the
5809 quoting of the values was not always being done properly. Specifically:
5810 (a) If the value started with a double quote, but contained no whitespace,
5812 (b) If the value contained whitespace other than a space character (i.e.
5813 tabs or newlines or carriage returns) it was not quoted.
5814 This fix has been applied to the mysql and pgsql lookups by writing a
5815 separate quoting function and calling it from the lookup code. The fix
5816 should probably also be applied to nisplus, ibase and oracle lookups, but
5817 since I cannot test any of those, I have not disturbed their existing code.
5819 2. A hit in the callout cache for a specific address caused a log line with no
5820 reason for rejecting RCPT. Now it says "Previous (cached) callout
5821 verification failure".
5823 3. There was an off-by-one bug in the queryprogram router. An over-long
5824 return line was truncated at 256 instead of 255 characters, thereby
5825 overflowing its buffer with the terminating zero. As well as fixing this, I
5826 have increased the buffer size to 1024 (and made a note to document this).
5828 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5829 when Exim is waiting for an SMTP response from a remote server, Exim
5830 restarts its select() call on the socket, thereby resetting its timeout.
5831 This is not a problem when such interrupts are rare. Somebody set up a cron
5832 job to run exiwhat every 2 minutes, which is less than the normal select()
5833 timeout (5 or 10 minutes). This meant that the select() timeout never
5834 kicked in because it was always reset. I have fixed this by comparing the
5835 time when an interrupt arrives with the time at the start of the first call
5836 to select(). If more time than the timeout has elapsed, the interrupt is
5837 treated as a timeout.
5839 5. Some internal re-factoring in preparation for the addition of Sieve
5840 extensions (by MH). In particular, the "personal" test is moved to a
5841 separate function, and given an option for scanning Cc: and Bcc: (which is
5842 not set for Exim filters).
5844 6. When Exim created an email address using the login of the caller as the
5845 local part (e.g. when creating a From: or Sender: header line), it was not
5846 quoting the local part when it contained special characters such as @.
5848 7. Installed new OpenBSD configuration files.
5850 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5851 try to make them clearer.
5853 9. Callout options, other than the timeout value, were being ignored when
5854 verifying sender addresses in header lines. For example, when using
5856 verify = header_sender/callout=no_cache
5858 the cache was (incorrectly) being used.
5860 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5861 only the Cygwin environment.
5863 11. When return_path_on_delivery was set as a log selector, if different remote
5864 addresses in the same message used different return paths and parallel
5865 remote delivery occurred, the wrong values would sometimes be logged.
5866 (Whenever a remote delivery process finished, the return path value from
5867 the most recently started remote delivery process was logged.)
5869 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5870 lines when AUTH and/or TLS are in use. This is the "received protocol"
5871 field. Exim used to use "asmtp" for authenticated SMTP, without any
5872 indication (in the protocol name) for TLS use. Now it follows the RFC and
5873 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5874 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5875 names appear in log lines as well as in Received: header lines.
5877 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5878 extensions, and comparison tests, and to fix some bugs.
5880 14. Changes to the "personal" filter test:
5882 (1) The test was buggy in that it was just doing the equivalent of
5883 "contains" tests on header lines. For example, if a user's address was
5884 anne@some.where, the "personal" test would incorrectly be true for
5886 To: susanne@some.where
5888 This test is now done by extracting each address from the header in turn,
5889 and checking the entire address. Other tests that are part of "personal"
5890 are now done using regular expressions (for example, to check local parts
5891 of addresses in From: header lines).
5893 (2) The list of non-personal local parts in From: addresses has been
5894 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5895 taken from the Sieve specification recommendations.
5897 (3) If the message contains any header line starting with "List-" it is
5898 treated as non-personal.
5900 (4) The test for "circular" in the Subject: header line has been removed
5901 because it now seems ill-conceived.
5903 15. Minor typos in src/EDITME comments corrected.
5905 16. Installed latest exipick from John Jetmore.
5907 17. If headers_add on a router specified a text string that was too long for
5908 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5909 of string_sprintf() is now avoided.
5911 18. $message_body_size was not set (it was always zero) when running the DATA
5912 ACL and the local_scan() function.
5914 19. For the "mail" command in an Exim filter, no default was being set for
5915 the once_repeat time, causing a random time value to be used if "once" was
5916 specified. (If the value happened to be <= 0, no repeat happened.) The
5917 default is now 0s, meaning "never repeat". The "vacation" command was OK
5918 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5919 (I found it when inspecting the code).
5921 20. There is now an overall timeout for performing a callout verification. It
5922 defaults to 4 times the callout timeout, which applies to individual SMTP
5923 commands during the callout. The overall timeout applies when there is more
5924 than one host that can be tried. The timeout is checked before trying the
5925 next host. This prevents very long delays if there are a large number of
5926 hosts and all are timing out (e.g. when the network connections are timing
5927 out). The value of the overall timeout can be changed by specifying an
5928 additional sub-option for "callout", called "maxwait". For example:
5930 verify = sender/callout=5s,maxwait=20s
5932 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5933 to the end before writing, but this should make it even safer).
5935 22. Exim was forgetting that it had advertised PIPELINING for the second and
5936 subsequent messages on an SMTP connection. It was also not resetting its
5937 memory on STARTTLS and an internal HELO.
5939 23. When Exim logs an SMTP synchronization error within a session, it now
5940 records whether PIPELINING has been advertised or not.
5942 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5943 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5944 rather than long int.
5946 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5948 26. Added the never_mail option to autoreply.
5954 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5955 crash if the getsockname() call failed; this can happen if a connection is
5956 closed very soon after it is established. The problem was simply in the
5957 order in which certain operations were done, causing Exim to try to write
5958 to the SMTP stream before it had set up the file descriptor. The bug has
5959 been fixed by making things happen in the correct order.
5965 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5966 before the connection was closed, thus losing the rejection response.
5968 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5969 some early Solaris releases, but causes trouble in current releases where
5970 socklen_t is defined.
5972 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5975 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5978 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5979 that releasing the top of it at the end releases what was used for sub-
5980 expansions (unless the block got too big). However, discard this block if
5981 the first thing is a variable or header, so that we can use its block when
5982 it is dynamic (useful for very large $message_headers, for example).
5984 6. Lookups now cache *every* query, not just the most recent. A new, separate
5985 store pool is used for this. It can be recovered when all lookup caches are
5986 flushed. Lookups now release memory at the end of their result strings.
5987 This has involved some general refactoring of the lookup sources.
5989 7. Some code has been added to the store_xxx() functions to reduce the amount
5990 of flapping under certain conditions.
5992 8. log_incoming_interface used to affect only the <= reception log lines. Now
5993 it causes the local interface and port to be added to several more SMTP log
5994 lines, for example "SMTP connection from", and rejection lines.
5996 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5998 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6000 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6001 could be overwritten at the end of the current message (or the start of a
6002 new message if it was set in a HELO ACL). The value is now preserved for
6003 the duration of the SMTP connection.
6005 12. If a transport had a headers_rewrite setting, and a matching header line
6006 contained an unqualified address, that address was qualified, even if it
6007 did not match any rewriting rules. The underlying bug was that the values
6008 of the flags that permit the existence of unqualified sender and recipient
6009 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6010 non-local messages, and by -bnq for local messages) were not being
6011 preserved with the message after it was received.
6013 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6014 "next input=" as part of the text comprising the host identity instead of
6015 the correct text. The code was using the same buffer for two different
6016 strings. However, depending on which order the printing function evaluated
6017 its arguments, the bug did not always show up. Under Linux, for example, my
6018 test suite worked just fine.
6020 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6021 doesn't work with some older versions of Perl. It has been changed to "my",
6022 which in any case is probably the better facility to use.
6024 15. A really picky compiler found some instances of statements for creating
6025 error messages that either had too many or two few arguments for the format
6028 16. The size of the buffer for calls to the DNS resolver has been increased
6029 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6030 for addresses that have a lot of PTR records. This alleviates a problem; it
6031 does not fully solve it.
6033 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6034 buffer now truncates the list and logs the incident, which is the same
6035 action as happens when Exim is looking up a host name and its aliases.
6036 Previously in this situation something unpredictable would happen;
6037 sometimes it was "internal error: store_reset failed".
