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
118 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
119 had introduced a string-copy (for ensuring NUL-termination) which was not
120 appropriate for that case, which can include embedded NUL bytes in the
121 block of data. Investigation showed the copy to actually be needless, the
122 data being length-specified.
124 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
125 done during a receiving connection, and both used TLS, global info was
126 used rather than per-connection info for tracking the state of data
127 queued for transmission. This could result in a connection hang.
129 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
130 Previously, when delivering serveral messages down a single connection
131 only the first would provide a SIZE. This was due to the size information
132 not being properly tracked.
134 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
135 TAI (at 37 seconds currently), pretend to be in UTC for time-related
136 expansion and logging. Previously, spurious values such as a future
137 minute could be seen.
139 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
140 it could crash from a null-deref. This could also affect the
141 ${addresses: } operator and ${readsock } item.
143 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
144 message following a mime one, the variable was not reset.
146 JH/30 When an pipelined-connect fails at the first response, assume incorrect
147 cached capability (perhaps the peer reneged?) and immediately retry in
150 JH/31 Fix spurious detection of timeout while writing to transport filter.
152 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
153 an attempt to copy the string was made before checking it.
159 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
160 side, to decode a TLS 1.3 packet capture.
162 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
163 Previously the default library behaviour applied, sending two, each in
166 JH/03 Debug output for ACL now gives the config file name and line number for
169 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
171 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
173 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
174 buffer overrun for (non-chunking) other transports.
176 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
177 TLS1.3, means that a server rejecting a client certificate is not visible
178 to the client until the first read of encrypted data (typically the
179 response to EHLO). Add detection for that case and treat it as a failed
180 TLS connection attempt, so that the normal retry-in-clear can work (if
181 suitably configured).
183 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
184 and/or domain. Found and fixed by Jason Betts.
186 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
187 configuration). If a CNAME target was not a wellformed name pattern, a
190 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
191 the OS reports them interleaved with other addresses.
193 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
194 used both for input and for a verify callout, both encrypted, SMTP
195 responses being sent by the server could be lost. This resulted in
196 dropped connections and sometimes bounces generated by a peer sending
199 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
200 string. Previously it was possible to cause undefined behaviour in a
201 library routine (usually a crash). Found by "zerons".
203 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
206 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
207 API was removed, so update to use the newer ones.
209 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
210 any timeout set, is taking a long time. Previously we would hang on to a
211 rotated logfile "forever" if the input was arriving with long gaps
212 (a previous attempt to fix addressed lack, for a long time, of initial
215 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
216 shared (NFS) environment. The length of the tempfile name is now
217 4 + 16 ("hdr.$message_exim_id") which might break on file
218 systems which restrict the file name length to lower values.
221 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
222 shared (NFS) environment.
224 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
225 did for all versions <4.90). Notably -M, -m, --invert, -I may be
228 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
229 on some platforms for bit 31.
231 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
232 to changes apparently associated with TLS1.3 handling some of the APIs
233 previously used were either nonfunctional or inappropriate. Strings
234 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
235 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
236 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
237 This affects log line X= elements, the $tls_{in,out}_cipher variables,
238 and the use of specific cipher names in the encrypted= ACL condition.
240 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
242 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
243 verification result was not updated unless hosts_require_ocsp applied.
245 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
246 queue_list_requires_admin set to false, non-admin users were denied the
249 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
250 directory-of-certs mode. Previously they were advertised despite the
253 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
254 A single TCP connection by a client will now hold a TLS connection open
255 for multiple message deliveries, by default. Previoud the default was to
258 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
259 default. If built with the facility, DANE will be used. The facility
260 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
262 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
263 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
264 must be defined and you must still, unless you define DISABLE_TLS, manage
265 the the include-dir and library-file requirements that go with that
266 choice. Non-TLS builds are still supported.
268 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
271 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
272 default. If the platform supports and has the facility enabled, it will
273 be requested on all coneections.
275 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
276 controlled by the build-time option SUPPORT_PIPE_CONNECT.
278 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
280 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
281 success-DSN messages. Previously the From: header was always the default
282 one for these; the option was ignored.
284 JH/28 Fix the timeout on smtp response to apply to the whole response.
285 Previously it was reset for every read, so a teergrubing peer sending
286 single bytes within the time limit could extend the connection for a
287 long time. Credit to Qualsys Security Advisory Team for the discovery.
289 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
290 delivery address, which leaked information of the results of local
291 forwarding. Change to the original envelope recipient address, per
294 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
295 requested. Previously not bounce was generated and a log entry of
296 error ignored was made.
298 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
300 JH/32 Introduce a general tainting mechanism for values read from the input
301 channel, and values derived from them. Refuse to expand any tainted
302 values, to catch one form of exploit.
304 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
305 was unused and the unexpanded text used for the test. Found and
306 fixed by Ruben Jenster.
308 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
309 an attempt to use a TLS library read routine dereffed a nul pointer,
312 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
313 them in our smtp response.
315 JH/36 Have the general router option retry_use_local_part default to true when
316 any of the restrictive preconditions are set (to anything). Previously it
317 was only for check_local user. The change removes one item of manual
318 configuration which is required for proper retries when a remote router
319 handles a subset of addresses for a domain.
321 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
322 link count into consideration.
324 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
325 caused the extension of big_buffer, the following lines were ignored.
327 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
328 accordance with RFC 2308. Previously there was no expiry, so a longlived
329 receive process (eg. due to ACL delays) versus a short SOA value could
332 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
334 JH/39 Promote DMARC support to mainline.
336 JH/40 Bug 2452: Add a References: header to DSNs.
338 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
339 parameters. The relevant library call is documented as "Deprecated: This
340 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
341 3.6.0, DH parameters are negotiated following RFC7919."
343 HS/06 Change the default of dnssec_request_domains to "*"
345 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
346 carried on and emitted a BDAT command, even when PIPELINING was not
349 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
350 buffer was used for the filename, resulting in a trap when tainted
351 arguments (eg. $domain) were used.
353 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
354 recommended to avoid a possible server-load attack. The feature can be
355 re-enabled via the openssl_options main cofiguration option.
357 JH/45 local_scan API: documented the current smtp_printf() call. This changed
358 for version 4.90 - adding a "more data" boolean to the arguments.
359 Bumped the ABI version number also, this having been missed previously;
360 release versions 4.90 to 4.92.3 inclusive were effectively broken in
361 respect of usage of smtp_printf() by either local_scan code or libraries
362 accessed via the ${dlfunc } expansion item. Both will need coding
363 adjustment for any calls to smtp_printf() to match the new function
364 signature; a FALSE value for the new argument is always safe.
366 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
367 the file-offset (which the Linux syscall does, and exim expects); this
368 resulted in an indefinite loop.
370 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
371 to do ARC verification. The Authentication-Results: header line added
372 by the configuration then had no ARC item.
378 JH/01 Remove code calling the customisable local_scan function, unless a new
379 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
381 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
382 non-signal-safe functions being used.
384 JH/03 Bug 2269: When presented with a received message having a stupidly large
385 number of DKIM-Signature headers, disable DKIM verification to avoid
386 a resource-consumption attack. The limit is set at twenty.
388 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
389 report of oldest_pass in ${authres } in consequence, and separate out
390 some descriptions of reasons for verification fail.
392 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
393 files in the spool were present and unlocked. A queue-runner could spot
394 them, resulting in a duplicate delivery. Fix that by doing the unlock
395 after the unlink. Investigation by Tim Stewart. Take the opportunity to
396 add more error-checking on spoolfile handling while that code is being
399 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
400 No known attacks, no CVE, this is defensive hardening.
402 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
403 a queue-runner could start a delivery while other operations were ongoing.
404 Cutthrough delivery was a common victim, resulting in duplicate delivery.
405 Found and investigated by Tim Stewart. Fix by using the open message data
406 file handle rather than opening another, and not locally closing it (which
407 releases a lock) for that case, while creating the temporary .eml format
408 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
410 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
411 $sender_verify_failure/$recipient_verify_failure to "random".
413 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
416 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
417 Previously this would segfault.
419 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
422 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
423 like zero, since the resolver should be doing this for us, But we need one
424 as a CNAME but no MX presence gets the CNAME returned; we need to check
425 that doesn't point to an MX to declare it "no MX returned" rather than
426 "error, loop". A new main option is added so the older capability of
427 following some limited number of chain links is maintained.
429 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
431 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
432 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
433 now-deprecated earlier definitions used only specified the range up to TLS
434 1.2 (in the older-version library docs).
436 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
438 JH/15 Rework TLS client-side context management. Stop using a global, and
439 explicitly pass a context around. This enables future use of TLS for
440 connections to service-daemons (eg. malware scanning) while a client smtp
441 connection is using TLS; with cutthrough connections this is quite likely.
443 JH/16 Fix ARC verification to do AS checks in reverse order.
445 JH/17 Support a "tls" option on the ${readsocket } expansion item.
447 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
448 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
449 Previously the "utf8" would be re-prepended for every additional message.
451 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
452 Previously thery were accepted, resulting in issues when attempting to
453 forward messages to a non-supporting MTA.
455 PP/02 Let -n work with printing macros too, not just options.
457 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
458 one parent address was copied, and bogus data was used at delivery-logging
459 time. Either a crash (after delivery) or bogus log data could result.
460 Discovery and analysis by Tim Stewart.
462 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
463 Previously if the string ended mid-character, we did not insert the
464 promised '?' replacement.
466 PP/04 Documentation: current string operators work on bytes, not codepoints.
468 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
469 should pack well giving a smaller memory footprint so better caching and
470 therefore performance. Group the declarations where this can't be done so
471 that the byte-sized flag variables are not interspersed among pointer
472 variables, giving a better chance of good packing by the compiler.
474 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
475 non-null, to avoid issues with sites running BATV. Previously reports were
476 sent with an empty envelope sender so looked like bounces.
478 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
479 The ignore_error flag wasn't being returned from the filter subprocess so
480 was not set for later routers. Investigation and fix by Matthias Kurz.
482 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
483 and a msg:complete for the whole, when a message is manually removed using
484 -Mrm. Developement by Matthias Kurz, hacked on by JH.
486 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
487 a "Gnu special" function, asprintf() in the DB utility binary builds; I
488 hope that is portable enough.
490 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
491 requiring a known-CA anchor certificate; make it now rely entirely on the
492 TLSA as an anchor. Checking the name on the leaf cert against the name
493 on the A-record for the host is still done for TA (but not for EE mode).
495 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
496 included in delivery lines for non-proxied connections, when compiled with
497 SUPPORT_SOCKS and running with proxy logging enabled.
499 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
500 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
501 move the existing event to fire before the normal logging of message
502 failure so that custom logging is bracketed by normal logging.
504 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
505 msg:fail:internal event. Developement by Matthias Kurz.
507 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
508 far too small for todays use of crypto signatures stored there. Go all
509 the way to the max DNS message size of 64kB, even though this might be
510 overmuch for IOT constrained device use.
512 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
513 copy a string over itself. The library routine is documented as not
514 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
516 JH/32 For main options check_spool_space and check_inode_space, where the
517 platform supports 64b integers, support more than the previous 2^31 kB
518 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
519 the previous G, M, k.
521 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
522 $authenticated_fail_id variable on authentication failure. Previously
525 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
526 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
527 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
528 GNUTLS_SEC_PARAM_MEDIUM.
530 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
531 side. Previously we would continue as if no SNI had been received.
533 JH/36 Harden the handling of string-lists. When a list consisted of a sole
534 "<" character, which should be a list-separator specification, we walked
535 off past the nul-terimation.
537 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
538 causes) even when the retry time is not yet met. Previously they were
539 not, meaning that when (say) an account was over-quota and temp-rejecting,
540 and multiple senders' messages were queued, only one sender would get
541 notified on each configured delay_warning cycle.
543 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
545 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
546 cipher-suites, an error can be left on the stack even for a succeeding
547 accept; this results in impossible error messages when a later operation
550 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
551 return error codes indicating retry. Under TLS1.3 this becomes required.
553 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
554 it only wrote the new authenticators, resulting in a lack of tracking of
555 peer changes of ESMTP extensions until the next cache flush.
557 JH/41 Fix the loop reading a message header line to check for integer overflow,
558 and more-often against header_maxsize. Previously a crafted message could
559 induce a crash of the recive process; now the message is cleanly rejected.
561 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
562 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
568 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
569 When redis_servers is set to a list of > 1 element, and the Redis servers
570 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
571 case of MOVED into a DEFER case instead, thus moving the query onto the
572 next server in the list. For a cluster of N elements, all N servers must
573 be defined in redis_servers.
575 GF/02 Catch and remove uninitialized value warning in exiqsumm
576 Check for existence of @ARGV before looking at $ARGV[0]
578 JH/01 Replace the store_release() internal interface with store_newblock(),
579 which internalises the check required to safely use the old one, plus
580 the allocate and data copy operations duplicated in both (!) of the
581 extant use locations.
583 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
584 modifier. This matches the restriction on the commandline.
586 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
587 Previously only the last row was returned.
589 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
590 we assumed that tags in the header were well-formed, and parsed the
591 element content after inspecting only the first char of the tag.
592 Assumptions at that stage could crash the receive process on malformed
595 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
596 While running the DKIM ACL we operate on the Permanent memory pool so that
597 variables created with "set" persist to the DATA ACL. Also (at any time)
598 DNS lookups that fail create cache records using the Permanent pool. But
599 expansions release any allocations made on the current pool - so a dnsdb
600 lookup expansion done in the DKIM ACL releases the memory used for the
601 DNS negative-cache, and bad things result. Solution is to switch to the
602 Main pool for expansions.
603 While we're in that code, add checks on the DNS cache during store_reset,
604 active in the testsuite.
605 Problem spotted, and debugging aided, by Wolfgang Breyha.
607 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
608 When none of the hosts presented to a transport match an already-open
609 connection, close it and proceed with the list. Previously we would
610 queue the message. Spotted by Lena with Yahoo, probably involving
613 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
614 Previously a spurious "250 OK id=" response was appended to the proper
617 JH/08 The "support for" informational output now, which built with Content
618 Scanning support, has a line for the malware scanner interfaces compiled
619 in. Interface can be individually included or not at build time.
621 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
622 by the template makefile "src/EDITME". The "STREAM" support for an older
623 ClamAV interface method is removed.
625 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
626 rows affected is given instead).
628 JH/11 The runtime Berkeley DB library version is now additionally output by
629 "exim -d -bV". Previously only the compile-time version was shown.
631 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
632 SMTP connection. Previously, when one had more receipients than the
633 first, an abortive onward connection was made. Move to full support for
634 multiple onward connections in sequence, handling cutthrough connection
635 for all multi-message initiating connections.
637 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
638 routers. Previously, a multi-recipient message would fail to match the
639 onward-connection opened for the first recipient, and cause its closure.
641 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
642 a timeout on read on a GnuTLS initiating connection, resulting in the
643 initiating connection being dropped. This mattered most when the callout
644 was marked defer_ok. Fix to keep the two timeout-detection methods
647 JH/15 Relax results from ACL control request to enable cutthrough, in
648 unsupported situations, from error to silently (except under debug)
649 ignoring. This covers use with PRDR, frozen messages, queue-only and
652 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
654 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
655 metadata, resulting in a crash in free().
657 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
658 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
659 Broken also in d185889f4, with init system revamp.
661 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
662 Previously we abruptly closed the connection after reading a malware-
663 found indication; now we go on to read the "scan ok" response line,
666 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
667 ACL. Previously, a crash would result.
669 JH/19 Speed up macro lookups during configuration file read, by skipping non-
670 macro text after a replacement (previously it was only once per line) and
671 by skipping builtin macros when searching for an uppercase lead character.
673 JH/20 DANE support moved from Experimental to mainline. The Makefile control
674 for the build is renamed.
676 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
677 was allocated for every new TLS startup, meaning one per message. Fix
678 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
680 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
681 reported the original. Fix to report (as far as possible) the ACL
682 result replacing the original.
684 JH/23 Fix memory leak during multi-message connections using STARTTLS under
685 OpenSSL. Certificate information is loaded for every new TLS startup,
686 and the resources needed to be freed.
688 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
690 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
693 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
694 DATA response info to the (existing) per-recipient response info for
695 the "C=" log element. It can have useful tracking info from the
696 destination system. Patch from Simon Arlott.
698 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
699 length value. Previously this would segfault.
701 HS/02 Support Avast multiline protoocol, this allows passing flags to
702 newer versions of the scanner.
704 JH/28 Ensure that variables possibly set during message acceptance are marked
705 dead before release of memory in the daemon loop. This stops complaints
706 about them when the debug_store option is enabled. Discovered specifically
707 for sender_rate_period, but applies to a whole set of variables.
708 Do the same for the queue-runner and queue-list loops, for variables set
709 from spool message files. Do the same for the SMTP per-message loop, for
710 certain variables indirectly set in ACL operations.
712 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
713 as a multi-recipient message from a mailinglist manager). The coding had
714 an arbitrary cutoff number of characters while checking for more input;
715 enforced by writing a NUL into the buffer. This corrupted long / fast
716 input. The problem was exposed more widely when more pipelineing of SMTP
717 responses was introduced, and one Exim system was feeding another.
718 The symptom is log complaints of SMTP syntax error (NUL chars) on the
719 receiving system, and refused recipients seen by the sending system
720 (propating to people being dropped from mailing lists).
721 Discovered and pinpointed by David Carter.
723 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
724 replaced by the ${authresults } expansion.
726 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
728 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
729 allows proper process termination in container environments.
731 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
732 Previously the "final dot" had a newline after it; ensure it is CR,LF.
734 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
735 and "err_perm", deprecated since 4.83 when the RFC-defined words
736 "temperror" and "permerror" were introduced.
738 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
739 transport-filters or DKIM-signing. The restriction was lost in the
740 consolidation of verify-callout and delivery SMTP handling.
741 Extend the restriction to also cover ARC-signing.
743 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
744 in defer=pass mode supply a 450 to the initiator. Previously the message
747 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
748 tls_require_ciphers is used as before.
750 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
751 "pass_unscanned". Only tmpfails from the scanner are written to
752 the paniclog, as they may require admin intervention (permission
753 denied, license issues). Other scanner errors (like decompression
754 bombs) do not cause a paniclog entry.
756 JH/36 Fix reinitialisation of DKIM logging variable between messages.
757 Previously it was possible to log spurious information in receive log
760 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
761 triggered odd behaviour from Outlook Express clients.
763 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
766 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
767 since the IETF WG has not yet settled on that versus the original
768 "bare" representation.
770 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
771 Previously the millisecond value corrupted the output.
772 Fix also for syslog_pid=no and log_selector +pid, for which the pid
773 corrupted the output.
779 JH/01 Rework error string handling in TLS interface so that the caller in
780 more cases is responsible for logging. This permits library-sourced
781 string to be attached to addresses during delivery, and collapses
782 pairs of long lines into single ones.
784 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
785 during configuration. Wildcards are allowed and expanded.
787 JH/02 Rework error string handling in DKIM to pass more info back to callers.
788 This permits better logging.
790 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
791 do not close it down and have the child transport start it up again on
792 the passed-on TCP connection. Instead, proxy the child (and any
793 subsequent ones) for TLS via a unix-domain socket channel. Logging is
794 affected: the continued delivery log lines do not have any DNSSEC, TLS
795 Certificate or OCSP information. TLS cipher information is still logged.
797 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
798 identical IP addresses on different listening ports. Will also affect
801 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
802 add noisy ifdef guards to special-case this sillyness.
803 Patch from Bernd Kuhls.
805 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
806 than 255 are no longer allowed.
808 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
809 Disable the session-cache too, which might reduce our load. Since we
810 currrectly use a new context for every connection, both as server and
811 client, there is no benefit for these.
812 GnuTLS appears to not support tickets server-side by default (we don't
813 call gnutls_session_ticket_enable_server()) but client side is enabled
814 by default on recent versions (3.1.3 +) unless the PFS priority string
817 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
818 <https://reproducible-builds.org/specs/source-date-epoch/>.
820 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
821 the check for any unsuccessful recipients did not notice the limit, and
822 erroneously found still-pending ones.
824 JH/08 Pipeline CHUNKING command and data together, on kernels that support
825 MSG_MORE. Only in-clear (not on TLS connections).
827 JH/09 Avoid using a temporary file during transport using dkim. Unless a
828 transport-filter is involved we can buffer the headers in memory for
829 creating the signature, and read the spool data file once for the
830 signature and again for transmission.
832 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
833 4.77 as the kernel support then wasn't solid, having issues in 64bit
834 mode. Now, it's been long enough. Add support for FreeBSD also.
836 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
837 case where the routing stage had gathered several addresses to send to
838 a host before calling the transport for the first, we previously failed
839 to close down TLS in the old transport process before passing the TCP
840 connection to the new process. The new one sent a STARTTLS command
841 which naturally failed, giving a failed delivery and bloating the retry
842 database. Investigation and fix prototype from Wolfgang Breyha.
844 JH/12 Fix check on SMTP command input synchronisation. Previously there were
845 false-negatives in the check that the sender had not preempted a response
846 or prompt from Exim (running as a server), due to that code's lack of
847 awareness of the SMTP input buffering.
849 PP/04 Add commandline_checks_require_admin option.
850 Exim drops privileges sanely, various checks such as -be aren't a
851 security problem, as long as you trust local users with access to their
852 own account. When invoked by services which pass untrusted data to
853 Exim, this might be an issue. Set this option in main configuration
854 AND make fixes to the calling application, such as using `--` to stop
857 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
858 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
859 responses to those, into a single TLS record each way (this usually means
860 a single packet). As a side issue, smtp_enforce_sync now works on TLS
863 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
864 affects you only if you're dancing at the edge of the param size limits.
865 If you are, and this message makes sense to you, then: raise the
866 configured limit or use OpenSSL 1.1. Nothing we can do for older
869 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
870 cmdline element to get the documented default one. Previously it was
873 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
874 are unsuable for non-message-specific reasons - eg. connection timeout,
875 banner-time rejection.
877 JH/16 Fix logging of delivery remote port, when specified by router, under
880 PP/06 Repair manualroute's ability to take options in any order, even if one
881 is the name of a transport.
884 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
886 JH/17 Change the list-building routines interface to use the expanding-string
887 triplet model, for better allocation and copying behaviour.
889 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
890 Previously it was constructed the first time a possibly-matching string
891 was met in the configuration file input during startup; now it is done
894 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
895 compatible one, to avoid the (poorly documented) possibility of a config
896 file in the working directory redirecting the DB files, possibly correpting
897 some existing file. CVE-2017-10140 assigned for BDB.
899 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
900 cache-hot. Previously, although the result was properly cached, the
901 initial verify call returned a defer.
903 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
904 the main verify for receipient in uncached-mode.
906 JH/22 Retire historical build files to an "unsupported" subdir. These are
907 defined as "ones for which we have no current evidence of testing".
909 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
910 if present. Previously it was ignored.
912 JH/24 Start using specified-initialisers in C structure init coding. This is
913 a C99 feature (it's 2017, so now considered safe).
915 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
916 if was a fixed-sized field and bitmask ops via macros; it is now more
919 PP/07 GitHub PR 56: Apply MariaDB build fix.
920 Patch provided by Jaroslav Å karvada.
922 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
923 during Coverity cleanups [4.87 JH/47]
924 Diagnosis and fix provided by Michael Fischer v. Mollard.
926 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
927 the right size to place the terminating semicolon on its own folded
928 line, the header hash was calculated to an incorrect value thanks to
929 the (relaxed) space the fold became.
931 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
932 and confused the parent.
934 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
935 which could crash as a result. This could lead to undeliverable messages.
937 JH/28 Logging: "next input sent too soon" now shows where input was truncated
940 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
941 matters on fast-turnover and PID-randomising systems, which were getting
942 out-of-order delivery.
944 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
945 a possibly-overlapping copy. The symptom was that "Remote host closed
946 connection in response to HELO" was logged instead of the actual 4xx
949 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
950 Previously only that bufferd was discarded, resulting in SYMTP command
953 JH/32 DKIM: when a message has multiple signatures matching an identity given
954 in dkim_verify_signers, run the dkim acl once for each. Previously only
955 one run was done. Bug 2189.
957 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
958 "panic the current process" to "deliberately defer". The panic log is
959 still written with the problem list name; the mail and reject logs now
960 get a temp-reject line for the message that was being handled, saying
961 something like "domains check lookup or other defer". The SMTP 451
962 message is still "Temporary local problem".
964 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
965 A crafted sequence of BDAT commands could result in in-use memory beeing
966 freed. CVE-2017-16943.
968 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
969 from SMTP input. Previously it was always done; now only done for DATA
970 and not BDAT commands. CVE-2017-16944.
972 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
973 to the message (such as an overlong header line). Previously this was
974 not done and we did not exit BDAT mode. Followon from the previous item
975 though a different problem.
981 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
982 than -2003 did; needs libidn2 in addition to libidn.
984 JH/02 The path option on a pipe transport is now expanded before use.
986 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
987 Patch provided by "Björn", documentation fix added too.
989 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
990 missing a wire-to-host endian conversion.
992 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
993 close after a BDAT command line could be taken as a following command,
994 giving a synch failure. Fix by only checking for synch immediately
995 before acknowledging the chunk.
997 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
998 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
999 macro. Patches provided by Josh Soref.
1001 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1002 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1003 by IANA as a service extension. However, John Klensin suggests that we
1006 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1007 the dkim code may be unix-mode line endings rather than smtp wire-format
1008 CRLF, so prepend a CR to any bare LF.
1010 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1011 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1013 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1014 the first were themselves being wrongly included in the feed into dkim
1015 processing; with most chunk sizes in use this resulted in an incorrect
1016 body hash calculated value.
1018 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1019 DKIM signature block, for verification. Although advised against by
1020 standards it is specifically not ruled illegal.
1022 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1024 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1025 missing a body hash (the bh= tag).
1027 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1028 It seems that HAProxy sends the Proxy Protocol information in clear and
1029 only then does a TLS startup, so do the same.
1031 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1032 TCP connections (such as for Spamd) unless the daemon successfully set
1033 Fast Open mode on its listening sockets. This fixes breakage seen on
1034 too-old kernels or those not configured for Fast Open, at the cost of
1035 requiring both directions being enabled for TFO, and TFO never being used
1036 by non-daemon-related Exim processes.
1038 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1039 endings, at least on the first header line. Try to canonify any that get
1040 past that check, despite the cost.
1042 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1043 now limited to an arbitrary five deep, while parsing addresses with the
1044 strip_excess_angle_brackets option enabled.
1046 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1047 instead leave the unprompted TLS handshake in socket buffer for the
1048 TLS library to consume.
1050 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1052 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1054 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1055 region in one operation, for speed, and this leaves assigned pointers
1056 dangling. Add checks run only under the testsuite which checks all
1057 variables at a store-reset and panics on a dangling pointer; add code
1058 explicitly nulling out all the variables discovered. Fixes one known
1059 bug: a transport crash, where a dangling pointer for $sending_ip_address
1060 originally assigned in a verify callout, is re-used.
1062 PP/06 Drop '.' from @INC in various Perl scripts.
1064 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1066 PP/08 Reduce a number of compilation warnings under clang; building with
1067 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1068 should be warning-free.
1070 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1072 HS/01 Fix portability problems introduced by PP/08 for platforms where
1073 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1075 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1076 chunk. This allows us to accept broken chunked messages. We need a more
1077 general solution here.
1079 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1080 already-broken messages in the queue.
1082 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1084 JH/19 Fix reference counting bug in routing-generated-address tracking.
1090 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1091 supports it and a size is available (ie. the sending peer gave us one).
1093 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1094 years of being deprecated). The replacements are the ACLs
1095 acl_smtp_mime and acl_not_smtp_mime.
1097 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1098 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1099 or even an in-clear connection were permitted. Now, if the host lookup
1100 was dnssec and dane was requested then the host is only used if the
1101 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1102 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1103 if one fails this test.
1104 This means that a poorly-configured remote DNS will make it incommunicado;
1105 but it protects against a DNS-interception attack on it.
1107 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1108 non-noisy when a race steals the message being considered.
1110 JH/05 If main configuration option tls_certificate is unset, generate a
1111 self-signed certificate for inbound TLS connections.
1113 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1114 in rewrites and routers.
1116 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1117 and logged a warning sing 4.83; now they are a configuration file error.
1119 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1120 (lacking @domain). Apply the same qualification processing as RCPT.
1122 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1124 JH/10 Support ${sha256:} applied to a string (as well as the previous
1127 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1128 a cutthrough deliver is pending, as we always want to make a connection.