6039 18. If a server dropped the connection unexpectedly when an Exim client was
6040 using GnuTLS and trying to read a response, the client delivery process
6041 crashed while trying to generate an error log message.
6043 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6044 string, for example:
6046 warn message = H1: something\nH2: something
6048 the text was added as a single header line from Exim's point of view
6049 though it ended up OK in the delivered message. However, searching for the
6050 second and subsequent header lines using $h_h2: did not work. This has been
6051 fixed. Similarly, if a system filter added multiple headers in this way,
6052 the routers could not see them.
6054 20. Expanded the error message when iplsearch is called with an invalid key to
6055 suggest using net-iplsearch in a host list.
6057 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6058 ACLs are no longer actually imposed (and a message to that effect is
6061 22. If a "gecos" field in a passwd entry contained escaped characters, in
6062 particular, if it contained a \" sequence, Exim got it wrong when building
6063 a From: or a Sender: header from that name. A second bug also caused
6064 incorrect handling when an unquoted " was present following a character
6065 that needed quoting.
6067 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6068 was not being matched caselessly.
6070 24. Arranged for all hyphens in the exim.8 source to be escaped with
6073 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6074 properly. Recipient callout cache records were still being keyed to include
6075 the sender, even when use_sender was set false. This led to far more
6076 callouts that were necessary. The sender is no longer included in the key
6077 when use_sender is false.
6079 26. Added "control = submission" modifier to ACLs.
6081 27. Added the ${base62d: operator to decode base 62 numbers.
6083 28. dnsdb lookups can now access SRV records.
6085 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6086 the configuration file.
6088 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6089 (-v) mode. This makes the output for a verbose queue run more intelligible.
6091 31. Added a use_postmaster feature to recipient callouts.
6093 32. Added the $body_zerocount variable, containing the number of binary zero
6094 bytes in the message body.
6096 33. The time of last modification of the "new" subdirectory is now used as the
6097 "mailbox time last read" when there is a quota error for a maildir
6100 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6102 35. Added +ignore_unknown as a special item in host lists.
6104 36. Code for decoding IPv6 addresses in host lists is now included, even if
6105 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6106 address was recognized as an IP address, but was then not correctly decoded
6107 into binary, causing unexpected and incorrect effects when compared with
6114 1. Very minor rewording of debugging text in manualroute to say "list of
6115 hosts" instead of "hostlist".
6117 2. If verify=header_syntax was set, and a header line with an unqualified
6118 address (no domain) and a large number of spaces between the end of the
6119 name and the colon was received, the reception process suffered a buffer
6120 overflow, and (when I tested it) crashed. This was caused by some obsolete
6121 code that should have been removed. The fix is to remove it!
6123 3. When running in the test harness, delay a bit after writing a bounce
6124 message to get a bit more predictability in the log output.
6126 4. Added a call to search_tidyup() just before forking a reception process. In
6127 theory, someone could use a lookup in the expansion of smtp_accept_max_
6128 per_host which, without the tidyup, could leave open a database connection.
6130 5. Added the variables $recipient_data and $sender_data which get set from a
6131 lookup success in an ACL "recipients" or "senders" condition, or a router
6132 "senders" option, similar to $domain_data and $local_part_data.
6134 6. Moved the writing of debug_print from before to after the "senders" test
6137 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6138 problems for message scanning, either using a data ACL, or using
6139 local_scan() because the Received: header was not generated till after they
6140 were called (in order to set the time as the time of reception completion).
6141 I have revised the way this works. The header is now generated after the
6142 body is received, but before the ACL or local_scan() are called. After they
6143 are run, the timestamp in the header is updated.
6149 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6150 before starting a queue runner without re-exec. This happened only when
6151 deliver_drop_privilege was set or when the Exim user was set to root. The
6152 effect of the bug was that timeouts during subsequent deliveries caused
6153 crashes instead of being properly handled. The handler is now left at its
6154 default (and expected) setting.
6156 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6157 message, again when deliver_drop_privilege is set or Exim is run as root.
6158 The bug described in (1) was not present in this case, but the tidying up
6159 of the other signals was missing. I have made the two cases consistent.
6161 3. The ignore_target_hosts setting on a manualroute router was being ignored
6162 for hosts that were looked up using the /MX notation.
6164 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6167 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6168 operated on the sender address. After changing the $sender_address to <>
6169 for the sender address verify, Exim was re-instated it as the original
6170 (before rewriting) address, but remembering that it had rewritten it, so it
6171 wasn't rewriting it again. This bug also had the effect of breaking the
6172 sender address verification caching when the sender address was rewritten.
6174 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6175 This has been changed so that if the ip literal address matches
6176 ignore_target_hosts, the router declines.
6178 7. Added expansion conditions match_domain, match_address, and match_local_
6179 part (NOT match_host).
6181 8. The placeholder for the Received: header didn't have a length field set.
6183 9. Added code to Exim itself and to exim_lock to test for a specific race
6184 condition that could lead to file corruption when using MBX delivery. The
6185 issue is with the lockfile that is created in /tmp. If this file is removed
6186 after a process has opened it but before that process has acquired a lock,
6187 there is the potential for a second process to recreate the file and also
6188 acquire a lock. This could lead to two Exim processes writing to the file
6189 at the same time. The added code performs the same test as UW imapd; it
6190 checks after acquiring the lock that its file descriptor still refers to
6191 the same named file.
6193 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6194 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6197 11. Added the smtp_active_hostname option. If used, this will typically be made
6198 to depend on the incoming interface address. Because $interface_address is
6199 not set up until the daemon has forked a reception process, error responses
6200 that can happen earlier (such as "too many connections") no longer contain
6203 12. If an expansion in a condition on a "warn" statement fails because a lookup
6204 defers, the "warn" statement is abandoned, and the next ACL statement is
6205 processed. Previously this caused the whole ACL to be aborted.
6207 13. Added the iplsearch lookup type.
6209 14. Added ident_timeout as a log selector.
6211 15. Added tls_certificate_verified as a log selector.
6213 16. Added a global option tls_require_ciphers (compare the smtp transport
6214 option of the same name). This controls incoming TLS connections.
6216 17. I finally figured out how to make tls_require_ciphers do a similar thing
6217 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6218 before starting the TLS session.
6220 18. Tabs are now shown as \t in -bP output.
6222 19. If the log selector return_path_on_delivery was set, Exim crashed when
6223 bouncing a message because it had too many Received: header lines.
6225 20. If two routers both had headers_remove settings, and the first one included
6226 a superfluous trailing colon, the final name in the first list and the
6227 first name in the second list were incorrectly joined into one item (with a
6228 colon in the middle).
6234 1. Added -C and -D options to the exinext utility, mainly to make it easier
6235 to include in the automated testing, but these could be helpful when
6236 multiple configurations are in use.
6238 2. The exinext utility was not formatting the output nicely when there was
6239 an alternate port involved in the retry record key, nor when there was a
6240 message id as well (for retries that were specific to a specific message
6241 and a specific host). It was also confused by IPv6 addresses, because of
6242 the additional colons they contain. I have fixed the IPv4 problem, and
6243 patched it up to do a reasonable job for IPv6.
6245 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6246 delivery, the log line now contains "pipelined" if PIPELINING was used.
6248 4. An SMTP transport process used to panic and die if the bind() call to set
6249 an explicit outgoing interface failed. This has been changed; it is now
6250 treated in the same way as a connect() failure.
6252 5. A reference to $sender_host_name in the part of a conditional expansion
6253 that was being skipped was still causing a DNS lookup. This no longer
6256 6. The def: expansion condition was not recognizing references to header lines
6257 that used bh_ and bheader_.
6259 7. Added the _cache feature to named lists.
6261 8. The code for checking quota_filecount in the appendfile transport was
6262 allowing one more file than it should have been.
6264 9. For compatibility with Sendmail, the command line option
6272 and sets the incoming protocol and host name (for trusted callers). The
6273 host name and its colon can be omitted when only the protocol is to be set.
6274 Note the Exim already has two private options, -pd and -ps, that refer to
6275 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6276 "s", but I don't think that's a major issue.
6278 10. A number of refactoring changes to the code, none of which should affect
6281 (a) The number of logging options was getting close to filling up the
6282 32-bit word that was used as a bit map. I have split them into two classes:
6283 those that are passed in the argument to log_write(), and those that are
6284 only ever tested independently outside of that function. These are now in
6285 separate 32-bit words, so there is plenty of room for expansion again.
6286 There is no change in the user interface or the logging behaviour.
6288 (b) When building, for example, log lines, the code previously used a
6289 macro that called string_cat() twice, in order to add two strings. This is
6290 not really sufficiently general. Furthermore, there was one instance where
6291 it was actually wrong because one of the argument was used twice, and in
6292 one call a function was used. (As it happened, calling the function twice
6293 did not affect the overall behaviour.) The macro has been replaced by a
6294 function that can join an arbitrary number of extra strings onto a growing
6297 (c) The code for expansion conditions now uses a table and a binary chop
6298 instead of a serial search (which was left over from when there were very
6299 few conditions). Also, it now recognizes conditions like "pam" even when
6300 the relevant support is not compiled in: a suitably worded error message is
6301 given if an attempt is made to use such a condition.
6303 11. Added ${time_interval:xxxxx}.
6305 12. A bug was causing one of the ddress fields not to be passed back correctly
6306 from remote delivery subprocesses. The field in question was not being
6307 subsequently used, so this caused to problems in practice.
6309 13. Added new log selectors queue_time and deliver_time.
6311 14. Might have fixed a bug in maildirsizefile handling that threw up
6312 "unexpected character" debug warnings, and recalculated the data
6313 unnecessarily. In any case, I expanded the warning message to give more
6316 15. Added the message "Restricted characters in address" to the statements in
6317 the default ACL that block characters like @ and % in local parts.
6319 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6320 Three changes have been made:
6322 (a) There was a serious bug; a negative response to MAIL caused the whole
6323 recipient domain to be cached as invalid, thereby blocking all messages
6324 to all local parts at the same domain, from all senders. This bug has
6325 been fixed. The domain is no longer cached after a negative response to
6326 MAIL if the sender used is not empty.
6328 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6331 (c) A new callout option, "use_sender" has been added for people who want
6332 the modified behaviour.
6338 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6341 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6342 indeed breaks things for older releases.
6344 3. Added additional logging to the case where there is a problem reading data
6345 from a filter that is running in a subprocess using a pipe, in order to
6346 try to track down a specific problem.
6348 4. Testing facility fudge: when running in the test harness and attempting
6349 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6350 getting "No route to host". Convert this to a timeout.
6352 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6355 6. Some OS don't have socklen_t but use size_t instead. This affects the
6356 fifth argument of getsockopt() amongst other things. This is now
6357 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6358 can be set for individual OS. I have set it for SunOS5, OSF1, and
6359 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6360 some earlier ones do not.
6362 7. Change 4.30/15 was not doing the test caselessly.
6364 8. The standard form for an IPv6 address literal was being rejected by address
6365 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6366 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6367 this, as well as the form without the "IPv6" on the front (but only when
6368 address literals are enabled, of course).
6370 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6372 10. Exim crashed if a message with an empty sender address specified by -f
6373 encountered a router with an errors_to setting. This could be provoked only
6374 by a command such as
6378 where an empty string was supplied; "<>" did not hit this bug.
6380 11. Installed PCRE release 4.5.
6382 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6383 remained set. It is now erased.
6385 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6386 times from message ids (which are base 36 rather than the normal 62).
6388 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6389 were being counted as actual protocol errors, and logged if the log
6390 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6391 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6392 and DATA following a set of rejected RCPTs do not count as protocol errors.
6393 In other words, Exim assumes they were pipelined, though this may not
6394 actually be the case. Of course, in all cases the client gets an
6395 appropriate error code.
6397 15. If a lookup fails in an ACL condition, a message about the failure may
6398 be available; it is used if testing the ACL cannot continue, because most
6399 such messages specify what the cause of the deferral is. However, some
6400 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6401 that caused an old message to be retained and used if a later statement
6402 caused a defer, replacing the real cause of the deferral.
6404 16. If an IP address had so many PTR records that the DNS lookup buffer
6405 was not large enough to hold them, Exim could crash while trying to process
6406 the truncated data. It now detects and logs this case.
6408 17. Further to 4.21/58, another change has been made: if (and only if) the
6409 first line of a message (the first header line) ends with CRLF, a bare LF
6410 in a subsequent header line has a space inserted after it, so as not to
6411 terminate the header.
6413 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6414 unnecessarily, used atoi() instead of strtol(), and didn't check the
6415 termination when getting file sizes from file names by regex.
6417 19. Completely re-implemented the support for maildirsize files, in the light
6418 of a number of problems with the previous contributed implementation
6419 (4.30/29). In particular:
6421 . If the quota is zero, the maildirsize file is maintained, but no quota is
6424 . If the maildir directory does not exist, it is created before any attempt
6425 to write a maildirsize file.
6427 . The quota value in the file is just a cache; if the quota is changed in
6428 the transport, the new value overrides.
6430 . A regular expression is available for excluding directories from the
6433 20. The autoreply transport checks the characters in options that define the
6434 message's headers; it allows continued headers, but it was checking with
6435 isspace() after an embedded newline instead of explicitly looking for a
6438 21. If all the "regular" hosts to which an address was routed had passed their
6439 expiry times, and had not reached their retry times, the address was
6440 bounced, even if fallback hosts were defined. Now Exim should go on to try
6443 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6444 equivalent code in the SMTP transport. Some hosts send humungous responses
6445 to HELO/EHLO, more than 1024 it seems.
6447 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6448 gives compiler warnings in some environments. I've now done it "properly",
6451 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6452 (because of problems with the built-in one) was declared to return uschar *
6453 instead of char *, causing compiler failure.
6455 25. Fixed a file descriptor leak when processing alias/forward files.
6457 26. Fixed a minor format string issue in dbfn.c.
6459 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6461 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6462 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6465 29. When a sender address is verified, it is cached, to save repeating the test
6466 when there is more than one recipient in a message. However, when the
6467 verification involves a callout, it is possible for different callout
6468 options to be set for different recipients. It is too complicated to keep
6469 track of this in the cache, so now Exim always runs a verification when a
6470 callout is required, relying on the callout cache for the optimization.
6471 The overhead is duplication of the address routing, but this should not be
6474 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6475 to be verified with callout=postmaster, and the main callout worked but the
6476 postmaster check failed, the verification correctly failed. However, if a
6477 subsequent RCPT command asked for sender verification *without* the
6478 postmaster check, incorrect caching caused this verification also to fail,
6481 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6482 it was not caching the DNS options (qualify_single, search_parents) that
6483 were used when the lookup failed. A subsequent lookup with different
6484 options therefore always gave the same answer, though there were cases
6485 where it should not have. (Example: a "domains = !$mx_any" option on a
6486 dnslookup router: the "domains" option is always processed without any
6487 widening, but the router might have qualify_single set.) Now Exim uses the
6488 cached value only when the same options are set.
6490 32. Added John Jetmore's "exipick" utility to the distribution.
6492 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6493 than a timeout (e.g. a certificate is required, and is not provided), an
6494 Exim server now closes the connection immediately. Previously it waited for
6495 the client to close - but if the client is SSL, it seems that they each
6496 wait for each other, leading to a delay before one of them times out.
6498 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6499 maintained 0.8.x compatibility because I don't think many are using it, and
6500 it is clearly obsolete.
6502 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6505 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6506 client certificate was expired. A simple patch fixes this, though I don't
6507 understand the full logic of why the verify callback is called multiple
6510 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6511 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6512 which causes problems with some clients (such as the Certicom SSL Plus
6513 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6514 disables the coutermeasure allowing Eudora to connect."
6516 38. Exim was not checking that a write() to a log file succeeded. This could
6517 lead to Bad Things if a log got too big, in particular if it hit a file
6518 size limit. Exim now panics and dies if it cannot write to a log file, just
6519 as it does if it cannot open a log file.
6521 39. Modified OS/Makefile-Linux so that it now contains
6523 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6525 The two -D definitions ensure that Exim is compiled with large file
6526 support, which makes it possible to handle log files that are bigger than
6529 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6530 instance) a domain was checked against a named list that involved a lookup,
6531 causing $domain_data to be set, then another domain was checked against the
6532 same list, then the first domain was re-checked, the value of $domain_data
6533 after the final check could be wrong. In particular, if the second check
6534 failed, it could be set empty. This bug probably also applied to
6537 41. The strip_trailing_dot option was not being applied to the address given
6538 with the -f command-line option.