1129 This also avoids re-routing the message when later placing the cutthrough
1130 connection after a verify cache hit.
1131 Do not update it with the verify result either.
1133 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1134 when routing results in more than one destination address.
1136 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1137 signing (which inhibits the cutthrough capability). Previously only
1138 the presence of an option was tested; now an expansion evaluating as
1139 empty is permissible (obviously it should depend only on data available
1140 when the cutthrough connection is made).
1142 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1143 the relevant preceding SMTP command did not note the pipelining mode.
1145 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1146 Previously they were not counted.
1148 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1149 as one having no matching records. Previously we deferred the message
1150 that needed the lookup.
1152 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1153 distinguished as "(=".
1155 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1156 for missing MX records. Previously it only worked for missing A records.
1158 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1160 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1161 after the data-go-ahead and data-ack. Patch from Jason Betts.
1163 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1164 even for a "none" policy. Patch from Tony Meyer.
1166 JH/22 Fix continued use of a connection for further deliveries. If a port was
1167 specified by a router, it must also match for the delivery to be
1170 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1171 When one of the recipients of a message was unsuitable for the connection
1172 (has no matching addresses), we lost track of needing to mark it
1173 deferred. As a result mail would be lost.
1175 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1177 JH/25 Decoding ACL controls is now done using a binary search; the source code
1178 takes up less space and should be simpler to maintain. Merge the ACL
1179 condition decode tables also, with similar effect.
1181 JH/26 Fix problem with one_time used on a redirect router which returned the
1182 parent address unchanged. A retry would see the parent address marked as
1183 delivered, so not attempt the (identical) child. As a result mail would
1186 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1187 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1188 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1191 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1192 Main config options check_{log,spool}_{inodes,space} are now
1193 100 inodes, 10MB unless set otherwise in the configuration.
1195 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1196 Previously it only applied to the main-section connection policy
1199 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1201 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1202 by me. Added RFC7919 DH primes as an alternative.
1204 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1205 are not in the system include path.
1207 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1208 GnuTLS, when a session startup failed (eg because the client disconnected)
1209 Exim did stdio operations after fclose. This was exposed by a recent
1210 change which nulled out the file handle after the fclose.
1212 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1213 signed directly by the cert-signing cert, rather than an intermediate
1214 OCSP-signing cert. This is the model used by LetsEncrypt.
1216 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1218 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1219 an incoming connection.
1221 HS/02 Bug 1802: Do not half-close the connection after sending a request
1224 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1225 fallback to "prime256v1".
1227 JH/34 SECURITY: Use proper copy of DATA command in error message.
1228 Could leak key material. Remotely exploitable. CVE-2016-9963.
1234 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1235 and 3.4.4 - once the server is enabled to respond to an OCSP request
1236 it does even when not requested, resulting in a stapling non-aware
1237 client dropping the TLS connection.
1239 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1240 support variable-length bit vectors. No functional change.
1242 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1243 The I= interface field on outgoing lines is now after the H= remote
1244 host field, same as incoming lines. There is a separate
1245 outgoing_interface log selector which allows you to disable the
1248 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1249 If not running log_selector +smtp_connection the mainlog would be held
1250 open indefinitely after a "too many connections" event, including to a
1251 deleted file after a log rotate. Leave the per net connection logging
1252 leaving it open for efficiency as that will be quickly detected by the
1253 check on the next write.
1255 HS/01 Bug 1671: Fix post transport crash.
1256 Processing the wait-<transport> messages could crash the delivery
1257 process if the message IDs didn't exist for some reason. When
1258 using 'split_spool_directory=yes' the construction of the spool
1259 file name failed already, exposing the same netto behaviour.
1261 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1262 mime_regex ACL conditions.
1264 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1265 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1266 to HELO, local diagnostic string.
1268 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1269 log line to debug. Even when configured with a tls authenticator many
1270 client connections are expected to not authenticate in this way, so
1271 an authenticate fail is not an error.
1273 HS/02 Add the Exim version string to the process info. This way exiwhat
1274 gives some more detail about the running daemon.
1276 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1277 matter for fast-change records such as DNSBLs.
1279 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1280 retry record, even if constant. There may be multiple transports with
1281 different interface settings and the retry behaviour needs to be kept
1284 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1286 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1288 JH/10 Bug 840: fix log_defer_output option of pipe transport
1290 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1291 hosts_randomize. This matters a lot when combined with mua_wrapper.
1293 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1294 ${quote_pgsql:<string>} operator.
1296 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1298 JH/14 Change header file name for freeradius-client. Relevant if compiling
1299 with Radius support; from the Gentoo tree and checked under Fedora.
1301 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1303 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1304 option transport_filter, meaning no filtering.
1306 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1308 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1309 defaults to "*" (all hosts). The variable is now available when not built
1310 with TLS, default unset, mainly to enable keeping the testsuite sane.
1311 If a server certificate is not supplied (via tls_certificate) an error is
1312 logged, and clients will find TLS connections fail on startup. Presumably
1313 they will retry in-clear.
1314 Packagers of Exim are strongly encouraged to create a server certificate
1315 at installation time.
1317 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1318 with the $config_file variable.
1320 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1321 in transport context, after the attempt, and per-recipient. The latter type
1322 is per host attempted. The event data is the error message, and the errno
1323 information encodes the lookup type (A vs. MX) used for the (first) host,
1324 and the trailing two digits of the smtp 4xx response.
1326 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1327 to write to mainlog (or rejectlog, paniclog) in the window between file
1328 creation and permissions/ownership being changed. Particularly affects
1329 installations where exicyclog is run as root, rather than exim user;
1330 result is that the running daemon panics and dies.
1332 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1334 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1335 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1336 "pri" and "weight". Note that the previous implicit priority given by the
1337 list order is no longer honoured.
1339 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1340 for DKIM processing.
1342 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1343 by defining SUPPORT_SOCKS.
1345 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1346 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1347 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1348 variables are renamed to proxy_{local,external}_{address,port}.
1350 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1351 for a build by defining SUPPORT_I18N
1353 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1354 of the query string, and make ${quote_redis:} do that quoting.
1356 JH/29 Move Events support from Experimental to mainline, enabled by default
1357 and removable for a build by defining DISABLE_EVENT.
1359 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1361 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1362 cached by the daemon.
1364 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1365 by defining LOOKUP_REDIS. The libhiredis library is required.
1367 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1368 keys are given for lookup.
1370 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1371 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1372 only supported when built with TLS support. The PolarSSL SHA routines
1373 are still used when the TLS library is too old for convenient support.
1375 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1376 openssl_options), for security. OpenSSL forces this from version 1.1.0
1377 server-side so match that on older versions.
1379 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1380 allocation for $value could be released as the expansion processing
1381 concluded, but leaving the global pointer active for it.
1383 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1384 and to use the domains and local_parts ACL conditions.
1386 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1387 incorrectly not doubled on cutthrough transmission, hence seen as a
1388 body-termination at the receiving system - resulting in truncated mails.
1389 Commonly the sender saw a TCP-level error, and retransmitted the message
1390 via the normal store-and-forward channel. This could result in duplicates
1391 received - but deduplicating mailstores were liable to retain only the
1392 initial truncated version.
1394 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1396 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1398 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1399 we're in there, support oversigning also; bug 1309.
1401 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1403 HS/04 Add support for keep_environment and add_environment options.
1405 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1406 either intentional arithmetic overflow during PRNG, or testing config-
1409 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1410 delivery resulted in actual delivery. Cancel cutthrough before DATA
1413 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1414 rejects a recipient: pass the reject to the originator.
1416 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1417 Many were false-positives and ignorable, but it's worth fixing the
1420 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1421 for the new environment-manipulation done at startup. Move the routines
1422 from being local to tls.c to being global via the os.c file.
1424 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1425 an extract embedded as result-arg for a map, the first arg for extract
1426 is unavailable so we cannot tell if this is a numbered or keyed
1427 extraction. Accept either.
1433 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1436 JH/02 The smtp transport option "multi_domain" is now expanded.
1438 JH/03 The smtp transport now requests PRDR by default, if the server offers
1441 JH/04 Certificate name checking on server certificates, when exim is a client,
1442 is now done by default. The transport option tls_verify_cert_hostnames
1443 can be used to disable this per-host. The build option
1444 EXPERIMENTAL_CERTNAMES is withdrawn.
1446 JH/05 The value of the tls_verify_certificates smtp transport and main options
1447 default to the word "system" to access the system default CA bundle.
1448 For GnuTLS, only version 3.0.20 or later.
1450 JH/06 Verification of the server certificate for a TLS connection is now tried
1451 (but not required) by default. The verification status is now logged by
1452 default, for both outbound TLS and client-certificate supplying inbound
1455 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1458 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1459 Status Notification (bounce) messages are now MIME format per RFC 3464.
1460 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1461 under the control of the dsn_advertise_hosts option, and routers may
1462 have a dsn_lasthop option.
1464 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1465 default, modifiable by a malware= option. The list separator for
1466 the options can now be changed in the usual way. Bug 68.
1468 JH/10 The smtp_receive_timeout main option is now expanded before use.
1470 JH/11 The incoming_interface log option now also enables logging of the
1471 local interface on delivery outgoing connections.
1473 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1474 if the interface and destination host and port all match.
1476 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1479 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1480 Patch from Andrew Lewis.
1482 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1483 now supports optional time-restrictions, weighting, and priority
1484 modifiers per server. Patch originally by <rommer@active.by>.
1486 JH/16 The spamd_address main option now supports a mixed list of local
1487 and remote servers. Remote servers can be IPv6 addresses, and
1488 specify a port-range.
1490 JH/17 Bug 68: The spamd_address main option now supports an optional
1491 timeout value per server.
1493 JH/18 Bug 1581: Router and transport options headers_add/remove can
1494 now have the list separator specified.
1496 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1499 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1502 JH/21 Support for the A6 type of dns record is withdrawn.
1504 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1505 rather than the verbs used.
1507 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1508 from 255 to 1024 chars.
1510 JH/24 Verification callouts now attempt to use TLS by default.
1512 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1513 are generic router options now. The defaults didn't change.
1515 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1516 Original patch from Alexander Shikoff, worked over by JH.
1518 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1519 files if ZCAT_COMMAND is not executable.
1521 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1523 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1525 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1526 Normally benign, it bites when the pair was led to by a CNAME;
1527 modern usage is to not canonicalize the domain to a CNAME target
1528 (and we were inconsistent anyway for A-only vs AAAA+A).
1530 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1532 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1533 when evaluating $sender_host_dnssec.
1535 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1536 $sender_helo_dnssec variable.
1538 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1540 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1542 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1544 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1545 documented as working, but never had. Support all but $spam_report.
1547 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1548 added for tls authenticator.
1550 HS/03 Add perl_taintmode main config option
1556 TL/01 When running the test suite, the README says that variables such as
1557 no_msglog_check are global and can be placed anywhere in a specific
1558 test's script, however it was observed that placement needed to be near
1559 the beginning for it to behave that way. Changed the runtest perl
1560 script to read through the entire script once to detect and set these
1561 variables, reset to the beginning of the script, and then run through
1562 the script parsing/test process like normal.
1564 TL/02 The BSD's have an arc4random API. One of the functions to induce
1565 adding randomness was arc4random_stir(), but it has been removed in
1566 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1567 function when detected.
1569 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1570 cause callback expansion.
1572 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1573 syntax errors in an expansion can be treated as a string instead of
1574 logging or causing an error, due to the internal use of bool_lax
1575 instead of bool when processing it.
1577 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1578 server certificates when making smtp deliveries.
1580 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1582 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1584 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1586 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1587 Merged patch from Sebastian Wiedenroth.
1589 JH/05 Fix results-pipe from transport process. Several recipients, combined
1590 with certificate use, exposed issues where response data items split
1591 over buffer boundaries were not parsed properly. This eventually
1592 resulted in duplicates being sent. This issue only became common enough
1593 to notice due to the introduction of connection certificate information,
1594 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1596 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1597 size buffer was used, resulting in syntax errors when an expansion
1600 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1601 version 3.3.6 or later.
1603 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1604 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1605 both become "event_action", the variables become $event_name, $event_data
1606 and $event_defer_errno. There is a new variable $verify_mode, usable in
1607 routers, transports and related events. The tls:cert event is now also
1608 raised for inbound connections, if the main configuration event_action
1611 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1612 early OCSP support, but no stapling (appears to be less than 1.0.0).
1614 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1615 server certificate names available under the smtp transport option
1616 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1619 JH/10 Time-related extraction expansions from certificates now use the main
1620 option "timezone" setting for output formatting, and are consistent
1621 between OpenSSL and GnuTLS compilations. Bug 1541.
1623 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1624 encoded parameter in the incoming message. Bug 1558.
1626 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1627 include certificate info, eximon was claiming there were spoolfile
1630 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1632 JH/14 Log delivery-related information more consistently, using the sequence
1633 "H=<name> [<ip>]" wherever possible.
1635 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1636 are problematic for Debian distribution, omit them from the release
1639 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1641 JH/16 Fix string representation of time values on 64bit time_t architectures.
1644 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1645 output list separator was used.
1650 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1651 checkers that were complaining about end of non-void function with no
1654 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1655 This was a regression introduced in 4.83 by another bugfix.
1657 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1659 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1660 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1666 TF/01 Correctly close the server side of TLS when forking for delivery.
1668 When a message was received over SMTP with TLS, Exim failed to clear up
1669 the incoming connection properly after forking off the child process to
1670 deliver the message. In some situations the subsequent outgoing
1671 delivery connection happened to have the same fd number as the incoming
1672 connection previously had. Exim would try to use TLS and fail, logging
1673 a "Bad file descriptor" error.
1675 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1676 utilities have not been installed.
1678 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1679 temporary space as the ACL may create new global variables.
1681 TL/01 LDAP support uses per connection or global context settings, depending
1682 upon the detected version of the libraries at build time.
1684 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1685 to extract and use the src ip:port in logging and expansions as if it
1686 were a direct connection from the outside internet. PPv2 support was
1687 updated based on HAProxy spec change in May 2014.
1689 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1691 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1692 Properly escape header and check for NULL return.
1694 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1697 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1699 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1700 characters in header names, implemented as a verify condition.
1701 Contributed by Michael Fischer v. Mollard.
1703 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1704 results permerror and temperror. Previous values are deprecated but
1705 still accepted. In a future release, err_perm and err_temp will be
1706 completely removed, which will be a backward incompatibility if the
1707 ACL tests for either of these two old results. Patch contributed by
1708 user bes-internal on the mailing list.
1710 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1712 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1713 selectors, in both main and reject logs.
1715 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1718 JH/07 Add malware type "sock" for talking to simple daemon.
1720 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1722 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1723 routers/transports under cutthrough routing.
1725 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1726 numbers. Touch up "bool" conditional to keep the same definition.
1728 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1730 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1732 JH/12 Expand items in router/transport headers_add or headers_remove lists
1733 individually rather than the list as a whole. Bug 1452.
1735 Required for reasonable handling of multiple headers_ options when
1736 they may be empty; requires that headers_remove items with embedded
1737 colons must have them doubled (or the list-separator changed).
1739 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1740 view the policy declared in the DMARC record. Currently, $dmarc_status
1741 is a combined value of both the record presence and the result of the
1744 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1746 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1747 dnslookup router and the smtp transport (applying to the forward
1750 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1751 of ldap servers used for a specific lookup. Patch provided by Heiko
1754 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1755 New variable $lookup_dnssec_authenticated for observability.
1757 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1758 Patch submitted by Lars Timman.
1760 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1762 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1763 Requires trusted mode and valid format message id, aborts otherwise.
1764 Patch contributed by Heiko Schlichting.
1766 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1767 certextract with support for various fields. Bug 1358.
1769 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1770 is requested by default, modifiable by smtp transport option
1773 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1774 operate on certificate variables to give certificate fingerprints
1775 Also new ${sha256:cert_variable}.
1777 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1779 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1780 Christian Aistleitner.
1782 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1784 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1785 file. Patch from Wolfgang Breyha.
1787 JH/25 Expand the coverage of the delivery $host and $host_address to
1788 client authenticators run in verify callout. Bug 1476.
1790 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1791 align with daemon_smtp_ports. Bug 72.
1793 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1794 support and error reporting did not work properly.
1796 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1797 and is readable. Patch from Andrew Colin Kissa.
1799 TL/14 Enhance documentation of ${run expansion and how it parses the
1800 commandline after expansion, particularly in the case when an
1801 unquoted variable expansion results in an empty value.
1803 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1805 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1806 by renaming away from it. Addresses GH issue 15, reported by
1809 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1811 TL/15 SECURITY: prevent double expansion in math comparison functions
1812 (can expand unsanitized data). Not remotely exploitable.
1819 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1821 PP/02 Make -n do something, by making it not do something.
1822 When combined with -bP, the name of an option is not output.
1824 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1827 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1828 $sender_host_name and config options to manage this, and basic check
1831 PP/05 DSCP support for outbound connections and control modifier for inbound.
1833 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1834 (Only plugin which currently uses this is kerberos4, which nobody should
1835 be using, but we should make it available and other future plugins might
1836 conceivably use it, even though it would break NAT; stuff *should* be
1837 using channel bindings instead).
1839 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1840 name; added for Sendmail compatibility; requires admin caller.
1841 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1842 just ignore it); requires trusted caller.
1843 Also parse but ignore: -Ac -Am -X<logfile>
1846 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1848 TL/02 Add +smtp_confirmation as a default logging option.
1850 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1851 Patch by Magnus Holmgren from 2007-02-20.
1853 TL/04 Bugzilla 1281 - Spec typo.
1854 Bugzilla 1283 - Spec typo.
1855 Bugzilla 1290 - Spec grammar fixes.
1857 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1859 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1861 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1862 mailing list by by Dmitry Isaikin.
1864 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1866 JH/02 Support "G" suffix to numbers in ${if comparisons.
1868 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1870 NM/01 Bugzilla 1197 - Spec typo
1871 Bugzilla 1196 - Spec examples corrections
1873 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1875 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1876 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1879 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1880 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1882 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1883 "acl {{name}{arg}...}", and optional args on acl condition
1886 JH/05 Permit multiple router/transport headers_add/remove lines.
1888 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1890 JH/07 Avoid using a waiting database for a single-message-only transport.
1891 Performance patch from Paul Fisher. Bugzilla 1262.
1893 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1896 JH/09 Add $headers_added variable, with content from use of ACL modifier
1897 add_header (but not yet added to the message). Bugzilla 199.
1899 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1900 Pulled from Bugzilla 817 by Wolfgang Breyha.
1902 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1904 (nb: this is the same fix as in Exim 4.80.1)
1906 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1909 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1912 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1914 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1915 advertise SMTP AUTH mechanism to us, instead of a generic
1916 protocol violation error. Also, make Exim more robust to bad
1917 data from the Dovecot auth socket.
1919 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1921 When a queue runner is handling a message, Exim first routes the
1922 recipient addresses, during which it prunes them based on the retry
1923 hints database. After that it attempts to deliver the message to
1924 any remaining recipients. It then updates the hints database using
1927 So if a recipient address works intermittently, it can get repeatedly
1928 deferred at routing time. The retry hints record remains fresh so the
1929 address never reaches the final cutoff time.
1931 This is a fairly common occurrence when a user is bumping up against
1932 their storage quota. Exim had some logic in its local delivery code
1933 to deal with this. However it did not apply to per-recipient defers
1934 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1936 This change adds a proper retry rule check during routing so that the
1937 final cutoff time is checked against the message's age. We only do
1938 this check if there is an address retry record and there is not a
1939 domain retry record; this implies that previous attempts to handle
1940 the address had the retry_use_local_parts option turned on. We use
1941 this as an approximation for the destination being like a local
1942 delivery, as in LMTP.
1944 I suspect this new check makes the old local delivery cutoff check
1945 redundant, but I have not verified this so I left the code in place.
1947 TF/02 Correct gecos expansion when From: is a prefix of the username.
1949 Test 0254 submits a message to Exim with the header
1953 When I ran the test suite under the user fanf2, Exim expanded
1954 the header to contain my full name, whereas it should have added
1955 a Resent-Sender: header. It erroneously treats any prefix of the
1956 username as equal to the username.
1958 This change corrects that bug.
1960 GF/01 DCC debug and logging tidyup
1961 Error conditions log to paniclog rather than rejectlog.
1962 Debug lines prefixed by "DCC: " to remove any ambiguity.
1964 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1966 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1967 Bug spotted by Jeremy Harris; was flawed since initial commit.
1968 Would have resulted in OCSP responses post-SNI triggering an Exim
1969 NULL dereference and crash.
1971 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1973 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1974 Bug detection, analysis and fix by Samuel Thibault.
1975 Bugzilla 1331, Debian bug #698092.
1977 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1979 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1980 Server implementation by Todd Lyons, client by JH.
1981 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1982 config variable "prdr_enable" controls whether the server
1983 advertises the facility. If the client requests PRDR a new
1984 acl_data_smtp_prdr ACL is called once for each recipient, after
1985 the body content is received and before the acl_smtp_data ACL.
1986 The client is controlled by both of: a hosts_try_prdr option
1987 on the smtp transport, and the server advertisement.
1988 Default client logging of deliveries and rejections involving
1989 PRDR are flagged with the string "PRDR".
1991 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1992 fclose(). Diagnosis by Todd Lyons.
1994 PP/17 Update configure.default to handle IPv6 localhost better.
1995 Patch by Alain Williams (plus minor tweaks).
1998 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1999 This is now consistent with GnuTLS, and is now documented: the
2000 previous undocumented portable approach to treating the option as
2001 unset was to force an expansion failure. That still works, and
2002 an empty string is now equivalent.
2004 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2005 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2006 not performing validation itself.
2008 PP/20 Added force_command boolean option to pipe transport.
2009 Patch from Nick Koston, of cPanel Inc.
2011 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2014 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2016 PP/21 Fix eximon continuous updating with timestamped log-files.
2017 Broken in a format-string cleanup in 4.80, missed when I repaired the
2018 other false fix of the same issue.
2019 Report and fix from Heiko Schlichting.
2022 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2023 Report from Prashanth Katuri.
2025 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2026 It's SecureTransport, so affects any MacOS clients which use the
2027 system-integrated TLS libraries, including email clients.
2029 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2030 using a MIME ACL for non-SMTP local injection.
2031 Report and assistance in diagnosis by Warren Baker.
2033 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2035 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2037 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2038 last id that failed so it may be referenced in subsequent ACL's.
2040 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2043 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2044 ldap library initialization, allowing self-signed CA's to be
2045 used. Also properly sets require_cert option later in code by
2046 using NULL (global ldap config) instead of ldap handle (per
2047 session). Bug diagnosis and testing by alxgomz.
2049 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2050 the src/util/ subdirectory.
2052 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2053 renamed to Transport Post Delivery Action by Jeremy Harris, as
2056 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2057 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2058 redis_servers = needs to be configured which will be used by the redis
2059 lookup. Patch from Warren Baker, of The Packet Hub.
2061 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2063 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2064 hostname or reverse DNS when processing a host list. Used suggestions
2065 from multiple comments on this bug.
2067 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2069 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2070 Missed a few lines, added it to make the runtest require no keyboard
2073 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2074 contains upper case chars. Make router use caseful_local_part.
2076 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2077 support when GnuTLS has been built with p11-kit.
2083 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2085 This, or similar/improved, will also be change PP/11 of 4.82.
2091 PP/01 Handle short writes when writing local log-files.
2092 In practice, only affects FreeBSD (8 onwards).
2093 Bugzilla 1053, with thanks to Dmitry Isaikin.
2095 NM/01 Bugzilla 949 - Documentation tweak
2097 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2100 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2102 PP/02 Implemented gsasl authenticator.
2104 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2106 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2107 `pkg-config foo` for cflags/libs.
2109 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2110 with rest of GSASL and with heimdal_gssapi.
2112 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2113 `pkg-config foo` for cflags/libs for the TLS implementation.
2115 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2116 properties get this fed in as external SSF. A number of robustness
2117 and debugging improvements to the cyrus_sasl authenticator.
2119 PP/08 cyrus_sasl server now expands the server_realm option.
2121 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2122 Patch by Jeremy Harris.
2124 PP/10 Added dbmjz lookup type.
2126 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2128 PP/12 MAIL args handles TAB as well as SP, for better interop with
2129 non-compliant senders.
2130 Analysis and variant patch by Todd Lyons.
2132 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2133 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2134 Patch from Dirk Mueller <dmueller@suse.com>
2136 PP/13 tls_peerdn now print-escaped for spool files.
2137 Observed some $tls_peerdn in wild which contained \n, which resulted
2138 in spool file corruption.
2140 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2141 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2142 or write after TLS renegotiation, which otherwise led to messages
2145 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2146 as a tracking header (ie: a signed header comes before the signature).
2147 Patch from Wolfgang Breyha.
2149 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2150 comma-sep list; embedded commas doubled.
2152 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2154 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2156 Report and patch from Dmitry Banschikov.
2158 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2159 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2160 failures. We appear to now support TLS1.1+ with Exim.
2162 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2163 lets Exim select keys and certificates based upon TLS SNI from client.
2164 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2165 before an outbound SMTP session. New log_selector, +tls_sni.
2167 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2168 NULL dereference. Report and patch from Alun Jones.
2170 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2171 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2172 Not seeing resolver debug output on NetBSD, but suspect this is a
2173 resolver implementation change.
2175 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2176 Left warnings. Added "eximon gdb" invocation mode.
2178 PP/21 Defaulting "accept_8bitmime" to true, not false.
2180 PP/22 Added -bw for inetd wait mode support.
2182 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2183 locate the relevant includes and libraries. Made this the default.
2185 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2186 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2188 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2189 This may cause build issues on older platforms.
2191 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2192 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2193 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2194 Added SNI support via GnuTLS too.
2195 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2197 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2199 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2200 Applied second patch from Janne, implementing suggestion to default
2201 multiple-strings-in-record handling to match SPF spec.
2203 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2205 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2206 read-only, out of scope).
2207 Patch from Wolfgang Breyha, report from Stuart Northfield.
2209 PP/29 Fix three issues highlighted by clang analyser static analysis.
2210 Only crash-plausible issue would require the Cambridge-specific
2211 iplookup router and a misconfiguration.
2212 Report from Marcin Mirosław.
2214 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2216 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2217 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2218 As part of this, removing so much warning spew let me fix some minor
2219 real issues in debug logging.
2221 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2222 assignment on my part. Fixed.
2224 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2225 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2226 Janne Snabb (who went above and beyond: thank you).
2228 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2229 string otherwise requires a connection and a bunch more work and it's
2230 relatively easy to get wrong. Should also expose TLS library linkage
2233 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2234 64-bit ${eval} (JH/03).
2236 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2237 GNU libc to support some of the 64-bit stuff, should not lead to
2238 conflicts. Defined before os.h is pulled in, so if a given platform
2239 needs to override this, it can.
2241 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2242 protection layer was required, which is not implemented.
2243 Bugzilla 1254, patch from Wolfgang Breyha.
2245 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2246 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2247 tls_dhparam take prime identifiers. Also unbreak combination of
2248 OpenSSL+DH_params+TLSSNI.
2250 PP/39 Disable SSLv2 by default in OpenSSL support.
2256 PP/01 Solaris build fix for Oracle's LDAP libraries.
2257 Bugzilla 1109, patch from Stephen Usher.
2259 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2261 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2264 TF/02 Fix a couple more cases where we did not log the error message
2265 when unlink() failed. See also change 4.74-TF/03.
2267 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2268 lock up or crash if it happened to be inside a call to libc when it
2269 got a SIGUSR1 from exiwhat.
2271 The SIGUSR1 handler appends the current process status to the process
2272 log which is later printed by exiwhat. It used to use the general
2273 purpose logging code to do this, but several functions it calls are
2274 not safe for signals.
2276 The new output code in the SIGUSR1 handler is specific to the process
2277 log, and simple enough that it's easy to inspect for signal safety.
2278 Removing some special cases also simplifies the general logging code.
2279 Removing the spurious timestamps from the process log simplifies
2282 TF/04 Improved ratelimit ACL condition.
2284 The /noupdate option has been deprecated in favour of /readonly which
2285 has clearer semantics. The /leaky, /strict, and /readonly update modes
2286 are mutually exclusive. The update mode is no longer included in the
2287 database key; it just determines when the database is updated. (This
2288 means that when you upgrade Exim will forget old rate measurements.)