6540 42. The code for reading a message's header from the spool was incrementing
6541 $received_count, but never initializing it. This meant that the value was
6542 incorrect (doubled) while delivering a message in the same process in which
6543 it was received. In the most common configuration of Exim, this never
6544 happens - a fresh exec is done - but it can happen when
6545 deliver_drop_privilege is set.
6547 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6548 it now includes up to 150 characters of the unexpected data in the log
6551 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6552 and building data strings. The size of both of these buffers was 10 000
6553 bytes - far larger than anybody would *ever* want, thought I. Needless to
6554 say, somebody hit the limit. I have increased the maximum line length to
6555 20 000 and the maximum data length of concatenated lines to 100 000. I have
6556 also fixed two bugs, because there was no checking on these buffers. Tsk,
6557 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6558 buffer is too small.
6560 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6561 lsearch lookups. Now it does.
6563 46. When parsing a route_list item in a manualroute router, a fixed-length
6564 buffer was used for the list of hosts. I made this 1024 bytes long,
6565 thinking that nobody would ever have a list of hosts that long. Wrong.
6566 Somebody had a whole pile of complicated expansion conditions, and the
6567 string was silently truncated, leading to an expansion error. It turns out
6568 that it is easier to change to an unlimited length (owing to other changes
6569 that have happened since this code was originally written) than to build
6570 structure for giving a limitation error. The length of the item that
6571 expands into the list of hosts is now unlimited.
6573 47. The lsearch lookup could not handle data where the length of text line was
6574 more than 4095 characters. Such lines were truncated, leading to shortened
6575 data being returned. It should now handle lines of any length.
6577 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6578 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6581 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6582 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6583 no longer happens in comment lines. A list of replacements is now placed
6584 at the head of all of the source files, except those whose only change is
6585 to replace PERL_COMMAND in the very first #! line.
6587 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6588 messages on the queue, with a bottom-up merge sort, using code contributed
6589 by Michael Haardt. This should make operations like -bp somewhat faster on
6590 large queues. It won't affect queue runners, except when queue_run_in_order
6593 51. Installed eximstats 1.31 in the distribution.
6595 52. Added support for SRV lookups to the dnslookup router.
6597 53. If an ACL referred to $message_body or $message_body_end, the value was not
6598 reset for any messages that followed in the same SMTP session.
6600 54. The store-handling optimization for building very long strings was not
6601 differentiating between the different store pools. I don't think this
6602 actually made any difference in practice, but I've tidied it.
6604 55. While running the routers to verify a sender address, $sender_address
6605 was still set to the sender address. This is wrong, because when routing to
6606 send a bounce to the sender, it would be empty. Therefore, I have changed
6607 it so that, while verifying a sender address, $sender_address is set to <>.
6608 (There is no change to what happens when verifying a recipient address.)
6610 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6611 target A or AAAA records (if not already returned) without resetting the
6612 qualify_single or search_parents options of the DNS resolver. These are
6613 inappropriate in this case because the targets of MX and SRV records must
6614 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6615 target that, when qualified, matched something in the local domain. These
6616 two options are now turned off when doing these lookups.
6618 57. It seems that at least some releases of Reiserfs (which does not have the
6619 concept of a fixed number of inodes) returns zero and not -1 for the
6620 number of available inodes. This interacted badly with check_spool_inodes,
6621 which assumed that -1 was the "no such thing" setting. What I have done is
6622 to check that the total number of inodes is greater than zero before doing
6623 the test of how many are available.
6625 58. When a "warn" ACL statement has a log_message modifier, the message is
6626 remembered, and not repeated. This is to avoid a lot of repetition when a
6627 message has many recipients that cause the same warning to be written.
6628 However, Exim was preserving the list of already written lines for an
6629 entire SMTP session, which doesn't seem right. The memory is now reset if a
6630 new message is started.
6632 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6633 debugging output unless log_rewrite was also set.
6635 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6636 of (int)(handle) when we know that handle contains (void *)(-1).
6638 61. The Exim daemon panic-logs an error return when it closes the incoming
6639 connection. However "connection reset by peer" seems to be common, and
6640 isn't really an error worthy of noting specially, so that particular error
6643 62. When Exim is trying to find all the local interfaces, it used to panic and
6644 die if the ioctl to get the interface flags failed. However, it seems that
6645 on at least one OS (Solaris 9) it is possible to have an interface that is
6646 included in the list of interfaces, but for which you get a failure error
6647 for this call. This happens when the interface is not "plumbed" into a
6648 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6649 failure of the "get flags" call assumes that the interface is down.
6651 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6652 makes life easier for people who are doing arithmetic on fields extracted
6653 from dates, where you often get leading zeros that should not be
6654 interpreted as octal.
6656 64. Added qualify_domain to the redirect router, to override the global
6659 65. If a pathologically long header line contained very many addresses (the
6660 report of this problem mentioned 10 000) and each of them was rewritten,
6661 Exim could use up a very large amount of memory. (It kept on making new
6662 copies of the header line as it rewrote, and never released the old ones.)
6663 At the expense of a bit more processing, the header rewriting function has
6664 been changed so that it no longer eats memory in this way.
6666 66. The generation of the Received: header has been moved from the time that a
6667 message starts to be received, to the time that it finishes. The timestamp
6668 in the Received: header should now be very close to that of the <= log
6669 line. There are two side-effects of this change:
6671 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6672 logged header lines no longer include the local Received: line, because
6673 it has not yet been created. The same applies to a copy of the message
6674 that is returned to a non-SMTP sender when a message is rejected.
6676 (b) When a filter file is tested using -bf, no additional Received: header
6677 is added to the test message. After some thought, I decided that this
6680 This change does not affect the value of $received_for. It is still set
6681 after address rewriting, but before local_scan() is called.
6683 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6685 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6686 gave an unhelpful panic error message, and a defer error. I have managed to
6687 change this behaviour so that it now rejects any supplied certificate,
6688 which seems right, as the list of acceptable certificates is empty.
6690 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6691 gave an unhelpful defer error. I have not managed to make this reject any
6692 supplied certificates, but the error message it gives is "no certificate
6693 supplied", which is not helpful.
6695 70. exigrep's output now also includes lines that are not associated with any
6696 message, but which match the given pattern. Implemented by a patch from
6697 Martin Sluka, which also tidied up the Perl a bit.
6699 71. Recipient callout verification, like sender verification, was using <> in
6700 the MAIL FROM command. This isn't really the right thing, since the actual
6701 sender may affect whether the remote host accepts the recipient or not. I
6702 have changed it to use the actual sender in the callout; this means that
6703 the cache record is now keyed on a recipient/sender pair, not just the
6704 recipient address. There doesn't seem to be a real danger of callout loops,
6705 since a callout by the remote host to check the sender would use <>.
6706 [SEE ABOVE: changed after hitting problems.]
6708 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6709 temporary errors. However, in the case of such a code being given after
6710 the end of a data transmission (i.e. after ".") Exim was failing to write
6711 a retry record for the message. (Yes, there was some broken host that was
6712 actually sending 8xx at this point.)
6714 73. An unknown lookup type in a host list could cause Exim to panic-die when
6715 the list was checked. (An example that provoked this was putting <; in the
6716 middle of a list instead of at the start.) If this happened during a DATA
6717 ACL check, a -D file could be left lying around. This kind of configuration
6718 error no longer causes Exim to die; instead it causes a defer error. The
6719 incident is still logged to the main and panic logs.
6721 74. Buglet left over from Exim 3 conversion. The message "too many messages
6722 in one connection" was written to the rejectlog but not the mainlog, except
6723 when address rewriting (yes!) was being logged.
6725 75. Added write_rejectlog option.
6727 76. When a system filter was run not as root (that is, when system_filter_user
6728 was set), the values of the $n variables were not being returned to the
6729 main process; thus, they were not subsequently available in the $sn
6732 77. Added +return_path_on_delivery log selector.
6734 78. A connection timeout was being treated differently from recipients deferred
6735 when testing hosts_max_try with a message that was older than the host's
6736 retry timeout. (The host should not be counted, thus allowing all hosts to
6737 be tried at least once before bouncing.) This may have been the cause of an
6738 occasionally reported bug whereby a message would remain on the queue
6739 longer than the retry timeout, but would be bounced if a delivery was
6740 forced. I say "may" because I never totally pinned down the problem;
6741 setting up timeout/retry tests is difficult. See also the next item.