2290 Exim now checks that the per_* options are used with an update mode that
2291 makes sense for the current ACL. For example, when Exim is processing a
2292 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2293 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2294 must specify per_mail/readonly. If you omit the update mode it defaults to
2295 /leaky where that makes sense (as before) or /readonly where required.
2297 The /noupdate option is now undocumented but still supported for
2298 backwards compatibility. It is equivalent to /readonly except that in
2299 ACLs where /readonly is required you may specify /leaky/noupdate or
2300 /strict/noupdate which are treated the same as /readonly.
2302 A useful new feature is the /count= option. This is a generalization
2303 of the per_byte option, so that you can measure the throughput of other
2304 aggregate values. For example, the per_byte option is now equivalent
2305 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2307 The per_rcpt option has been generalized using the /count= mechanism
2308 (though it's more complicated than the per_byte equivalence). When it is
2309 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2310 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2311 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2312 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2313 Note that using per_rcpt with a non-readonly update mode in more than
2314 one ACL will cause the recipients to be double-counted. (The per_mail
2315 and per_byte options don't have this problem.)
2317 The handling of very low rates has changed slightly. If the computed rate
2318 is less than the event's count (usually one) then this event is the first
2319 after a long gap. In this case the rate is set to the same as this event's
2320 count, so that the first message of a spam run is counted properly.
2322 The major new feature is a mechanism for counting the rate of unique
2323 events. The new per_addr option counts the number of different
2324 recipients that someone has sent messages to in the last time period. It
2325 behaves like per_rcpt if all the recipient addresses are different, but
2326 duplicate recipient addresses do not increase the measured rate. Like
2327 the /count= option this is a general mechanism, so the per_addr option
2328 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2329 example, measure the rate that a client uses different sender addresses
2330 with the options per_mail/unique=$sender_address. There are further
2331 details in the main documentation.
2333 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2335 TF/06 Removed a few PCRE remnants.
2337 TF/07 Automatically extract Exim's version number from tags in the git
2338 repository when doing development or release builds.
2340 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2341 Bugzilla 879. Patch from Paul Fisher.
2343 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2344 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2347 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2349 PP/05 Variable $av_failed, true if the AV scanner deferred.
2350 Bugzilla 1078. Patch from John Horne.
2352 PP/06 Stop make process more reliably on build failure.
2353 Bugzilla 1087. Patch from Heiko Schlittermann.
2355 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2356 Bugzilla 1089. Patch from Heiko Schlittermann.
2358 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2359 Bugzilla 1131. Patch from Holger Weiß.
2361 PP/09 Handle IPv6 addresses with SPF.
2362 Bugzilla 860. Patch from Wolfgang Breyha.
2364 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2366 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2369 PP/11 match_* no longer expand right-hand-side by default.
2370 New compile-time build option, EXPAND_LISTMATCH_RHS.
2371 New expansion conditions, "inlist", "inlisti".
2373 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2375 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2377 PP/14 fix log_write() format string regression from TF/03.
2378 Bugzilla 1152. Patch from Dmitry Isaikin.
2384 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2386 PP/02 Harmonised TLS library version reporting; only show if debugging.
2387 Layout now matches that introduced for other libraries in 4.74 PP/03.
2389 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2391 PP/04 New "dns_use_edns0" global option.
2393 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2396 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2397 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2399 TK/01 Updated PolarSSL code to 0.14.2.
2400 Bugzilla 1097. Patch from Andreas Metzler.
2402 PP/07 Catch divide-by-zero in ${eval:...}.
2403 Fixes bugzilla 1102.
2405 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2408 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2409 format-string attack -- SECURITY: remote arbitrary code execution.
2411 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2412 time unintentionally subject to list matching rules, letting the header
2413 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2414 arbitrary string expansion). This allowed for information disclosure.
2416 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2417 INT_MIN/-1 -- value coerced to INT_MAX.
2423 NM/01 Workaround for PCRE version dependency in version reporting
2426 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2427 This fixes portability to compilers other than gcc, notably
2428 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2430 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2431 makefiles for portability to HP-UX and POSIX correctness.
2433 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2434 Also via indented variable definition in the Makefile.
2435 (Debugging by Oliver Heesakkers).
2437 PP/02 Restore caching of spamd results with expanded spamd_address.
2438 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2440 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2441 Improves build reliability. Fix from: Frank Elsner
2443 NM/02 Fix wide character breakage in the rfc2047 coding
2444 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2446 NM/03 Allow underscore in dnslist lookups
2447 Fixes bug 1026. Patch from Graeme Fowler
2449 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2450 Code patches from Adam Ciarcinski of NetBSD.
2452 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2455 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2456 is logged, to avoid truncation. Patch from John Horne.
2458 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2459 Patch from Jakob Hirsch.
2461 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2462 SQL string expansion failure details.
2463 Patch from Andrey Oktyabrski.
2465 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2466 Patch from Simon Arlott.
2468 PP/09 New lookups functionality failed to compile on old gcc which rejects
2469 extern declarations in function scope.
2470 Patch from Oliver Fleischmann
2472 PP/10 Use sig_atomic_t for flags set from signal handlers.
2473 Check getgroups() return and improve debugging.
2474 Fixed developed for diagnosis in bug 927 (which turned out to be
2477 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2478 Patch from Mark Zealey.
2480 PP/12 Bugzilla 1056: Improved spamd server selection.
2481 Patch from Mark Zealey.
2483 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2484 Based on patch from Heiko Schlittermann.
2486 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2487 Patch from Uwe Doering, sign-off by Michael Haardt.
2489 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2490 variable declaration deep within a block. Bug and patch from
2493 PP/15 lookups-Makefile IRIX compatibility coercion.
2495 PP/16 Make DISABLE_DKIM build knob functional.
2497 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2498 Patch by Simon Arlott
2500 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2501 variable argument macros. Our copy now differs from upstream.
2507 TF/01 Failure to get a lock on a hints database can have serious
2508 consequences so log it to the panic log.
2510 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2511 controlled using the smtp_confirmation log selector.
2513 TF/03 Include the error message when we fail to unlink a spool file.
2515 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2516 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2517 for maintaining out-of-tree patches for some time.
2519 PP/01 Bugzilla 139: Documentation and portability issues.
2520 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2521 Handle per-OS dynamic-module compilation flags.
2523 PP/02 Let /dev/null have normal permissions.
2524 The 4.73 fixes were a little too stringent and complained about the
2525 permissions on /dev/null. Exempt it from some checks.
2526 Reported by Andreas M. Kirchwitz.
2528 PP/03 Report version information for many libraries, including
2529 Exim version information for dynamically loaded libraries. Created
2530 version.h, now support a version extension string for distributors
2531 who patch heavily. Dynamic module ABI change.
2533 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2534 privilege escalation vulnerability whereby the Exim run-time user
2535 can cause root to append content of the attacker's choosing to
2538 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2541 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2542 If dropping privileges for untrusted macros, we disabled normal logging
2543 on the basis that it would fail; for the Exim run-time user, this is not
2544 the case, and it resulted in successful deliveries going unlogged.
2545 Fixed. Reported by Andreas Metzler.
2551 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2552 only prepend for the Resent-* case. Fixes regression introduced in
2553 Exim 4.70 by NM/22 for Bugzilla 607.
2555 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2556 increasing numbers of administrators be bitten by this.
2558 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2560 PP/03 Bugzilla 994: added openssl_options main configuration option.
2562 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2564 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2566 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2568 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2569 without a peer certificate, leading to a segfault because of an
2570 assumption that peers always have certificates. Be a little more
2571 paranoid. Problem reported by Martin Tscholak.
2573 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2574 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2575 NB: ClamAV planning to remove STREAM in "middle of 2010".
2576 CL also introduces -bmalware, various -d+acl logging additions and
2577 more caution in buffer sizes.
2579 PP/09 Implemented reverse_ip expansion operator.
2581 PP/10 Bugzilla 937: provide a "debug" ACL control.
2583 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2585 PP/12 Bugzilla 973: Implement --version.
2587 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2589 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2591 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2593 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2594 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2595 ignore trailing whitespace.
2597 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2599 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2602 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2603 Notification from Dr Andrew Aitchison.
2605 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2606 ExtendedDetectionInfo response format.
2607 Notification from John Horne.
2609 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2612 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2613 XSL and documented dependency on system catalogs, with examples of how
2616 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2619 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2620 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2621 configuration file which is writeable by the Exim user or group.
2623 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2624 of configuration files to cover files specified with the -C option if
2625 they are going to be used with root privileges, not just the default
2628 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2629 option (effectively making it always true).
2631 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2632 files to be used while preserving root privileges.
2634 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2635 that rogue child processes cannot use them.
2637 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2638 run-time user, instead of root.
2640 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2641 Exim run-time user without dropping privileges.
2643 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2644 result string, instead of calling string_vformat() twice with the same
2647 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2648 for other users. Others should always drop root privileges if they use
2649 -C on the command line, even for a whitelisted configure file.
2651 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2653 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2659 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2660 $data_path, and $header_path variables; fixed documentation bugs and
2663 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2664 exipick to access non-standard spools, including the "frozen" queue
2667 NM/01 Bugzilla 965: Support mysql stored procedures.
2668 Patch from Alain Williams
2670 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2672 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2673 Patch from Andreas Metzler
2675 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2676 Patch from Kirill Miazine
2678 NM/05 Bugzilla 671: Added umask to procmail example.
2680 JJ/03 installed exipick 20100323.0, fixing doc bug
2682 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2683 directory. Notification and patch from Dan Rosenberg.
2685 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2687 TK/02 Improve log output when DKIM signing operation fails.
2689 MH/01 Treat the transport option dkim_domain as a colon separated
2690 list, not as a single string, and sign the message with each element,
2691 omitting multiple occurences of the same signer.
2693 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2694 Bugzilla 985, 986. Patch by Simon Arlott
2696 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2697 Patch by Simon Arlott
2699 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2700 MBX locking. Notification from Dan Rosenberg.
2706 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2708 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2710 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2712 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2714 NM/04 Bugzilla 915: Flag broken perl installation during build.
2720 TK/01 Added patch by Johannes Berg that expands the main option
2721 "spamd_address" if it starts with a dollar sign.
2723 TK/02 Write list of recipients to X-Envelope-Sender header when building
2724 the mbox-format spool file for content scanning (suggested by Jakob
2727 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2728 (http://www.dcc-servers.net/) support via dccifd. Activated by
2729 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2731 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2732 by Mark Daniel Reidel <mr@df.eu>.
2734 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2735 When building exim an external PCRE library is now needed -
2736 PCRE is a system library on the majority of modern systems.
2737 See entry on PCRE_LIBS in EDITME file.
2739 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2740 conversation. Added nologin parameter to request.
2741 Patch contributed by Kirill Miazine.
2743 TF/01 Do not log submission mode rewrites if they do not change the address.
2745 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2747 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2748 log files in place. Contributed by Roberto Lima.
2750 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2752 TF/03 Bugzilla 615: When checking the local_parts router precondition
2753 after a local_part_suffix or local_part_prefix option, Exim now
2754 does not use the address's named list lookup cache, since this
2755 contains cached lookups for the whole local part.
2757 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2758 Robert Millan. Documentation is in experimental-spec.txt.
2760 TF/04 Bugzilla 668: Fix parallel build (make -j).
2762 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2764 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2765 Patch provided by Jan Srzednicki.
2767 TF/05 Leading white space used to be stripped from $spam_report which
2768 wrecked the formatting. Now it is preserved.
2770 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2771 that they are available at delivery time.
2773 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2775 TF/08 TLS error reporting now respects the incoming_interface and
2776 incoming_port log selectors.
2778 TF/09 Produce a more useful error message if an SMTP transport's hosts
2779 setting expands to an empty string.
2781 NM/06 Bugzilla 744: EXPN did not work under TLS.
2782 Patch provided by Phil Pennock.
2784 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2785 Patch provided by Richard Godbee.
2787 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2788 acl_smtp_notquit, added index entry.
2790 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2791 Patch provided by Eugene Bujak.
2793 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2794 accept(). Patch provided by Maxim Dounin.
2796 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2797 Patch provided by Phil Pennock.
2799 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2801 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2802 Patch provided by Brad "anomie" Jorsch.
2804 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2805 Patch provided by Dean Brooks.
2807 TK/05 Add native DKIM support (does not depend on external libraries).
2809 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2810 Patch provided by Graeme Fowler.
2812 NM/16 Bugzilla 851: Documentation example syntax fix.
2814 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2816 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2819 NM/19 Bugzilla 745: TLS version reporting.
2820 Patch provided by Phil Pennock.
2822 NM/20 Bugzilla 167: bool: condition support.
2823 Patch provided by Phil Pennock.
2825 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2826 clients. Patch provided by Phil Pennock.
2828 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2829 Patch provided by Brad "anomie" Jorsch.
2831 NM/23 Bugzilla 687: Fix misparses in eximstats.
2832 Patch provided by Heiko Schlittermann.
2834 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2835 Patch provided by Heiko Schlittermann.
2837 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2838 plus update to original patch.
2840 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2842 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2843 Patch provided by David Brownlee.
2845 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2847 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2849 NM/30 Bugzilla 888: TLS documentation bugfixes.
2851 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2853 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2854 Unlike the original bugzilla I have changed all shell scripts in src tree.
2856 NM/33 Bugzilla 898: Transport filter timeout fix.
2857 Patch by Todd Rinaldo.
2859 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2860 Patch by Serge Demonchaux.
2862 NM/35 Bugzilla 39: Base64 decode bug fixes.
2863 Patch by Jakob Hirsch.
2865 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2867 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2869 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2871 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2877 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2878 ALT-N's libdkim that I have put here:
2879 http://duncanthrax.net/exim-experimental/
2881 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2882 were called 'true' and it seems that C99 defines that as a reserved
2883 keyword to be used with 'bool' variable types. That means you could
2884 not include C99-style headers which use bools without triggering
2885 build errors in sieve.c.
2887 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2888 as mailq or other aliases. Changed the --help handling significantly
2889 to do whats expected. exim_usage() emits usage/help information.
2891 SC/01 Added the -bylocaldomain option to eximstats.
2893 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2895 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2897 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2903 PH/01 Another patch from the Sieve maintainer.
2905 PH/02 When an IPv6 address is converted to a string for single-key lookup
2906 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2907 dots are used instead of colons so that keys in lsearch files need not
2908 contain colons. This was done some time before quoting was made available
2909 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2910 (notated using the quote facility) so as to distinguish them from IPv4
2911 keys. This meant that lookups for IP addresses in host lists did not work
2912 for iplsearch lookups.
2914 This has been fixed by arranging for IPv6 addresses to be expressed with
2915 colons if the lookup type is iplsearch. This is not incompatible, because
2916 previously such lookups could never work.
2918 The situation is now rather anomalous, since one *can* have colons in
2919 ordinary lsearch keys. However, making the change in all cases is
2920 incompatible and would probably break a number of configurations.
2922 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2925 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2926 conversion specification without a maximum field width, thereby enabling
2927 a rogue spamd server to cause a buffer overflow. While nobody in their
2928 right mind would setup Exim to query an untrusted spamd server, an
2929 attacker that gains access to a server running spamd could potentially
2930 exploit this vulnerability to run arbitrary code as the Exim user.
2932 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2933 $primary_hostname instead of what libspf2 thinks the hosts name is.
2935 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2936 a directory entry by the name of the lookup key. Previously, if a
2937 symlink pointed to a non-existing file or a file in a directory that
2938 Exim lacked permissions to read, a lookup for a key matching that
2939 symlink would fail. Now it is enough that a matching directory entry
2940 exists, symlink or not. (Bugzilla 503.)
2942 PH/03 The body_linecount and body_zerocount variables are now exported in the
2945 PH/04 Added the $dnslist_matched variable.
2947 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2948 This means they are set thereafter only if the connection becomes
2951 PH/06 Added the client_condition to authenticators so that some can be skipped
2952 by clients under certain conditions.
2954 PH/07 The error message for a badly-placed control=no_multiline_responses left
2955 "_responses" off the end of the name.
2957 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2959 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2960 (without spaces) instead of just copying the configuration text.
2962 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2964 PH/11 Added $max_received_linelength.
2966 PH/12 Added +ignore_defer and +include_defer to host lists.
2968 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2969 way in which PCRE > 7.0 is built.
2971 PH/14 Implemented queue_only_load_latch.
2973 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2974 MAIL command. The effect was to mangle the value on 64-bit systems.
2976 PH/16 Another patch from the Sieve maintainer.
2978 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2980 PH/18 If a system quota error occurred while trying to create the file for
2981 a maildir delivery, the message "Mailbox is full" was not appended to the
2982 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2983 only to a quota excession during the actual writing of the file.
2985 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2986 characters?) which causes problems in log lines. The DN values are now
2987 passed through string_printing() before being added to log lines.
2989 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2990 and InterBase are left for another time.)
2992 PH/21 Added message_body_newlines option.
2994 PH/22 Guard against possible overflow in moan_check_errorcopy().
2996 PH/23 POSIX allows open() to be a macro; guard against that.
2998 PH/24 If the recipient of an error message contained an @ in the local part
2999 (suitably quoted, of course), incorrect values were put in $domain and
3000 $local_part during the evaluation of errors_copy.
3006 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3007 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3010 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3011 issue a MAIL command.
3013 PH/02 In an ACL statement such as
3015 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3017 if a client was not listed at all, or was listed with a value other than
3018 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3019 the condition was not true (as it should be), so access was not denied.
3020 The bug was that the ! inversion was incorrectly passed on to the second
3021 item. This has been fixed.
3023 PH/03 Added additional dnslists conditions == and =& which are different from
3024 = and & when the dns lookup returns more than one IP address.
3026 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3027 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3029 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3030 FSYNC, which compiles an option called disable_fsync that allows for
3031 bypassing fsync(). The documentation is heavily laced with warnings.
3033 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3035 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3036 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3037 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3038 including adding "make clean"; (3) Added -fPIC when compiling the test
3039 dynamically loaded module, to get rid of a warning.
3041 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3042 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3043 = 0s. The bug is otherwise harmless.
3045 PH/07 There was a bug in the dovecot authenticator such that the value of
3046 $auth1 could be overwritten, and so not correctly preserved, after a
3047 successful authentication. This usually meant that the value preserved by
3048 the server_setid option was incorrect.
3050 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3052 PH/09 Installed PCRE release 7.0.
3054 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3055 run for batched SMTP input. It is now run at the start of every message
3056 in the batch. While fixing this I discovered that the process information
3057 (output by running exiwhat) was not always getting set for -bs and -bS
3058 input. This is fixed, and it now also says "batched" for BSMTP.
3060 PH/11 Added control=no_pipelining.
3062 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3063 patch, slightly modified), and move the expansion of helo_data till after
3064 the connection is made in the smtp transport (so it can use these
3067 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3069 PH/14 Added log_selector = +pid.
3071 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3073 PH/16 Add ${if forany and ${if forall.
3075 PH/17 Added dsn_from option to vary the From: line in DSNs.
3077 PH/18 Flush SMTP output before performing a callout, unless control =
3078 no_callout_flush is set.
3080 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3081 was true (the default) a successful delivery failed to delete the retry
3082 item, thus causing premature timeout of the address. The bug is now
3085 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3087 PH/21 Long custom messages for fakedefer and fakereject are now split up
3088 into multiline responses in the same way that messages for "deny" and
3089 other ACL rejections are.
3091 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3092 with slight modification.
3094 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3095 draft, changing the syntax and factoring some duplicate code".
3097 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3098 for deliveries of the second and subsequent messages over the same SMTP
3101 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3102 ${reduce, with only minor "tidies".
3104 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3106 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3107 expansion side effects.
3109 PH/27 When a message times out after an over-quota error from an Exim-imposed
3110 quota, the bounce message says "mailbox is full". This message was not
3111 being given when it was a system quota that was exceeded. It now should
3114 MH/03 Made $recipients available in local_scan(). local_scan() already has
3115 better access to the recipient list through recipients_list[], but
3116 $recipients can be useful in postmaster-provided expansion strings.
3118 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3119 in the case of a MAIL command with additional options following the
3120 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3121 were accidentally chopped off.
3123 PH/29 SMTP synchronization checks are implemented when a command is read -
3124 there is a check that no more input is waiting when there shouldn't be
3125 any. However, for some commands, a delay in an ACL can mean that it is
3126 some time before the response is written. In this time, more input might
3127 arrive, invalidly. So now there are extra checks after an ACL has run for
3128 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3129 pipelining has not been advertised.
3131 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3133 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3134 returned in the SMTP response when smtp_return_error_details was false.
3135 This has been fixed.
3137 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3138 instead of the C I/O that was originally supplied, because problems were
3139 reported on Solaris.
3141 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3142 Exim which did not show up earlier: it was assuming that a call to
3143 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3144 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3145 was a macro that became an assignment, so it seemed to work. This has
3146 changed to a proper function call with a void return, hence the compile
3147 error. Exim's code has been fixed.
3149 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3152 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3154 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3156 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3157 because it turns out to be too big or there is a timeout) there is no
3158 "Completed" line in the log. When some messages of this type were
3159 selected by exigrep, they were listed as "not completed". Others were
3160 picked up by some special patterns. I have improved the selection
3161 criteria to be more general.
3163 PH/38 The host_find_failed option in the manualroute router can now be set
3164 to "ignore", to completely ignore a host whose IP address cannot be
3165 found. If all hosts are ignored, the behaviour is controlled by the new
3166 host_all_ignored option.
3168 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3169 homing or because of multiple MX records with /mx) generated more than
3170 one IP address, and the following item turned out to be the local host,
3171 all the secondary addresses of the first item were incorrectly removed
3172 from the list, along with the local host and any following hosts (which
3173 is what is supposed to happen).
3175 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3176 whoever called Exim into the -H file. In the case of the daemon it was
3177 behaving confusingly. When first started, it used values for whoever
3178 started the daemon, but after a SIGHUP it used the Exim user (because it
3179 calls itself on a restart). I have changed the code so that it now always
3182 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3183 message are rejected with the same error (e.g. no authentication or bad
3184 sender address), and a DATA command is nevertheless sent (as can happen
3185 with PIPELINING or a stupid MUA), the error message that was given to the
3186 RCPT commands is included in the rejection of the DATA command. This is
3187 intended to be helpful for MUAs that show only the final error to their
3190 PH/42 Another patch from the Sieve maintainer.
3192 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3193 Eximstats - Fixed some broken HTML links and added missing column headers
3195 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3196 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3198 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3200 PH/43 Yet another patch from the Sieve maintainer.
3202 PH/44 I found a way to check for a TCP/IP connection going away before sending
3203 the response to the final '.' that terminates a message, but only in the
3204 case where the client has not sent further data following the '.'
3205 (unfortunately, this is allowed). However, in many cases there won't be
3206 any further data because there won't be any more messages to send. A call
3207 to select() can be used: if it shows that the input is "ready", there is
3208 either input waiting, or the socket has been closed. An attempt to read
3209 the next input character can distinguish the two cases. Previously, Exim
3210 would have sent an OK response which the client would never have see.
3211 This could lead to message repetition. This fix should cure that, at
3212 least in a lot of common cases.
3214 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3215 advertised in response to EHLO.
3221 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3222 fixed by 4.65/MH/01 (is this a record?) are fixed:
3224 (i) An empty string was always treated as zero by the numeric comparison
3225 operators. This behaviour has been restored.
3227 (ii) It is documented that the numeric comparison operators always treat
3228 their arguments as decimal numbers. This was broken in that numbers
3229 starting with 0 were being interpreted as octal.
3231 While fixing these problems I realized that there was another issue that
3232 hadn't been noticed. Values of message_size_limit (both the global option
3233 and the transport option) were treated as octal if they started with 0.
3234 The documentation was vague. These values are now always treated as
3235 decimal, and I will make that clear in the documentation.
3241 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3242 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3245 MH/01 Don't check that the operands of numeric comparison operators are
3246 integers when their expansion is in "skipping" mode (fixes bug
3247 introduced by 4.64-PH/07).
3249 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3250 child addresses, Exim now panics and dies. Previously, because the count
3251 is held in a short int, deliveries were likely to be lost. As such a
3252 large number of recipients for a single message is ridiculous
3253 (performance will be very, very poor), I have chosen to impose a limit
3254 rather than extend the field.
3260 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3261 leftover -K file (the existence of which was triggered by #402).
3262 While we were at it, introduced process PID as part of the -K
3263 filename. This should rule out race conditions when creating
3266 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3267 processing considerably. Previous code took too long for large mails,
3268 triggering a timeout which in turn triggers #401.
3270 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3271 in the DK code in transports.c. sendfile() is not really portable,
3272 hence the _LINUX specificness.
3274 TF/01 In the add_headers option to the mail command in an Exim filter,
3275 there was a bug that Exim would claim a syntax error in any
3276 header after the first one which had an odd number of characters
3279 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3280 callout verification, Exim cached a "reject" for the entire domain. This
3281 is correct for most verifications, but it is not correct for a recipient
3282 verification with use_sender or use_postmaster set, because in that case
3283 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3284 case of MAIL FROM:<> rejection from other early rejections (e.g.
3285 rejection of HELO). When verifying a recipient using a non-null MAIL
3286 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3287 Whatever the result of the callout, the value of the domain cache is
3288 left unchanged (for any other kind of callout, getting as far as trying
3289 RCPT means that the domain itself is ok).
3291 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3294 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3295 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3296 session, the daemon ignores ECONNECT errors and logs others; it now
3297 ignores EPIPE as well.
3299 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3300 (quoted-printable decoding).
3302 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3303 later the small subsequent patch to fix an introduced bug.
3305 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3307 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3309 PH/08 An error is now given if message_size_limit is specified negative.
3311 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3312 to be given (somewhat) arbitrary names.
3314 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3317 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3318 miscellaneous code fixes
3320 PH/10 Added the log_reject_target ACL modifier to specify where to log
3323 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3324 hostname. This is wrong, because it relates to the incoming message (and
3325 probably the interface on which it is arriving) and not to the outgoing
3326 callout (which could be using a different interface). This has been
3327 changed to use the value of the helo_data option from the smtp transport
3328 instead - this is what is used when a message is actually being sent. If
3329 there is no remote transport (possible with a router that sets up host
3330 addresses), $smtp_active_hostname is used.
3332 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3333 tweaks were necessary in order to get it to work (see also 21 below):
3334 (a) The code assumed that strncpy() returns a negative number on buffer
3335 overflow, which isn't the case. Replaced with Exim's string_format()
3337 (b) There were several signed/unsigned issues. I just did the minimum
3338 hacking in of casts. There is scope for a larger refactoring.
3339 (c) The code used strcasecmp() which is not a standard C function.
3340 Replaced with Exim's strcmpic() function.
3341 (d) The code set only $1; it now sets $auth1 as well.
3342 (e) A simple test gave the error "authentication client didn't specify
3343 service in request". It would seem that Dovecot has changed its
3344 interface. Fortunately there's a specification; I followed it and
3345 changed what the client sends and it appears to be working now.
3347 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3350 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3351 address A is aliased to B and C, where B exists and C does not. Without
3352 -v the output is "A verified" because verification stops after a
3353 successful redirection if more than one address is generated. However,
3354 with -v the child addresses are also verified. Exim was outputting "A
3355 failed to verify" and then showing the successful verification for C,
3356 with its parentage. It now outputs "B failed to verify", showing B's
3357 parentage before showing the successful verification of C.
3359 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3360 look up a TXT record in a specific list after matching in a combined
3363 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3364 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3365 they consult the DNS. I had assumed they would set it the way they
3366 wanted; and indeed my experiments on Linux seem to show that in some
3367 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3368 To be on the safe side, however, I have now made the interface to
3369 host_find_byname() similar to host_find_bydns(), with an argument
3370 containing the DNS resolver options. The host_find_byname() function now
3371 sets these options at its start, just as host_find_bydns() does. The smtp
3372 transport options dns_qualify_single and dns_search_parents are passed to
3373 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3374 of host_find_byname() use the default settings of RES_DEFNAMES
3375 (qualify_single) but not RES_DNSRCH (search_parents).
3377 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3378 spool_read_header() do less string testing, by means of a preliminary
3379 switch on the second character of optional "-foo" lines. (This is
3380 overdue, caused by the large number of possibilities that now exist.
3381 Originally there were few.) While I was there, I also converted the
3382 str(n)cmp tests so they don't re-test the leading "-" and the first
3383 character, in the hope this might squeeze out yet more improvement.