6743 79. The ultimate address timeout was not being applied to errors that involved
6744 a combination of host plus message (for example, a timeout on a MAIL
6745 command). When an address resolved to a number of possible hosts, and they
6746 were not all tried for each delivery (e.g. because of hosts_max_try), a
6747 message could remain on the queue longer than the retry timeout.
6749 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6752 81. Fixed an obscure SMTP outgoing bug which required at least the following
6753 conditions: (a) there was another message waiting for the same server;
6754 (b) the server returned 5xx to all RCPT commands in the first message so
6755 that the message was not completed; (c) the server dropped the connection
6756 or gave a negative response to the RSET that Exim sends to abort the
6757 transaction. The observed case was a dropped connection after DATA that had
6758 been sent in pipelining mode. That is, the server had advertised PIPELINING
6759 but was not implementing it correctly. The effect of the bug was incorrect
6760 behaviour, such as trying another host, and this could lead to a crash.
6766 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6767 and daemon.c were passed as pointers to ints; they should have been
6768 pointers to socklen_t variables (which are typically unsigned ints).
6770 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6773 3. Fixed a really odd bug that affected only the testing scheme; patching a
6774 certain fixed string in the binary changed the value of another string that
6775 happened to be identical to the end of the original first string.
6777 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6778 name", it returns that address as the IP address. On some operating
6779 systems (e.g. Solaris), it also passes back the IP address string as the
6780 "host name". However, on others (e.g. Linux), it passes back an empty
6781 string. Exim wasn't checking for this, and was changing the host name to an
6782 empty string, assuming it had been canonicalized.
6784 5. Although rare, it is permitted to have more than one PTR record for a given
6785 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6786 all the names associated with an address, because they do in Solaris.
6787 However, it seems that they do not in Linux for data that comes from the
6788 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6789 I found this out when I moved to a new Linux workstation and tried to run
6790 the Exim test suite.
6792 To get round this problem I have changed the code so that it now does its
6793 own call to the DNS to look up PTR records when searching for a host name.
6794 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6795 addresses that are only in /etc/hosts are still found.
6797 This behaviour is, however, controlled by an option called host_lookup_
6798 order, which defaults to "bydns:byaddr". If people want to use the other
6799 order, or indeed, just use one or the other means of lookup, they can
6800 specify it in this variable.
6802 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6803 some operating systems, this comes back from gethostbyaddr() as an empty
6804 string, and this is what Exim used to test for. However, it seems that in
6805 other systems, "." is yielded. Exim now tests for this case too.
6807 7. The values of check_spool_space and check_log_space are now held internally
6808 as a number of kilobytes instead of an absolute number of bytes. If a
6809 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6810 kilobyte. This means that much larger values can be stored.
6812 8. Exim monitor: an attempt to get the action menu when not actually pointing
6813 at a message produces an empty menu entitled "No message selected". This
6814 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6815 no entries in it ("Shell widget menu has zero width and/or height"). So I
6816 have added a single, blank menu entry in this case.
6818 9. Added ${quote_local_part.
6820 10. MIME decoding is now applied to the contents of Subject: header lines when
6823 11. Now that a reference to $sender_host_address automatically causes a reverse
6824 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6825 host lookup before query-style lookups in lists that might use this
6826 variable. This has therefore been abolished, and the "net-" prefix is no
6827 longer necessary for query-style lookups.
6829 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6830 have been a typo for LFLAGS, so it has been changed.
6832 13. The install script calls Exim with "-C /dev/null" in order to find the
6833 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6834 to be output. However, since Exim outputs its version number before the
6835 error, it didn't break the script. It just looked ugly. I fixed this by
6836 always allowing "-C /dev/null" if the caller is root.
6838 14. Ignore overlarge ACL variable number when reading spool file - insurance
6839 against a later release with more variables having written the file.
6841 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6842 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6843 this, as well as the form without the "IPv6" on the front.
6845 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6846 OS/Makefile-Darwin file.
6848 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6849 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6851 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6852 variable. However, some OS use ints for inodes. I've added cast to long int
6853 to get rid of the compiler warning.
6855 19. I had forgotten to lock out "/../" in configuration file names when
6856 ALT_CONFIG_PREFIX was set.
6858 20. Routers used for verification do not need to specify transports. However,
6859 if such a router generated a host list, and callout was configured, Exim
6860 crashed, because it could not find a port number from the (non-existent)
6861 transport. It now assumes port 25 in this circumstance.
6863 21. Added the -t option to exigrep.
6865 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6866 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6867 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6868 them is set, LOOKUP_LSEARCH is forced.
6870 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6872 24. Added sender and host information to the "rejected by local_scan()" log
6873 line; previously there was no indication of these.
6875 25. Added .include_if_exists.
6877 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6878 Linux. It turns out that not all file systems support this. Apparently some
6879 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6880 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6881 not supported on the file descriptor, is now ignored when Exim is trying to
6882 sync a directory. This applies only to Linux.
6884 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6886 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6887 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6890 29. Added support for maildirsize files from supplied patch (modified a bit).
6892 30. The use of :fail: followed by an empty string could lead Exim to respond to
6893 sender verification failures with (e.g.):
6895 550 Verification failed for <xxx>
6896 550 Sender verify failed
6898 where the first response line was missing the '-' that indicates it is not
6899 the final line of the response.
6901 31. The loop for finding the name of the user that called Exim had a hardwired
6902 limit of 10; it now uses the value of finduser_retries, which is used for
6903 all other user lookups.
6905 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6908 33. Exim was neglecting to zero errno before one call of strtol() when
6909 expanding a string and expecting an integer value. On some systems this
6910 resulted in spurious "integer overflow" errors. Also, it was casting the
6911 result into an int without checking.
6913 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6914 did not work. The code looks as if it has *never* worked, though it appears
6915 to have been documented since at least release 1.62. I have made it work.
6917 35. The "timeout_DNS" error in retry rules, also documented since at least
6918 1.62, also never worked. As it isn't clear exactly what this means, and
6919 clearly it isn't a major issue, I have abolished the feature by treating it
6920 as "timeout", and writing a warning to the main and panic logs.
6922 36. The display of retry rules for -brt wasn't always showing the error code
6925 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6926 timeout_connect_A, timeout_connect_MX.
6928 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6929 to the empty sender.
6931 39. The daemon was not analysing the content of -oX till after it had closed
6932 stderr and disconnected from the controlling terminal. This meant that any
6933 syntax errors were only noted on the panic log, and the return code from
6934 the command was 0. By re-arranging the code a little, I've made the
6935 decoding happen first, so such errors now appear on stderr, and the return
6936 code is 1. However, the actual setting up of the sockets still happens in
6937 the disconnected process, so errors there are still only recorded on the
6940 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6941 connections (as happens on some IP stacks) was logged at start up time as
6942 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6943 it from "IPv6 and IPv4", which means that two separate sockets are being
6946 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6947 says whether AF_INET or AF_INET6 was passed as an argument.
6949 42. Exiwhat output was messed up when time zones were included in log
6952 43. Exiwhat now gives more information about the daemon's listening ports,
6953 and whether -tls-on-connect was used.
6955 44. The "port" option of the smtp transport is now expanded.
6957 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6958 silently ignored. Now an error message is written to the main and panic
6961 46. There's a new ACL modifier called "logwrite" which writes to a log file
6962 as soon as it is encountered.
6964 47. Added $local_user_uid and $local_user_gid at routing time.
6966 48. Exim crashed when trying to verify a sender address that was being
6969 49. Exim was recognizing only a space character after ".include". It now also
6970 recognizes a tab character.
6972 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6973 extracting the relevant information from the specification. The man page no
6974 longer contains scrambled data for the -d option, and I've added a section
6975 at the front about calling Exim under different names.
6977 51. Added "extra_headers" argument to the "mail" command in filter files.
6979 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6982 53. Installed eximstats 1.29.
6984 54. Added transport_filter_timeout as a generic transport option.
6986 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6987 Cc: header lines. This was required by RFC 822, but it not required by RFC
6990 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6991 incoming messages that did not have them. Now it does so only if the
6992 message originates locally, that is, if there is no associated remote host
6993 address. When Resent- header lines are present, this applies to the Resent-
6994 lines rather than the non-Resent- lines.