3385 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3386 flag allowing group syntax was set by the header_syntax check but not
3387 turned off, possible causing trouble later; (2) The flag was not being
3388 set at all for the header_verify test, causing "group"-style headers to
3389 be rejected. I have now set it in this case, and also caused header_
3390 verify to ignore an empty address taken from a group. While doing this, I
3391 came across some other cases where the code for allowing group syntax
3392 while scanning a header line wasn't quite right (mostly, not resetting
3393 the flag correctly in the right place). These bugs could have caused
3394 trouble for malformed header lines. I hope it is now all correct.
3396 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3397 with the "reply" argument non-NULL. The code, however (which originally
3398 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3399 but it didn't always do it. This confused somebody who was copying the
3400 code for some other use. I have removed all the tests.
3402 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3403 feature that was used to support insecure browsers during the U.S. crypto
3404 embargo. It requires special client support, and Exim is probably the
3405 only MTA that supported it -- and would never use it because real RSA is
3406 always available. This code has been removed, because it had the bad
3407 effect of slowing Exim down by computing (never used) parameters for the
3408 RSA_EXPORT functionality.
3410 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3411 authenticator to fail if there's a tab character in the incoming data
3412 (there should never be unless someone is messing about, as it's supposed
3413 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3414 if the connection is using TLS or if the remote IP is the same as the
3415 local IP, and the "valid-client-cert option" if a client certificate has
3418 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3419 authenticators. This can be used for authorization after authentication
3420 succeeds. (In the case of plaintext, it servers for both authentication
3423 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3424 if any retry times were supplied.
3426 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3427 connection, where there is no client IP address to check. In this
3428 situation, the verify now always succeeds.
3430 PH/25 Applied John Jetmore's -Mset patch.
3432 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3434 PH/27 In a string expansion for a processed (not raw) header when multiple
3435 headers of the same name were present, leading whitespace was being
3436 removed from all of them, but trailing whitespace was being removed only
3437 from the last one. Now trailing whitespace is removed from each header
3438 before concatenation. Completely empty headers in a concatenation (as
3439 before) are ignored.
3441 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3442 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3444 PH/29 [Removed. This was a change that I later backed out, and forgot to
3445 correct the ChangeLog entry (that I had efficiently created) before
3446 committing the later change.]
3448 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3449 address errors (4xx response to RCPT) over the same connection as other
3450 messages routed to the same hosts. Such deliveries are always "forced",
3451 so retry times are not inspected. This resulted in far too many retries
3452 for the affected addresses. The effect occurred only when there were more
3453 hosts than the hosts_max_try setting in the smtp transport when it had
3454 the 4xx errors. Those hosts that it had tried were not added to the list
3455 of hosts for which the message was waiting, so if all were tried, there
3456 was no problem. Two fixes have been applied:
3458 (i) If there are any address or message errors in an SMTP delivery, none
3459 of the hosts (tried or untried) are now added to the list of hosts
3460 for which the message is waiting, so the message should not be a
3461 candidate for sending over the same connection that was used for a
3462 successful delivery of some other message. This seems entirely
3463 reasonable: after all the message is NOT "waiting for some host".
3464 This is so "obvious" that I'm not sure why it wasn't done
3465 previously. Hope I haven't missed anything, but it can't do any
3466 harm, as the worst effect is to miss an optimization.
3468 (ii) If, despite (i), such a delivery is accidentally attempted, the
3469 routing retry time is respected, so at least it doesn't keep
3470 hammering the server.
3472 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3473 in ${readsocket because some servers need this prod.
3475 PH/32 Added some extra debug output when updating a wait-xxx database.
3477 PH/33 The hint "could be header name not terminated by colon", which has been
3478 given for certain expansion errors for a long time, was not being given
3479 for the ${if def:h_colon_omitted{... case.
3481 PH/34 The spec says: "With one important exception, whenever a domain list is
3482 being scanned, $domain contains the subject domain." There was at least
3483 one case where this was not true.
3485 PH/35 The error "getsockname() failed: connection reset by peer" was being
3486 written to the panic log as well as the main log, but it isn't really
3487 panic-worthy as it just means the connection died rather early on. I have
3488 removed the panic log writing for the ECONNRESET error when getsockname()
3491 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3492 runs only) independently of the message's sender address. This meant
3493 that, if the 4xx error was in fact related to the sender, a different
3494 message to the same recipient with a different sender could confuse
3495 things. In particular, this can happen when sending to a greylisting
3496 server, but other circumstances could also provoke similar problems.
3497 I have changed the default so that the retry time for these errors is now
3498 based a combination of the sender and recipient addresses. This change
3499 can be overridden by setting address_retry_include_sender=false in the
3502 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3503 remote server are returned as part of bounce messages. This was not
3504 happening for LMTP over a pipe (the lmtp transport), but now it is the
3505 same for both kinds of LMTP.
3507 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3508 in header lines that were in fact CNAMEs. This is no longer the case.
3510 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3511 and queue runs started by the daemon processed all messages. This has
3512 been fixed so that -R and -S can now usefully be given with -q<time>.
3514 PH/40 Import PCRE release 6.7 (fixes some bugs).
3516 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3518 PH/42 Give an error if -q is specified more than once.
3520 PH/43 Renamed the variables $interface_address and $interface_port as
3521 $received_ip_address and $received_port, to make it clear that these
3522 values apply to message reception, and not to the outgoing interface when
3523 a message is delivered. (The old names remain recognized, of course.)
3525 PH/44 There was no timeout on the connect() call when using a Unix domain
3526 socket in the ${readsocket expansion. There now is.
3528 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3529 be meaningful with "accept".
3531 SC/01 Eximstats V1.43
3532 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3534 SC/02 Eximstats V1.44
3535 Use a glob alias rather than an array ref in the generated
3536 parser. This improves both readability and performance.
3538 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3539 Collect SpamAssassin and rejection statistics.
3540 Don't display local sender or destination tables unless
3541 there is data to show.
3542 Added average volumes into the top table text output.
3544 SC/04 Eximstats V1.46
3545 Collect data on the number of addresses (recipients)
3546 as well as the number of messages.
3548 SC/05 Eximstats V1.47
3549 Added 'Message too big' to the list of mail rejection
3550 reasons (thanks to Marco Gaiarin).
3552 SC/06 Eximstats V1.48
3553 Mainlog lines which have GMT offsets and are too short to
3554 have a flag are now skipped.
3556 SC/07 Eximstats V1.49 (Alain Williams)
3557 Added the -emptyok flag.
3559 SC/08 Eximstats V1.50
3560 Fixes for obtaining the IP address from reject messages.
3562 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3563 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3564 whitespace changes from 4.64-PH/27
3566 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3569 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3570 are found, allow negative numbers in numeric criteria)
3572 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3574 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3575 to match changes made in 4.64-PH/43
3577 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3579 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3580 "domain" is actually an IP address, and force a failure. This locks out
3581 those revolvers/nameservers that support "A-for-A" lookups, in
3582 contravention of the specifications.
3584 PH/48 When a host name was looked up from an IP address, and the subsequent
3585 forward lookup of the name timed out, the host name was left in
3586 $sender_host_name, contrary to the specification.
3588 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3589 restricted to single-key lookups, Exim was not diagnosing an error if
3590 * or *@ was used with a query-style lookup.
3592 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3594 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3595 long ago, but noone interested enough thought of it. Let's just say that
3596 the "1.1" means that there are some new functions that weren't there at
3597 some point in the past.
3599 PH/51 Error processing for expansion failure of helo_data from an smtp
3600 transport during callout processing was broken.
3602 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3603 tested/used via the -bh/-bhc/-bs options.
3605 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3606 bug, fixed in subsequent PCRE releases).
3608 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3609 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3611 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3617 SC/01 Use a glob alias rather than an array ref in eximstats generated
3618 parser. This improves both readability and performance.
3620 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3621 Don't display local sender or destination tables in eximstats unless
3622 there is data to show.
3623 Added average volumes into the eximstats top table text output.
3625 SC/03 Collect data on the number of addresses (recipients) as well
3626 as the number of messages in eximstats.
3628 TF/01 Correct an error in the documentation for the redirect router. Exim
3629 does (usually) call initgroups() when daemonizing.
3631 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3632 with consistent privilege compared to when running as a daemon.
3634 TF/03 Note in the spec that $authenticated_id is not set for local
3635 submissions from trusted users.
3637 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3638 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3640 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3641 by adding some example configuration directives to the default
3642 configuration file. A little bit of work is required to uncomment the
3643 directives and define how usernames and passwords are checked, but
3644 there is now a framework to start from.
3646 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3647 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3648 without this. I don't know how relevant this is to other LDAP libraries.
3650 PH/02 Add the verb name to the "unknown ACL verb" error.
3652 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3654 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3656 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3657 directory not expanded when it should be if an expanded home directory
3658 was set for the address (which is overridden by the transport).
3660 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3663 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3664 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3665 because it is too late at that time, and has no effect.
3667 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3668 security issue with \' (bugzilla #107). I could not use the
3669 PQescapeStringConn() function, because it needs a PGconn value as one of
3672 PH/08 When testing addresses using -bt, indicate those final addresses that
3673 are duplicates that would not cause an additional delivery. At least one
3674 person was confused, thinking that -bt output corresponded to deliveries.
3675 (Suppressing duplicates isn't a good idea as you lose the information
3676 about possibly different redirections that led to the duplicates.)
3678 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3679 systems where poll() doesn't work, in particular OS X.
3681 PH/10 Added more information to debugging output for retry time not reached.
3683 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3684 operations in malware.c.
3686 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3689 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3690 syslog_duplication set false, log lines that would normally be written
3691 both the the main log and to the reject log were not written to syslog at
3694 PH/14 In the default configuration, change the use of "message" in ACL warn
3695 statements to "add_header".
3697 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3698 not followed by a command (e.g. "seen endif").
3700 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3701 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3704 PH/17 Added extra conditions to the default value of delay_warning_condition
3708 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3709 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3710 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3713 The Auto-Submitted: and various List- headers are standardised, whereas I
3714 don't think Precedence: ever was.
3716 PH/18 Refactored debugging code in route_finduser() to show more information,
3717 in particular, the error code if getpwnam() issues one.
3719 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3720 This is apparently needed in addition to the PH/07 change above to avoid
3721 any possible encoding problems.
3723 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3724 but not after initializing Perl.
3726 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3727 output them only if debugging. By default they are written stderr,
3728 apparently, which is not desirable.
3730 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3733 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3736 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3738 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3739 authenticated or an ident call has been made. Suppress the default
3740 values for $authenticated_id and $authenticated_sender (but permit -oMai
3741 and -oMas) when testing with -bh.
3743 PH/24 Re-jigged the order of the tests in the default configuration so that the
3744 tests for valid domains and recipients precede the DNS black list and CSA
3745 tests, on the grounds that those ones are more expensive.
3747 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3748 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3749 This bug exists in every version of Exim that I still have, right back to
3752 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3753 However, an attempt to turn on case-sensitivity in a regex key by
3754 including (?-i) didn't work because the subject string was already
3755 lowercased, and the effects were non-intuitive. It turns out that a
3756 one-line patch can be used to allow (?-i) to work as expected.
3762 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3763 other effects) broke the use of negated acl sub-conditions.
3765 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3768 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3769 "Deny" causes Exim to reject the incoming connection with a 554 error.
3770 Unfortunately, if there is a major crisis, such as a disk failure,
3771 tcp-wrappers gives "deny", whereas what one would like would be some
3772 kind of temporary error. A kludge has been added to help with this.
3773 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3774 554 error is used if errno is still zero or contains ENOENT (which occurs
3775 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3778 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3780 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3781 errors. Otherwise a message that provokes a temporary error (when other
3782 messages do not) can cause a whole host to time out.
3784 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3785 addresses were routed directly to files or pipes from a redirect router.
3786 File deliveries just didn't batch; pipe deliveries might have suffered
3789 PH/06 A failure to get a lock for a hints database would erroneously always say
3790 "Failed to get write lock", even when it was really a read lock.
3792 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3793 of 0600. This has been changed to use the value of the lockfile_mode
3794 option (which defaults to 0600).
3796 PH/08 Applied small patch from the Sieve maintainer.
3798 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3799 folder from quota calculations, a direct delivery into this folder messed
3800 up the contents of the maildirsize file. This was because the regex was
3801 used only to exclude .Trash (or whatever) when the size of the mailbox
3802 was calculated. There was no check that a delivery was happening into an
3803 excluded directory. This bug has been fixed by ignoring all quota
3804 processing for deliveries into excluded directories.
3806 PH/10 Added the maildirfolder_create_regex option to appendfile.
3812 PH/01 The code for finding all the local interface addresses on a FreeBSD
3813 system running IPv6 was broken. This may well have applied to all BSD
3814 systems, as well as to others that have similar system calls. The broken
3815 code found IPv4 interfaces correctly, but gave incorrect values for the
3816 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3817 that it would not match correctly against @[] and not recognize the IPv6
3820 PH/02 The ipliteral router was not recognizing addresses of the form user@
3821 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3823 PH/03 Added disable_ipv6.
3825 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3826 decoded form, because it is most often used to construct To: headers
3827 lines in autoreplies, and the decoded form may well be syntactically
3828 invalid. However, $reply_address has leading white space removed, and all
3829 newlines turned into spaces so that the autoreply transport does not
3832 PH/05 If group was specified without a user on a router, and no group or user
3833 was specified on a transport, the group from the router was ignored.
3835 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3836 for visible compile-time settings that can be used to change these
3837 numbers, for those that want even more. Backwards compatibility with old
3838 spool files has been maintained. However, going back to a previous Exim
3839 release will lost any variables that are in spool files.
3841 PH/07 Two small changes when running in the test harness: increase delay when
3842 passing a TCP/IP connection to a new process, in case the original
3843 process has to generate a bounce, and remove special handling of
3844 127.0.0.2 (sic), which is no longer necessary.
3846 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3847 be the same on different OS.
3849 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3852 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3853 whether --show-vars was specified or not
3855 JJ/02 exipick: Added support for new ACL variable spool format introduced
3858 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3859 syntactically invalid From: or Reply-to: line, and a filter used this to
3860 generate an autoreply, and therefore failed to obtain an address for the
3861 autoreply, Exim could try to deliver to a non-existent relative file
3862 name, causing unrelated and misleading errors. What now happens is that
3863 it logs this as a hard delivery error, but does not attempt to create a
3866 PH/11 The exinext utility has a -C option for testing purposes, but although
3867 the given file was scanned by exinext itself; it wasn't being passed on
3868 when Exim was called.
3870 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3871 an end-of-file indication when reading a command response.
3873 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3874 compiled. In many other places in Exim, IPv6 addresses are always
3875 recognized, so I have changed this. It also means that IPv4 domain
3876 literals of the form [IPV4:n.n.n.n] are now always recognized.
3878 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3879 used if the router is not running as root, for example, when verifying at
3880 ACL time, or when using -bh. The debugging output from this situation was
3881 non-existent - all you got was a failure to exec. I have made two
3884 (a) Failures to set uid/gid, the current directory, or a process leader
3885 in a subprocess such as that created by queryprogram now generate
3886 suitable debugging output when -d is set.
3888 (b) The queryprogram router detects when it is not running as root,
3889 outputs suitable debugging information if -d is set, and then runs
3890 the subprocess without attempting to change uid/gid.
3892 PH/15 Minor change to Makefile for building test_host (undocumented testing
3895 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3896 additional section of a DNS packet that returns MX or SRV records.
3897 Instead, it always explicitly searches for A/AAAA records. This avoids
3898 major problems that occur when a DNS server includes only records of one
3899 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3900 fixed another bug: if SRV records were looked up and the corresponding
3901 address records were *not* found in the additional section, the port
3902 values from the SRV records were lost.
3904 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3905 using the correct key (the original address) when searching the retry
3906 rules in order to find which one to use for generating the retry hint.
3908 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3909 adding the default one. Similarly, if it contains a Reply-To: header, the
3910 errors_reply_to option, if set, is not used.
3912 PH/19 When calculating a retry time, Exim used to measure the "time since
3913 failure" by looking at the "first failed" field in the retry record. Now
3914 it does not use this if it is later than than the arrival time of the
3915 message. Instead it uses the arrival time. This makes for better
3916 behaviour in cases where some deliveries succeed, thus re-setting the
3917 "first failed" field. An example is a quota failure for a huge message
3918 when small messages continue to be delivered. Without this change, the
3919 "time since failure" will always be short, possible causing more frequent
3920 delivery attempts for the huge message than are intended.
3921 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3923 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3924 $1, $2, $3) because the numerical variables can be reset during some
3925 expansion items (e.g. "match"), thereby losing the authentication data.
3927 PH/21 Make -bV show the size of off_t variables so that the test suite can
3928 decide whether to run tests for quotas > 2G.
3930 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3931 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3932 filecount value is greater than 2G or if a quota value is greater than 2G
3933 on a system where the size of off_t is not greater than 4, a panic error
3936 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3937 never match. The debug and -bh output now contains an explicit error
3938 message indicating a malformed IPv4 address or mask.
3940 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3941 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3942 PH/23 above applies.
3944 PH/25 Do not write to syslog when running in the test harness. The only
3945 occasion when this arises is a failure to open the main or panic logs
3946 (for which there is an explicit test).
3948 PH/26 Added the /no_tell option to "control=freeze".
3950 PH/27 If a host name lookup failed very early in a connection, for example, if
3951 the IP address matched host_lookup and the reverse lookup yielded a name
3952 that did not have a forward lookup, an error message of the form "no IP
3953 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3954 could be logged. Now it outputs the IP address instead of "NULL".
3956 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3957 allows the sender and the authenticated sender to be set when
3958 submitting a message from within Exim. Since child_open_exim() is
3959 documented for local_scan(), the new function should be too.
3961 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3962 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3963 results in an empty string is now treated as unset.
3965 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3967 PH/31 Added sender_verify_fail logging option.
3969 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3970 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3971 tidied the source and removed it altogether.
3973 PH/33 When a queue run was abandoned because the load average was too high, a
3974 log line was always written; now it is written only if the queue_run log
3975 selector is set. In addition, the log line for abandonment now contains
3976 information about the queue run such as the pid. This is always present
3977 in "start" and "stop" lines but was omitted from the "abandon" line.
3979 PH/34 Omit spaces between a header name and the colon in the error message that
3980 is given when verify = headers_syntax fails (if there are lots of them,
3981 the message gets confusing).
3983 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3984 names, as there are now some PTR records that contain slashes. This check
3985 is only to protect against broken name servers that fall over on strange
3986 characters, so the fact that it applies to all lookups doesn't matter.
3988 PH/36 Now that the new test suite is complete, we can remove some of the
3989 special code in Exim that was needed for the old test suite. For example,
3990 sorting DNS records because real resolvers return them in an arbitrary
3991 order. The new test suite's fake resolver always returns records in the
3994 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3995 bounces) except when queue_only is set, to avoid logging races between
3996 the different processes.
3998 PH/38 Panic-die if .include specifies a non-absolute path.
4000 PH/39 A tweak to the "H" retry rule from its user.
4002 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4003 a label. They prevented compilation on older perls.
4005 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4006 a warning to be raised on newish perls.
4008 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4009 on queue. Changes to match documented behaviour of showing count of
4010 messages matching specified criteria.
4012 PH/40 Changed the default ident timeout from 30s to 5s.
4014 PH/41 Added support for the use of login_cap features, on those BSD systems
4015 that have them, for controlling the resources used by pipe deliveries.
4017 PH/42 The content-scanning code uses fopen() to create files in which to put
4018 message data. Previously it was not paying any attention to the mode of
4019 the files. Exim runs with umask(0) because the rest of the code creates
4020 files with open(), and sets the required mode explicitly. Thus, these
4021 files were ending up world-writeable. This was not a big issue, because,
4022 being within the spool directory, they were not world-accessible. I have
4023 created a function called modefopen, which takes an additional mode
4024 argument. It sets umask(777), creates the file, chmods it to the required
4025 mode, then resets the umask. All the relevant calls to fopen() in the
4026 content scanning code have been changed to use this function.
4028 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4029 to 24 hours. This avoids potential overflow problems when processing G
4030 and H retry rules. I suspect nobody ever tinkers with this value.
4032 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4034 PH/45 When the plaintext authenticator is running as a client, the server's
4035 challenges are checked to ensure they are valid base64 strings. By
4036 default, the authentication attempt is cancelled if an invalid string is
4037 received. Setting client_ignore_invalid_base64 true ignores these errors.
4038 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4039 they are received. Thus, the responses can be made to depend on the
4040 challenges. If an invalid string is ignored, an empty string is placed in
4043 PH/46 Messages that are created by the autoreply transport now contains a
4044 References: header, in accordance with RFCs 2822 and 3834.
4046 PH/47 Added authenticated_sender_force to the smtp transport.
4048 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4050 PH/49 Installed latest patch from the Sieve maintainer.
4052 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4053 was also set, the appendfile transport was unnecessarily scanning a
4054 directory of message files (e.g. for maildir delivery) to find the count
4055 of files (along with the size), even though it did not need this
4056 information. It now does the scan only if it needs to find either the
4057 size of the count of files.
4059 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4061 PH/52 Two bugs concerned with error handling when the smtp transport is
4064 (i) Exim was not creating retry information for temporary errors given
4065 for individual recipients after the DATA command when the smtp transport
4066 was used in LMTP mode. This meant that they could be retried too
4067 frequently, and not timed out correctly.
4069 (ii) Exim was setting the flag that allows error details to be returned
4070 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4071 recipients that were returned after the DATA command.
4073 PH/53 This is related to PH/52, but is more general: for any failing address,
4074 when detailed error information was permitted to be returned to the
4075 sender, but the error was temporary, then after the final timeout, only
4076 "retry timeout exceeded" was returned. Now it returns the full error as
4077 well as "retry timeout exceeded".
4079 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4080 do this, and (what is worse) MTAs that accept it.
4082 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4083 will now be deprecated.
4085 PH/56 New os.c-cygwin from the Cygwin maintainer.
4087 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4088 formats (previously only available in exim formats via -bpr, -bpru,
4089 and -bpra. Now also available in native and exiqgrep formats)
4091 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4092 with very large, slow to parse queues
4094 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4096 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4098 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4099 responses to authentication challenges, though it was showing the
4100 challenges; (ii) I've removed the CR characters from the debug output for
4103 PH/58 Allow for the insertion of a newline as well as a space when a string
4104 is turned into more than one encoded-word during RFC 2047 encoding. The
4105 Sieve code now uses this.
4107 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4108 data_4xx, lost_connection, tls_required.
4110 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4111 message was being sent as an SMTP response.
4113 PH/61 Add -l and -k options to exicyclog.
4115 PH/62 When verifying, if an address was redirected to one new address, so that
4116 verification continued, and the new address failed or deferred after
4117 having set something in $address_data, the value of $address_data was not
4118 passed back to the ACL. This was different to the case when no
4119 redirection occurred. The value is now passed back in both cases.
4121 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4122 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4123 use login_cap.h, so on its own it isn't the distinguishing feature. The
4124 new name refers directly to the setclassresources() function.
4126 PH/65 Added configuration files for NetBSD3.
4128 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4130 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4131 is preferred over IPv4.
4133 PH/68 The bounce_return_message and bounce_return_body options were not being
4134 honoured for bounces generated during the reception of non-SMTP messages.
4135 In particular, this applied to messages rejected by the ACL. This bug has
4136 been fixed. However, if bounce_return_message is true and bounce_return_
4137 body is false, the headers that are returned for a non-SMTP message
4138 include only those that have been read before the error was detected.
4139 (In the case of an ACL rejection, they have all been read.)
4141 PH/69 The HTML version of the specification is now built in a directory called
4142 spec_html instead of spec.html, because the latter looks like a path with
4143 a MIME-type, and this confuses some software.
4145 PH/70 Catch two compiler warnings in sieve.c.
4147 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4148 function verify_get_ident() calls ip_connect() to connect a socket, but
4149 if the "connect()" function timed out, ip_connect() used to close the
4150 socket. However, verify_get_ident() also closes the socket later, and in
4151 between Exim writes to the log, which may get opened at this point. When
4152 the socket was closed in ip_connect(), the log could get the same file
4153 descriptor number as the socket. This naturally causes chaos. The fix is
4154 not to close the socket in ip_connect(); the socket should be closed by
4155 the function that creates it. There was only one place in the code where
4156 this was missing, in the iplookup router, which I don't think anybody now
4157 uses, but I've fixed it anyway.
4159 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4160 well as to direct DNS lookups. Otherwise the handling of names in host
4161 lists is inconsistent and therefore confusing.
4167 PH/01 Two changes to the default runtime configuration:
4169 (1) Move the checks for relay_from_hosts and authenticated clients from
4170 after to before the (commented out) DNS black list checks.
4172 (2) Add control=submission to the relay_from_hosts and authenticated
4173 clients checks, on the grounds that messages accepted by these
4174 statements are most likely to be submissions.
4176 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4178 (1) Generate an error if the third argument for the ${prvs expansion is
4181 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4184 (3) Reset the variables that are obtained from the first argument of
4185 ${prvscheck and used in the second argument before leaving the code,
4186 because their memory is reclaimed, so using them afterwards may do
4189 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4190 one (it's much easier than Tom thought :-).
4192 (5) Because of (4), we can now allow for the use of $prvscheck_result
4193 inside the third argument.
4195 PH/03 For some reason, the default setting of PATH when running a command from
4196 a pipe transport was just "/usr/bin". I have changed it to
4199 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4200 anything to be listed in the output from -bV.
4202 PH/05 When a filter generated an autoreply, the entire To: header line was
4203 quoted in the delivery log line, like this:
4205 => >A.N.Other <ano@some.domain> <original@ddress> ...
4207 This has been changed so that it extracts the operative address. There
4208 may be more than one such address. If so, they are comma-separated, like
4211 => >ano@some.domain,ona@other.domain <original@ddress> ...
4213 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4214 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4215 not being looked up in the rDNS to get a host name, Exim was showing the
4216 IP address twice in Received: lines, even though the IP addresses were
4217 identical. For example:
4219 Received: from [1.2.3.4] (helo=[1.2.3.4])
4221 However, if the real host name was known, it was omitting the HELO data
4222 if it matched the actual IP address. This has been tidied up so that it
4223 doesn't show the same IP address twice.
4225 PH/07 When both +timestamp and +memory debugging was on, the value given by
4226 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4227 called by the debug printing, thereby overwriting the timestamp buffer.
4228 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4230 PH/08 When the original message was included in an autoreply transport, it
4231 always said "this is a copy of the message, including all the headers",
4232 even if body_only or headers_only was set. It now gives an appropriate
4235 PH/09 Applied a patch from the Sieve maintainer which:
4237 o fixes some comments
4238 o adds the (disabled) notify extension core
4239 o adds some debug output for the result of if/elsif tests
4240 o points to the current vacation draft in the documentation
4241 and documents the missing references header update
4245 o fixes a bug in processing the envelope test (when testing
4246 multiple envelope elements, the last element determined the
4249 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4250 Electronic Mail") by including:
4252 Auto-submitted: auto-generated
4254 in the messages that it generates (bounce messages and others, such as
4255 warnings). In the case of bounce messages for non-SMTP messages, there was
4256 also a typo: it was using "Auto_submitted" (underscore instead of
4257 hyphen). Since every message generated by Exim is necessarily in response
4258 to another message, thes have all been changed to:
4260 Auto-Submitted: auto-replied
4262 in accordance with these statements in the RFC:
4264 The auto-replied keyword:
4266 - SHOULD be used on messages sent in direct response to another
4267 message by an automatic process,
4269 - MUST NOT be used on manually-generated messages,
4271 - MAY be used on Delivery Status Notifications (DSNs) and Message
4272 Disposition Notifications (MDNs),
4274 - MUST NOT be used on messages generated by automatic or periodic
4275 processes, except for messages which are automatic responses to
4278 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4279 to the default Received: header definition.
4281 PH/12 Added log selector acl_warn_skipped (default on).
4283 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4284 variables because (a) they are in the wrong storage pool and (b) even if
4285 they were copied, it wouldn't work properly because of the caching.
4287 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4288 checking when decoding. Apparently there are clients that generate
4289 overlong encoded strings. Why am I not surprised?
4291 PH/15 If the first argument of "${if match_address" was not empty, but did not
4292 contain an "@" character, Exim crashed. Now it writes a panic log message
4293 and treats the condition as false.
4295 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4297 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4298 extension "envelope-auth". The code is finished and in agreement with
4299 other implementations, but there is no documentation so far and in fact,
4300 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4301 not changing the active code.