6996 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6997 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6999 58. Messages for configuration errors now include the name of the main
7000 configuration file - useful now that there may be more than one file in a
7001 list (.included file names were always shown).
7003 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7004 for those rare installations that do not start the daemon as root or run it
7005 setuid root. I've cut out the call to initgroups() if the daemon is not
7008 60. The Exim user and group can now be bound into the binary as text strings
7009 that are looked up at the start of Exim's processing.
7011 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7013 62. Added $mailstore_basename variable.
7015 63. Installed patch to sieve.c from Michael Haardt.
7017 64. When Exim failed to open the panic log after failing to open the main log,
7018 the original message it was trying to log was written to stderr and debug
7019 output, but if they were not available (the usual case in production), it
7020 was lost. Now it is written to syslog before the two lines that record the
7021 failures to open the logs.
7023 65. Users' Exim filters run in subprocesses under the user's uid. It is
7024 possible for a "deliver" command or an alias in a "personal" command to
7025 provoke an address rewrite. If logging of address rewriting is configured,
7026 this fails because the process is not running as root or exim. There may be
7027 a better way of dealing with this, but for the moment (because 4.30 needs
7028 to be released), I have disabled address rewrite logging when running a
7029 filter in a non-root, non-exim process.
7035 1. The buildconfig auxiliary program wasn't quoting the value set for
7036 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7037 not defined. This bug was masked in 4.22 by the effect that was fixed in
7040 2. Some messages that were rejected after a message id was allocated were
7041 shown as "incomplete" by exigrep. It no longer does this for messages that
7042 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7044 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7045 have allow_domain_literals set, the ID did not get logged in the <= line.
7046 Domain literals are now always recognized in Message-ID: header lines.
7048 4. The first argument for a ${extract expansion item is the key name or field
7049 number. Leading and trailing spaces in this item were not being ignored,
7050 causing some misleading effects.
7052 5. When deliver_drop_privilege was set, single queue runner processes started
7053 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7054 same command in the process it spins off) were not dropping privilege.
7056 6. When the daemon running as "exim" started a queue runner, it always
7057 re-executed Exim in the spun-off process. This is a waste of effort when
7058 deliver_drop_privilege is set. The new process now just calls the
7059 queue-runner function directly.
7065 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7068 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7069 ignored. Though the use of -oP was forcing the writing of a pid file, it
7070 was always written to the default place.
7072 3. If the message "no IP address found for host xxxx" is generated during
7073 incoming verification, it is now followed by identification of the incoming
7074 connection (so you can more easily find what provoked it).
7076 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7078 5. Added some features to "harden" Exim a bit more against certain attacks:
7080 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7081 be put in Local/Makefile. This is like the never_users runtime option,
7082 but it cannot be overridden. The default setting is "root".
7084 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7085 prefix string with which any file named in a -C command line option
7088 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7089 is retained for -C and -D only if the caller of Exim is root. Without
7090 it, the exim user may also use -C and -D and retain privilege.
7092 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7093 command line option is disabled.
7095 6. Macro names set by the -D option must start with an upper case letter, just
7096 like macro names defined in the configuration file.
7098 7. Added "dereference=" facility to LDAP.
7100 8. Two instances of the typo "uknown" in the source files are fixed.
7102 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7103 the Configure-Makefile script screwed up while processing it.
7105 10. Incorporated PCRE 4.4.
7107 11. The SMTP synchronization check was not operating right at the start of an
7108 SMTP session. For example, it could not catch a HELO sent before the client
7109 waited for the greeting. There is now a check for outstanding input at the
7110 point when the greeting is written. Because of the duplex, asynchronous
7111 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7112 way, but not yet received, when the check is performed.
7114 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7115 on TCP/IP sockets, because this apparently causes some broken clients to
7118 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7119 unchanged) from the Cygwin maintainer.
7121 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7122 was supported instead of testing for mbx. Effectively a typo.
7124 15. The spa authenticator server code was not checking that the input it
7125 received was valid base64.
7127 16. The debug output line for the "set" modifier in ACLs was not showing the
7128 name of the variable that was being set.
7130 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7132 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7133 lookup on its own. Something else was needed to trigger the lookup. For
7134 example, a match in host_lookup or the need for a host name in a host list.
7135 Now, if $sender_host_name is referenced and the host name has not yet been
7136 looked up, a lookup is performed. If the lookup fails, the variable remains
7137 empty, and $host_lookup_failed is set to "1".
7139 19. Added "eqi" as a case-independent comparison operator.
7141 20. The saslauthd authentication condition could segfault if neither service
7142 nor realm was specified.
7144 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7145 error message that was logged was misleading, and incoming SMTP
7146 connections were dropped. The message is now more accurate, and temporary
7147 errors are given to SMTP connections.
7149 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7150 (except RSET) with a 421 error, until QUIT is received. However, it was
7151 failing to send a response to QUIT.
7153 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7154 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7155 ACL is now run after the helo_try_verify_hosts code.
7157 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7158 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7159 case-independent, so other case variants are also recognized). Apparently
7160 some systems use these upper case variants.
7162 25. If more than two messages were waiting for the same host, and a transport
7163 filter was specified for the transport, Exim sent two messages over the
7164 same TCP/IP connection, and then failed with "socket operation on non-
7165 socket" when it tried to send the third.
7167 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7169 27. The extern definition of crypt16() in expand.c was not being excluded when
7170 the OS had its own crypt16() function.
7172 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7173 as a preferred synonym for return_size_limit, both as an option and as an
7176 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7178 30. Changed the default configuration ACL to relax the local part checking rule
7179 for addresses that are not in any local domains. For these addresses,
7180 slashes and pipe symbols are allowed within local parts, but the sequence
7181 /../ is explicitly forbidden.
7183 31. SPA server authentication was not clearing the challenge buffer before
7186 32. log_message in a "warn" ACL statement was writing to the reject log as
7187 well as to the main log, which contradicts the documentation and doesn't
7188 seem right (because no rejection is happening). So I have stopped it.
7190 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7191 However, I am unable to do any testing of this.
7193 34. Fixed an infelicity in the appendfile transport. When checking directories
7194 for a mailbox, to see if any needed to be created, it was accidentally
7195 using path names with one or more superfluous leading slashes; tracing
7196 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7198 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7199 contents are added to the log line that is written for every discarded
7200 recipient. (Previously a log_message setting was ignored.)
7202 36. The ${quote: operator now quotes the string if it is empty.
7204 37. The install script runs exim in order to find its version number. If for
7205 some reason other than non-existence or emptiness, which it checks, it
7206 could not run './exim', it was installing it with an empty version number,
7207 i.e. as "exim-". This error state is now caught, and the installation is
7210 38. An argument was missing from the function that creates an error message
7211 when Exim fails to connect to the socket for saslauthd authentication.
7212 This could cause Exim to crash, or give a corrupted message.
7214 39. Added isip, isip4, and isip6 to ${if conditions.
7216 40. The ACL variables $acl_xx are now saved with the message, and can be
7217 accessed later in routers, transports, and filters.
7219 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7220 strings in the file are not string-expanded.
7222 42. If a MAIL command specified a SIZE value that was too large to fit into an
7223 int variable, the check against message_size_limit failed. Such values are
7224 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7225 day this will have to be increased, but I don't think I want to be around
7226 when emails are that large.
7233 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7234 iconv() is not standard in FreeBSD.
7236 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7237 IPv6 enabled. The observed symptom was a segmentation fault on return from
7238 the function os_common_find_running_interfaces() in src/os.c.
7240 3. In the check_special_case() function in daemon.c I had used "errno" as an
7241 argument name, which causes warnings on some systems. This was basically a
7242 typo, since it was named "eno" in the comments!
7244 4. The code that waits for the clock to tick (at a resolution of some fraction
7245 of a second) so as to ensure message-id uniqueness was always waiting for
7246 at least one whole tick, when it could have waited for less. [This is
7247 almost certainly not relevant at current processor speeds, where it is
7248 unlikely to ever wait at all. But we try to future-proof.]
7250 5. The function that sleeps for a time interval that includes fractions of a
7251 second contained a race. It did not block SIGALRM between setting the
7252 timer, and suspending (a couple of lines later). If the interval was short
7253 and the sigsuspend() was delayed until after it had expired, the suspension
7254 never ended. On busy systems this could lead to processes getting stuck for
7257 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7258 process, before it forks any delivery processes. The open lookup caching
7259 mechanism meant that the open file or database connection was passed into
7260 the delivery process. The problem was that delivery processes always tidy
7261 up cached lookup data. This could cause a problem for the next delivery
7262 process started by the queue runner, because the external queue runner
7263 process does not know about the closure. So the next delivery process
7264 still has data in the lookup cache. In the case of a file lookup, there was
7265 no problem because closing a file descriptor in a subprocess doesn't affect
7266 the parent. However, if the lookup was caching a connection to a database,
7267 the connection was closed, and the second delivery process was likely to
7268 see errors such as "PGSQL: query failed: server closed the connection
7269 unexpectedly". The problem has been fixed by closing all cached lookups
7270 in a queue runner before running a delivery process.
7272 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7273 seem to have the "const" qualifier which it has on other OS. I've
7276 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7277 *of the same type* (client or server) with the same public name that an
7278 error should be diagnosed.
7280 9. When Exim looked up a host name for an IP address, but failed to find the
7281 original IP address when looking up the host name (a safety check), it
7282 output the message "<ip address> does not match any IP for NULL", which was
7283 confusing, to say the least. The bug was that the host name should have
7284 appeared instead of "NULL".
7286 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7287 user that is built into the binary, and the -C or -D options is used, root
7288 privilege is dropped before the configuration file is read. In addition,
7289 logging is switched to stderr instead of the normal log files. If the
7290 configuration then re-defines the Exim user, the unprivileged environment
7291 is probably not what is expected, so Exim logs a panic warning message (but
7294 However, if deliver_drop_privilege is set, the unprivileged state may well
7295 be exactly what is intended, so the warning has been cut out in that case,
7296 and Exim is allowed to try to write to its normal log files.
7302 1. smtp_return_error_details was not giving details for temporary sender
7303 or receiver verification errors.
7305 2. Diagnose a configuration error if two authenticators have the same public
7308 3. Exim used not to create the message log file for a message until the first
7309 delivery attempt. This could be confusing when incoming messages were held
7310 for policy or load reasons. The message log file is now created at the time
7311 the message is received, and an initial "Received" line is written to it.
7313 4. The automatically generated man page for command line options had a minor
7314 bug that caused no ill effects; however, a more serious problem was that
7315 the procedure for building the man page automatically didn't always
7316 operate. Consequently, release 4.20 contains an out-of-date version. This
7317 shouldn't happen again.
7319 5. When building Exim with embedded Perl support, the script that builds the
7320 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7321 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7323 6. The freeze_tell option was not being used for messages that were frozen on
7324 arrival, either by an ACL or by local_scan().
7326 7. Added the smtp_incomplete_transaction log selector.
7328 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7329 was accepting AUTH without a new EHLO.
7331 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7332 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7335 10. Logging of TCP/IP connections (when configured) now happens in the main
7336 daemon process instead of the child process, so that the TCP/IP connection
7337 count is more accurate (but it can never be perfect).
7339 11. The use of "drop" in a nested ACL was not being handled correctly in the
7340 outer ACL. Now, if condition failure induced by the nested "drop" causes
7341 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7342 or "require"), the connection is dropped.
7344 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7345 that yield "discard" can now be used with an "accept" or a "discard" verb,
7346 but an error is generated for any others (because I can't see a useful way
7347 to define what should happen).
7349 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7350 are now processed in the same way as lines in the Exim configuration file.
7351 In particular, continuation lines are supported.
7353 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7355 15. Added -ti meaning -t -i.
7357 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7358 domains, and warn by logging if others are found.
7360 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7361 returned from GIFCONF when Exim is trying to find the list of interfaces on
7362 a host. The code in os.c has been modified to copy each ifreq to an aligned
7363 structure in all cases.
7365 Also, in some cases, the returned ifreq's were being copied to a 'struct
7366 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7367 means the last couple of bytes of an IPv6 address could be chopped if the
7368 ifreq contained only a normal sockaddr (14 bytes storage).
7370 18. Named domain lists were not supported in the hosts_treat_as_local option.
7371 An entry such as +xxxx was not recognized, and was treated as a literal
7374 19. Ensure that header lines added by a DATA ACL are included in the reject log
7375 if the ACL subsequently rejects the message.
7377 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7378 MD5 (which is deprecated).
7380 21. When testing a filter file using -bf, Exim was writing a message when it
7381 took the sender from a "From " line in the message, but it was not doing so
7382 when it took $return_path from a Return-Path: header line. It now does.
7384 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7385 with a valid header line field name (a series of printing characters
7386 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7388 23. Changed "disc" in the source to "disk" to conform to the documentation and
7389 the book and for uniformity.
7391 24. Ignore Sendmail's -Ooption=value command line item.
7393 25. When execve() failed while trying to run a command in a pipe transport,
7394 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7395 could be confused with a return value of 69 from the command itself. This
7396 has been changed to 127, the value the shell returns if it is asked to run
7397 a non-existent command. The wording for the related log line suggests a
7398 non-existent command as the problem.
7400 26. If received_header_text expands to an empty string, do not add a Received:
7401 header line to the message. (Well, it adds a token one on the spool, but
7402 marks it "old" so that it doesn't get used or transmitted.)
7404 27. Installed eximstats 1.28 (addition of -nt option).
7406 28. There was no check for failure on the call to getsockname() in the daemon
7407 code. This can fail if there is a shortage of resources on the system, with
7408 ENOMEM, for example. A temporary error is now given on failure.
7410 29. Contrary to the C standard, it seems that in some environments, the
7411 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7412 program. Exim now does this explicitly; it affects the formatting of
7413 timestamps using strftime().
7415 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7416 complaints. I've now initialized all the variables, to avoid this.
7418 32. Header lines added by a system filter were not being "seen" during
7419 transport-time rewrites.
7421 33. The info_callback() function passed to OpenSSL is set up with type void
7422 (*)(SSL *, int, int), as described somewhere. However, when calling the
7423 function (actually a macro) that sets it up, the type void(*)() is
7424 expected. I've put in a cast to prevent warnings from picky compilers.
7426 34. If a DNS black list lookup found a CNAME record, but there were no A
7427 records associated with the domain it pointed at, Exim crashed.
7429 35. If a DNS black list lookup returned more than one A record, Exim ignored
7430 all but the first. It now scans all returned addresses if a particular IP
7431 value is being sought. In this situation, the contents of the
7432 $dnslist_value variable are a list of all the addresses, separated by a
7435 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7436 to accept a host name and all its aliases if the forward lookup for any of
7437 them yielded the IP address of the incoming connection. Now it accepts only
7438 those names whose forward lookup yields the correct IP address. Any other
7439 names are discarded. This closes a loophole whereby a rogue DNS
7440 administrator could create reverse DNS records to break through a
7441 wildcarded host restriction in an ACL.
7443 37. If a user filter or a system filter that ran in a subprocess used any of
7444 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7445 the wrong values were passed to the pipe command ($thisaddress had the
7446 value of $0, $0 had the value of $1, etc). This bug was introduced by
7447 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7449 38. Improved the line breaking for long SMTP error messages from ACLs.
7450 Previously, if there was no break point between 40 and 75 characters, Exim
7451 left the rest of the message alone. Two changes have been made: (a) I've
7452 reduced the minimum length to 35 characters; (b) if it can't find a break
7453 point between 35 and 75 characters, it looks ahead and uses the first one
7454 that it finds. This may give the occasional overlong line, but at least the
7455 remaining text gets split now.
7457 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7458 file descriptors might be low, and that setting 1000 would always raise it.
7459 It turns out that in some environments, the limit is already over 1000 and
7460 that lowering it causes trouble. So now Exim takes care not to decrease it.
7462 40. When delivering a message, the value of $return_path is set to $sender_
7463 address at the start of routing (routers may change the value). By an
7464 oversight, this default was not being set up when an address was tested by
7465 -bt or -bv, which affected the outcome if any router or filter referred to
7468 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7469 log only when "message" or "log_message" are set. However, if one of the
7470 conditions was an address verification, or a call to a nested ACL, the
7471 messages generated by the underlying test were being passed through. This
7472 no longer happens. The underlying message is available in $acl_verify_
7473 message for both "message" and "log_message" expansions, so it can be
7474 passed through if needed.