4303 Print executed "if" and "elsif" statements when debugging is used. This
4304 helps a great deal to understand what a filter does.
4306 Document more things not specified clearly in RFC3028. I had all this
4307 sorted out, when out of a sudden new issues came to my mind. Oops."
4309 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4312 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4313 canonical form (as documented). However, after a host name lookup from
4314 the IP address, check_host() was doing a simple string comparison with
4315 addresses acquired from the DNS when checking that the found name did
4316 have the original IP as one of its addresses. Since any found IPv6
4317 addresses are likely to be in abbreviated form, the comparison could
4318 fail. Luckily, there already exists a function for doing the comparison
4319 by converting both addresses to binary, so now that is used instead of
4320 the text comparison.
4322 PH/20 There was another similar case to PH/19, when a complete host name was
4323 given in a host list; looking up its IP address could give an abbreviated
4324 form, whereas the current host's name might or might not be abbreviated.
4325 The same fix has been applied.
4331 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4332 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4335 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4336 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4338 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4340 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4341 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4342 Also, for "Auto-Submitted", treat anything other than "no" as
4343 non-personal, in accordance with RFC 3834. (Previously it treated
4344 anything starting "auto-" as non-personal.)
4346 TF/01 The control=submission/name=... option had a problem with syntax
4347 errors if the name included a slash character. The /name= option
4348 now slurps the rest of the string, so it can include any characters
4349 but it must come last in the list of options (after /sender_retain
4352 PH/05 Some modifications to the interface to the fake nameserver for the new
4360 TK/01 Added the "success_on_redirect" address verification option. See
4361 NewStuff for rationale and an example.
4363 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4365 PH/02 Patch to exigrep to allow it to work on syslog lines.
4367 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4368 fread() to skip over the body file's header line, because in Cygwin the
4369 header line is locked and is inaccessible.
4371 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4372 co-exist for some time) to make it clear that it is the Exim ID that is
4373 referenced, not the Message-ID: header line.
4375 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4376 string_format() function, because snprintf() does not exist on all
4379 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4380 ${stat: expansion item.
4382 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4383 protocol synchronization error", to keep the pedants happy.
4385 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4386 well as for IRIX systems, when gcc is being used. See the host.c source
4389 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4391 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4394 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4395 transport and to the smtp transport in LMTP mode.
4397 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4399 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4400 run a filter in a subprocess. This could lead to confusion in subsequent
4401 lookups in the parent process. There should also be a search_tidyup() at
4402 the end of the subprocess.
4404 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4405 only if the host matched helo_try_verify_hosts, which caused the
4406 verification to occur when the EHLO/HELO command was issued. The ACL just
4407 tested the remembered result. Now, if a previous verification attempt has
4408 not happened, "verify = helo" does it there and then.
4410 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4412 TK/03 Fix log output including CR from clamd.
4414 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4415 exist provoked a memory error which could cause a segfault.
4417 PH/15 Installed PCRE 6.2
4419 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4421 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4422 of the problem. Specifically, suggested +O2 rather than +O1 for the
4425 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4427 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4428 record, the port was not correctly logged when the outgoing_port log
4429 selector was set (it logged the transort's default port).
4431 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4432 fallback_hosts, and "hosts" in the smtp transport.
4434 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4435 host errors such as "Connection refused".
4437 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4438 authentication with radiusclient 0.4.9:
4440 - Error returned from rc_read_config was caught wrongly
4441 - Username/password not passed on to radius server due to wrong length.
4443 The presumption is that some radiusclient API changes for 4.51/PH/17
4444 were not taken care of correctly. The code is still untested by me (my
4445 Linux distribution still has 0.3.2 of radiusclient), but it was
4446 contributed by a Radius user.
4448 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4449 expanding the "port" option of the smtp transport.
4451 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4452 while reading a MIME header. Thanks to Tom Hughes for a patch.
4454 PH/24 Include config.h inside local_scan.h so that configuration settings are
4457 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4458 that in an ACL for RCPT (for example), you can examine exactly what was
4461 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4462 commands, but it was not correctly comparing the address with the actual
4463 client host address. Thus, it would show the EHLO address in Received:
4464 header lines when this was not necessary.
4466 PH/27 Added the % operator to ${eval:}.
4468 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4469 it should be ignoring failures (because with -C, for example, it has lost
4470 privilege). It wasn't ignoring creation failures other than "already
4473 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4476 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4477 because an input line was too long, either on its own, or by virtue of
4478 too many continuations, the temporary file was not being removed, and the
4479 return code was incorrect.
4481 PH/31 Missing "BOOL" in function definition in filtertest.c.
4483 PH/32 Applied Sieve patches from the maintainer.
4485 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4487 PH/33 Added "verify = not_blind".
4489 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4490 Local/Makefile (with some defaults set). These are used in built scripts
4491 such as exicyclog, but they have never been used in the exim_install
4492 script (though there are many overriding facilities there). I have
4493 arranged that the exim_install script now takes note of these two
4496 PH/35 Installed configuration files for Dragonfly.
4498 PH/36 When a locally submitted message by a trusted user did not contain a
4499 From: header, and the sender address was obtained from -f or from an SMTP
4500 MAIL command, and the trusted user did not use -F to supply a sender
4501 name, $originator_name was incorrectly used when constructing a From:
4502 header. Furthermore, $originator_name was used for submission mode
4503 messages from external hosts without From: headers in a similar way,
4504 which is clearly wrong.
4506 PH/37 Added control=suppress_local_fixups.
4508 PH/38 When log_selector = +received_sender was set, and the addition of the
4509 sender made the log line's construction buffer exactly full, or one byte
4510 less than full, an overflow happened when the terminating "\n" was
4513 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4514 when the result of a list match is failure because a DNS lookup failed.
4516 PH/40 RM_COMMAND is now used in the building process.
4518 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4519 the "build-* directories that it finds.
4521 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4522 address in a domain literal was a prefix of an interface address.
4524 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4525 when verifying a sender address, unless rewrite_headers is false.
4527 PH/44 Wrote a long comment about why errors_to addresses are verified as
4528 recipients, not senders.
4530 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4531 the ratelimit ACL was added.
4533 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4535 PH/46 Added extra information about PostgreSQL errors to the error string.
4537 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4538 avoiding the need to install special zones in a real server. This is
4539 backwards compatible; if it can't find the fake resolver, it drops back.
4540 Thus, both old and new test suites can be run.
4542 TF/02 Added util/ratelimit.pl
4544 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4545 clock is set back in time.
4547 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4548 Candler <B.Candler@pobox.com>.
4550 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4551 OK for addresses that are the result of redirections.
4553 PH/48 A number of further additions for the benefit of the new test suite,
4554 including a fake gethostbyname() that interfaces to the fake DNS resolver
4557 TF/06 The fix for widen_domains has also been applied to qualify_single and
4558 search_parents which are the other dnslookup options that can cause
4561 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4564 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4566 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4567 service" if it encountered an MX record with an empty target hostname.
4568 The message is now "an MX or SRV record indicated no SMTP service".
4570 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4571 if the DNS of the sending site is misconfigured. This is quite a
4572 common situation. This change restores the behaviour of treating a
4573 helo verification defer as a failure.
4575 PH/51 If self=fail was set on a router, the bounce message did not include the
4576 actual error message.
4582 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4584 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4585 command itself ended in error, the underlying message about the transport
4586 filter timeout was being overwritten with the pipe command error. Now the
4587 underlying error message should be appended to the second error message.
4589 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4591 PH/02 Reduce the amount of output that "make" produces by default. Full output
4592 can still be requested.
4594 PH/03 The warning log line about a condition test deferring for a "warn" verb
4595 was being output only once per connection, rather than after each
4596 occurrence (because it was using the same function as for successful
4597 "warn" verbs). This seems wrong, so I have changed it.
4599 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4600 it should not have, which might have caused a crash in the right
4601 circumstances, but probably never did.
4603 PH/04 Installed a modified version of Tony Finch's patch to make submission
4604 mode fix the return path as well as the Sender: header line, and to
4605 add a /name= option so that you can make the user's friendly name appear
4608 TF/03 Added the control = fakedefer ACL modifier.
4610 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4611 Mark Lowes for thorough testing.
4613 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4615 TK/03 Merged latest SRS patch from Miles Wilton.
4617 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4618 with the definition in sysexits.h (which is #included earlier).
4619 Fortunately, Exim does not actually use EX_OK. The code used to try to
4620 preserve the sysexits.h value, by assuming that macro definitions were
4621 scanned for macro replacements. I have been disabused of this notion,
4622 so now the code just undefines EX_OK before #including unistd.h.
4624 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4625 in the smtp transport. When a block could not be written in a single
4626 write() function, the timeout was being re-applied to each part-write.
4627 This seems wrong - if the receiver was accepting one byte at a time it
4628 would take for ever. The timeout is now adjusted when this happens. It
4629 doesn't have to be particularly precise.
4631 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4632 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4634 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4635 without a domain if the check to <postmaster@domain> fails.
4637 SC/01 Eximstats: added -xls and the ability to specify output files
4638 (patch written by Frank Heydlauf).
4640 SC/02 Eximstats: use FileHandles for outputting results.
4642 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4644 SC/04 Eximstats: fixed display of large numbers with -nvr option
4646 SC/05 Eximstats: fixed merging of reports with empty tables.
4648 SC/06 Eximstats: added the -include_original_destination flag
4650 SC/07 Eximstats: removed tabs and trailing whitespace.
4652 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4654 TK/06 MBOX spool code: Add real "From " MBOX separator line
4655 so the .eml file is really in mbox format (even though
4656 most programs do not really care). Patch from Alex Miller.
4658 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4659 The latter is generated from $received_to and is only set if the
4660 message has one envelope recipient. SA can use these headers,
4661 obviously out-of-the-box. Patch from Alex Miller.
4663 PH/08 The ${def test on a variable was returning false if the variable's
4664 value was "0", contrary to what the specification has always said!
4665 The result should be true unless the variable is empty.
4667 PH/09 The syntax error of a character other than { following "${if
4668 def:variable_name" (after optional whitespace) was not being diagnosed.
4669 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4670 accidental colon was present, for example, could give incorrect results.
4672 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4673 result is used (not including appendfile, where other changes are about
4676 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4677 This involved changing a lot of size variables from int to off_t. It
4678 should work with maildirs and everything.
4680 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4681 spamd dying while we are connected to it.
4683 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4686 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4687 function for BDB 4.3.
4689 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4691 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4692 the sources for PCRE have been re-arranged and more files are now
4695 PH/15 The code I had for printing potentially long long variables in PH/11
4696 above was not the best (it lost precision). The length of off_t variables
4697 is now inspected at build time, and an appropriate printing format (%ld
4698 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4699 to be "long long int" or "long int". This is needed for the internal
4700 formatting function string_vformat().
4702 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4703 the configuration file to be ":syslog", then the script "guesses" where
4704 the logs files are, rather than using the compiled in default. In our
4705 case the guess is not the same as the compiled default, so the script
4706 suddenly stopped working when I started to use syslog. The patch checks
4707 to see if log_file_path is "". If so, it attempts to read it from exim
4708 with no configuration file to get the compiled in version, before it
4709 falls back to the previous guessing code."
4711 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4712 implementing BATV in an Exim configuration. See NewStuff for the gory
4715 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4716 Makefile that are specific to HP-UX.
4718 PH/18 If the "use_postmaster" option was set for a recipient callout together
4719 with the "random" option, the postmaster address was used as the MAIL
4720 FROM address for the random test, but not for the subsequent recipient
4721 test. It is now used for both.
4723 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4724 patch removes a few documentation additions to RFC 3028, because the
4725 latest draft now contains them. It adds the new en;ascii-case comparator
4726 and a new error check for 8bit text in MIME parts. Comparator and
4727 require names are now matched exactly. I enabled the subaddress
4728 extension, but it is not well tested yet (read: it works for me)."
4730 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4731 rework some of the code of TK/09 above to avoid the hardwired use of
4732 "%lld" and "long long". Replaced the call to snprintf() with a call to
4735 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4736 records point to non-existent hosts", "retry timeout exceeded", and
4737 "retry time not reached for any host after a long failure period".
4739 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4740 experimental DomainKeys support:
4742 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4743 (2) On an error such as an illegally used "control", the wrong name for
4744 the control was given.
4746 These problems did NOT occur unless DomainKeys support was compiled.
4748 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4750 PH/24 Added ${if match_ip condition.
4752 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4753 databases so that it will be absolutely obvious if a crash occurs in the
4754 DB library. This is a regular occurrence (often caused by mis-matched
4757 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4758 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4759 picked up on a user's system that detects such things. There doesn't seem
4760 to be a gcc warning option for this - only an attribute that has to be
4761 put on the function's prototype. It seems that in Fedora Core 4 they have
4762 set this on a number of new functions. No doubt there will be more in due
4765 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4766 specify a transport. However, if an address that was verified by such a
4767 router was the subject of a callout, Exim crashed because it tried to
4768 read the rcpt_include_affixes from the non-existent transport. Now it
4769 just assumes that the setting of that option is false. This bug was
4770 introduced by 4.51/PH/31.
4772 PH/28 Changed -d+all to exclude +memory, because that information is very
4773 rarely of interest, but it makes the output a lot bigger. People tend to
4774 do -d+all out of habit.
4776 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4777 code in os-type was giving problems when libc.so lives in lib64, like on
4780 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4781 aren't the modern standard, and it seems that some systems' include files
4782 don't always have them. Exim was already checking for some of the newer
4783 ones like T_AAAA, and defining it itself. I've added checks for all the
4784 record types that Exim uses.
4786 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4787 not automatically generating a new one, as it is supposed to. This
4788 prevented TLS from working. If the file did exist, but contained invalid
4789 data, a new version was generated, as expected. It was only the case of a
4790 non-existent file that was broken.
4792 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4793 with a change in libdomainkeys > 0.64.
4795 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4796 from DNS. If the selector record carries the flag, it now has
4797 precedence over the domain-wide flag.
4799 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4801 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4802 the use of a port name that isn't defined in /etc/services) occurred, the
4803 message was deferred as in a normal delivery, and thus remained on the
4804 spool, instead of being failed because of the mua_wrapper setting. This
4805 is now fixed, and I tidied up some of the mua_wrapper messages at the
4808 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4809 the messages in a hash of arrays rather than using individual hashes.
4810 This is a bit cleaner and results in dramatic memory savings, albeit
4811 at a slight CPU cost.
4813 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4814 as requested by Marc Sherman.
4816 SC/10 Eximstats: added histograms for user specified patterns as requested
4819 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4821 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4822 fopen() in the content-scanning modules that did not already have it.
4828 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4829 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4831 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4833 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4835 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4836 to test Sieve filters that use "vacation".
4838 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4839 that changes the way the GnuTLS parameters are stored in the cache file.
4840 The new format can be generated externally. For backward compatibility,
4841 if the data in the cache doesn't make sense, Exim assumes it has read an
4842 old-format file, and it generates new data and writes a new file. This
4843 means that you can't go back to an older release without removing the
4846 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4847 work if there are any delivery delays because "one_time" forces the
4848 parent to be marked "delivered", so its unseen clone is never tried
4849 again. For this reason, Exim now forbids the simultaneous setting of
4852 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4853 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4854 entry if you want to know the details. The fix, however, neglected to
4855 consider the case when local delivery batching is involved. The test for
4856 "previously delivered" was not happening when checking to see if an
4857 address could be batched with a previous (undelivered) one; under
4858 certain circumstances this could lead to multiple deliveries to the same
4861 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4862 in its include files, and this causes problems building Exim.
4864 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4865 header_syntax) but Exim was just ignoring anything given after a slash.
4866 In particular, this caused confusion with an attempt to use "verify =
4867 reverse_host_lookup/defer_ok". An error is now given when options are
4868 supplied for verify items that do not have them. (Maybe reverse_host_
4869 lookup should have a defer_ok option, but that's a different point.)
4871 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4872 defined by RFC 821) to 2048, because there were problems with some AUTH
4873 commands, and RFC 1869 says the size should be increased for extended
4874 SMTP commands that take arguments.
4876 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4879 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4880 "unknown" error; now it says that the functionality isn't in the binary.
4882 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4883 an address' error message when a string expansion fails (syntax or
4884 whatever). Otherwise the password may appear in the log. Following change
4885 PH/42 below, there is no longer a chance of it appearing in a bounce
4888 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4890 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4891 of its addresses was ever tried. (Bugzilla bug #2.)
4893 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4894 the result incorrectly in the debug output. (It correctly added a newline
4895 to what was transported.)
4897 TF/01 Added $received_time.
4899 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4900 commented out examples of how to interface to a virus scanner and to
4901 SpamAssassin. Also added commented examples of av_scanner and
4902 spamd_address settings.
4904 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4905 and controls are allowed in which ACLs. There were a couple of minor
4906 errors. Some of the entries in the conditions table (which is a table of
4907 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4908 negation of where the condition IS allowed.
4910 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4912 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4913 header file does not have a version number, so I've had to invent a new
4914 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4915 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4916 radiusclient), but it was contributed by a Radius user.
4918 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4919 files or directories, for OpenSSL.
4921 PH/19 When an Exim process that is running as root has to create an Exim log
4922 file, it does so in a subprocess that runs as exim:exim so as to get the
4923 ownership right at creation (otherwise, other Exim processes might see
4924 the file with the wrong ownership). There was no test for failure of this
4925 fork() call, which would lead to the process getting stuck as it waited
4926 for a non-existent subprocess. Forks do occasionally fail when resources
4927 run out. I reviewed all the other calls to fork(); they all seem to check
4930 PH/20 When checking for unexpected SMTP input at connect time (before writing
4931 the banner), Exim was not dealing correctly with a non-positive return
4932 from the read() function. If the client had disconnected by this time,
4933 the result was a log entry for a synchronization error with an empty
4934 string after "input=" when read() returned zero. If read() returned -1
4935 (an event I could not check), uninitialized data bytes were printed.
4936 There were reports of junk text (parts of files, etc) appearing after
4939 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4941 PH/22 Added support for macro redefinition, and (re)definition in between
4942 driver and ACL definitions.
4944 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4945 forgetting to use the resulting value; it was using the unexpanded value.
4947 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4948 hadn't been configured. The fix is from Juergen Kreileder, who
4949 understands it better than I do:
4951 "Here's what I see happening with three configured cyrus_sasl
4952 authenticators configured (plain, login, cram-md5):
4954 On startup auth_cyrus_sasl_init() gets called for each of these.
4955 This means three calls to sasl_listmech() without a specified mech_list.
4956 => SASL tests which mechs of all available mechs actually work
4957 => three warnings about OTP not working
4958 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4960 With the patch, sasl_listmech() also gets called three times. But now
4961 SASL's mech_list option is set to the server_mech specified in the the
4962 authenticator. Or in other words, the answer from sasl_listmech()
4963 gets limited to just the mech you're testing for (which is different
4965 => the return list contains just 'plain' or 'login', 'cram-md5' or
4966 nothing depending on the value of ob->server_mech.
4968 I've just tested the patch: Authentication still works fine,
4969 unavailable mechs specified in the exim configuration are still
4970 caught, and the auth.log warnings about OTP are gone."
4972 PH/25 When debugging is enabled, the contents of the command line are added
4973 to the debugging output, even when log_selector=+arguments is not
4976 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4977 answer is "GNU", and only if the return is "GNU/something" is the answer
4980 PH/27 $acl_verify_message is now set immediately after the failure of a
4981 verification in an ACL, and so is available in subsequent modifiers. In
4982 particular, the message can be preserved by coding like this:
4984 warn !verify = sender
4985 set acl_m0 = $acl_verify_message
4987 Previously, $acl_verify_message was set only while expanding "message"
4988 and "log_message" when a very denied access.
4990 PH/28 Modified OS/os.c-Linux with
4992 -#ifndef OS_LOAD_AVERAGE
4993 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4995 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4996 nomenclature these days.)
4998 PH/29 Installed patch from the Sieve maintainer that adds the options
4999 sieve_useraddress and sieve_subaddress to the redirect router.
5001 PH/30 In these circumstances:
5002 . Two addresses routed to the same list of hosts;
5003 . First host does not offer TLS;
5004 . First host accepts first address;
5005 . First host gives temporary error to second address;
5006 . Second host offers TLS and a TLS session is established;
5007 . Second host accepts second address.
5008 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5009 and peerdn, if requested) that were in fact used only for the second
5012 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5013 attention to any local part prefix or suffix that was matched by the
5014 router that accepted the address. It now behaves in the same way as it
5015 does for delivery: the affixes are removed from the local part unless
5016 rcpt_include_affixes is set on the transport.
5018 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5019 timeout during the DATA phase of an incoming message.
5021 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5022 applied a patch sanctioned by the Sieve maintainer.
5024 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5025 the uid or gid is negative. A case of a negative gid caused this to be
5026 noticed. The fix allows for either to be negative.
5028 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5029 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5032 PH/36 The change PH/12 above was broken. Fixed it.
5034 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5035 the grounds that routing the same address twice would always produce the
5036 same answer. This might have been true once, but it is certainly no
5037 longer true now. Routing a child address may depend on the previous
5038 routing that produced that child. Some complicated redirection strategies
5039 went wrong when messages had multiple recipients, and made Exim's
5040 behaviour dependent on the order in which the addresses were given.
5042 I have moved the duplicate checking until after the routing is complete.
5043 Exim scans the addresses that are assigned to local and remote
5044 transports, and removes any duplicates. This means that more work will be
5045 done, as duplicates will always all be routed, but duplicates are
5046 presumably rare, so I don't expect this is of any significance.
5048 For deliveries to pipes, files, and autoreplies, the duplicate checking
5049 still happens during the routing process, since they are not going to be
5052 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5053 It corrects a timeout issue with spamd. This is Ian's comment: "The
5054 background is that sometimes spamd either never reads data from a
5055 connection it has accepted, or it never writes response data. The exiscan
5056 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5057 blindly assumes that writes won't block so it may never time out."
5059 PH/39 Allow G after quota size as well as K and M.
5061 PH/40 The value set for $authenticated_id in an authenticator may not contain
5062 binary zeroes or newlines because the value is written to log lines and
5063 to spool files. There was no check on this. Now the value is run through
5064 the string_printing() function so that such characters are converted to
5065 printable escape sequences.
5067 PH/41 $message_linecount is a new variable that contains the total number of
5068 lines in the message. Compare $body_linecount, which is the count for the
5071 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5072 bounce and delay warning messages, except in certain special cases, which
5075 (a) An SMTP error message from a remote host;
5076 (b) A message specified in a :fail: redirection;
5077 (c) A message specified in a "fail" command in a system filter;
5078 (d) A message specified in a FAIL return from the queryprogram router;
5079 (e) A message specified by the cannot_route_message router option.
5081 In these cases only, Exim does include the error details in bounce and
5082 warning messages. There are also a few cases where bland messages such
5083 as "unrouteable address" or "local delivery error" are given.
5085 PH/43 $value is now also set for the "else" part of a ${run expansion.
5087 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5088 being worked on, but at least Exim now implements the latest version to
5091 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5092 process to complete was treated as a delivery failure, a timeout while
5093 writing the message to the pipe was logged, but erroneously treated as a
5094 successful delivery. Such timeouts include transport filter timeouts. For
5095 consistency with the overall process timeout, these timeouts are now
5096 treated as errors, giving rise to delivery failures by default. However,
5097 there is now a new Boolean option for the pipe transport called
5098 timeout_defer, which, if set TRUE, converts the failures into defers for
5099 both kinds of timeout. A transport filter timeout is now identified in
5102 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5103 systems where "make" and "gmake" are different, calling "gmake" at top
5104 level broke things. I've arranged for the value of $(MAKE) to be passed
5105 from the Makefile to this script so that it can call the same version of
5109 A note about Exim versions 4.44 and 4.50
5110 ----------------------------------------
5112 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5113 changes of various kinds. As a consequence, a big documentation update was
5114 needed. This delayed the release for rather longer than seemed good, especially
5115 in the light of a couple of (minor) security issues. Therefore, the changes
5116 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5117 release. So 4.44 and 4.50 are in effect two different branches that both start
5120 I have left the 4.50 change log unchanged; it contains all the changes since
5121 4.43. The change log for 4.44 is below; many of its items are identical to
5122 those for 4.50. This seems to be the most sensible way to preserve the
5123 historical information.
5129 1. Minor wording change to the doc/README.SIEVE file.
5131 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5132 computation of the current number of files was incorrect.
5134 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5135 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5138 4. Give more explanation in the error message when the command for a transport
5139 filter fails to execute.
5141 5. There are several places where Exim runs a non-Exim command in a
5142 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5143 was being done only for the command run by the queryprogram router. It is
5144 now done for all such subprocesses. The other cases are: ${run, transport
5145 filters, and the commands run by the lmtp and pipe transports.
5147 6. Added CONFIGURE_GROUP build-time option.
5149 7. Some older OS have a limit of 256 on the maximum number of file
5150 descriptors. Exim was using setrlimit() to set 1000 as a large value
5151 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5152 systems. I've change it so that if it can't get 1000, it tries for 256.
5154 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5155 was an oversight, and furthermore, ever since the addition of extra
5156 controls (e.g. 4.43/32), the checks on when to allow different forms of
5157 "control" were broken. There should now be diagnostics for all cases when a
5158 control that does not make sense is encountered.
5160 9. Added the /retain_sender option to "control=submission".
5162 10. $recipients is now available in the predata ACL (oversight).
5164 11. Tidy the search cache before the fork to do a delivery from a message
5165 received from the command line. Otherwise the child will trigger a lookup
5166 failure and thereby defer the delivery if it tries to use (for example) a
5167 cached ldap connection that the parent has called unbind on.
5169 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5170 of $address_data from the recipient verification was clobbered by the
5171 sender verification.
5173 13. The value of address_data from a sender verification is now available in
5174 $sender_address_data in subsequent conditions in the ACL statement.
5176 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5178 15. Added a new option "connect=<time>" to callout options, to set a different
5181 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5182 was its contents. (It was OK if the option was not defined at all.)
5184 17. A "Completed" log line is now written for messages that are removed from
5185 the spool by the -Mrm option.
5187 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5188 information about exactly what failed.
5190 19. Added -dd to debug only the daemon process.
5192 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5193 handles timeouts, both on the server side and network timeouts. Renamed the
5194 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5196 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5197 It is now set to "smtps".
5199 22. $host_address is now set to the target address during the checking of
5200 ignore_target_hosts.
5202 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5203 being passed; this would have caused $sender_host_name to have been used if
5204 matching the list had actually called for a host name (not very likely,
5205 since this list is usually IP addresses). A host name is now passed as
5208 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5209 code that specifies a non-restarting handler (typically sigaction() in
5210 modern systems) in an attempt to fix a rare and obscure crash bug.
5212 25. Narrowed the window for a race in the daemon that could cause it to ignore
5213 SIGCHLD signals. This is not a major problem, because they are used only to
5214 wake it up if nothing else does.
5216 26. A malformed maildirsize file could cause Exim to calculate negative values
5217 for the mailbox size or file count. Odd effects could occur as a result.
5218 The maildirsize information is now recalculated if the size or filecount
5221 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5222 support for a long time. Removed HAVE_SYS_VFS_H.
5224 28. Installed the latest version of exipick from John Jetmore.
5226 29. In an address list, if the pattern was not a regular expression, an empty
5227 subject address (from a bounce message) matched only if the pattern was an
5228 empty string. Non-empty patterns were not even tested. This was the wrong
5229 because it is perfectly reasonable to use an empty address as part of a
5230 database query. An empty address is now tested by patterns that are
5231 lookups. However, all the other forms of pattern expect the subject to
5232 contain a local part and a domain, and therefore, for them, an empty
5233 address still always fails if the pattern is not itself empty.
5235 30. Exim went into a mad DNS loop when attempting to do a callout where the
5236 host was specified on an smtp transport, and looking it up yielded more
5237 than one IP address.
5239 31. Re-factored the code for checking spool and log partition space into a
5240 function that finds that data and another that does the check. The former
5241 is then used to implement four new variables: $spool_space, $log_space,
5242 $spool_inodes, and $log_inodes.
5244 32. The RFC2047 encoding function was originally intended for short strings
5245 such as real names; it was not keeping to the 75-character limit for
5246 encoded words that the RFC imposes. It now respects the limit, and
5247 generates multiple encoded words if necessary. To be on the safe side, I
5248 have increased the buffer size for the ${rfc2047: expansion operator from
5251 33. It is now permitted to omit both strings after an "if" condition; if the
5252 condition is true, the result is "true". As before, when the second string
5253 is omitted, a false condition yields an empty string. This makes it less
5254 cumbersome to write custom ACL and router conditions.
5256 34. Failure to deliver a bounce message always caused it to be frozen, even if
5257 there was an errors_to setting on the router. The errors_to setting is now
5260 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5261 canonical form (fully expanded) before being placed in
5262 $sender_host_address.
5264 36. The table in the code that translates DNS record types into text (T_A to
5265 "A" for instance) was missing entries for NS and CNAME. It is just possible
5266 that this could have caused confusion if both these types were looked up
5267 for the same domain, because the text type is used as part of Exim's
5268 per-process caching. But the chance of anyone hitting this buglet seems
5271 37. The dnsdb lookup has been extended in a number of ways.
5273 (1) There is a new type, "zns", which walks up the domain tree until it
5274 finds some nameserver records. It should be used with care.
5276 (2) There is a new type, "mxh", which is like "mx" except that it returns
5277 just the host names, not the priorities.
5279 (3) It is now possible to give a list of domains (or IP addresses) to be
5280 looked up. The behaviour when one of the lookups defers can be
5281 controlled by a keyword.
5283 (4) It is now possible to specify the separator character for use when
5284 multiple records are returned.
5286 38. The dnslists ACL condition has been extended: it is now possible to supply
5287 a list of IP addresses and/or domains to be looked up in a particular DNS
5290 39. Added log_selector=+queue_time_overall.
5292 40. When running the queue in the test harness, wait just a tad after forking a
5293 delivery process, to get repeatability of debugging output.
5295 41. Include certificate and key file names in error message when GnuTLS fails
5296 to set them up, because the GnuTLS error message doesn't include the name
5297 of the failing file when there is a problem reading it.
5299 42. Allow both -bf and -bF in the same test run.
5301 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5303 44. The "Exiscan patch" is now merged into the mainline Exim source.
5305 45. Sometimes the final signoff response after QUIT could fail to get
5306 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5307 < 0 before doing a fflush(). This bug looks as though it goes back to the
5308 introduction of TLS in release 3.20, but "sometimes" must have been rare
5309 because the tests only now provoked it.
5311 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5312 (this can affect the format of dates).
5314 47. exim_tidydb, when checking for the continued existence of a message for
5315 which it has found a message-specific retry record, was not finding
5316 messages that were in split spool directories. Consequently, it was
5317 deleting retry records that should have stayed in existence.
5319 48. Steve fixed some bugs in eximstats.
5321 49. The SPA authentication driver was not abandoning authentication and moving
5322 on to the next authenticator when an expansion was forced to fail,
5323 contradicting the general specification for all authenticators. Instead it
5324 was generating a temporary error. It now behaves as specified.
5326 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5327 (the order specifies the preference for clients). The order is now AES256,
5328 AES128, 3DES, ARCFOUR128.
5330 51. Small patch to Sieve code - explicitly set From: when generating an
5333 52. Exim crashed if a remote delivery caused a very long error message to be
5334 recorded - for instance if somebody sent an entire SpamAssassin report back
5335 as a large number of 550 error lines. This bug was coincidentally fixed by
5336 increasing the size of one of Exim's internal buffers (big_buffer) that
5337 happened as part of the Exiscan merge. However, to be on the safe side, I
5338 have made the code more robust (and fixed the comments that describe what
5341 53. Now that there can be additional text after "Completed" in log lines (if
5342 the queue_time_overall log selector is set), a one-byte patch to exigrep
5343 was needed to allow it to recognize "Completed" as not the last thing in
5346 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5347 patch that reportedly fixes this has been added. I am not expert enough to
5348 create a test for it. This is what the patch creator wrote:
5350 "I found a little strange behaviour of ldap code when working with
5351 Windows 2003 AD Domain, where users was placed in more than one
5352 Organization Units. When I tried to give exim partial DN, the exit code
5353 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5354 But simultaneously result of request was absolutely normal ldap result,
5355 so I produce this patch..."
5357 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5358 REFERENCE, so I have modified the code to exclude the patch when that macro
5361 55. Some experimental protocols are using DNS PTR records for new purposes. The
5362 keys for these records are domain names, not reversed IP addresses. The
5363 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5364 leaves it alone. Component reversal etc. now happens only for IP addresses.
5367 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5369 57. Double the size of the debug message buffer (to 2048) so that more of very
5370 long debug lines gets shown.
5372 58. The exicyclog utility now does better if the number of log files to keep
5373 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5375 59. Two changes related to the smtp_active_hostname option:
5377 (1) $smtp_active_hostname is now available as a variable.
5378 (2) The default for smtp_banner uses $smtp_active_hostname instead
5379 of $primary_hostname.
5381 60. The host_aton() function is supposed to be passed a string that is known
5382 to be a valid IP address. However, in the case of IPv6 addresses, it was
5383 not checking this. This is a hostage to fortune. Exim now panics and dies
5384 if the condition is not met. A case was found where this could be provoked
5385 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5386 components; fortuitously, this particular loophole had already been fixed
5387 by change 4.50/55 above.
5389 If there are any other similar loopholes, the new check in host_aton()
5390 itself should stop them being exploited. The report I received stated that
5391 data on the command line could provoke the exploit when Exim was running as
5392 exim, but did not say which command line option was involved. All I could
5393 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5394 running as the user.
5397 61. There was a buffer overflow vulnerability in the SPA authentication code
5398 (which came originally from the Samba project). I have added a test to the
5399 spa_base64_to_bits() function which I hope fixes it.
5402 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5403 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5405 63. The daemon start-up calls getloadavg() while still root for those OS that
5406 need the first call to be done as root, but it missed one case: when
5407 deliver_queue_load_max is set with deliver_drop_privilege. This is
5408 necessary for the benefit of the queue runner, because there is no re-exec
5409 when deliver_drop_privilege is set.
5411 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5412 This has been fixed.
5414 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5415 was in use, was not putting the data itself into the right store pool;
5416 consequently, it could be overwritten for a subsequent message in the same
5417 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5420 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5422 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5423 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5424 the function were treating the return as a boolean value, which happened to
5425 work because 0=false and not-0=true, but is not correct code.
5427 68. The host_aton() function was not handling scoped IPv6 addresses (those
5428 with, for example, "%eth0" on the end) correctly.
5430 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5431 negated items (that is, ~something) in unsigned ints. Some compilers
5432 apparently mutter when there is no cast.
5434 70. If an address verification called from an ACL failed, and did not produce a
5435 user-specific message (i.e. there was only a "system" message), nothing was
5436 put in $acl_verify_message. In this situation, it now puts the system
5439 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5440 session; change 4.31/43 added the unwanted input to the log line - except
5441 that it did not do this in the start of session case. It now does.
5443 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5444 This is wrong and can cause the other end to generate a synchronization
5445 error if it is another Exim or anything else that does the synchronization
5446 check. A QUIT command is no longer sent after a timeout.
5448 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5449 during host lookups.
5451 74. The defer_ok option of callout verification was not working if it was used
5452 when verifying addresses in header lines, that is, for this case:
5454 verify = header_sender/callout=defer_ok
5456 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5457 those file descriptors could be used for SMTP connections. If anything
5458 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5459 could be sent to the SMTP client, causing chaos. The daemon now opens
5460 stdin, stdout, and stderr to /dev/null when it puts itself into the
5463 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5464 log by default. The user can override this with suitable Perl magic.
5466 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5467 the log message when discard triggers, was not working for the DATA ACL or
5468 for the non-SMTP ACL.
5470 78. Error message wording change in sieve.c.
5472 79. If smtp_accept_max_per_host was set, the number of connections could be
5473 restricted to fewer than expected, because the daemon was trying to set up
5474 a new connection before checking whether the processes handling previous
5475 connections had finished. The check for completed processes is now done
5476 earlier. On busy systems, this bug wouldn't be noticed because something
5477 else would have woken the daemon, and it would have reaped the completed
5480 80. If a message was submitted locally by a user whose login name contained one
5481 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5482 It caused a spool format error. I have fixed the spool reading code. A
5483 related problem was that the "from" clause in the Received: line became
5484 illegal because of the space(s). It is now covered by ${quote_local_part.
5486 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5489 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5491 83. Patch from Sieve maintainer to fix unterminated string problem in
5492 "vacation" handling.
5494 84. Some minor changes to the Linux configuration files to help with other
5495 OS variants using glibc.
5497 85. One more patch for Sieve to update vacation handling to latest spec.
5500 ----------------------------------------------------
5501 See the note above about the 4.44 and 4.50 releases.
5502 ----------------------------------------------------
5508 1. Change 4.43/35 introduced a bug that caused file counts to be
5509 incorrectly computed when quota_filecount was set in an appendfile
5512 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5513 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5516 3. Give more explanation in the error message when the command for a transport
5517 filter fails to execute.
5519 4. There are several places where Exim runs a non-Exim command in a
5520 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5521 was being done only for the command run by the queryprogram router. It is
5522 now done for all such subprocesses. The other cases are: ${run, transport
5523 filters, and the commands run by the lmtp and pipe transports.
5525 5. Some older OS have a limit of 256 on the maximum number of file
5526 descriptors. Exim was using setrlimit() to set 1000 as a large value
5527 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5528 systems. I've change it so that if it can't get 1000, it tries for 256.
5530 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5531 was an oversight, and furthermore, ever since the addition of extra
5532 controls (e.g. 4.43/32), the checks on when to allow different forms of
5533 "control" were broken. There should now be diagnostics for all cases when a
5534 control that does not make sense is encountered.
5536 7. $recipients is now available in the predata ACL (oversight).
5538 8. Tidy the search cache before the fork to do a delivery from a message
5539 received from the command line. Otherwise the child will trigger a lookup
5540 failure and thereby defer the delivery if it tries to use (for example) a
5541 cached ldap connection that the parent has called unbind on.
5543 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5544 of $address_data from the recipient verification was clobbered by the
5545 sender verification.
5547 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5548 was its contents. (It was OK if the option was not defined at all.)
5550 11. A "Completed" log line is now written for messages that are removed from
5551 the spool by the -Mrm option.
5553 12. $host_address is now set to the target address during the checking of
5554 ignore_target_hosts.
5556 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5557 being passed; this would have caused $sender_host_name to have been used if
5558 matching the list had actually called for a host name (not very likely,
5559 since this list is usually IP addresses). A host name is now passed as
5562 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5563 code that specifies a non-restarting handler (typically sigaction() in
5564 modern systems) in an attempt to fix a rare and obscure crash bug.
5566 15. Narrowed the window for a race in the daemon that could cause it to ignore
5567 SIGCHLD signals. This is not a major problem, because they are used only to
5568 wake it up if nothing else does.
5570 16. A malformed maildirsize file could cause Exim to calculate negative values
5571 for the mailbox size or file count. Odd effects could occur as a result.
5572 The maildirsize information is now recalculated if the size or filecount
5575 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5576 support for a long time. Removed HAVE_SYS_VFS_H.
5578 18. Updated exipick to current release from John Jetmore.
5580 19. Allow an empty sender to be matched against a lookup in an address list.
5581 Previously the only cases considered were a regular expression, or an
5584 20. Exim went into a mad DNS lookup loop when doing a callout where the
5585 host was specified on the transport, if the DNS lookup yielded more than
5588 21. The RFC2047 encoding function was originally intended for short strings
5589 such as real names; it was not keeping to the 75-character limit for
5590 encoded words that the RFC imposes. It now respects the limit, and
5591 generates multiple encoded words if necessary. To be on the safe side, I
5592 have increased the buffer size for the ${rfc2047: expansion operator from
5595 22. Failure to deliver a bounce message always caused it to be frozen, even if
5596 there was an errors_to setting on the router. The errors_to setting is now
5599 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5600 canonical form (fully expanded) before being placed in
5601 $sender_host_address.
5603 24. Updated eximstats to version 1.33
5605 25. Include certificate and key file names in error message when GnuTLS fails
5606 to set them up, because the GnuTLS error message doesn't include the name
5607 of the failing file when there is a problem reading it.
5609 26. Expand error message when OpenSSL has problems setting up cert/key files.
5612 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5613 (this can affect the format of dates).
5615 28. exim_tidydb, when checking for the continued existence of a message for
5616 which it has found a message-specific retry record, was not finding
5617 messages that were in split spool directories. Consequently, it was
5618 deleting retry records that should have stayed in existence.
5620 29. eximstats updated to version 1.35
5621 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5622 1.35 - bugfix such that pie charts by volume are generated correctly
5624 30. The SPA authentication driver was not abandoning authentication and moving
5625 on to the next authenticator when an expansion was forced to fail,
5626 contradicting the general specification for all authenticators. Instead it
5627 was generating a temporary error. It now behaves as specified.
5629 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5630 (the order specifies the preference for clients). The order is now AES256,
5631 AES128, 3DES, ARCFOUR128.
5633 31. Small patch to Sieve code - explicitly set From: when generating an
5636 32. Exim crashed if a remote delivery caused a very long error message to be
5637 recorded - for instance if somebody sent an entire SpamAssassin report back
5638 as a large number of 550 error lines. This bug was coincidentally fixed by
5639 increasing the size of one of Exim's internal buffers (big_buffer) that
5640 happened as part of the Exiscan merge. However, to be on the safe side, I
5641 have made the code more robust (and fixed the comments that describe what
5644 33. Some experimental protocols are using DNS PTR records for new purposes. The
5645 keys for these records are domain names, not reversed IP addresses. The
5646 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5647 leaves it alone. Component reversal etc. now happens only for IP addresses.
5650 34. The host_aton() function is supposed to be passed a string that is known
5651 to be a valid IP address. However, in the case of IPv6 addresses, it was
5652 not checking this. This is a hostage to fortune. Exim now panics and dies
5653 if the condition is not met. A case was found where this could be provoked
5654 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5655 components; fortuitously, this particular loophole had already been fixed
5656 by change 4.50/55 or 4.44/33 above.
5658 If there are any other similar loopholes, the new check in host_aton()
5659 itself should stop them being exploited. The report I received stated that
5660 data on the command line could provoke the exploit when Exim was running as
5661 exim, but did not say which command line option was involved. All I could
5662 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5663 running as the user.
5666 35. There was a buffer overflow vulnerability in the SPA authentication code
5667 (which came originally from the Samba project). I have added a test to the
5668 spa_base64_to_bits() function which I hope fixes it.
5671 36. The daemon start-up calls getloadavg() while still root for those OS that
5672 need the first call to be done as root, but it missed one case: when
5673 deliver_queue_load_max is set with deliver_drop_privilege. This is
5674 necessary for the benefit of the queue runner, because there is no re-exec
5675 when deliver_drop_privilege is set.
5677 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5678 was in use, was not putting the data itself into the right store pool;
5679 consequently, it could be overwritten for a subsequent message in the same
5680 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5683 38. Sometimes the final signoff response after QUIT could fail to get
5684 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5685 < 0 before doing a fflush(). This bug looks as though it goes back to the
5686 introduction of TLS in release 3.20, but "sometimes" must have been rare
5687 because the tests only now provoked it.
5693 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5694 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5695 Now it returns an integer. A number of calls were still expecting a T/F
5696 return. Fortuitously, in all cases, the tests worked in OK situations,
5697 which is the norm. However, things would have gone wrong on any write
5698 failures on the smtp file descriptor. This function is used when sending
5699 messages over SMTP and also when doing verify callouts.
5701 2. When Exim is called to do synchronous delivery of a locally submitted
5702 message (the -odf or -odi options), it no longer closes stderr before doing
5705 3. Implemented the mua_wrapper option.
5707 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5709 5. Implemented the functions header_remove(), header_testname(),
5710 header_add_at_position(), and receive_remove_recipient(), and exported them
5713 6. If an ACL "warn" statement specified the addition of headers, Exim already
5714 inserted X-ACL-Warn: at the start if there was no header name. However, it
5715 was not making this test for the second and subsequent header lines if
5716 there were newlines in the string. This meant that an invalid header could
5717 be inserted if Exim was badly configured.
5719 7. Allow an ACL "warn" statement to add header lines at the start or after all
5720 the Received: headers, as well as at the end.
5722 8. Added the rcpt_4xx retry error code.
5724 9. Added postmaster_mailfrom=xxx to callout verification option.
5726 10. Added mailfrom=xxxx to the callout verification option, for verify=
5729 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5730 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5732 12. Inserted some casts to stop certain compilers warning when using pointer
5733 differences as field lengths or precisions in printf-type calls (mostly
5734 affecting debugging statements).
5736 13. Added optional readline() support for -be (dynamically loaded).
5738 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5739 same clock tick as a message's arrival, so that its received time was the
5740 same as the "first fail" time on the retry record, and that message
5741 remained on the queue past the ultimate address timeout, every queue runner
5742 would try a delivery (because it was past the ultimate address timeout) but
5743 after another failure, the ultimate address timeout, which should have then
5744 bounced the address, did not kick in. This was a "< instead of <=" error;
5745 in most cases the first failure would have been in the next clock tick
5746 after the received time, and all would be well.
5748 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5749 being recognized when the domain list was tested by the match_domain
5750 condition in an expansion string.
5752 16. Added the ${str2b64: operator.
5754 17. Exim was always calling setrlimit() to set a large limit for the number of
5755 processes, without checking whether the existing limit was already
5756 adequate. (It did check for the limit on file descriptors.) Furthermore,
5757 errors from getrlimit() and setrlimit() were being ignored. Now they are
5758 logged to the main and panic logs, but Exim does carry on, to try to do its
5759 job under whatever limits there are.
5761 18. Imported PCRE 5.0.
5763 19. Trivial typo in log message " temporarily refused connection" (the leading
5766 20. If the log selector return_path_on_delivery was set and an address was
5767 redirected to /dev/null, the delivery process crashed because it assumed
5768 that a return path would always be set for a "successful" delivery. In this
5769 case, the whole delivery is bypassed as an optimization, and therefore no
5772 21. Internal re-arrangement: the function for sending a challenge and reading
5773 a response while authentication was assuming a zero-terminated challenge
5774 string. It's now changed to take a pointer and a length, to allow for
5775 binary data in such strings.
5777 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5779 23. Exim was not respecting finduser_retries when seeking the login of the
5780 uid under which it was called; it was always trying 10 times. (The default
5781 setting of finduser_retries is zero.) Also, it was sleeping after the final
5782 failure, which is pointless.
5784 24. Implemented tls_on_connect_ports.
5786 25. Implemented acl_smtp_predata.
5788 26. If the domain in control=submission is set empty, Exim assumes that the
5789 authenticated id is a complete email address when it generates From: or
5790 Sender: header lines.
5792 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5793 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5794 chown and chgrp in /bin and hostname in /usr/bin.
5796 28. Exim was keeping the "process log" file open after each use, just as it
5797 does for the main log. This opens the possibility of it remaining open for
5798 long periods when the USR1 signal hits a daemon. Occasional processlog
5799 errors were reported, that could have been caused by this. Anyway, it seems
5800 much more sensible not to leave this file open at all, so that is what now
5803 29. The long-running daemon process does not normally write to the log once it
5804 has entered its main loop, and it closes the log before doing so. This is
5805 so that log files can straightforwardly be renamed and moved. However,
5806 there are a couple of unusual error situations where the daemon does write
5807 log entries, and I had neglected to close the log afterwards.
5809 30. The text of an SMTP error response that was received during a remote
5810 delivery was being truncated at 512 bytes. This is too short for some of
5811 the long messages that one sometimes sees. I've increased the limit to
5814 31. It is now possible to make retry rules that apply only when a message has a
5815 specific sender, in particular, an empty sender.
5817 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5818 it possible to be selective about when SMTP synchronization is enforced.
5820 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5822 32. Implemented hosts_connection_nolog.
5824 33. Added an ACL for QUIT.
5826 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5829 35. Added mailbox_size and mailbox_filecount to appendfile.
5831 36. Added control = no_multiline_responses to ACLs.
5833 37. There was a bug in the logic of the code that waits for the clock to tick
5834 in the case where the clock went backwards by a substantial amount such
5835 that the microsecond fraction of "now" was more than the microsecond
5836 fraction of "then" (but the whole seconds number was less).
5838 38. Added support for the libradius Radius client library this is found on
5839 FreeBSD (previously only the radiusclient library was supported).
5845 1. When certain lookups returned multiple values in the form name=value, the
5846 quoting of the values was not always being done properly. Specifically:
5847 (a) If the value started with a double quote, but contained no whitespace,
5849 (b) If the value contained whitespace other than a space character (i.e.
5850 tabs or newlines or carriage returns) it was not quoted.
5851 This fix has been applied to the mysql and pgsql lookups by writing a
5852 separate quoting function and calling it from the lookup code. The fix
5853 should probably also be applied to nisplus, ibase and oracle lookups, but
5854 since I cannot test any of those, I have not disturbed their existing code.
5856 2. A hit in the callout cache for a specific address caused a log line with no
5857 reason for rejecting RCPT. Now it says "Previous (cached) callout
5858 verification failure".
5860 3. There was an off-by-one bug in the queryprogram router. An over-long
5861 return line was truncated at 256 instead of 255 characters, thereby
5862 overflowing its buffer with the terminating zero. As well as fixing this, I
5863 have increased the buffer size to 1024 (and made a note to document this).
5865 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5866 when Exim is waiting for an SMTP response from a remote server, Exim
5867 restarts its select() call on the socket, thereby resetting its timeout.
5868 This is not a problem when such interrupts are rare. Somebody set up a cron
5869 job to run exiwhat every 2 minutes, which is less than the normal select()
5870 timeout (5 or 10 minutes). This meant that the select() timeout never
5871 kicked in because it was always reset. I have fixed this by comparing the
5872 time when an interrupt arrives with the time at the start of the first call
5873 to select(). If more time than the timeout has elapsed, the interrupt is
5874 treated as a timeout.
5876 5. Some internal re-factoring in preparation for the addition of Sieve
5877 extensions (by MH). In particular, the "personal" test is moved to a
5878 separate function, and given an option for scanning Cc: and Bcc: (which is
5879 not set for Exim filters).
5881 6. When Exim created an email address using the login of the caller as the
5882 local part (e.g. when creating a From: or Sender: header line), it was not
5883 quoting the local part when it contained special characters such as @.
5885 7. Installed new OpenBSD configuration files.
5887 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5888 try to make them clearer.
5890 9. Callout options, other than the timeout value, were being ignored when
5891 verifying sender addresses in header lines. For example, when using
5893 verify = header_sender/callout=no_cache
5895 the cache was (incorrectly) being used.
5897 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5898 only the Cygwin environment.
5900 11. When return_path_on_delivery was set as a log selector, if different remote
5901 addresses in the same message used different return paths and parallel
5902 remote delivery occurred, the wrong values would sometimes be logged.
5903 (Whenever a remote delivery process finished, the return path value from
5904 the most recently started remote delivery process was logged.)
5906 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5907 lines when AUTH and/or TLS are in use. This is the "received protocol"
5908 field. Exim used to use "asmtp" for authenticated SMTP, without any
5909 indication (in the protocol name) for TLS use. Now it follows the RFC and
5910 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5911 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5912 names appear in log lines as well as in Received: header lines.
5914 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5915 extensions, and comparison tests, and to fix some bugs.
5917 14. Changes to the "personal" filter test:
5919 (1) The test was buggy in that it was just doing the equivalent of
5920 "contains" tests on header lines. For example, if a user's address was
5921 anne@some.where, the "personal" test would incorrectly be true for
5923 To: susanne@some.where
5925 This test is now done by extracting each address from the header in turn,
5926 and checking the entire address. Other tests that are part of "personal"
5927 are now done using regular expressions (for example, to check local parts
5928 of addresses in From: header lines).
5930 (2) The list of non-personal local parts in From: addresses has been
5931 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5932 taken from the Sieve specification recommendations.
5934 (3) If the message contains any header line starting with "List-" it is
5935 treated as non-personal.
5937 (4) The test for "circular" in the Subject: header line has been removed
5938 because it now seems ill-conceived.
5940 15. Minor typos in src/EDITME comments corrected.
5942 16. Installed latest exipick from John Jetmore.
5944 17. If headers_add on a router specified a text string that was too long for
5945 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5946 of string_sprintf() is now avoided.
5948 18. $message_body_size was not set (it was always zero) when running the DATA
5949 ACL and the local_scan() function.
5951 19. For the "mail" command in an Exim filter, no default was being set for
5952 the once_repeat time, causing a random time value to be used if "once" was
5953 specified. (If the value happened to be <= 0, no repeat happened.) The
5954 default is now 0s, meaning "never repeat". The "vacation" command was OK
5955 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5956 (I found it when inspecting the code).
5958 20. There is now an overall timeout for performing a callout verification. It
5959 defaults to 4 times the callout timeout, which applies to individual SMTP
5960 commands during the callout. The overall timeout applies when there is more
5961 than one host that can be tried. The timeout is checked before trying the
5962 next host. This prevents very long delays if there are a large number of
5963 hosts and all are timing out (e.g. when the network connections are timing
5964 out). The value of the overall timeout can be changed by specifying an
5965 additional sub-option for "callout", called "maxwait". For example:
5967 verify = sender/callout=5s,maxwait=20s
5969 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5970 to the end before writing, but this should make it even safer).
5972 22. Exim was forgetting that it had advertised PIPELINING for the second and
5973 subsequent messages on an SMTP connection. It was also not resetting its
5974 memory on STARTTLS and an internal HELO.
5976 23. When Exim logs an SMTP synchronization error within a session, it now
5977 records whether PIPELINING has been advertised or not.
5979 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5980 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5981 rather than long int.
5983 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5985 26. Added the never_mail option to autoreply.
5991 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5992 crash if the getsockname() call failed; this can happen if a connection is
5993 closed very soon after it is established. The problem was simply in the
5994 order in which certain operations were done, causing Exim to try to write
5995 to the SMTP stream before it had set up the file descriptor. The bug has
5996 been fixed by making things happen in the correct order.
6002 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6003 before the connection was closed, thus losing the rejection response.
6005 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6006 some early Solaris releases, but causes trouble in current releases where
6007 socklen_t is defined.
6009 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6012 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6015 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6016 that releasing the top of it at the end releases what was used for sub-
6017 expansions (unless the block got too big). However, discard this block if
6018 the first thing is a variable or header, so that we can use its block when
6019 it is dynamic (useful for very large $message_headers, for example).
6021 6. Lookups now cache *every* query, not just the most recent. A new, separate
6022 store pool is used for this. It can be recovered when all lookup caches are
6023 flushed. Lookups now release memory at the end of their result strings.
6024 This has involved some general refactoring of the lookup sources.
6026 7. Some code has been added to the store_xxx() functions to reduce the amount
6027 of flapping under certain conditions.
6029 8. log_incoming_interface used to affect only the <= reception log lines. Now
6030 it causes the local interface and port to be added to several more SMTP log
6031 lines, for example "SMTP connection from", and rejection lines.
6033 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6035 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6037 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6038 could be overwritten at the end of the current message (or the start of a
6039 new message if it was set in a HELO ACL). The value is now preserved for
6040 the duration of the SMTP connection.
6042 12. If a transport had a headers_rewrite setting, and a matching header line
6043 contained an unqualified address, that address was qualified, even if it
6044 did not match any rewriting rules. The underlying bug was that the values
6045 of the flags that permit the existence of unqualified sender and recipient
6046 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6047 non-local messages, and by -bnq for local messages) were not being
6048 preserved with the message after it was received.
6050 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6051 "next input=" as part of the text comprising the host identity instead of
6052 the correct text. The code was using the same buffer for two different
6053 strings. However, depending on which order the printing function evaluated
6054 its arguments, the bug did not always show up. Under Linux, for example, my
6055 test suite worked just fine.
6057 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6058 doesn't work with some older versions of Perl. It has been changed to "my",
6059 which in any case is probably the better facility to use.
6061 15. A really picky compiler found some instances of statements for creating
6062 error messages that either had too many or two few arguments for the format
6065 16. The size of the buffer for calls to the DNS resolver has been increased
6066 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6067 for addresses that have a lot of PTR records. This alleviates a problem; it
6068 does not fully solve it.
6070 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6071 buffer now truncates the list and logs the incident, which is the same
6072 action as happens when Exim is looking up a host name and its aliases.
6073 Previously in this situation something unpredictable would happen;
6074 sometimes it was "internal error: store_reset failed".
6076 18. If a server dropped the connection unexpectedly when an Exim client was
6077 using GnuTLS and trying to read a response, the client delivery process
6078 crashed while trying to generate an error log message.
6080 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6081 string, for example:
6083 warn message = H1: something\nH2: something
6085 the text was added as a single header line from Exim's point of view
6086 though it ended up OK in the delivered message. However, searching for the
6087 second and subsequent header lines using $h_h2: did not work. This has been
6088 fixed. Similarly, if a system filter added multiple headers in this way,
6089 the routers could not see them.
6091 20. Expanded the error message when iplsearch is called with an invalid key to
6092 suggest using net-iplsearch in a host list.
6094 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6095 ACLs are no longer actually imposed (and a message to that effect is
6098 22. If a "gecos" field in a passwd entry contained escaped characters, in
6099 particular, if it contained a \" sequence, Exim got it wrong when building
6100 a From: or a Sender: header from that name. A second bug also caused
6101 incorrect handling when an unquoted " was present following a character
6102 that needed quoting.
6104 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6105 was not being matched caselessly.
6107 24. Arranged for all hyphens in the exim.8 source to be escaped with
6110 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6111 properly. Recipient callout cache records were still being keyed to include
6112 the sender, even when use_sender was set false. This led to far more
6113 callouts that were necessary. The sender is no longer included in the key
6114 when use_sender is false.
6116 26. Added "control = submission" modifier to ACLs.
6118 27. Added the ${base62d: operator to decode base 62 numbers.
6120 28. dnsdb lookups can now access SRV records.
6122 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6123 the configuration file.
6125 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6126 (-v) mode. This makes the output for a verbose queue run more intelligible.
6128 31. Added a use_postmaster feature to recipient callouts.
6130 32. Added the $body_zerocount variable, containing the number of binary zero
6131 bytes in the message body.
6133 33. The time of last modification of the "new" subdirectory is now used as the
6134 "mailbox time last read" when there is a quota error for a maildir
6137 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6139 35. Added +ignore_unknown as a special item in host lists.
6141 36. Code for decoding IPv6 addresses in host lists is now included, even if
6142 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6143 address was recognized as an IP address, but was then not correctly decoded
6144 into binary, causing unexpected and incorrect effects when compared with
6151 1. Very minor rewording of debugging text in manualroute to say "list of
6152 hosts" instead of "hostlist".
6154 2. If verify=header_syntax was set, and a header line with an unqualified
6155 address (no domain) and a large number of spaces between the end of the
6156 name and the colon was received, the reception process suffered a buffer
6157 overflow, and (when I tested it) crashed. This was caused by some obsolete
6158 code that should have been removed. The fix is to remove it!
6160 3. When running in the test harness, delay a bit after writing a bounce
6161 message to get a bit more predictability in the log output.
6163 4. Added a call to search_tidyup() just before forking a reception process. In
6164 theory, someone could use a lookup in the expansion of smtp_accept_max_
6165 per_host which, without the tidyup, could leave open a database connection.
6167 5. Added the variables $recipient_data and $sender_data which get set from a
6168 lookup success in an ACL "recipients" or "senders" condition, or a router
6169 "senders" option, similar to $domain_data and $local_part_data.
6171 6. Moved the writing of debug_print from before to after the "senders" test
6174 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6175 problems for message scanning, either using a data ACL, or using
6176 local_scan() because the Received: header was not generated till after they
6177 were called (in order to set the time as the time of reception completion).
6178 I have revised the way this works. The header is now generated after the
6179 body is received, but before the ACL or local_scan() are called. After they
6180 are run, the timestamp in the header is updated.
6186 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6187 before starting a queue runner without re-exec. This happened only when
6188 deliver_drop_privilege was set or when the Exim user was set to root. The
6189 effect of the bug was that timeouts during subsequent deliveries caused
6190 crashes instead of being properly handled. The handler is now left at its
6191 default (and expected) setting.
6193 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6194 message, again when deliver_drop_privilege is set or Exim is run as root.
6195 The bug described in (1) was not present in this case, but the tidying up
6196 of the other signals was missing. I have made the two cases consistent.
6198 3. The ignore_target_hosts setting on a manualroute router was being ignored
6199 for hosts that were looked up using the /MX notation.
6201 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6204 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6205 operated on the sender address. After changing the $sender_address to <>
6206 for the sender address verify, Exim was re-instated it as the original
6207 (before rewriting) address, but remembering that it had rewritten it, so it
6208 wasn't rewriting it again. This bug also had the effect of breaking the
6209 sender address verification caching when the sender address was rewritten.
6211 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6212 This has been changed so that if the ip literal address matches
6213 ignore_target_hosts, the router declines.
6215 7. Added expansion conditions match_domain, match_address, and match_local_
6216 part (NOT match_host).
6218 8. The placeholder for the Received: header didn't have a length field set.
6220 9. Added code to Exim itself and to exim_lock to test for a specific race
6221 condition that could lead to file corruption when using MBX delivery. The
6222 issue is with the lockfile that is created in /tmp. If this file is removed
6223 after a process has opened it but before that process has acquired a lock,
6224 there is the potential for a second process to recreate the file and also
6225 acquire a lock. This could lead to two Exim processes writing to the file
6226 at the same time. The added code performs the same test as UW imapd; it
6227 checks after acquiring the lock that its file descriptor still refers to
6228 the same named file.
6230 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6231 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6234 11. Added the smtp_active_hostname option. If used, this will typically be made
6235 to depend on the incoming interface address. Because $interface_address is
6236 not set up until the daemon has forked a reception process, error responses
6237 that can happen earlier (such as "too many connections") no longer contain
6240 12. If an expansion in a condition on a "warn" statement fails because a lookup
6241 defers, the "warn" statement is abandoned, and the next ACL statement is
6242 processed. Previously this caused the whole ACL to be aborted.
6244 13. Added the iplsearch lookup type.
6246 14. Added ident_timeout as a log selector.
6248 15. Added tls_certificate_verified as a log selector.
6250 16. Added a global option tls_require_ciphers (compare the smtp transport
6251 option of the same name). This controls incoming TLS connections.
6253 17. I finally figured out how to make tls_require_ciphers do a similar thing
6254 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6255 before starting the TLS session.
6257 18. Tabs are now shown as \t in -bP output.
6259 19. If the log selector return_path_on_delivery was set, Exim crashed when
6260 bouncing a message because it had too many Received: header lines.
6262 20. If two routers both had headers_remove settings, and the first one included
6263 a superfluous trailing colon, the final name in the first list and the
6264 first name in the second list were incorrectly joined into one item (with a
6265 colon in the middle).
6271 1. Added -C and -D options to the exinext utility, mainly to make it easier
6272 to include in the automated testing, but these could be helpful when
6273 multiple configurations are in use.
6275 2. The exinext utility was not formatting the output nicely when there was
6276 an alternate port involved in the retry record key, nor when there was a
6277 message id as well (for retries that were specific to a specific message
6278 and a specific host). It was also confused by IPv6 addresses, because of
6279 the additional colons they contain. I have fixed the IPv4 problem, and
6280 patched it up to do a reasonable job for IPv6.
6282 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6283 delivery, the log line now contains "pipelined" if PIPELINING was used.
6285 4. An SMTP transport process used to panic and die if the bind() call to set
6286 an explicit outgoing interface failed. This has been changed; it is now
6287 treated in the same way as a connect() failure.
6289 5. A reference to $sender_host_name in the part of a conditional expansion
6290 that was being skipped was still causing a DNS lookup. This no longer
6293 6. The def: expansion condition was not recognizing references to header lines
6294 that used bh_ and bheader_.
6296 7. Added the _cache feature to named lists.
6298 8. The code for checking quota_filecount in the appendfile transport was
6299 allowing one more file than it should have been.
6301 9. For compatibility with Sendmail, the command line option
6309 and sets the incoming protocol and host name (for trusted callers). The
6310 host name and its colon can be omitted when only the protocol is to be set.
6311 Note the Exim already has two private options, -pd and -ps, that refer to
6312 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6313 "s", but I don't think that's a major issue.
6315 10. A number of refactoring changes to the code, none of which should affect
6318 (a) The number of logging options was getting close to filling up the
6319 32-bit word that was used as a bit map. I have split them into two classes:
6320 those that are passed in the argument to log_write(), and those that are
6321 only ever tested independently outside of that function. These are now in
6322 separate 32-bit words, so there is plenty of room for expansion again.
6323 There is no change in the user interface or the logging behaviour.
6325 (b) When building, for example, log lines, the code previously used a
6326 macro that called string_cat() twice, in order to add two strings. This is
6327 not really sufficiently general. Furthermore, there was one instance where
6328 it was actually wrong because one of the argument was used twice, and in
6329 one call a function was used. (As it happened, calling the function twice
6330 did not affect the overall behaviour.) The macro has been replaced by a
6331 function that can join an arbitrary number of extra strings onto a growing
6334 (c) The code for expansion conditions now uses a table and a binary chop
6335 instead of a serial search (which was left over from when there were very
6336 few conditions). Also, it now recognizes conditions like "pam" even when
6337 the relevant support is not compiled in: a suitably worded error message is
6338 given if an attempt is made to use such a condition.
6340 11. Added ${time_interval:xxxxx}.
6342 12. A bug was causing one of the ddress fields not to be passed back correctly
6343 from remote delivery subprocesses. The field in question was not being
6344 subsequently used, so this caused to problems in practice.
6346 13. Added new log selectors queue_time and deliver_time.
6348 14. Might have fixed a bug in maildirsizefile handling that threw up
6349 "unexpected character" debug warnings, and recalculated the data
6350 unnecessarily. In any case, I expanded the warning message to give more
6353 15. Added the message "Restricted characters in address" to the statements in
6354 the default ACL that block characters like @ and % in local parts.
6356 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6357 Three changes have been made:
6359 (a) There was a serious bug; a negative response to MAIL caused the whole
6360 recipient domain to be cached as invalid, thereby blocking all messages
6361 to all local parts at the same domain, from all senders. This bug has
6362 been fixed. The domain is no longer cached after a negative response to
6363 MAIL if the sender used is not empty.
6365 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6368 (c) A new callout option, "use_sender" has been added for people who want
6369 the modified behaviour.
6375 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6378 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6379 indeed breaks things for older releases.
6381 3. Added additional logging to the case where there is a problem reading data
6382 from a filter that is running in a subprocess using a pipe, in order to
6383 try to track down a specific problem.
6385 4. Testing facility fudge: when running in the test harness and attempting
6386 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6387 getting "No route to host". Convert this to a timeout.
6389 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6392 6. Some OS don't have socklen_t but use size_t instead. This affects the
6393 fifth argument of getsockopt() amongst other things. This is now
6394 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6395 can be set for individual OS. I have set it for SunOS5, OSF1, and
6396 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6397 some earlier ones do not.
6399 7. Change 4.30/15 was not doing the test caselessly.
6401 8. The standard form for an IPv6 address literal was being rejected by address
6402 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6403 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6404 this, as well as the form without the "IPv6" on the front (but only when
6405 address literals are enabled, of course).
6407 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6409 10. Exim crashed if a message with an empty sender address specified by -f
6410 encountered a router with an errors_to setting. This could be provoked only
6411 by a command such as
6415 where an empty string was supplied; "<>" did not hit this bug.
6417 11. Installed PCRE release 4.5.
6419 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6420 remained set. It is now erased.
6422 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6423 times from message ids (which are base 36 rather than the normal 62).
6425 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6426 were being counted as actual protocol errors, and logged if the log
6427 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6428 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6429 and DATA following a set of rejected RCPTs do not count as protocol errors.
6430 In other words, Exim assumes they were pipelined, though this may not
6431 actually be the case. Of course, in all cases the client gets an
6432 appropriate error code.
6434 15. If a lookup fails in an ACL condition, a message about the failure may
6435 be available; it is used if testing the ACL cannot continue, because most
6436 such messages specify what the cause of the deferral is. However, some
6437 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6438 that caused an old message to be retained and used if a later statement
6439 caused a defer, replacing the real cause of the deferral.
6441 16. If an IP address had so many PTR records that the DNS lookup buffer
6442 was not large enough to hold them, Exim could crash while trying to process
6443 the truncated data. It now detects and logs this case.
6445 17. Further to 4.21/58, another change has been made: if (and only if) the
6446 first line of a message (the first header line) ends with CRLF, a bare LF
6447 in a subsequent header line has a space inserted after it, so as not to
6448 terminate the header.
6450 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6451 unnecessarily, used atoi() instead of strtol(), and didn't check the
6452 termination when getting file sizes from file names by regex.
6454 19. Completely re-implemented the support for maildirsize files, in the light
6455 of a number of problems with the previous contributed implementation
6456 (4.30/29). In particular:
6458 . If the quota is zero, the maildirsize file is maintained, but no quota is
6461 . If the maildir directory does not exist, it is created before any attempt
6462 to write a maildirsize file.
6464 . The quota value in the file is just a cache; if the quota is changed in
6465 the transport, the new value overrides.
6467 . A regular expression is available for excluding directories from the
6470 20. The autoreply transport checks the characters in options that define the
6471 message's headers; it allows continued headers, but it was checking with
6472 isspace() after an embedded newline instead of explicitly looking for a
6475 21. If all the "regular" hosts to which an address was routed had passed their
6476 expiry times, and had not reached their retry times, the address was
6477 bounced, even if fallback hosts were defined. Now Exim should go on to try
6480 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6481 equivalent code in the SMTP transport. Some hosts send humungous responses
6482 to HELO/EHLO, more than 1024 it seems.
6484 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6485 gives compiler warnings in some environments. I've now done it "properly",
6488 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6489 (because of problems with the built-in one) was declared to return uschar *
6490 instead of char *, causing compiler failure.
6492 25. Fixed a file descriptor leak when processing alias/forward files.
6494 26. Fixed a minor format string issue in dbfn.c.
6496 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6498 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6499 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6502 29. When a sender address is verified, it is cached, to save repeating the test
6503 when there is more than one recipient in a message. However, when the
6504 verification involves a callout, it is possible for different callout
6505 options to be set for different recipients. It is too complicated to keep
6506 track of this in the cache, so now Exim always runs a verification when a
6507 callout is required, relying on the callout cache for the optimization.
6508 The overhead is duplication of the address routing, but this should not be
6511 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6512 to be verified with callout=postmaster, and the main callout worked but the
6513 postmaster check failed, the verification correctly failed. However, if a
6514 subsequent RCPT command asked for sender verification *without* the
6515 postmaster check, incorrect caching caused this verification also to fail,
6518 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6519 it was not caching the DNS options (qualify_single, search_parents) that
6520 were used when the lookup failed. A subsequent lookup with different
6521 options therefore always gave the same answer, though there were cases
6522 where it should not have. (Example: a "domains = !$mx_any" option on a
6523 dnslookup router: the "domains" option is always processed without any
6524 widening, but the router might have qualify_single set.) Now Exim uses the
6525 cached value only when the same options are set.
6527 32. Added John Jetmore's "exipick" utility to the distribution.
6529 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6530 than a timeout (e.g. a certificate is required, and is not provided), an
6531 Exim server now closes the connection immediately. Previously it waited for
6532 the client to close - but if the client is SSL, it seems that they each
6533 wait for each other, leading to a delay before one of them times out.
6535 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6536 maintained 0.8.x compatibility because I don't think many are using it, and
6537 it is clearly obsolete.
6539 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6542 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6543 client certificate was expired. A simple patch fixes this, though I don't
6544 understand the full logic of why the verify callback is called multiple
6547 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6548 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6549 which causes problems with some clients (such as the Certicom SSL Plus
6550 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6551 disables the coutermeasure allowing Eudora to connect."
6553 38. Exim was not checking that a write() to a log file succeeded. This could
6554 lead to Bad Things if a log got too big, in particular if it hit a file
6555 size limit. Exim now panics and dies if it cannot write to a log file, just
6556 as it does if it cannot open a log file.
6558 39. Modified OS/Makefile-Linux so that it now contains
6560 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6562 The two -D definitions ensure that Exim is compiled with large file
6563 support, which makes it possible to handle log files that are bigger than
6566 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6567 instance) a domain was checked against a named list that involved a lookup,
6568 causing $domain_data to be set, then another domain was checked against the
6569 same list, then the first domain was re-checked, the value of $domain_data
6570 after the final check could be wrong. In particular, if the second check
6571 failed, it could be set empty. This bug probably also applied to
6574 41. The strip_trailing_dot option was not being applied to the address given
6575 with the -f command-line option.
6577 42. The code for reading a message's header from the spool was incrementing
6578 $received_count, but never initializing it. This meant that the value was
6579 incorrect (doubled) while delivering a message in the same process in which
6580 it was received. In the most common configuration of Exim, this never
6581 happens - a fresh exec is done - but it can happen when
6582 deliver_drop_privilege is set.
6584 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6585 it now includes up to 150 characters of the unexpected data in the log
6588 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6589 and building data strings. The size of both of these buffers was 10 000
6590 bytes - far larger than anybody would *ever* want, thought I. Needless to
6591 say, somebody hit the limit. I have increased the maximum line length to
6592 20 000 and the maximum data length of concatenated lines to 100 000. I have
6593 also fixed two bugs, because there was no checking on these buffers. Tsk,
6594 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6595 buffer is too small.
6597 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6598 lsearch lookups. Now it does.
6600 46. When parsing a route_list item in a manualroute router, a fixed-length
6601 buffer was used for the list of hosts. I made this 1024 bytes long,
6602 thinking that nobody would ever have a list of hosts that long. Wrong.
6603 Somebody had a whole pile of complicated expansion conditions, and the
6604 string was silently truncated, leading to an expansion error. It turns out
6605 that it is easier to change to an unlimited length (owing to other changes
6606 that have happened since this code was originally written) than to build
6607 structure for giving a limitation error. The length of the item that
6608 expands into the list of hosts is now unlimited.
6610 47. The lsearch lookup could not handle data where the length of text line was
6611 more than 4095 characters. Such lines were truncated, leading to shortened
6612 data being returned. It should now handle lines of any length.
6614 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6615 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6618 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6619 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6620 no longer happens in comment lines. A list of replacements is now placed
6621 at the head of all of the source files, except those whose only change is
6622 to replace PERL_COMMAND in the very first #! line.
6624 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6625 messages on the queue, with a bottom-up merge sort, using code contributed
6626 by Michael Haardt. This should make operations like -bp somewhat faster on
6627 large queues. It won't affect queue runners, except when queue_run_in_order
6630 51. Installed eximstats 1.31 in the distribution.
6632 52. Added support for SRV lookups to the dnslookup router.
6634 53. If an ACL referred to $message_body or $message_body_end, the value was not
6635 reset for any messages that followed in the same SMTP session.
6637 54. The store-handling optimization for building very long strings was not
6638 differentiating between the different store pools. I don't think this
6639 actually made any difference in practice, but I've tidied it.
6641 55. While running the routers to verify a sender address, $sender_address
6642 was still set to the sender address. This is wrong, because when routing to
6643 send a bounce to the sender, it would be empty. Therefore, I have changed
6644 it so that, while verifying a sender address, $sender_address is set to <>.
6645 (There is no change to what happens when verifying a recipient address.)
6647 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6648 target A or AAAA records (if not already returned) without resetting the
6649 qualify_single or search_parents options of the DNS resolver. These are
6650 inappropriate in this case because the targets of MX and SRV records must
6651 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6652 target that, when qualified, matched something in the local domain. These
6653 two options are now turned off when doing these lookups.
6655 57. It seems that at least some releases of Reiserfs (which does not have the
6656 concept of a fixed number of inodes) returns zero and not -1 for the
6657 number of available inodes. This interacted badly with check_spool_inodes,
6658 which assumed that -1 was the "no such thing" setting. What I have done is
6659 to check that the total number of inodes is greater than zero before doing
6660 the test of how many are available.
6662 58. When a "warn" ACL statement has a log_message modifier, the message is
6663 remembered, and not repeated. This is to avoid a lot of repetition when a
6664 message has many recipients that cause the same warning to be written.
6665 However, Exim was preserving the list of already written lines for an
6666 entire SMTP session, which doesn't seem right. The memory is now reset if a
6667 new message is started.
6669 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6670 debugging output unless log_rewrite was also set.
6672 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6673 of (int)(handle) when we know that handle contains (void *)(-1).
6675 61. The Exim daemon panic-logs an error return when it closes the incoming
6676 connection. However "connection reset by peer" seems to be common, and
6677 isn't really an error worthy of noting specially, so that particular error
6680 62. When Exim is trying to find all the local interfaces, it used to panic and
6681 die if the ioctl to get the interface flags failed. However, it seems that
6682 on at least one OS (Solaris 9) it is possible to have an interface that is
6683 included in the list of interfaces, but for which you get a failure error
6684 for this call. This happens when the interface is not "plumbed" into a
6685 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6686 failure of the "get flags" call assumes that the interface is down.
6688 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6689 makes life easier for people who are doing arithmetic on fields extracted
6690 from dates, where you often get leading zeros that should not be
6691 interpreted as octal.
6693 64. Added qualify_domain to the redirect router, to override the global
6696 65. If a pathologically long header line contained very many addresses (the
6697 report of this problem mentioned 10 000) and each of them was rewritten,
6698 Exim could use up a very large amount of memory. (It kept on making new
6699 copies of the header line as it rewrote, and never released the old ones.)
6700 At the expense of a bit more processing, the header rewriting function has
6701 been changed so that it no longer eats memory in this way.
6703 66. The generation of the Received: header has been moved from the time that a
6704 message starts to be received, to the time that it finishes. The timestamp
6705 in the Received: header should now be very close to that of the <= log
6706 line. There are two side-effects of this change:
6708 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6709 logged header lines no longer include the local Received: line, because
6710 it has not yet been created. The same applies to a copy of the message
6711 that is returned to a non-SMTP sender when a message is rejected.
6713 (b) When a filter file is tested using -bf, no additional Received: header
6714 is added to the test message. After some thought, I decided that this
6717 This change does not affect the value of $received_for. It is still set
6718 after address rewriting, but before local_scan() is called.
6720 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6722 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6723 gave an unhelpful panic error message, and a defer error. I have managed to
6724 change this behaviour so that it now rejects any supplied certificate,
6725 which seems right, as the list of acceptable certificates is empty.
6727 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6728 gave an unhelpful defer error. I have not managed to make this reject any
6729 supplied certificates, but the error message it gives is "no certificate
6730 supplied", which is not helpful.
6732 70. exigrep's output now also includes lines that are not associated with any
6733 message, but which match the given pattern. Implemented by a patch from
6734 Martin Sluka, which also tidied up the Perl a bit.
6736 71. Recipient callout verification, like sender verification, was using <> in
6737 the MAIL FROM command. This isn't really the right thing, since the actual
6738 sender may affect whether the remote host accepts the recipient or not. I
6739 have changed it to use the actual sender in the callout; this means that
6740 the cache record is now keyed on a recipient/sender pair, not just the
6741 recipient address. There doesn't seem to be a real danger of callout loops,
6742 since a callout by the remote host to check the sender would use <>.
6743 [SEE ABOVE: changed after hitting problems.]
6745 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6746 temporary errors. However, in the case of such a code being given after
6747 the end of a data transmission (i.e. after ".") Exim was failing to write
6748 a retry record for the message. (Yes, there was some broken host that was
6749 actually sending 8xx at this point.)
6751 73. An unknown lookup type in a host list could cause Exim to panic-die when
6752 the list was checked. (An example that provoked this was putting <; in the
6753 middle of a list instead of at the start.) If this happened during a DATA
6754 ACL check, a -D file could be left lying around. This kind of configuration
6755 error no longer causes Exim to die; instead it causes a defer error. The
6756 incident is still logged to the main and panic logs.
6758 74. Buglet left over from Exim 3 conversion. The message "too many messages
6759 in one connection" was written to the rejectlog but not the mainlog, except
6760 when address rewriting (yes!) was being logged.
6762 75. Added write_rejectlog option.
6764 76. When a system filter was run not as root (that is, when system_filter_user
6765 was set), the values of the $n variables were not being returned to the
6766 main process; thus, they were not subsequently available in the $sn
6769 77. Added +return_path_on_delivery log selector.
6771 78. A connection timeout was being treated differently from recipients deferred
6772 when testing hosts_max_try with a message that was older than the host's
6773 retry timeout. (The host should not be counted, thus allowing all hosts to
6774 be tried at least once before bouncing.) This may have been the cause of an
6775 occasionally reported bug whereby a message would remain on the queue
6776 longer than the retry timeout, but would be bounced if a delivery was
6777 forced. I say "may" because I never totally pinned down the problem;
6778 setting up timeout/retry tests is difficult. See also the next item.
6780 79. The ultimate address timeout was not being applied to errors that involved
6781 a combination of host plus message (for example, a timeout on a MAIL
6782 command). When an address resolved to a number of possible hosts, and they
6783 were not all tried for each delivery (e.g. because of hosts_max_try), a
6784 message could remain on the queue longer than the retry timeout.
6786 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6789 81. Fixed an obscure SMTP outgoing bug which required at least the following
6790 conditions: (a) there was another message waiting for the same server;
6791 (b) the server returned 5xx to all RCPT commands in the first message so
6792 that the message was not completed; (c) the server dropped the connection
6793 or gave a negative response to the RSET that Exim sends to abort the
6794 transaction. The observed case was a dropped connection after DATA that had
6795 been sent in pipelining mode. That is, the server had advertised PIPELINING
6796 but was not implementing it correctly. The effect of the bug was incorrect
6797 behaviour, such as trying another host, and this could lead to a crash.
6803 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6804 and daemon.c were passed as pointers to ints; they should have been
6805 pointers to socklen_t variables (which are typically unsigned ints).
6807 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6810 3. Fixed a really odd bug that affected only the testing scheme; patching a
6811 certain fixed string in the binary changed the value of another string that
6812 happened to be identical to the end of the original first string.
6814 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6815 name", it returns that address as the IP address. On some operating
6816 systems (e.g. Solaris), it also passes back the IP address string as the
6817 "host name". However, on others (e.g. Linux), it passes back an empty
6818 string. Exim wasn't checking for this, and was changing the host name to an
6819 empty string, assuming it had been canonicalized.
6821 5. Although rare, it is permitted to have more than one PTR record for a given
6822 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6823 all the names associated with an address, because they do in Solaris.
6824 However, it seems that they do not in Linux for data that comes from the
6825 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6826 I found this out when I moved to a new Linux workstation and tried to run
6827 the Exim test suite.
6829 To get round this problem I have changed the code so that it now does its
6830 own call to the DNS to look up PTR records when searching for a host name.
6831 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6832 addresses that are only in /etc/hosts are still found.
6834 This behaviour is, however, controlled by an option called host_lookup_
6835 order, which defaults to "bydns:byaddr". If people want to use the other
6836 order, or indeed, just use one or the other means of lookup, they can
6837 specify it in this variable.
6839 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6840 some operating systems, this comes back from gethostbyaddr() as an empty
6841 string, and this is what Exim used to test for. However, it seems that in
6842 other systems, "." is yielded. Exim now tests for this case too.
6844 7. The values of check_spool_space and check_log_space are now held internally
6845 as a number of kilobytes instead of an absolute number of bytes. If a
6846 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6847 kilobyte. This means that much larger values can be stored.
6849 8. Exim monitor: an attempt to get the action menu when not actually pointing
6850 at a message produces an empty menu entitled "No message selected". This
6851 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6852 no entries in it ("Shell widget menu has zero width and/or height"). So I
6853 have added a single, blank menu entry in this case.
6855 9. Added ${quote_local_part.
6857 10. MIME decoding is now applied to the contents of Subject: header lines when
6860 11. Now that a reference to $sender_host_address automatically causes a reverse
6861 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6862 host lookup before query-style lookups in lists that might use this
6863 variable. This has therefore been abolished, and the "net-" prefix is no
6864 longer necessary for query-style lookups.
6866 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6867 have been a typo for LFLAGS, so it has been changed.
6869 13. The install script calls Exim with "-C /dev/null" in order to find the
6870 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6871 to be output. However, since Exim outputs its version number before the
6872 error, it didn't break the script. It just looked ugly. I fixed this by
6873 always allowing "-C /dev/null" if the caller is root.
6875 14. Ignore overlarge ACL variable number when reading spool file - insurance
6876 against a later release with more variables having written the file.
6878 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6879 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6880 this, as well as the form without the "IPv6" on the front.
6882 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6883 OS/Makefile-Darwin file.
6885 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6886 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6888 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6889 variable. However, some OS use ints for inodes. I've added cast to long int
6890 to get rid of the compiler warning.
6892 19. I had forgotten to lock out "/../" in configuration file names when
6893 ALT_CONFIG_PREFIX was set.
6895 20. Routers used for verification do not need to specify transports. However,
6896 if such a router generated a host list, and callout was configured, Exim
6897 crashed, because it could not find a port number from the (non-existent)
6898 transport. It now assumes port 25 in this circumstance.
6900 21. Added the -t option to exigrep.
6902 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6903 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6904 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6905 them is set, LOOKUP_LSEARCH is forced.
6907 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6909 24. Added sender and host information to the "rejected by local_scan()" log
6910 line; previously there was no indication of these.
6912 25. Added .include_if_exists.
6914 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6915 Linux. It turns out that not all file systems support this. Apparently some
6916 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6917 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6918 not supported on the file descriptor, is now ignored when Exim is trying to
6919 sync a directory. This applies only to Linux.
6921 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6923 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6924 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6927 29. Added support for maildirsize files from supplied patch (modified a bit).
6929 30. The use of :fail: followed by an empty string could lead Exim to respond to
6930 sender verification failures with (e.g.):
6932 550 Verification failed for <xxx>
6933 550 Sender verify failed
6935 where the first response line was missing the '-' that indicates it is not
6936 the final line of the response.
6938 31. The loop for finding the name of the user that called Exim had a hardwired
6939 limit of 10; it now uses the value of finduser_retries, which is used for
6940 all other user lookups.
6942 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6945 33. Exim was neglecting to zero errno before one call of strtol() when
6946 expanding a string and expecting an integer value. On some systems this
6947 resulted in spurious "integer overflow" errors. Also, it was casting the
6948 result into an int without checking.
6950 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6951 did not work. The code looks as if it has *never* worked, though it appears
6952 to have been documented since at least release 1.62. I have made it work.
6954 35. The "timeout_DNS" error in retry rules, also documented since at least
6955 1.62, also never worked. As it isn't clear exactly what this means, and
6956 clearly it isn't a major issue, I have abolished the feature by treating it
6957 as "timeout", and writing a warning to the main and panic logs.
6959 36. The display of retry rules for -brt wasn't always showing the error code
6962 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6963 timeout_connect_A, timeout_connect_MX.
6965 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6966 to the empty sender.
6968 39. The daemon was not analysing the content of -oX till after it had closed
6969 stderr and disconnected from the controlling terminal. This meant that any
6970 syntax errors were only noted on the panic log, and the return code from
6971 the command was 0. By re-arranging the code a little, I've made the
6972 decoding happen first, so such errors now appear on stderr, and the return
6973 code is 1. However, the actual setting up of the sockets still happens in
6974 the disconnected process, so errors there are still only recorded on the
6977 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6978 connections (as happens on some IP stacks) was logged at start up time as
6979 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6980 it from "IPv6 and IPv4", which means that two separate sockets are being
6983 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6984 says whether AF_INET or AF_INET6 was passed as an argument.
6986 42. Exiwhat output was messed up when time zones were included in log
6989 43. Exiwhat now gives more information about the daemon's listening ports,
6990 and whether -tls-on-connect was used.
6992 44. The "port" option of the smtp transport is now expanded.
6994 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6995 silently ignored. Now an error message is written to the main and panic
6998 46. There's a new ACL modifier called "logwrite" which writes to a log file
6999 as soon as it is encountered.
7001 47. Added $local_user_uid and $local_user_gid at routing time.
7003 48. Exim crashed when trying to verify a sender address that was being
7006 49. Exim was recognizing only a space character after ".include". It now also
7007 recognizes a tab character.
7009 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7010 extracting the relevant information from the specification. The man page no
7011 longer contains scrambled data for the -d option, and I've added a section
7012 at the front about calling Exim under different names.
7014 51. Added "extra_headers" argument to the "mail" command in filter files.
7016 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7019 53. Installed eximstats 1.29.
7021 54. Added transport_filter_timeout as a generic transport option.
7023 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7024 Cc: header lines. This was required by RFC 822, but it not required by RFC
7027 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7028 incoming messages that did not have them. Now it does so only if the
7029 message originates locally, that is, if there is no associated remote host
7030 address. When Resent- header lines are present, this applies to the Resent-
7031 lines rather than the non-Resent- lines.
7033 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7034 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7036 58. Messages for configuration errors now include the name of the main
7037 configuration file - useful now that there may be more than one file in a
7038 list (.included file names were always shown).
7040 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7041 for those rare installations that do not start the daemon as root or run it
7042 setuid root. I've cut out the call to initgroups() if the daemon is not
7045 60. The Exim user and group can now be bound into the binary as text strings
7046 that are looked up at the start of Exim's processing.
7048 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7050 62. Added $mailstore_basename variable.
7052 63. Installed patch to sieve.c from Michael Haardt.
7054 64. When Exim failed to open the panic log after failing to open the main log,
7055 the original message it was trying to log was written to stderr and debug
7056 output, but if they were not available (the usual case in production), it
7057 was lost. Now it is written to syslog before the two lines that record the
7058 failures to open the logs.
7060 65. Users' Exim filters run in subprocesses under the user's uid. It is
7061 possible for a "deliver" command or an alias in a "personal" command to
7062 provoke an address rewrite. If logging of address rewriting is configured,
7063 this fails because the process is not running as root or exim. There may be
7064 a better way of dealing with this, but for the moment (because 4.30 needs
7065 to be released), I have disabled address rewrite logging when running a
7066 filter in a non-root, non-exim process.
7072 1. The buildconfig auxiliary program wasn't quoting the value set for
7073 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7074 not defined. This bug was masked in 4.22 by the effect that was fixed in
7077 2. Some messages that were rejected after a message id was allocated were
7078 shown as "incomplete" by exigrep. It no longer does this for messages that
7079 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7081 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7082 have allow_domain_literals set, the ID did not get logged in the <= line.
7083 Domain literals are now always recognized in Message-ID: header lines.
7085 4. The first argument for a ${extract expansion item is the key name or field
7086 number. Leading and trailing spaces in this item were not being ignored,
7087 causing some misleading effects.
7089 5. When deliver_drop_privilege was set, single queue runner processes started
7090 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7091 same command in the process it spins off) were not dropping privilege.
7093 6. When the daemon running as "exim" started a queue runner, it always
7094 re-executed Exim in the spun-off process. This is a waste of effort when
7095 deliver_drop_privilege is set. The new process now just calls the
7096 queue-runner function directly.
7102 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7105 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7106 ignored. Though the use of -oP was forcing the writing of a pid file, it
7107 was always written to the default place.
7109 3. If the message "no IP address found for host xxxx" is generated during
7110 incoming verification, it is now followed by identification of the incoming
7111 connection (so you can more easily find what provoked it).
7113 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7115 5. Added some features to "harden" Exim a bit more against certain attacks:
7117 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7118 be put in Local/Makefile. This is like the never_users runtime option,
7119 but it cannot be overridden. The default setting is "root".
7121 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7122 prefix string with which any file named in a -C command line option
7125 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7126 is retained for -C and -D only if the caller of Exim is root. Without
7127 it, the exim user may also use -C and -D and retain privilege.
7129 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7130 command line option is disabled.
7132 6. Macro names set by the -D option must start with an upper case letter, just
7133 like macro names defined in the configuration file.
7135 7. Added "dereference=" facility to LDAP.
7137 8. Two instances of the typo "uknown" in the source files are fixed.
7139 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7140 the Configure-Makefile script screwed up while processing it.
7142 10. Incorporated PCRE 4.4.
7144 11. The SMTP synchronization check was not operating right at the start of an
7145 SMTP session. For example, it could not catch a HELO sent before the client
7146 waited for the greeting. There is now a check for outstanding input at the
7147 point when the greeting is written. Because of the duplex, asynchronous
7148 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7149 way, but not yet received, when the check is performed.
7151 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7152 on TCP/IP sockets, because this apparently causes some broken clients to
7155 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7156 unchanged) from the Cygwin maintainer.
7158 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7159 was supported instead of testing for mbx. Effectively a typo.
7161 15. The spa authenticator server code was not checking that the input it
7162 received was valid base64.
7164 16. The debug output line for the "set" modifier in ACLs was not showing the
7165 name of the variable that was being set.
7167 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7169 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7170 lookup on its own. Something else was needed to trigger the lookup. For
7171 example, a match in host_lookup or the need for a host name in a host list.
7172 Now, if $sender_host_name is referenced and the host name has not yet been
7173 looked up, a lookup is performed. If the lookup fails, the variable remains
7174 empty, and $host_lookup_failed is set to "1".
7176 19. Added "eqi" as a case-independent comparison operator.
7178 20. The saslauthd authentication condition could segfault if neither service
7179 nor realm was specified.
7181 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7182 error message that was logged was misleading, and incoming SMTP
7183 connections were dropped. The message is now more accurate, and temporary
7184 errors are given to SMTP connections.
7186 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7187 (except RSET) with a 421 error, until QUIT is received. However, it was
7188 failing to send a response to QUIT.
7190 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7191 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7192 ACL is now run after the helo_try_verify_hosts code.
7194 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7195 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7196 case-independent, so other case variants are also recognized). Apparently
7197 some systems use these upper case variants.
7199 25. If more than two messages were waiting for the same host, and a transport
7200 filter was specified for the transport, Exim sent two messages over the
7201 same TCP/IP connection, and then failed with "socket operation on non-
7202 socket" when it tried to send the third.
7204 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7206 27. The extern definition of crypt16() in expand.c was not being excluded when
7207 the OS had its own crypt16() function.
7209 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7210 as a preferred synonym for return_size_limit, both as an option and as an
7213 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7215 30. Changed the default configuration ACL to relax the local part checking rule
7216 for addresses that are not in any local domains. For these addresses,
7217 slashes and pipe symbols are allowed within local parts, but the sequence
7218 /../ is explicitly forbidden.
7220 31. SPA server authentication was not clearing the challenge buffer before
7223 32. log_message in a "warn" ACL statement was writing to the reject log as
7224 well as to the main log, which contradicts the documentation and doesn't
7225 seem right (because no rejection is happening). So I have stopped it.
7227 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7228 However, I am unable to do any testing of this.
7230 34. Fixed an infelicity in the appendfile transport. When checking directories
7231 for a mailbox, to see if any needed to be created, it was accidentally
7232 using path names with one or more superfluous leading slashes; tracing
7233 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7235 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7236 contents are added to the log line that is written for every discarded
7237 recipient. (Previously a log_message setting was ignored.)
7239 36. The ${quote: operator now quotes the string if it is empty.
7241 37. The install script runs exim in order to find its version number. If for
7242 some reason other than non-existence or emptiness, which it checks, it
7243 could not run './exim', it was installing it with an empty version number,
7244 i.e. as "exim-". This error state is now caught, and the installation is
7247 38. An argument was missing from the function that creates an error message
7248 when Exim fails to connect to the socket for saslauthd authentication.
7249 This could cause Exim to crash, or give a corrupted message.
7251 39. Added isip, isip4, and isip6 to ${if conditions.
7253 40. The ACL variables $acl_xx are now saved with the message, and can be
7254 accessed later in routers, transports, and filters.
7256 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7257 strings in the file are not string-expanded.
7259 42. If a MAIL command specified a SIZE value that was too large to fit into an
7260 int variable, the check against message_size_limit failed. Such values are
7261 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7262 day this will have to be increased, but I don't think I want to be around
7263 when emails are that large.
7270 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7271 iconv() is not standard in FreeBSD.
7273 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7274 IPv6 enabled. The observed symptom was a segmentation fault on return from
7275 the function os_common_find_running_interfaces() in src/os.c.
7277 3. In the check_special_case() function in daemon.c I had used "errno" as an
7278 argument name, which causes warnings on some systems. This was basically a
7279 typo, since it was named "eno" in the comments!
7281 4. The code that waits for the clock to tick (at a resolution of some fraction
7282 of a second) so as to ensure message-id uniqueness was always waiting for
7283 at least one whole tick, when it could have waited for less. [This is
7284 almost certainly not relevant at current processor speeds, where it is
7285 unlikely to ever wait at all. But we try to future-proof.]
7287 5. The function that sleeps for a time interval that includes fractions of a
7288 second contained a race. It did not block SIGALRM between setting the
7289 timer, and suspending (a couple of lines later). If the interval was short
7290 and the sigsuspend() was delayed until after it had expired, the suspension
7291 never ended. On busy systems this could lead to processes getting stuck for
7294 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7295 process, before it forks any delivery processes. The open lookup caching
7296 mechanism meant that the open file or database connection was passed into
7297 the delivery process. The problem was that delivery processes always tidy
7298 up cached lookup data. This could cause a problem for the next delivery
7299 process started by the queue runner, because the external queue runner
7300 process does not know about the closure. So the next delivery process
7301 still has data in the lookup cache. In the case of a file lookup, there was
7302 no problem because closing a file descriptor in a subprocess doesn't affect
7303 the parent. However, if the lookup was caching a connection to a database,
7304 the connection was closed, and the second delivery process was likely to
7305 see errors such as "PGSQL: query failed: server closed the connection
7306 unexpectedly". The problem has been fixed by closing all cached lookups
7307 in a queue runner before running a delivery process.
7309 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7310 seem to have the "const" qualifier which it has on other OS. I've
7313 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7314 *of the same type* (client or server) with the same public name that an
7315 error should be diagnosed.
7317 9. When Exim looked up a host name for an IP address, but failed to find the
7318 original IP address when looking up the host name (a safety check), it
7319 output the message "<ip address> does not match any IP for NULL", which was
7320 confusing, to say the least. The bug was that the host name should have
7321 appeared instead of "NULL".
7323 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7324 user that is built into the binary, and the -C or -D options is used, root
7325 privilege is dropped before the configuration file is read. In addition,
7326 logging is switched to stderr instead of the normal log files. If the
7327 configuration then re-defines the Exim user, the unprivileged environment
7328 is probably not what is expected, so Exim logs a panic warning message (but
7331 However, if deliver_drop_privilege is set, the unprivileged state may well
7332 be exactly what is intended, so the warning has been cut out in that case,
7333 and Exim is allowed to try to write to its normal log files.
7339 1. smtp_return_error_details was not giving details for temporary sender
7340 or receiver verification errors.
7342 2. Diagnose a configuration error if two authenticators have the same public
7345 3. Exim used not to create the message log file for a message until the first
7346 delivery attempt. This could be confusing when incoming messages were held
7347 for policy or load reasons. The message log file is now created at the time
7348 the message is received, and an initial "Received" line is written to it.
7350 4. The automatically generated man page for command line options had a minor
7351 bug that caused no ill effects; however, a more serious problem was that
7352 the procedure for building the man page automatically didn't always
7353 operate. Consequently, release 4.20 contains an out-of-date version. This
7354 shouldn't happen again.
7356 5. When building Exim with embedded Perl support, the script that builds the
7357 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7358 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7360 6. The freeze_tell option was not being used for messages that were frozen on
7361 arrival, either by an ACL or by local_scan().
7363 7. Added the smtp_incomplete_transaction log selector.
7365 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7366 was accepting AUTH without a new EHLO.
7368 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7369 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7372 10. Logging of TCP/IP connections (when configured) now happens in the main
7373 daemon process instead of the child process, so that the TCP/IP connection
7374 count is more accurate (but it can never be perfect).
7376 11. The use of "drop" in a nested ACL was not being handled correctly in the
7377 outer ACL. Now, if condition failure induced by the nested "drop" causes
7378 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7379 or "require"), the connection is dropped.
7381 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7382 that yield "discard" can now be used with an "accept" or a "discard" verb,
7383 but an error is generated for any others (because I can't see a useful way
7384 to define what should happen).
7386 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7387 are now processed in the same way as lines in the Exim configuration file.
7388 In particular, continuation lines are supported.
7390 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7392 15. Added -ti meaning -t -i.
7394 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7395 domains, and warn by logging if others are found.
7397 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7398 returned from GIFCONF when Exim is trying to find the list of interfaces on
7399 a host. The code in os.c has been modified to copy each ifreq to an aligned
7400 structure in all cases.
7402 Also, in some cases, the returned ifreq's were being copied to a 'struct
7403 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7404 means the last couple of bytes of an IPv6 address could be chopped if the
7405 ifreq contained only a normal sockaddr (14 bytes storage).
7407 18. Named domain lists were not supported in the hosts_treat_as_local option.
7408 An entry such as +xxxx was not recognized, and was treated as a literal
7411 19. Ensure that header lines added by a DATA ACL are included in the reject log
7412 if the ACL subsequently rejects the message.
7414 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7415 MD5 (which is deprecated).
7417 21. When testing a filter file using -bf, Exim was writing a message when it
7418 took the sender from a "From " line in the message, but it was not doing so
7419 when it took $return_path from a Return-Path: header line. It now does.
7421 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7422 with a valid header line field name (a series of printing characters
7423 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7425 23. Changed "disc" in the source to "disk" to conform to the documentation and
7426 the book and for uniformity.
7428 24. Ignore Sendmail's -Ooption=value command line item.
7430 25. When execve() failed while trying to run a command in a pipe transport,
7431 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7432 could be confused with a return value of 69 from the command itself. This
7433 has been changed to 127, the value the shell returns if it is asked to run
7434 a non-existent command. The wording for the related log line suggests a
7435 non-existent command as the problem.
7437 26. If received_header_text expands to an empty string, do not add a Received:
7438 header line to the message. (Well, it adds a token one on the spool, but
7439 marks it "old" so that it doesn't get used or transmitted.)
7441 27. Installed eximstats 1.28 (addition of -nt option).
7443 28. There was no check for failure on the call to getsockname() in the daemon
7444 code. This can fail if there is a shortage of resources on the system, with
7445 ENOMEM, for example. A temporary error is now given on failure.
7447 29. Contrary to the C standard, it seems that in some environments, the
7448 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7449 program. Exim now does this explicitly; it affects the formatting of
7450 timestamps using strftime().
7452 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7453 complaints. I've now initialized all the variables, to avoid this.
7455 32. Header lines added by a system filter were not being "seen" during
7456 transport-time rewrites.
7458 33. The info_callback() function passed to OpenSSL is set up with type void
7459 (*)(SSL *, int, int), as described somewhere. However, when calling the
7460 function (actually a macro) that sets it up, the type void(*)() is
7461 expected. I've put in a cast to prevent warnings from picky compilers.
7463 34. If a DNS black list lookup found a CNAME record, but there were no A
7464 records associated with the domain it pointed at, Exim crashed.
7466 35. If a DNS black list lookup returned more than one A record, Exim ignored
7467 all but the first. It now scans all returned addresses if a particular IP
7468 value is being sought. In this situation, the contents of the
7469 $dnslist_value variable are a list of all the addresses, separated by a
7472 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7473 to accept a host name and all its aliases if the forward lookup for any of
7474 them yielded the IP address of the incoming connection. Now it accepts only
7475 those names whose forward lookup yields the correct IP address. Any other
7476 names are discarded. This closes a loophole whereby a rogue DNS
7477 administrator could create reverse DNS records to break through a
7478 wildcarded host restriction in an ACL.
7480 37. If a user filter or a system filter that ran in a subprocess used any of
7481 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7482 the wrong values were passed to the pipe command ($thisaddress had the
7483 value of $0, $0 had the value of $1, etc). This bug was introduced by
7484 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7486 38. Improved the line breaking for long SMTP error messages from ACLs.
7487 Previously, if there was no break point between 40 and 75 characters, Exim
7488 left the rest of the message alone. Two changes have been made: (a) I've
7489 reduced the minimum length to 35 characters; (b) if it can't find a break
7490 point between 35 and 75 characters, it looks ahead and uses the first one
7491 that it finds. This may give the occasional overlong line, but at least the
7492 remaining text gets split now.
7494 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7495 file descriptors might be low, and that setting 1000 would always raise it.
7496 It turns out that in some environments, the limit is already over 1000 and
7497 that lowering it causes trouble. So now Exim takes care not to decrease it.
7499 40. When delivering a message, the value of $return_path is set to $sender_
7500 address at the start of routing (routers may change the value). By an
7501 oversight, this default was not being set up when an address was tested by
7502 -bt or -bv, which affected the outcome if any router or filter referred to
7505 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7506 log only when "message" or "log_message" are set. However, if one of the
7507 conditions was an address verification, or a call to a nested ACL, the
7508 messages generated by the underlying test were being passed through. This
7509 no longer happens. The underlying message is available in $acl_verify_
7510 message for both "message" and "log_message" expansions, so it can be
7511 passed through if needed.
7513 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7514 new expansion $bh_ to give the encoded byte string without charset
7515 translation. Translation happens only if iconv() is available; HAVE_ICONV
7516 indicates this at build time. HEADERS_CHARSET gives the charset to
7517 translate to; headers_charset can change it in the configuration, and
7518 "headers charset" can change it in an individual filter file.
7520 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7521 that creates RFC 2047 encoded "words" labels them as that charset instead
7522 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7523 expansion operator; (ii) when Exim creates a From: line for a local
7524 message; (iii) when a header line is rewritten to include a "phrase" part.
7526 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7527 buggy, causing it to skip the first lines of messages whose message ID
7528 ended in 'D'. This would not have bitten before Exim release 4.14, because
7529 message IDs were unlikely to end in 'D' before then. The effect was to have
7530 incorrect size information for certain domains.
7532 45. #include "config.h" was missing at the start of the crypt16.c module. This
7533 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7536 46. If there was a timeout during a "random" callout check, Exim treated it as
7537 a failure of the random address, and carried on sending RSET and the real
7538 address. If the delay was just some slowness somewhere, the response to the
7539 original RCPT would be taken as a response to RSET and so on, causing
7540 mayhem of various kinds.
7542 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7543 when I implemented it. It didn't allow for the fact that some option values
7544 may legitimately be negative (e.g. size_addition), and it didn't even do
7545 the right test for positive values.
7547 48. Domain names in DNS records are case-independent. Exim always looks them up
7548 in lower case. Some resolvers return domain names in exactly the case they
7549 appear in the zone file, that is, they may contain uppercase letters. Not
7550 all resolvers do this - some return always lower case. Exim was treating a
7551 change of case by a resolver as a change of domain, similar to a widening
7552 of a domain abbreviation. This triggered its re-routing code and so it was
7553 trying to route what was effectively the same domain again. This normally
7554 caused routing to fail (because the router wouldn't handle the domain
7555 twice). Now Exim checks for this case specially, and just changes the
7556 casing of the domain that it ultimately uses when it transmits the message
7559 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7562 50. If a filter generated a file delivery with a non-absolute name (possible if
7563 no home directory exists for the router), the forbid_file option was not
7566 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7567 the existing equality matching.
7569 52. Exim was using ints instead of ino_t variables in some places where it was
7570 dealing with inode numbers.
7572 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7573 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7574 called TMPDIR, and if it finds it is different, it changes its value.
7576 54. The smtp_printf() function is now made available to local_scan() so
7577 additional output lines can be written before returning. There is also an
7578 smtp_fflush() function to enable the detection of a dropped connection.
7579 The variables smtp_input and smtp_batched_input are exported to
7582 55. Changed the default runtime configuration: the message "Unknown user"
7583 has been removed from the ACL, and instead placed on the localuser router,
7584 using the cannot_route_message feature. This means that any verification
7585 failures that generate their own messages won't get overridden. Similarly,
7586 the "Unrouteable address" message that was in the ACL for unverifiable
7587 relay addresses has also been removed.
7589 56. Added hosts_avoid_esmtp to the smtp transport.
7591 57. The exicyclog script was not checking for the esoteric option
7592 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7593 will work only if exicyclog is run under the appropriate euid.
7595 58. Following a discussion on the list, the rules by which Exim recognises line
7596 endings on incoming messages have been changed. The -dropcr and drop_cr
7597 options are now no-ops, retained only for backwards compatibility. The
7598 following line terminators are recognized: LF CRLF CR. However, special
7599 processing applies to CR:
7601 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7602 nor a local message in the state where . is a terminator.
7604 (ii) If a bare CR is encountered in a header line, an extra space is added
7605 after the line terminator so as not to end the header. The reasoning
7606 behind this is that bare CRs in header lines are most likely either
7607 to be mistakes, or people trying to play silly games.
7609 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7610 was being incorrectly given as 18 bytes larger than it should have been.
7611 This is a VOB (very old bug).
7613 60. This may never have affected anything current, but just in case it has:
7614 When the local host is found other than at the start of a list of hosts,
7615 the local host, those with the same MX, and any that follow, are discarded.
7616 When the list in question was part of a longer list of hosts, the following
7617 hosts (not currently being processed) were also being discarded. This no
7618 longer happens. I'm not sure if this situation could ever has previously
7621 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7624 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7625 header. This is something that is recommended in a new Internet Draft, and
7626 is something that is documented as being done by Sendmail. There are two
7627 possible values. For messages generated by the autoreply transport, Exim
7630 Auto-Submitted: auto-replied
7632 whereas for all other generated messages (e.g. bounces) it adds
7634 Auto-Submitted: auto-generated
7636 63. The "personal" condition in filters now includes a test for the
7637 Auto-Submitted: header. If it contains the string "auto-" the message it
7638 not considered personal.
7640 64. Added rcpt_include_affixes as a generic transport option.
7642 65. Added queue_only_override (default true).
7644 66. Added the syslog_duplication option.
7646 67. If what should have been the first header line of a message consisted of
7647 a space followed by a colon, Exim was mis-interpreting it as a header line.
7648 It isn't of course - it is syntactically invalid and should therefore be
7649 treated as the start of the message body. The misbehaviour could have
7650 caused a number of strange effects, including loss of data in subsequent
7651 header lines, and spool format errors.
7653 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7654 client host had authenticated. This control can now be exercised by an ACL
7655 for more flexibility.
7657 69. By default, callouts do not happen when testing with -bh. There is now a
7658 variant, -bhc, which does actually run the callout code, including
7659 consulting and updating the callout cache.
7661 70. Added support for saslauthd authentication, courtesy of Alexander
7664 71. If statvfs() failed on the spool or log directories while checking their
7665 size for availability, Exim confusingly gave the error "space shortage".
7666 Furthermore, in debugging mode it crashed with a floating point exception.
7667 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7668 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7669 serious problem, Exim now writes to the main and panic logs when this
7670 happens, with details of the failure. It then refuses to accept the
7671 incoming message, giving the message "spool directory problem" or "log
7672 directory problem" with a 421 code for SMTP messages.
7674 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7675 0, 1, and 2 exist, because some OS complain for execs without them (see
7676 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7677 descriptors. However, the code omitted to check that the open succeeded,
7678 causing mysterious errors if for some reason the permissions on /dev/null
7679 got screwed. Now Exim writes a message to the main and panic logs, and
7680 bombs out if it can't open /dev/null.
7682 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7683 interact so that it is all more flexible. It is supposed to remain
7684 backwards compatible. Also added extra_local_interfaces.
7686 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7687 to bomb out with an assertion failure - to the client this appears as a
7688 connection drop. This problem occurs in the part of the code that was taken
7689 from the Samba project. Fortunately, the assertion is in a very simple
7690 function, so I have fixed this by reproducing the function inline in the
7691 one place where it is called, and arranging for authentication to fail
7692 instead of killing the process with assert().
7694 75. The SPA client code was not working when the server requested OEM rather
7695 than Unicode encoding.
7697 76. Added code to make require_files with a specific uid setting more usable in
7698 the case where statting the file as root fails - usually a non-root-mounted
7699 NFS file system. When this happens and the failure is EACCES, Exim now
7700 forks a subprocess and does the per-uid checking as the relevant uid.
7702 77. Added process_log_path.
7704 78. If log_file_path was not explicitly set, a setting of check_log_space or
7705 check_log_inodes was ignored.
7707 79. If a space check for the spool or log partitions fails, the incident is now
7708 logged. Of course, in the latter case the data may get lost...
7710 80. Added the %p formatting code to string_format() so that it can be used to
7711 print addresses in debug_print(). Adjusted all the address printing in the
7712 debugging in store.c to use %p rather than %d.
7714 81. There was a concern that a line of code in smtp_in.c could overflow a
7715 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7716 initially expressed, the concern was not well-founded, because trailing
7717 spaces are removed early. However, if the trailing spaces were followed by
7718 a NULL, they did not get removed, so the overflow was possible. Two fixes
7721 (a) I re-wrote the offending code in a cleaner fashion.
7722 (b) If an incoming SMTP command contains a NULL character, it is rejected
7725 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7726 runs initgroups(), so that if the Exim user is in any additional groups,
7727 they will be used during message reception.
7733 The change log for 4.20 and earlier releases has been archived.