7476 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7477 new expansion $bh_ to give the encoded byte string without charset
7478 translation. Translation happens only if iconv() is available; HAVE_ICONV
7479 indicates this at build time. HEADERS_CHARSET gives the charset to
7480 translate to; headers_charset can change it in the configuration, and
7481 "headers charset" can change it in an individual filter file.
7483 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7484 that creates RFC 2047 encoded "words" labels them as that charset instead
7485 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7486 expansion operator; (ii) when Exim creates a From: line for a local
7487 message; (iii) when a header line is rewritten to include a "phrase" part.
7489 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7490 buggy, causing it to skip the first lines of messages whose message ID
7491 ended in 'D'. This would not have bitten before Exim release 4.14, because
7492 message IDs were unlikely to end in 'D' before then. The effect was to have
7493 incorrect size information for certain domains.
7495 45. #include "config.h" was missing at the start of the crypt16.c module. This
7496 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7499 46. If there was a timeout during a "random" callout check, Exim treated it as
7500 a failure of the random address, and carried on sending RSET and the real
7501 address. If the delay was just some slowness somewhere, the response to the
7502 original RCPT would be taken as a response to RSET and so on, causing
7503 mayhem of various kinds.
7505 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7506 when I implemented it. It didn't allow for the fact that some option values
7507 may legitimately be negative (e.g. size_addition), and it didn't even do
7508 the right test for positive values.
7510 48. Domain names in DNS records are case-independent. Exim always looks them up
7511 in lower case. Some resolvers return domain names in exactly the case they
7512 appear in the zone file, that is, they may contain uppercase letters. Not
7513 all resolvers do this - some return always lower case. Exim was treating a
7514 change of case by a resolver as a change of domain, similar to a widening
7515 of a domain abbreviation. This triggered its re-routing code and so it was
7516 trying to route what was effectively the same domain again. This normally
7517 caused routing to fail (because the router wouldn't handle the domain
7518 twice). Now Exim checks for this case specially, and just changes the
7519 casing of the domain that it ultimately uses when it transmits the message
7522 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7525 50. If a filter generated a file delivery with a non-absolute name (possible if
7526 no home directory exists for the router), the forbid_file option was not
7529 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7530 the existing equality matching.
7532 52. Exim was using ints instead of ino_t variables in some places where it was
7533 dealing with inode numbers.
7535 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7536 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7537 called TMPDIR, and if it finds it is different, it changes its value.
7539 54. The smtp_printf() function is now made available to local_scan() so
7540 additional output lines can be written before returning. There is also an
7541 smtp_fflush() function to enable the detection of a dropped connection.
7542 The variables smtp_input and smtp_batched_input are exported to
7545 55. Changed the default runtime configuration: the message "Unknown user"
7546 has been removed from the ACL, and instead placed on the localuser router,
7547 using the cannot_route_message feature. This means that any verification
7548 failures that generate their own messages won't get overridden. Similarly,
7549 the "Unrouteable address" message that was in the ACL for unverifiable
7550 relay addresses has also been removed.
7552 56. Added hosts_avoid_esmtp to the smtp transport.
7554 57. The exicyclog script was not checking for the esoteric option
7555 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7556 will work only if exicyclog is run under the appropriate euid.
7558 58. Following a discussion on the list, the rules by which Exim recognises line
7559 endings on incoming messages have been changed. The -dropcr and drop_cr
7560 options are now no-ops, retained only for backwards compatibility. The
7561 following line terminators are recognized: LF CRLF CR. However, special
7562 processing applies to CR:
7564 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7565 nor a local message in the state where . is a terminator.
7567 (ii) If a bare CR is encountered in a header line, an extra space is added
7568 after the line terminator so as not to end the header. The reasoning
7569 behind this is that bare CRs in header lines are most likely either
7570 to be mistakes, or people trying to play silly games.
7572 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7573 was being incorrectly given as 18 bytes larger than it should have been.
7574 This is a VOB (very old bug).
7576 60. This may never have affected anything current, but just in case it has:
7577 When the local host is found other than at the start of a list of hosts,
7578 the local host, those with the same MX, and any that follow, are discarded.
7579 When the list in question was part of a longer list of hosts, the following
7580 hosts (not currently being processed) were also being discarded. This no
7581 longer happens. I'm not sure if this situation could ever has previously
7584 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7587 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7588 header. This is something that is recommended in a new Internet Draft, and
7589 is something that is documented as being done by Sendmail. There are two
7590 possible values. For messages generated by the autoreply transport, Exim
7593 Auto-Submitted: auto-replied
7595 whereas for all other generated messages (e.g. bounces) it adds
7597 Auto-Submitted: auto-generated
7599 63. The "personal" condition in filters now includes a test for the
7600 Auto-Submitted: header. If it contains the string "auto-" the message it
7601 not considered personal.
7603 64. Added rcpt_include_affixes as a generic transport option.
7605 65. Added queue_only_override (default true).
7607 66. Added the syslog_duplication option.
7609 67. If what should have been the first header line of a message consisted of
7610 a space followed by a colon, Exim was mis-interpreting it as a header line.
7611 It isn't of course - it is syntactically invalid and should therefore be
7612 treated as the start of the message body. The misbehaviour could have
7613 caused a number of strange effects, including loss of data in subsequent
7614 header lines, and spool format errors.
7616 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7617 client host had authenticated. This control can now be exercised by an ACL
7618 for more flexibility.
7620 69. By default, callouts do not happen when testing with -bh. There is now a
7621 variant, -bhc, which does actually run the callout code, including
7622 consulting and updating the callout cache.
7624 70. Added support for saslauthd authentication, courtesy of Alexander
7627 71. If statvfs() failed on the spool or log directories while checking their
7628 size for availability, Exim confusingly gave the error "space shortage".
7629 Furthermore, in debugging mode it crashed with a floating point exception.
7630 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7631 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7632 serious problem, Exim now writes to the main and panic logs when this
7633 happens, with details of the failure. It then refuses to accept the
7634 incoming message, giving the message "spool directory problem" or "log
7635 directory problem" with a 421 code for SMTP messages.
7637 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7638 0, 1, and 2 exist, because some OS complain for execs without them (see
7639 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7640 descriptors. However, the code omitted to check that the open succeeded,
7641 causing mysterious errors if for some reason the permissions on /dev/null
7642 got screwed. Now Exim writes a message to the main and panic logs, and
7643 bombs out if it can't open /dev/null.
7645 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7646 interact so that it is all more flexible. It is supposed to remain
7647 backwards compatible. Also added extra_local_interfaces.
7649 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7650 to bomb out with an assertion failure - to the client this appears as a
7651 connection drop. This problem occurs in the part of the code that was taken
7652 from the Samba project. Fortunately, the assertion is in a very simple
7653 function, so I have fixed this by reproducing the function inline in the
7654 one place where it is called, and arranging for authentication to fail
7655 instead of killing the process with assert().
7657 75. The SPA client code was not working when the server requested OEM rather
7658 than Unicode encoding.
7660 76. Added code to make require_files with a specific uid setting more usable in
7661 the case where statting the file as root fails - usually a non-root-mounted
7662 NFS file system. When this happens and the failure is EACCES, Exim now
7663 forks a subprocess and does the per-uid checking as the relevant uid.
7665 77. Added process_log_path.
7667 78. If log_file_path was not explicitly set, a setting of check_log_space or
7668 check_log_inodes was ignored.
7670 79. If a space check for the spool or log partitions fails, the incident is now
7671 logged. Of course, in the latter case the data may get lost...
7673 80. Added the %p formatting code to string_format() so that it can be used to
7674 print addresses in debug_print(). Adjusted all the address printing in the
7675 debugging in store.c to use %p rather than %d.
7677 81. There was a concern that a line of code in smtp_in.c could overflow a
7678 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7679 initially expressed, the concern was not well-founded, because trailing
7680 spaces are removed early. However, if the trailing spaces were followed by
7681 a NULL, they did not get removed, so the overflow was possible. Two fixes
7684 (a) I re-wrote the offending code in a cleaner fashion.
7685 (b) If an incoming SMTP command contains a NULL character, it is rejected
7688 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7689 runs initgroups(), so that if the Exim user is in any additional groups,
7690 they will be used during message reception.
7696 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob