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.
156 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
157 side, to decode a TLS 1.3 packet capture.
159 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
160 Previously the default library behaviour applied, sending two, each in
163 JH/03 Debug output for ACL now gives the config file name and line number for
166 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
168 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
170 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
171 buffer overrun for (non-chunking) other transports.
173 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
174 TLS1.3, means that a server rejecting a client certificate is not visible
175 to the client until the first read of encrypted data (typically the
176 response to EHLO). Add detection for that case and treat it as a failed
177 TLS connection attempt, so that the normal retry-in-clear can work (if
178 suitably configured).
180 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
181 and/or domain. Found and fixed by Jason Betts.
183 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
184 configuration). If a CNAME target was not a wellformed name pattern, a
187 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
188 the OS reports them interleaved with other addresses.
190 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
191 used both for input and for a verify callout, both encrypted, SMTP
192 responses being sent by the server could be lost. This resulted in
193 dropped connections and sometimes bounces generated by a peer sending
196 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
197 string. Previously it was possible to cause undefined behaviour in a
198 library routine (usually a crash). Found by "zerons".
200 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
203 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
204 API was removed, so update to use the newer ones.
206 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
207 any timeout set, is taking a long time. Previously we would hang on to a
208 rotated logfile "forever" if the input was arriving with long gaps
209 (a previous attempt to fix addressed lack, for a long time, of initial
212 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
213 shared (NFS) environment. The length of the tempfile name is now
214 4 + 16 ("hdr.$message_exim_id") which might break on file
215 systems which restrict the file name length to lower values.
218 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
219 shared (NFS) environment.
221 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
222 did for all versions <4.90). Notably -M, -m, --invert, -I may be
225 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
226 on some platforms for bit 31.
228 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
229 to changes apparently associated with TLS1.3 handling some of the APIs
230 previously used were either nonfunctional or inappropriate. Strings
231 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
232 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
233 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
234 This affects log line X= elements, the $tls_{in,out}_cipher variables,
235 and the use of specific cipher names in the encrypted= ACL condition.
237 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
239 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
240 verification result was not updated unless hosts_require_ocsp applied.
242 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
243 queue_list_requires_admin set to false, non-admin users were denied the
246 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
247 directory-of-certs mode. Previously they were advertised despite the
250 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
251 A single TCP connection by a client will now hold a TLS connection open
252 for multiple message deliveries, by default. Previoud the default was to
255 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
256 default. If built with the facility, DANE will be used. The facility
257 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
259 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
260 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
261 must be defined and you must still, unless you define DISABLE_TLS, manage
262 the the include-dir and library-file requirements that go with that
263 choice. Non-TLS builds are still supported.
265 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
268 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
269 default. If the platform supports and has the facility enabled, it will
270 be requested on all coneections.
272 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
273 controlled by the build-time option SUPPORT_PIPE_CONNECT.
275 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
277 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
278 success-DSN messages. Previously the From: header was always the default
279 one for these; the option was ignored.
281 JH/28 Fix the timeout on smtp response to apply to the whole response.
282 Previously it was reset for every read, so a teergrubing peer sending
283 single bytes within the time limit could extend the connection for a
284 long time. Credit to Qualsys Security Advisory Team for the discovery.
286 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
287 delivery address, which leaked information of the results of local
288 forwarding. Change to the original envelope recipient address, per
291 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
292 requested. Previously not bounce was generated and a log entry of
293 error ignored was made.
295 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
297 JH/32 Introduce a general tainting mechanism for values read from the input
298 channel, and values derived from them. Refuse to expand any tainted
299 values, to catch one form of exploit.
301 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
302 was unused and the unexpanded text used for the test. Found and
303 fixed by Ruben Jenster.
305 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
306 an attempt to use a TLS library read routine dereffed a nul pointer,
309 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
310 them in our smtp response.
312 JH/36 Have the general router option retry_use_local_part default to true when
313 any of the restrictive preconditions are set (to anything). Previously it
314 was only for check_local user. The change removes one item of manual
315 configuration which is required for proper retries when a remote router
316 handles a subset of addresses for a domain.
318 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
319 link count into consideration.
321 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
322 caused the extension of big_buffer, the following lines were ignored.
324 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
325 accordance with RFC 2308. Previously there was no expiry, so a longlived
326 receive process (eg. due to ACL delays) versus a short SOA value could
329 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
331 JH/39 Promote DMARC support to mainline.
333 JH/40 Bug 2452: Add a References: header to DSNs.
335 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
336 parameters. The relevant library call is documented as "Deprecated: This
337 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
338 3.6.0, DH parameters are negotiated following RFC7919."
340 HS/06 Change the default of dnssec_request_domains to "*"
342 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
343 carried on and emitted a BDAT command, even when PIPELINING was not
346 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
347 buffer was used for the filename, resulting in a trap when tainted
348 arguments (eg. $domain) were used.
350 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
351 recommended to avoid a possible server-load attack. The feature can be
352 re-enabled via the openssl_options main cofiguration option.
354 JH/45 local_scan API: documented the current smtp_printf() call. This changed
355 for version 4.90 - adding a "more data" boolean to the arguments.
356 Bumped the ABI version number also, this having been missed previously;
357 release versions 4.90 to 4.92.3 inclusive were effectively broken in
358 respect of usage of smtp_printf() by either local_scan code or libraries
359 accessed via the ${dlfunc } expansion item. Both will need coding
360 adjustment for any calls to smtp_printf() to match the new function
361 signature; a FALSE value for the new argument is always safe.
363 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
364 the file-offset (which the Linux syscall does, and exim expects); this
365 resulted in an indefinite loop.
367 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
368 to do ARC verification. The Authentication-Results: header line added
369 by the configuration then had no ARC item.
375 JH/01 Remove code calling the customisable local_scan function, unless a new
376 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
378 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
379 non-signal-safe functions being used.
381 JH/03 Bug 2269: When presented with a received message having a stupidly large
382 number of DKIM-Signature headers, disable DKIM verification to avoid
383 a resource-consumption attack. The limit is set at twenty.
385 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
386 report of oldest_pass in ${authres } in consequence, and separate out
387 some descriptions of reasons for verification fail.
389 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
390 files in the spool were present and unlocked. A queue-runner could spot
391 them, resulting in a duplicate delivery. Fix that by doing the unlock
392 after the unlink. Investigation by Tim Stewart. Take the opportunity to
393 add more error-checking on spoolfile handling while that code is being
396 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
397 No known attacks, no CVE, this is defensive hardening.
399 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
400 a queue-runner could start a delivery while other operations were ongoing.
401 Cutthrough delivery was a common victim, resulting in duplicate delivery.
402 Found and investigated by Tim Stewart. Fix by using the open message data
403 file handle rather than opening another, and not locally closing it (which
404 releases a lock) for that case, while creating the temporary .eml format
405 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
407 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
408 $sender_verify_failure/$recipient_verify_failure to "random".
410 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
413 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
414 Previously this would segfault.
416 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
419 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
420 like zero, since the resolver should be doing this for us, But we need one
421 as a CNAME but no MX presence gets the CNAME returned; we need to check
422 that doesn't point to an MX to declare it "no MX returned" rather than
423 "error, loop". A new main option is added so the older capability of
424 following some limited number of chain links is maintained.
426 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
428 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
429 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
430 now-deprecated earlier definitions used only specified the range up to TLS
431 1.2 (in the older-version library docs).
433 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
435 JH/15 Rework TLS client-side context management. Stop using a global, and
436 explicitly pass a context around. This enables future use of TLS for
437 connections to service-daemons (eg. malware scanning) while a client smtp
438 connection is using TLS; with cutthrough connections this is quite likely.
440 JH/16 Fix ARC verification to do AS checks in reverse order.
442 JH/17 Support a "tls" option on the ${readsocket } expansion item.
444 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
445 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
446 Previously the "utf8" would be re-prepended for every additional message.
448 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
449 Previously thery were accepted, resulting in issues when attempting to
450 forward messages to a non-supporting MTA.
452 PP/02 Let -n work with printing macros too, not just options.
454 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
455 one parent address was copied, and bogus data was used at delivery-logging
456 time. Either a crash (after delivery) or bogus log data could result.
457 Discovery and analysis by Tim Stewart.
459 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
460 Previously if the string ended mid-character, we did not insert the
461 promised '?' replacement.
463 PP/04 Documentation: current string operators work on bytes, not codepoints.
465 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
466 should pack well giving a smaller memory footprint so better caching and
467 therefore performance. Group the declarations where this can't be done so
468 that the byte-sized flag variables are not interspersed among pointer
469 variables, giving a better chance of good packing by the compiler.
471 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
472 non-null, to avoid issues with sites running BATV. Previously reports were
473 sent with an empty envelope sender so looked like bounces.
475 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
476 The ignore_error flag wasn't being returned from the filter subprocess so
477 was not set for later routers. Investigation and fix by Matthias Kurz.
479 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
480 and a msg:complete for the whole, when a message is manually removed using
481 -Mrm. Developement by Matthias Kurz, hacked on by JH.
483 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
484 a "Gnu special" function, asprintf() in the DB utility binary builds; I
485 hope that is portable enough.
487 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
488 requiring a known-CA anchor certificate; make it now rely entirely on the
489 TLSA as an anchor. Checking the name on the leaf cert against the name
490 on the A-record for the host is still done for TA (but not for EE mode).
492 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
493 included in delivery lines for non-proxied connections, when compiled with
494 SUPPORT_SOCKS and running with proxy logging enabled.
496 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
497 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
498 move the existing event to fire before the normal logging of message
499 failure so that custom logging is bracketed by normal logging.
501 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
502 msg:fail:internal event. Developement by Matthias Kurz.
504 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
505 far too small for todays use of crypto signatures stored there. Go all
506 the way to the max DNS message size of 64kB, even though this might be
507 overmuch for IOT constrained device use.
509 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
510 copy a string over itself. The library routine is documented as not
511 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
513 JH/32 For main options check_spool_space and check_inode_space, where the
514 platform supports 64b integers, support more than the previous 2^31 kB
515 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
516 the previous G, M, k.
518 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
519 $authenticated_fail_id variable on authentication failure. Previously
522 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
523 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
524 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
525 GNUTLS_SEC_PARAM_MEDIUM.
527 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
528 side. Previously we would continue as if no SNI had been received.
530 JH/36 Harden the handling of string-lists. When a list consisted of a sole
531 "<" character, which should be a list-separator specification, we walked
532 off past the nul-terimation.
534 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
535 causes) even when the retry time is not yet met. Previously they were
536 not, meaning that when (say) an account was over-quota and temp-rejecting,
537 and multiple senders' messages were queued, only one sender would get
538 notified on each configured delay_warning cycle.
540 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
542 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
543 cipher-suites, an error can be left on the stack even for a succeeding
544 accept; this results in impossible error messages when a later operation
547 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
548 return error codes indicating retry. Under TLS1.3 this becomes required.
550 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
551 it only wrote the new authenticators, resulting in a lack of tracking of
552 peer changes of ESMTP extensions until the next cache flush.
554 JH/41 Fix the loop reading a message header line to check for integer overflow,
555 and more-often against header_maxsize. Previously a crafted message could
556 induce a crash of the recive process; now the message is cleanly rejected.
558 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
559 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
565 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
566 When redis_servers is set to a list of > 1 element, and the Redis servers
567 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
568 case of MOVED into a DEFER case instead, thus moving the query onto the
569 next server in the list. For a cluster of N elements, all N servers must
570 be defined in redis_servers.
572 GF/02 Catch and remove uninitialized value warning in exiqsumm
573 Check for existence of @ARGV before looking at $ARGV[0]
575 JH/01 Replace the store_release() internal interface with store_newblock(),
576 which internalises the check required to safely use the old one, plus
577 the allocate and data copy operations duplicated in both (!) of the
578 extant use locations.
580 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
581 modifier. This matches the restriction on the commandline.
583 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
584 Previously only the last row was returned.
586 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
587 we assumed that tags in the header were well-formed, and parsed the
588 element content after inspecting only the first char of the tag.
589 Assumptions at that stage could crash the receive process on malformed
592 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
593 While running the DKIM ACL we operate on the Permanent memory pool so that
594 variables created with "set" persist to the DATA ACL. Also (at any time)
595 DNS lookups that fail create cache records using the Permanent pool. But
596 expansions release any allocations made on the current pool - so a dnsdb
597 lookup expansion done in the DKIM ACL releases the memory used for the
598 DNS negative-cache, and bad things result. Solution is to switch to the
599 Main pool for expansions.
600 While we're in that code, add checks on the DNS cache during store_reset,
601 active in the testsuite.
602 Problem spotted, and debugging aided, by Wolfgang Breyha.
604 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
605 When none of the hosts presented to a transport match an already-open
606 connection, close it and proceed with the list. Previously we would
607 queue the message. Spotted by Lena with Yahoo, probably involving
610 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
611 Previously a spurious "250 OK id=" response was appended to the proper
614 JH/08 The "support for" informational output now, which built with Content
615 Scanning support, has a line for the malware scanner interfaces compiled
616 in. Interface can be individually included or not at build time.
618 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
619 by the template makefile "src/EDITME". The "STREAM" support for an older
620 ClamAV interface method is removed.
622 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
623 rows affected is given instead).
625 JH/11 The runtime Berkeley DB library version is now additionally output by
626 "exim -d -bV". Previously only the compile-time version was shown.
628 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
629 SMTP connection. Previously, when one had more receipients than the
630 first, an abortive onward connection was made. Move to full support for
631 multiple onward connections in sequence, handling cutthrough connection
632 for all multi-message initiating connections.
634 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
635 routers. Previously, a multi-recipient message would fail to match the
636 onward-connection opened for the first recipient, and cause its closure.
638 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
639 a timeout on read on a GnuTLS initiating connection, resulting in the
640 initiating connection being dropped. This mattered most when the callout
641 was marked defer_ok. Fix to keep the two timeout-detection methods
644 JH/15 Relax results from ACL control request to enable cutthrough, in
645 unsupported situations, from error to silently (except under debug)
646 ignoring. This covers use with PRDR, frozen messages, queue-only and
649 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
651 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
652 metadata, resulting in a crash in free().
654 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
655 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
656 Broken also in d185889f4, with init system revamp.
658 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
659 Previously we abruptly closed the connection after reading a malware-
660 found indication; now we go on to read the "scan ok" response line,
663 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
664 ACL. Previously, a crash would result.
666 JH/19 Speed up macro lookups during configuration file read, by skipping non-
667 macro text after a replacement (previously it was only once per line) and
668 by skipping builtin macros when searching for an uppercase lead character.
670 JH/20 DANE support moved from Experimental to mainline. The Makefile control
671 for the build is renamed.
673 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
674 was allocated for every new TLS startup, meaning one per message. Fix
675 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
677 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
678 reported the original. Fix to report (as far as possible) the ACL
679 result replacing the original.
681 JH/23 Fix memory leak during multi-message connections using STARTTLS under
682 OpenSSL. Certificate information is loaded for every new TLS startup,
683 and the resources needed to be freed.
685 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
687 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
690 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
691 DATA response info to the (existing) per-recipient response info for
692 the "C=" log element. It can have useful tracking info from the
693 destination system. Patch from Simon Arlott.
695 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
696 length value. Previously this would segfault.
698 HS/02 Support Avast multiline protoocol, this allows passing flags to
699 newer versions of the scanner.
701 JH/28 Ensure that variables possibly set during message acceptance are marked
702 dead before release of memory in the daemon loop. This stops complaints
703 about them when the debug_store option is enabled. Discovered specifically
704 for sender_rate_period, but applies to a whole set of variables.
705 Do the same for the queue-runner and queue-list loops, for variables set
706 from spool message files. Do the same for the SMTP per-message loop, for
707 certain variables indirectly set in ACL operations.
709 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
710 as a multi-recipient message from a mailinglist manager). The coding had
711 an arbitrary cutoff number of characters while checking for more input;
712 enforced by writing a NUL into the buffer. This corrupted long / fast
713 input. The problem was exposed more widely when more pipelineing of SMTP
714 responses was introduced, and one Exim system was feeding another.
715 The symptom is log complaints of SMTP syntax error (NUL chars) on the
716 receiving system, and refused recipients seen by the sending system
717 (propating to people being dropped from mailing lists).
718 Discovered and pinpointed by David Carter.
720 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
721 replaced by the ${authresults } expansion.
723 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
725 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
726 allows proper process termination in container environments.
728 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
729 Previously the "final dot" had a newline after it; ensure it is CR,LF.
731 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
732 and "err_perm", deprecated since 4.83 when the RFC-defined words
733 "temperror" and "permerror" were introduced.
735 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
736 transport-filters or DKIM-signing. The restriction was lost in the
737 consolidation of verify-callout and delivery SMTP handling.
738 Extend the restriction to also cover ARC-signing.
740 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
741 in defer=pass mode supply a 450 to the initiator. Previously the message
744 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
745 tls_require_ciphers is used as before.
747 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
748 "pass_unscanned". Only tmpfails from the scanner are written to
749 the paniclog, as they may require admin intervention (permission
750 denied, license issues). Other scanner errors (like decompression
751 bombs) do not cause a paniclog entry.
753 JH/36 Fix reinitialisation of DKIM logging variable between messages.
754 Previously it was possible to log spurious information in receive log
757 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
758 triggered odd behaviour from Outlook Express clients.
760 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
763 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
764 since the IETF WG has not yet settled on that versus the original
765 "bare" representation.
767 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
768 Previously the millisecond value corrupted the output.
769 Fix also for syslog_pid=no and log_selector +pid, for which the pid
770 corrupted the output.
776 JH/01 Rework error string handling in TLS interface so that the caller in
777 more cases is responsible for logging. This permits library-sourced
778 string to be attached to addresses during delivery, and collapses
779 pairs of long lines into single ones.
781 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
782 during configuration. Wildcards are allowed and expanded.
784 JH/02 Rework error string handling in DKIM to pass more info back to callers.
785 This permits better logging.
787 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
788 do not close it down and have the child transport start it up again on
789 the passed-on TCP connection. Instead, proxy the child (and any
790 subsequent ones) for TLS via a unix-domain socket channel. Logging is
791 affected: the continued delivery log lines do not have any DNSSEC, TLS
792 Certificate or OCSP information. TLS cipher information is still logged.
794 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
795 identical IP addresses on different listening ports. Will also affect
798 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
799 add noisy ifdef guards to special-case this sillyness.
800 Patch from Bernd Kuhls.
802 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
803 than 255 are no longer allowed.
805 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
806 Disable the session-cache too, which might reduce our load. Since we
807 currrectly use a new context for every connection, both as server and
808 client, there is no benefit for these.
809 GnuTLS appears to not support tickets server-side by default (we don't
810 call gnutls_session_ticket_enable_server()) but client side is enabled
811 by default on recent versions (3.1.3 +) unless the PFS priority string
814 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
815 <https://reproducible-builds.org/specs/source-date-epoch/>.
817 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
818 the check for any unsuccessful recipients did not notice the limit, and
819 erroneously found still-pending ones.
821 JH/08 Pipeline CHUNKING command and data together, on kernels that support
822 MSG_MORE. Only in-clear (not on TLS connections).
824 JH/09 Avoid using a temporary file during transport using dkim. Unless a
825 transport-filter is involved we can buffer the headers in memory for
826 creating the signature, and read the spool data file once for the
827 signature and again for transmission.
829 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
830 4.77 as the kernel support then wasn't solid, having issues in 64bit
831 mode. Now, it's been long enough. Add support for FreeBSD also.
833 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
834 case where the routing stage had gathered several addresses to send to
835 a host before calling the transport for the first, we previously failed
836 to close down TLS in the old transport process before passing the TCP
837 connection to the new process. The new one sent a STARTTLS command
838 which naturally failed, giving a failed delivery and bloating the retry
839 database. Investigation and fix prototype from Wolfgang Breyha.
841 JH/12 Fix check on SMTP command input synchronisation. Previously there were
842 false-negatives in the check that the sender had not preempted a response
843 or prompt from Exim (running as a server), due to that code's lack of
844 awareness of the SMTP input buffering.
846 PP/04 Add commandline_checks_require_admin option.
847 Exim drops privileges sanely, various checks such as -be aren't a
848 security problem, as long as you trust local users with access to their
849 own account. When invoked by services which pass untrusted data to
850 Exim, this might be an issue. Set this option in main configuration
851 AND make fixes to the calling application, such as using `--` to stop
854 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
855 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
856 responses to those, into a single TLS record each way (this usually means
857 a single packet). As a side issue, smtp_enforce_sync now works on TLS
860 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
861 affects you only if you're dancing at the edge of the param size limits.
862 If you are, and this message makes sense to you, then: raise the
863 configured limit or use OpenSSL 1.1. Nothing we can do for older
866 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
867 cmdline element to get the documented default one. Previously it was
870 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
871 are unsuable for non-message-specific reasons - eg. connection timeout,
872 banner-time rejection.
874 JH/16 Fix logging of delivery remote port, when specified by router, under
877 PP/06 Repair manualroute's ability to take options in any order, even if one
878 is the name of a transport.
881 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
883 JH/17 Change the list-building routines interface to use the expanding-string
884 triplet model, for better allocation and copying behaviour.
886 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
887 Previously it was constructed the first time a possibly-matching string
888 was met in the configuration file input during startup; now it is done
891 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
892 compatible one, to avoid the (poorly documented) possibility of a config
893 file in the working directory redirecting the DB files, possibly correpting
894 some existing file. CVE-2017-10140 assigned for BDB.
896 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
897 cache-hot. Previously, although the result was properly cached, the
898 initial verify call returned a defer.
900 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
901 the main verify for receipient in uncached-mode.
903 JH/22 Retire historical build files to an "unsupported" subdir. These are
904 defined as "ones for which we have no current evidence of testing".
906 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
907 if present. Previously it was ignored.
909 JH/24 Start using specified-initialisers in C structure init coding. This is
910 a C99 feature (it's 2017, so now considered safe).
912 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
913 if was a fixed-sized field and bitmask ops via macros; it is now more
916 PP/07 GitHub PR 56: Apply MariaDB build fix.
917 Patch provided by Jaroslav Škarvada.
919 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
920 during Coverity cleanups [4.87 JH/47]
921 Diagnosis and fix provided by Michael Fischer v. Mollard.
923 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
924 the right size to place the terminating semicolon on its own folded
925 line, the header hash was calculated to an incorrect value thanks to
926 the (relaxed) space the fold became.
928 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
929 and confused the parent.
931 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
932 which could crash as a result. This could lead to undeliverable messages.
934 JH/28 Logging: "next input sent too soon" now shows where input was truncated
937 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
938 matters on fast-turnover and PID-randomising systems, which were getting
939 out-of-order delivery.
941 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
942 a possibly-overlapping copy. The symptom was that "Remote host closed
943 connection in response to HELO" was logged instead of the actual 4xx
946 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
947 Previously only that bufferd was discarded, resulting in SYMTP command
950 JH/32 DKIM: when a message has multiple signatures matching an identity given
951 in dkim_verify_signers, run the dkim acl once for each. Previously only
952 one run was done. Bug 2189.
954 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
955 "panic the current process" to "deliberately defer". The panic log is
956 still written with the problem list name; the mail and reject logs now
957 get a temp-reject line for the message that was being handled, saying
958 something like "domains check lookup or other defer". The SMTP 451
959 message is still "Temporary local problem".
961 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
962 A crafted sequence of BDAT commands could result in in-use memory beeing
963 freed. CVE-2017-16943.
965 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
966 from SMTP input. Previously it was always done; now only done for DATA
967 and not BDAT commands. CVE-2017-16944.
969 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
970 to the message (such as an overlong header line). Previously this was
971 not done and we did not exit BDAT mode. Followon from the previous item
972 though a different problem.
978 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
979 than -2003 did; needs libidn2 in addition to libidn.
981 JH/02 The path option on a pipe transport is now expanded before use.
983 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
984 Patch provided by "Björn", documentation fix added too.
986 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
987 missing a wire-to-host endian conversion.
989 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
990 close after a BDAT command line could be taken as a following command,
991 giving a synch failure. Fix by only checking for synch immediately
992 before acknowledging the chunk.
994 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
995 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
996 macro. Patches provided by Josh Soref.
998 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
999 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1000 by IANA as a service extension. However, John Klensin suggests that we
1003 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1004 the dkim code may be unix-mode line endings rather than smtp wire-format
1005 CRLF, so prepend a CR to any bare LF.
1007 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1008 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1010 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1011 the first were themselves being wrongly included in the feed into dkim
1012 processing; with most chunk sizes in use this resulted in an incorrect
1013 body hash calculated value.
1015 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1016 DKIM signature block, for verification. Although advised against by
1017 standards it is specifically not ruled illegal.
1019 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1021 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1022 missing a body hash (the bh= tag).
1024 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1025 It seems that HAProxy sends the Proxy Protocol information in clear and
1026 only then does a TLS startup, so do the same.
1028 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1029 TCP connections (such as for Spamd) unless the daemon successfully set
1030 Fast Open mode on its listening sockets. This fixes breakage seen on
1031 too-old kernels or those not configured for Fast Open, at the cost of
1032 requiring both directions being enabled for TFO, and TFO never being used
1033 by non-daemon-related Exim processes.
1035 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1036 endings, at least on the first header line. Try to canonify any that get
1037 past that check, despite the cost.
1039 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1040 now limited to an arbitrary five deep, while parsing addresses with the
1041 strip_excess_angle_brackets option enabled.
1043 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1044 instead leave the unprompted TLS handshake in socket buffer for the
1045 TLS library to consume.
1047 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1049 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1051 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1052 region in one operation, for speed, and this leaves assigned pointers
1053 dangling. Add checks run only under the testsuite which checks all
1054 variables at a store-reset and panics on a dangling pointer; add code
1055 explicitly nulling out all the variables discovered. Fixes one known
1056 bug: a transport crash, where a dangling pointer for $sending_ip_address
1057 originally assigned in a verify callout, is re-used.
1059 PP/06 Drop '.' from @INC in various Perl scripts.
1061 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1063 PP/08 Reduce a number of compilation warnings under clang; building with
1064 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1065 should be warning-free.
1067 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1069 HS/01 Fix portability problems introduced by PP/08 for platforms where
1070 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1072 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1073 chunk. This allows us to accept broken chunked messages. We need a more
1074 general solution here.
1076 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1077 already-broken messages in the queue.
1079 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1081 JH/19 Fix reference counting bug in routing-generated-address tracking.
1087 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1088 supports it and a size is available (ie. the sending peer gave us one).
1090 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1091 years of being deprecated). The replacements are the ACLs
1092 acl_smtp_mime and acl_not_smtp_mime.
1094 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1095 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1096 or even an in-clear connection were permitted. Now, if the host lookup
1097 was dnssec and dane was requested then the host is only used if the
1098 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1099 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1100 if one fails this test.
1101 This means that a poorly-configured remote DNS will make it incommunicado;
1102 but it protects against a DNS-interception attack on it.
1104 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1105 non-noisy when a race steals the message being considered.
1107 JH/05 If main configuration option tls_certificate is unset, generate a
1108 self-signed certificate for inbound TLS connections.
1110 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1111 in rewrites and routers.
1113 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1114 and logged a warning sing 4.83; now they are a configuration file error.
1116 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1117 (lacking @domain). Apply the same qualification processing as RCPT.
1119 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1121 JH/10 Support ${sha256:} applied to a string (as well as the previous
1124 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1125 a cutthrough deliver is pending, as we always want to make a connection.
1126 This also avoids re-routing the message when later placing the cutthrough
1127 connection after a verify cache hit.
1128 Do not update it with the verify result either.
1130 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1131 when routing results in more than one destination address.
1133 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1134 signing (which inhibits the cutthrough capability). Previously only
1135 the presence of an option was tested; now an expansion evaluating as
1136 empty is permissible (obviously it should depend only on data available
1137 when the cutthrough connection is made).
1139 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1140 the relevant preceding SMTP command did not note the pipelining mode.
1142 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1143 Previously they were not counted.
1145 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1146 as one having no matching records. Previously we deferred the message
1147 that needed the lookup.
1149 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1150 distinguished as "(=".
1152 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1153 for missing MX records. Previously it only worked for missing A records.
1155 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1157 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1158 after the data-go-ahead and data-ack. Patch from Jason Betts.
1160 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1161 even for a "none" policy. Patch from Tony Meyer.
1163 JH/22 Fix continued use of a connection for further deliveries. If a port was
1164 specified by a router, it must also match for the delivery to be
1167 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1168 When one of the recipients of a message was unsuitable for the connection
1169 (has no matching addresses), we lost track of needing to mark it
1170 deferred. As a result mail would be lost.
1172 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1174 JH/25 Decoding ACL controls is now done using a binary search; the source code
1175 takes up less space and should be simpler to maintain. Merge the ACL
1176 condition decode tables also, with similar effect.
1178 JH/26 Fix problem with one_time used on a redirect router which returned the
1179 parent address unchanged. A retry would see the parent address marked as
1180 delivered, so not attempt the (identical) child. As a result mail would
1183 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1184 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1185 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1188 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1189 Main config options check_{log,spool}_{inodes,space} are now
1190 100 inodes, 10MB unless set otherwise in the configuration.
1192 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1193 Previously it only applied to the main-section connection policy
1196 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1198 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1199 by me. Added RFC7919 DH primes as an alternative.
1201 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1202 are not in the system include path.
1204 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1205 GnuTLS, when a session startup failed (eg because the client disconnected)
1206 Exim did stdio operations after fclose. This was exposed by a recent
1207 change which nulled out the file handle after the fclose.
1209 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1210 signed directly by the cert-signing cert, rather than an intermediate
1211 OCSP-signing cert. This is the model used by LetsEncrypt.
1213 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1215 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1216 an incoming connection.
1218 HS/02 Bug 1802: Do not half-close the connection after sending a request
1221 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1222 fallback to "prime256v1".
1224 JH/34 SECURITY: Use proper copy of DATA command in error message.
1225 Could leak key material. Remotely exploitable. CVE-2016-9963.
1231 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1232 and 3.4.4 - once the server is enabled to respond to an OCSP request
1233 it does even when not requested, resulting in a stapling non-aware
1234 client dropping the TLS connection.
1236 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1237 support variable-length bit vectors. No functional change.
1239 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1240 The I= interface field on outgoing lines is now after the H= remote
1241 host field, same as incoming lines. There is a separate
1242 outgoing_interface log selector which allows you to disable the
1245 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1246 If not running log_selector +smtp_connection the mainlog would be held
1247 open indefinitely after a "too many connections" event, including to a
1248 deleted file after a log rotate. Leave the per net connection logging
1249 leaving it open for efficiency as that will be quickly detected by the
1250 check on the next write.
1252 HS/01 Bug 1671: Fix post transport crash.
1253 Processing the wait-<transport> messages could crash the delivery
1254 process if the message IDs didn't exist for some reason. When
1255 using 'split_spool_directory=yes' the construction of the spool
1256 file name failed already, exposing the same netto behaviour.
1258 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1259 mime_regex ACL conditions.
1261 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1262 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1263 to HELO, local diagnostic string.
1265 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1266 log line to debug. Even when configured with a tls authenticator many
1267 client connections are expected to not authenticate in this way, so
1268 an authenticate fail is not an error.
1270 HS/02 Add the Exim version string to the process info. This way exiwhat
1271 gives some more detail about the running daemon.
1273 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1274 matter for fast-change records such as DNSBLs.
1276 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1277 retry record, even if constant. There may be multiple transports with
1278 different interface settings and the retry behaviour needs to be kept
1281 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1283 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1285 JH/10 Bug 840: fix log_defer_output option of pipe transport
1287 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1288 hosts_randomize. This matters a lot when combined with mua_wrapper.
1290 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1291 ${quote_pgsql:<string>} operator.
1293 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1295 JH/14 Change header file name for freeradius-client. Relevant if compiling
1296 with Radius support; from the Gentoo tree and checked under Fedora.
1298 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1300 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1301 option transport_filter, meaning no filtering.
1303 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1305 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1306 defaults to "*" (all hosts). The variable is now available when not built
1307 with TLS, default unset, mainly to enable keeping the testsuite sane.
1308 If a server certificate is not supplied (via tls_certificate) an error is
1309 logged, and clients will find TLS connections fail on startup. Presumably
1310 they will retry in-clear.
1311 Packagers of Exim are strongly encouraged to create a server certificate
1312 at installation time.
1314 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1315 with the $config_file variable.
1317 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1318 in transport context, after the attempt, and per-recipient. The latter type
1319 is per host attempted. The event data is the error message, and the errno
1320 information encodes the lookup type (A vs. MX) used for the (first) host,
1321 and the trailing two digits of the smtp 4xx response.
1323 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1324 to write to mainlog (or rejectlog, paniclog) in the window between file
1325 creation and permissions/ownership being changed. Particularly affects
1326 installations where exicyclog is run as root, rather than exim user;
1327 result is that the running daemon panics and dies.
1329 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1331 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1332 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1333 "pri" and "weight". Note that the previous implicit priority given by the
1334 list order is no longer honoured.
1336 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1337 for DKIM processing.
1339 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1340 by defining SUPPORT_SOCKS.
1342 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1343 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1344 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1345 variables are renamed to proxy_{local,external}_{address,port}.
1347 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1348 for a build by defining SUPPORT_I18N
1350 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1351 of the query string, and make ${quote_redis:} do that quoting.
1353 JH/29 Move Events support from Experimental to mainline, enabled by default
1354 and removable for a build by defining DISABLE_EVENT.
1356 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1358 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1359 cached by the daemon.
1361 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1362 by defining LOOKUP_REDIS. The libhiredis library is required.
1364 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1365 keys are given for lookup.
1367 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1368 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1369 only supported when built with TLS support. The PolarSSL SHA routines
1370 are still used when the TLS library is too old for convenient support.
1372 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1373 openssl_options), for security. OpenSSL forces this from version 1.1.0
1374 server-side so match that on older versions.
1376 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1377 allocation for $value could be released as the expansion processing
1378 concluded, but leaving the global pointer active for it.
1380 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1381 and to use the domains and local_parts ACL conditions.
1383 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1384 incorrectly not doubled on cutthrough transmission, hence seen as a
1385 body-termination at the receiving system - resulting in truncated mails.
1386 Commonly the sender saw a TCP-level error, and retransmitted the message
1387 via the normal store-and-forward channel. This could result in duplicates
1388 received - but deduplicating mailstores were liable to retain only the
1389 initial truncated version.
1391 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1393 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1395 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1396 we're in there, support oversigning also; bug 1309.
1398 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1400 HS/04 Add support for keep_environment and add_environment options.
1402 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1403 either intentional arithmetic overflow during PRNG, or testing config-
1406 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1407 delivery resulted in actual delivery. Cancel cutthrough before DATA
1410 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1411 rejects a recipient: pass the reject to the originator.
1413 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1414 Many were false-positives and ignorable, but it's worth fixing the
1417 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1418 for the new environment-manipulation done at startup. Move the routines
1419 from being local to tls.c to being global via the os.c file.
1421 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1422 an extract embedded as result-arg for a map, the first arg for extract
1423 is unavailable so we cannot tell if this is a numbered or keyed
1424 extraction. Accept either.
1430 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1433 JH/02 The smtp transport option "multi_domain" is now expanded.
1435 JH/03 The smtp transport now requests PRDR by default, if the server offers
1438 JH/04 Certificate name checking on server certificates, when exim is a client,
1439 is now done by default. The transport option tls_verify_cert_hostnames
1440 can be used to disable this per-host. The build option
1441 EXPERIMENTAL_CERTNAMES is withdrawn.
1443 JH/05 The value of the tls_verify_certificates smtp transport and main options
1444 default to the word "system" to access the system default CA bundle.
1445 For GnuTLS, only version 3.0.20 or later.
1447 JH/06 Verification of the server certificate for a TLS connection is now tried
1448 (but not required) by default. The verification status is now logged by
1449 default, for both outbound TLS and client-certificate supplying inbound
1452 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1455 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1456 Status Notification (bounce) messages are now MIME format per RFC 3464.
1457 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1458 under the control of the dsn_advertise_hosts option, and routers may
1459 have a dsn_lasthop option.
1461 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1462 default, modifiable by a malware= option. The list separator for
1463 the options can now be changed in the usual way. Bug 68.
1465 JH/10 The smtp_receive_timeout main option is now expanded before use.
1467 JH/11 The incoming_interface log option now also enables logging of the
1468 local interface on delivery outgoing connections.
1470 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1471 if the interface and destination host and port all match.
1473 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1476 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1477 Patch from Andrew Lewis.
1479 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1480 now supports optional time-restrictions, weighting, and priority
1481 modifiers per server. Patch originally by <rommer@active.by>.
1483 JH/16 The spamd_address main option now supports a mixed list of local
1484 and remote servers. Remote servers can be IPv6 addresses, and
1485 specify a port-range.
1487 JH/17 Bug 68: The spamd_address main option now supports an optional
1488 timeout value per server.
1490 JH/18 Bug 1581: Router and transport options headers_add/remove can
1491 now have the list separator specified.
1493 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1496 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1499 JH/21 Support for the A6 type of dns record is withdrawn.
1501 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1502 rather than the verbs used.
1504 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1505 from 255 to 1024 chars.
1507 JH/24 Verification callouts now attempt to use TLS by default.
1509 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1510 are generic router options now. The defaults didn't change.
1512 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1513 Original patch from Alexander Shikoff, worked over by JH.
1515 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1516 files if ZCAT_COMMAND is not executable.
1518 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1520 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1522 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1523 Normally benign, it bites when the pair was led to by a CNAME;
1524 modern usage is to not canonicalize the domain to a CNAME target
1525 (and we were inconsistent anyway for A-only vs AAAA+A).
1527 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1529 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1530 when evaluating $sender_host_dnssec.
1532 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1533 $sender_helo_dnssec variable.
1535 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1537 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1539 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1541 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1542 documented as working, but never had. Support all but $spam_report.
1544 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1545 added for tls authenticator.
1547 HS/03 Add perl_taintmode main config option
1553 TL/01 When running the test suite, the README says that variables such as
1554 no_msglog_check are global and can be placed anywhere in a specific
1555 test's script, however it was observed that placement needed to be near
1556 the beginning for it to behave that way. Changed the runtest perl
1557 script to read through the entire script once to detect and set these
1558 variables, reset to the beginning of the script, and then run through
1559 the script parsing/test process like normal.
1561 TL/02 The BSD's have an arc4random API. One of the functions to induce
1562 adding randomness was arc4random_stir(), but it has been removed in
1563 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1564 function when detected.
1566 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1567 cause callback expansion.
1569 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1570 syntax errors in an expansion can be treated as a string instead of
1571 logging or causing an error, due to the internal use of bool_lax
1572 instead of bool when processing it.
1574 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1575 server certificates when making smtp deliveries.
1577 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1579 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1581 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1583 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1584 Merged patch from Sebastian Wiedenroth.
1586 JH/05 Fix results-pipe from transport process. Several recipients, combined
1587 with certificate use, exposed issues where response data items split
1588 over buffer boundaries were not parsed properly. This eventually
1589 resulted in duplicates being sent. This issue only became common enough
1590 to notice due to the introduction of connection certificate information,
1591 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1593 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1594 size buffer was used, resulting in syntax errors when an expansion
1597 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1598 version 3.3.6 or later.
1600 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1601 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1602 both become "event_action", the variables become $event_name, $event_data
1603 and $event_defer_errno. There is a new variable $verify_mode, usable in
1604 routers, transports and related events. The tls:cert event is now also
1605 raised for inbound connections, if the main configuration event_action
1608 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1609 early OCSP support, but no stapling (appears to be less than 1.0.0).
1611 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1612 server certificate names available under the smtp transport option
1613 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1616 JH/10 Time-related extraction expansions from certificates now use the main
1617 option "timezone" setting for output formatting, and are consistent
1618 between OpenSSL and GnuTLS compilations. Bug 1541.
1620 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1621 encoded parameter in the incoming message. Bug 1558.
1623 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1624 include certificate info, eximon was claiming there were spoolfile
1627 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1629 JH/14 Log delivery-related information more consistently, using the sequence
1630 "H=<name> [<ip>]" wherever possible.
1632 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1633 are problematic for Debian distribution, omit them from the release
1636 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1638 JH/16 Fix string representation of time values on 64bit time_t architectures.
1641 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1642 output list separator was used.
1647 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1648 checkers that were complaining about end of non-void function with no
1651 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1652 This was a regression introduced in 4.83 by another bugfix.
1654 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1656 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1657 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1663 TF/01 Correctly close the server side of TLS when forking for delivery.
1665 When a message was received over SMTP with TLS, Exim failed to clear up
1666 the incoming connection properly after forking off the child process to
1667 deliver the message. In some situations the subsequent outgoing
1668 delivery connection happened to have the same fd number as the incoming
1669 connection previously had. Exim would try to use TLS and fail, logging
1670 a "Bad file descriptor" error.
1672 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1673 utilities have not been installed.
1675 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1676 temporary space as the ACL may create new global variables.
1678 TL/01 LDAP support uses per connection or global context settings, depending
1679 upon the detected version of the libraries at build time.
1681 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1682 to extract and use the src ip:port in logging and expansions as if it
1683 were a direct connection from the outside internet. PPv2 support was
1684 updated based on HAProxy spec change in May 2014.
1686 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1688 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1689 Properly escape header and check for NULL return.
1691 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1694 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1696 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1697 characters in header names, implemented as a verify condition.
1698 Contributed by Michael Fischer v. Mollard.
1700 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1701 results permerror and temperror. Previous values are deprecated but
1702 still accepted. In a future release, err_perm and err_temp will be
1703 completely removed, which will be a backward incompatibility if the
1704 ACL tests for either of these two old results. Patch contributed by
1705 user bes-internal on the mailing list.
1707 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1709 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1710 selectors, in both main and reject logs.
1712 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1715 JH/07 Add malware type "sock" for talking to simple daemon.
1717 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1719 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1720 routers/transports under cutthrough routing.
1722 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1723 numbers. Touch up "bool" conditional to keep the same definition.
1725 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1727 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1729 JH/12 Expand items in router/transport headers_add or headers_remove lists
1730 individually rather than the list as a whole. Bug 1452.
1732 Required for reasonable handling of multiple headers_ options when
1733 they may be empty; requires that headers_remove items with embedded
1734 colons must have them doubled (or the list-separator changed).
1736 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1737 view the policy declared in the DMARC record. Currently, $dmarc_status
1738 is a combined value of both the record presence and the result of the
1741 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1743 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1744 dnslookup router and the smtp transport (applying to the forward
1747 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1748 of ldap servers used for a specific lookup. Patch provided by Heiko
1751 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1752 New variable $lookup_dnssec_authenticated for observability.
1754 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1755 Patch submitted by Lars Timman.
1757 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1759 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1760 Requires trusted mode and valid format message id, aborts otherwise.
1761 Patch contributed by Heiko Schlichting.
1763 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1764 certextract with support for various fields. Bug 1358.
1766 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1767 is requested by default, modifiable by smtp transport option
1770 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1771 operate on certificate variables to give certificate fingerprints
1772 Also new ${sha256:cert_variable}.
1774 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1776 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1777 Christian Aistleitner.
1779 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1781 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1782 file. Patch from Wolfgang Breyha.
1784 JH/25 Expand the coverage of the delivery $host and $host_address to
1785 client authenticators run in verify callout. Bug 1476.
1787 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1788 align with daemon_smtp_ports. Bug 72.
1790 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1791 support and error reporting did not work properly.
1793 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1794 and is readable. Patch from Andrew Colin Kissa.
1796 TL/14 Enhance documentation of ${run expansion and how it parses the
1797 commandline after expansion, particularly in the case when an
1798 unquoted variable expansion results in an empty value.
1800 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1802 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1803 by renaming away from it. Addresses GH issue 15, reported by
1806 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1808 TL/15 SECURITY: prevent double expansion in math comparison functions
1809 (can expand unsanitized data). Not remotely exploitable.
1816 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1818 PP/02 Make -n do something, by making it not do something.
1819 When combined with -bP, the name of an option is not output.
1821 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1824 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1825 $sender_host_name and config options to manage this, and basic check
1828 PP/05 DSCP support for outbound connections and control modifier for inbound.
1830 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1831 (Only plugin which currently uses this is kerberos4, which nobody should
1832 be using, but we should make it available and other future plugins might
1833 conceivably use it, even though it would break NAT; stuff *should* be
1834 using channel bindings instead).
1836 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1837 name; added for Sendmail compatibility; requires admin caller.
1838 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1839 just ignore it); requires trusted caller.
1840 Also parse but ignore: -Ac -Am -X<logfile>
1843 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1845 TL/02 Add +smtp_confirmation as a default logging option.
1847 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1848 Patch by Magnus Holmgren from 2007-02-20.
1850 TL/04 Bugzilla 1281 - Spec typo.
1851 Bugzilla 1283 - Spec typo.
1852 Bugzilla 1290 - Spec grammar fixes.
1854 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1856 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1858 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1859 mailing list by by Dmitry Isaikin.
1861 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1863 JH/02 Support "G" suffix to numbers in ${if comparisons.
1865 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1867 NM/01 Bugzilla 1197 - Spec typo
1868 Bugzilla 1196 - Spec examples corrections
1870 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1872 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1873 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1876 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1877 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1879 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1880 "acl {{name}{arg}...}", and optional args on acl condition
1883 JH/05 Permit multiple router/transport headers_add/remove lines.
1885 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1887 JH/07 Avoid using a waiting database for a single-message-only transport.
1888 Performance patch from Paul Fisher. Bugzilla 1262.
1890 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1893 JH/09 Add $headers_added variable, with content from use of ACL modifier
1894 add_header (but not yet added to the message). Bugzilla 199.
1896 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1897 Pulled from Bugzilla 817 by Wolfgang Breyha.
1899 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1901 (nb: this is the same fix as in Exim 4.80.1)
1903 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1906 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1909 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1911 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1912 advertise SMTP AUTH mechanism to us, instead of a generic
1913 protocol violation error. Also, make Exim more robust to bad
1914 data from the Dovecot auth socket.
1916 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1918 When a queue runner is handling a message, Exim first routes the
1919 recipient addresses, during which it prunes them based on the retry
1920 hints database. After that it attempts to deliver the message to
1921 any remaining recipients. It then updates the hints database using
1924 So if a recipient address works intermittently, it can get repeatedly
1925 deferred at routing time. The retry hints record remains fresh so the
1926 address never reaches the final cutoff time.
1928 This is a fairly common occurrence when a user is bumping up against
1929 their storage quota. Exim had some logic in its local delivery code
1930 to deal with this. However it did not apply to per-recipient defers
1931 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1933 This change adds a proper retry rule check during routing so that the
1934 final cutoff time is checked against the message's age. We only do
1935 this check if there is an address retry record and there is not a
1936 domain retry record; this implies that previous attempts to handle
1937 the address had the retry_use_local_parts option turned on. We use
1938 this as an approximation for the destination being like a local
1939 delivery, as in LMTP.
1941 I suspect this new check makes the old local delivery cutoff check
1942 redundant, but I have not verified this so I left the code in place.
1944 TF/02 Correct gecos expansion when From: is a prefix of the username.
1946 Test 0254 submits a message to Exim with the header
1950 When I ran the test suite under the user fanf2, Exim expanded
1951 the header to contain my full name, whereas it should have added
1952 a Resent-Sender: header. It erroneously treats any prefix of the
1953 username as equal to the username.
1955 This change corrects that bug.
1957 GF/01 DCC debug and logging tidyup
1958 Error conditions log to paniclog rather than rejectlog.
1959 Debug lines prefixed by "DCC: " to remove any ambiguity.
1961 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1963 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1964 Bug spotted by Jeremy Harris; was flawed since initial commit.
1965 Would have resulted in OCSP responses post-SNI triggering an Exim
1966 NULL dereference and crash.
1968 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1970 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1971 Bug detection, analysis and fix by Samuel Thibault.
1972 Bugzilla 1331, Debian bug #698092.
1974 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1976 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1977 Server implementation by Todd Lyons, client by JH.
1978 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1979 config variable "prdr_enable" controls whether the server
1980 advertises the facility. If the client requests PRDR a new
1981 acl_data_smtp_prdr ACL is called once for each recipient, after
1982 the body content is received and before the acl_smtp_data ACL.
1983 The client is controlled by both of: a hosts_try_prdr option
1984 on the smtp transport, and the server advertisement.
1985 Default client logging of deliveries and rejections involving
1986 PRDR are flagged with the string "PRDR".
1988 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1989 fclose(). Diagnosis by Todd Lyons.
1991 PP/17 Update configure.default to handle IPv6 localhost better.
1992 Patch by Alain Williams (plus minor tweaks).
1995 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1996 This is now consistent with GnuTLS, and is now documented: the
1997 previous undocumented portable approach to treating the option as
1998 unset was to force an expansion failure. That still works, and
1999 an empty string is now equivalent.
2001 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2002 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2003 not performing validation itself.
2005 PP/20 Added force_command boolean option to pipe transport.
2006 Patch from Nick Koston, of cPanel Inc.
2008 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2011 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2013 PP/21 Fix eximon continuous updating with timestamped log-files.
2014 Broken in a format-string cleanup in 4.80, missed when I repaired the
2015 other false fix of the same issue.
2016 Report and fix from Heiko Schlichting.
2019 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2020 Report from Prashanth Katuri.
2022 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2023 It's SecureTransport, so affects any MacOS clients which use the
2024 system-integrated TLS libraries, including email clients.
2026 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2027 using a MIME ACL for non-SMTP local injection.
2028 Report and assistance in diagnosis by Warren Baker.
2030 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2032 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2034 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2035 last id that failed so it may be referenced in subsequent ACL's.
2037 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2040 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2041 ldap library initialization, allowing self-signed CA's to be
2042 used. Also properly sets require_cert option later in code by
2043 using NULL (global ldap config) instead of ldap handle (per
2044 session). Bug diagnosis and testing by alxgomz.
2046 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2047 the src/util/ subdirectory.
2049 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2050 renamed to Transport Post Delivery Action by Jeremy Harris, as
2053 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2054 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2055 redis_servers = needs to be configured which will be used by the redis
2056 lookup. Patch from Warren Baker, of The Packet Hub.
2058 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2060 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2061 hostname or reverse DNS when processing a host list. Used suggestions
2062 from multiple comments on this bug.
2064 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2066 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2067 Missed a few lines, added it to make the runtest require no keyboard
2070 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2071 contains upper case chars. Make router use caseful_local_part.
2073 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2074 support when GnuTLS has been built with p11-kit.
2080 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2082 This, or similar/improved, will also be change PP/11 of 4.82.
2088 PP/01 Handle short writes when writing local log-files.
2089 In practice, only affects FreeBSD (8 onwards).
2090 Bugzilla 1053, with thanks to Dmitry Isaikin.
2092 NM/01 Bugzilla 949 - Documentation tweak
2094 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2097 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2099 PP/02 Implemented gsasl authenticator.
2101 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2103 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2104 `pkg-config foo` for cflags/libs.
2106 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2107 with rest of GSASL and with heimdal_gssapi.
2109 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2110 `pkg-config foo` for cflags/libs for the TLS implementation.
2112 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2113 properties get this fed in as external SSF. A number of robustness
2114 and debugging improvements to the cyrus_sasl authenticator.
2116 PP/08 cyrus_sasl server now expands the server_realm option.
2118 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2119 Patch by Jeremy Harris.
2121 PP/10 Added dbmjz lookup type.
2123 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2125 PP/12 MAIL args handles TAB as well as SP, for better interop with
2126 non-compliant senders.
2127 Analysis and variant patch by Todd Lyons.
2129 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2130 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2131 Patch from Dirk Mueller <dmueller@suse.com>
2133 PP/13 tls_peerdn now print-escaped for spool files.
2134 Observed some $tls_peerdn in wild which contained \n, which resulted
2135 in spool file corruption.
2137 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2138 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2139 or write after TLS renegotiation, which otherwise led to messages
2142 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2143 as a tracking header (ie: a signed header comes before the signature).
2144 Patch from Wolfgang Breyha.
2146 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2147 comma-sep list; embedded commas doubled.
2149 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2151 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2153 Report and patch from Dmitry Banschikov.
2155 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2156 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2157 failures. We appear to now support TLS1.1+ with Exim.
2159 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2160 lets Exim select keys and certificates based upon TLS SNI from client.
2161 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2162 before an outbound SMTP session. New log_selector, +tls_sni.
2164 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2165 NULL dereference. Report and patch from Alun Jones.
2167 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2168 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2169 Not seeing resolver debug output on NetBSD, but suspect this is a
2170 resolver implementation change.
2172 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2173 Left warnings. Added "eximon gdb" invocation mode.
2175 PP/21 Defaulting "accept_8bitmime" to true, not false.
2177 PP/22 Added -bw for inetd wait mode support.
2179 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2180 locate the relevant includes and libraries. Made this the default.
2182 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2183 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2185 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2186 This may cause build issues on older platforms.
2188 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2189 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2190 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2191 Added SNI support via GnuTLS too.
2192 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2194 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2196 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2197 Applied second patch from Janne, implementing suggestion to default
2198 multiple-strings-in-record handling to match SPF spec.
2200 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2202 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2203 read-only, out of scope).
2204 Patch from Wolfgang Breyha, report from Stuart Northfield.
2206 PP/29 Fix three issues highlighted by clang analyser static analysis.
2207 Only crash-plausible issue would require the Cambridge-specific
2208 iplookup router and a misconfiguration.
2209 Report from Marcin Mirosław.
2211 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2213 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2214 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2215 As part of this, removing so much warning spew let me fix some minor
2216 real issues in debug logging.
2218 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2219 assignment on my part. Fixed.
2221 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2222 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2223 Janne Snabb (who went above and beyond: thank you).
2225 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2226 string otherwise requires a connection and a bunch more work and it's
2227 relatively easy to get wrong. Should also expose TLS library linkage
2230 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2231 64-bit ${eval} (JH/03).
2233 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2234 GNU libc to support some of the 64-bit stuff, should not lead to
2235 conflicts. Defined before os.h is pulled in, so if a given platform
2236 needs to override this, it can.
2238 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2239 protection layer was required, which is not implemented.
2240 Bugzilla 1254, patch from Wolfgang Breyha.
2242 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2243 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2244 tls_dhparam take prime identifiers. Also unbreak combination of
2245 OpenSSL+DH_params+TLSSNI.
2247 PP/39 Disable SSLv2 by default in OpenSSL support.
2253 PP/01 Solaris build fix for Oracle's LDAP libraries.
2254 Bugzilla 1109, patch from Stephen Usher.
2256 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2258 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2261 TF/02 Fix a couple more cases where we did not log the error message
2262 when unlink() failed. See also change 4.74-TF/03.
2264 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2265 lock up or crash if it happened to be inside a call to libc when it
2266 got a SIGUSR1 from exiwhat.
2268 The SIGUSR1 handler appends the current process status to the process
2269 log which is later printed by exiwhat. It used to use the general
2270 purpose logging code to do this, but several functions it calls are
2271 not safe for signals.
2273 The new output code in the SIGUSR1 handler is specific to the process
2274 log, and simple enough that it's easy to inspect for signal safety.
2275 Removing some special cases also simplifies the general logging code.
2276 Removing the spurious timestamps from the process log simplifies
2279 TF/04 Improved ratelimit ACL condition.
2281 The /noupdate option has been deprecated in favour of /readonly which
2282 has clearer semantics. The /leaky, /strict, and /readonly update modes
2283 are mutually exclusive. The update mode is no longer included in the
2284 database key; it just determines when the database is updated. (This
2285 means that when you upgrade Exim will forget old rate measurements.)
2287 Exim now checks that the per_* options are used with an update mode that
2288 makes sense for the current ACL. For example, when Exim is processing a
2289 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2290 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2291 must specify per_mail/readonly. If you omit the update mode it defaults to
2292 /leaky where that makes sense (as before) or /readonly where required.
2294 The /noupdate option is now undocumented but still supported for
2295 backwards compatibility. It is equivalent to /readonly except that in
2296 ACLs where /readonly is required you may specify /leaky/noupdate or
2297 /strict/noupdate which are treated the same as /readonly.
2299 A useful new feature is the /count= option. This is a generalization
2300 of the per_byte option, so that you can measure the throughput of other
2301 aggregate values. For example, the per_byte option is now equivalent
2302 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2304 The per_rcpt option has been generalized using the /count= mechanism
2305 (though it's more complicated than the per_byte equivalence). When it is
2306 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2307 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2308 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2309 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2310 Note that using per_rcpt with a non-readonly update mode in more than
2311 one ACL will cause the recipients to be double-counted. (The per_mail
2312 and per_byte options don't have this problem.)
2314 The handling of very low rates has changed slightly. If the computed rate
2315 is less than the event's count (usually one) then this event is the first
2316 after a long gap. In this case the rate is set to the same as this event's
2317 count, so that the first message of a spam run is counted properly.
2319 The major new feature is a mechanism for counting the rate of unique
2320 events. The new per_addr option counts the number of different
2321 recipients that someone has sent messages to in the last time period. It
2322 behaves like per_rcpt if all the recipient addresses are different, but
2323 duplicate recipient addresses do not increase the measured rate. Like
2324 the /count= option this is a general mechanism, so the per_addr option
2325 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2326 example, measure the rate that a client uses different sender addresses
2327 with the options per_mail/unique=$sender_address. There are further
2328 details in the main documentation.
2330 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2332 TF/06 Removed a few PCRE remnants.
2334 TF/07 Automatically extract Exim's version number from tags in the git
2335 repository when doing development or release builds.
2337 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2338 Bugzilla 879. Patch from Paul Fisher.
2340 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2341 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2344 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2346 PP/05 Variable $av_failed, true if the AV scanner deferred.
2347 Bugzilla 1078. Patch from John Horne.
2349 PP/06 Stop make process more reliably on build failure.
2350 Bugzilla 1087. Patch from Heiko Schlittermann.
2352 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2353 Bugzilla 1089. Patch from Heiko Schlittermann.
2355 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2356 Bugzilla 1131. Patch from Holger Weiß.
2358 PP/09 Handle IPv6 addresses with SPF.
2359 Bugzilla 860. Patch from Wolfgang Breyha.
2361 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2363 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2366 PP/11 match_* no longer expand right-hand-side by default.
2367 New compile-time build option, EXPAND_LISTMATCH_RHS.
2368 New expansion conditions, "inlist", "inlisti".
2370 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2372 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2374 PP/14 fix log_write() format string regression from TF/03.
2375 Bugzilla 1152. Patch from Dmitry Isaikin.
2381 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2383 PP/02 Harmonised TLS library version reporting; only show if debugging.
2384 Layout now matches that introduced for other libraries in 4.74 PP/03.
2386 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2388 PP/04 New "dns_use_edns0" global option.
2390 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2393 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2394 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2396 TK/01 Updated PolarSSL code to 0.14.2.
2397 Bugzilla 1097. Patch from Andreas Metzler.
2399 PP/07 Catch divide-by-zero in ${eval:...}.
2400 Fixes bugzilla 1102.
2402 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2405 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2406 format-string attack -- SECURITY: remote arbitrary code execution.
2408 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2409 time unintentionally subject to list matching rules, letting the header
2410 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2411 arbitrary string expansion). This allowed for information disclosure.
2413 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2414 INT_MIN/-1 -- value coerced to INT_MAX.
2420 NM/01 Workaround for PCRE version dependency in version reporting
2423 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2424 This fixes portability to compilers other than gcc, notably
2425 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2427 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2428 makefiles for portability to HP-UX and POSIX correctness.
2430 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2431 Also via indented variable definition in the Makefile.
2432 (Debugging by Oliver Heesakkers).
2434 PP/02 Restore caching of spamd results with expanded spamd_address.
2435 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2437 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2438 Improves build reliability. Fix from: Frank Elsner
2440 NM/02 Fix wide character breakage in the rfc2047 coding
2441 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2443 NM/03 Allow underscore in dnslist lookups
2444 Fixes bug 1026. Patch from Graeme Fowler
2446 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2447 Code patches from Adam Ciarcinski of NetBSD.
2449 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2452 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2453 is logged, to avoid truncation. Patch from John Horne.
2455 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2456 Patch from Jakob Hirsch.
2458 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2459 SQL string expansion failure details.
2460 Patch from Andrey Oktyabrski.
2462 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2463 Patch from Simon Arlott.
2465 PP/09 New lookups functionality failed to compile on old gcc which rejects
2466 extern declarations in function scope.
2467 Patch from Oliver Fleischmann
2469 PP/10 Use sig_atomic_t for flags set from signal handlers.
2470 Check getgroups() return and improve debugging.
2471 Fixed developed for diagnosis in bug 927 (which turned out to be
2474 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2475 Patch from Mark Zealey.
2477 PP/12 Bugzilla 1056: Improved spamd server selection.
2478 Patch from Mark Zealey.
2480 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2481 Based on patch from Heiko Schlittermann.
2483 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2484 Patch from Uwe Doering, sign-off by Michael Haardt.
2486 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2487 variable declaration deep within a block. Bug and patch from
2490 PP/15 lookups-Makefile IRIX compatibility coercion.
2492 PP/16 Make DISABLE_DKIM build knob functional.
2494 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2495 Patch by Simon Arlott
2497 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2498 variable argument macros. Our copy now differs from upstream.
2504 TF/01 Failure to get a lock on a hints database can have serious
2505 consequences so log it to the panic log.
2507 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2508 controlled using the smtp_confirmation log selector.
2510 TF/03 Include the error message when we fail to unlink a spool file.
2512 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2513 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2514 for maintaining out-of-tree patches for some time.
2516 PP/01 Bugzilla 139: Documentation and portability issues.
2517 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2518 Handle per-OS dynamic-module compilation flags.
2520 PP/02 Let /dev/null have normal permissions.
2521 The 4.73 fixes were a little too stringent and complained about the
2522 permissions on /dev/null. Exempt it from some checks.
2523 Reported by Andreas M. Kirchwitz.
2525 PP/03 Report version information for many libraries, including
2526 Exim version information for dynamically loaded libraries. Created
2527 version.h, now support a version extension string for distributors
2528 who patch heavily. Dynamic module ABI change.
2530 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2531 privilege escalation vulnerability whereby the Exim run-time user
2532 can cause root to append content of the attacker's choosing to
2535 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2538 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2539 If dropping privileges for untrusted macros, we disabled normal logging
2540 on the basis that it would fail; for the Exim run-time user, this is not
2541 the case, and it resulted in successful deliveries going unlogged.
2542 Fixed. Reported by Andreas Metzler.
2548 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2549 only prepend for the Resent-* case. Fixes regression introduced in
2550 Exim 4.70 by NM/22 for Bugzilla 607.
2552 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2553 increasing numbers of administrators be bitten by this.
2555 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2557 PP/03 Bugzilla 994: added openssl_options main configuration option.
2559 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2561 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2563 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2565 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2566 without a peer certificate, leading to a segfault because of an
2567 assumption that peers always have certificates. Be a little more
2568 paranoid. Problem reported by Martin Tscholak.
2570 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2571 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2572 NB: ClamAV planning to remove STREAM in "middle of 2010".
2573 CL also introduces -bmalware, various -d+acl logging additions and
2574 more caution in buffer sizes.
2576 PP/09 Implemented reverse_ip expansion operator.
2578 PP/10 Bugzilla 937: provide a "debug" ACL control.
2580 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2582 PP/12 Bugzilla 973: Implement --version.
2584 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2586 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2588 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2590 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2591 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2592 ignore trailing whitespace.
2594 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2596 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2599 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2600 Notification from Dr Andrew Aitchison.
2602 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2603 ExtendedDetectionInfo response format.
2604 Notification from John Horne.
2606 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2609 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2610 XSL and documented dependency on system catalogs, with examples of how
2613 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2616 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2617 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2618 configuration file which is writeable by the Exim user or group.
2620 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2621 of configuration files to cover files specified with the -C option if
2622 they are going to be used with root privileges, not just the default
2625 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2626 option (effectively making it always true).
2628 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2629 files to be used while preserving root privileges.
2631 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2632 that rogue child processes cannot use them.
2634 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2635 run-time user, instead of root.
2637 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2638 Exim run-time user without dropping privileges.
2640 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2641 result string, instead of calling string_vformat() twice with the same
2644 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2645 for other users. Others should always drop root privileges if they use
2646 -C on the command line, even for a whitelisted configure file.
2648 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2650 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2656 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2657 $data_path, and $header_path variables; fixed documentation bugs and
2660 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2661 exipick to access non-standard spools, including the "frozen" queue
2664 NM/01 Bugzilla 965: Support mysql stored procedures.
2665 Patch from Alain Williams
2667 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2669 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2670 Patch from Andreas Metzler
2672 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2673 Patch from Kirill Miazine
2675 NM/05 Bugzilla 671: Added umask to procmail example.
2677 JJ/03 installed exipick 20100323.0, fixing doc bug
2679 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2680 directory. Notification and patch from Dan Rosenberg.
2682 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2684 TK/02 Improve log output when DKIM signing operation fails.
2686 MH/01 Treat the transport option dkim_domain as a colon separated
2687 list, not as a single string, and sign the message with each element,
2688 omitting multiple occurences of the same signer.
2690 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2691 Bugzilla 985, 986. Patch by Simon Arlott
2693 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2694 Patch by Simon Arlott
2696 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2697 MBX locking. Notification from Dan Rosenberg.
2703 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2705 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2707 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2709 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2711 NM/04 Bugzilla 915: Flag broken perl installation during build.
2717 TK/01 Added patch by Johannes Berg that expands the main option
2718 "spamd_address" if it starts with a dollar sign.
2720 TK/02 Write list of recipients to X-Envelope-Sender header when building
2721 the mbox-format spool file for content scanning (suggested by Jakob
2724 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2725 (http://www.dcc-servers.net/) support via dccifd. Activated by
2726 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2728 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2729 by Mark Daniel Reidel <mr@df.eu>.
2731 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2732 When building exim an external PCRE library is now needed -
2733 PCRE is a system library on the majority of modern systems.
2734 See entry on PCRE_LIBS in EDITME file.
2736 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2737 conversation. Added nologin parameter to request.
2738 Patch contributed by Kirill Miazine.
2740 TF/01 Do not log submission mode rewrites if they do not change the address.
2742 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2744 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2745 log files in place. Contributed by Roberto Lima.
2747 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2749 TF/03 Bugzilla 615: When checking the local_parts router precondition
2750 after a local_part_suffix or local_part_prefix option, Exim now
2751 does not use the address's named list lookup cache, since this
2752 contains cached lookups for the whole local part.
2754 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2755 Robert Millan. Documentation is in experimental-spec.txt.
2757 TF/04 Bugzilla 668: Fix parallel build (make -j).
2759 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2761 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2762 Patch provided by Jan Srzednicki.
2764 TF/05 Leading white space used to be stripped from $spam_report which
2765 wrecked the formatting. Now it is preserved.
2767 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2768 that they are available at delivery time.
2770 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2772 TF/08 TLS error reporting now respects the incoming_interface and
2773 incoming_port log selectors.
2775 TF/09 Produce a more useful error message if an SMTP transport's hosts
2776 setting expands to an empty string.
2778 NM/06 Bugzilla 744: EXPN did not work under TLS.
2779 Patch provided by Phil Pennock.
2781 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2782 Patch provided by Richard Godbee.
2784 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2785 acl_smtp_notquit, added index entry.
2787 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2788 Patch provided by Eugene Bujak.
2790 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2791 accept(). Patch provided by Maxim Dounin.
2793 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2794 Patch provided by Phil Pennock.
2796 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2798 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2799 Patch provided by Brad "anomie" Jorsch.
2801 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2802 Patch provided by Dean Brooks.
2804 TK/05 Add native DKIM support (does not depend on external libraries).
2806 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2807 Patch provided by Graeme Fowler.
2809 NM/16 Bugzilla 851: Documentation example syntax fix.
2811 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2813 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2816 NM/19 Bugzilla 745: TLS version reporting.
2817 Patch provided by Phil Pennock.
2819 NM/20 Bugzilla 167: bool: condition support.
2820 Patch provided by Phil Pennock.
2822 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2823 clients. Patch provided by Phil Pennock.
2825 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2826 Patch provided by Brad "anomie" Jorsch.
2828 NM/23 Bugzilla 687: Fix misparses in eximstats.
2829 Patch provided by Heiko Schlittermann.
2831 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2832 Patch provided by Heiko Schlittermann.
2834 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2835 plus update to original patch.
2837 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2839 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2840 Patch provided by David Brownlee.
2842 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2844 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2846 NM/30 Bugzilla 888: TLS documentation bugfixes.
2848 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2850 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2851 Unlike the original bugzilla I have changed all shell scripts in src tree.
2853 NM/33 Bugzilla 898: Transport filter timeout fix.
2854 Patch by Todd Rinaldo.
2856 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2857 Patch by Serge Demonchaux.
2859 NM/35 Bugzilla 39: Base64 decode bug fixes.
2860 Patch by Jakob Hirsch.
2862 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2864 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2866 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2868 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2874 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2875 ALT-N's libdkim that I have put here:
2876 http://duncanthrax.net/exim-experimental/
2878 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2879 were called 'true' and it seems that C99 defines that as a reserved
2880 keyword to be used with 'bool' variable types. That means you could
2881 not include C99-style headers which use bools without triggering
2882 build errors in sieve.c.
2884 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2885 as mailq or other aliases. Changed the --help handling significantly
2886 to do whats expected. exim_usage() emits usage/help information.
2888 SC/01 Added the -bylocaldomain option to eximstats.
2890 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2892 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2894 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2900 PH/01 Another patch from the Sieve maintainer.
2902 PH/02 When an IPv6 address is converted to a string for single-key lookup
2903 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2904 dots are used instead of colons so that keys in lsearch files need not
2905 contain colons. This was done some time before quoting was made available
2906 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2907 (notated using the quote facility) so as to distinguish them from IPv4
2908 keys. This meant that lookups for IP addresses in host lists did not work
2909 for iplsearch lookups.
2911 This has been fixed by arranging for IPv6 addresses to be expressed with
2912 colons if the lookup type is iplsearch. This is not incompatible, because
2913 previously such lookups could never work.
2915 The situation is now rather anomalous, since one *can* have colons in
2916 ordinary lsearch keys. However, making the change in all cases is
2917 incompatible and would probably break a number of configurations.
2919 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2922 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2923 conversion specification without a maximum field width, thereby enabling
2924 a rogue spamd server to cause a buffer overflow. While nobody in their
2925 right mind would setup Exim to query an untrusted spamd server, an
2926 attacker that gains access to a server running spamd could potentially
2927 exploit this vulnerability to run arbitrary code as the Exim user.
2929 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2930 $primary_hostname instead of what libspf2 thinks the hosts name is.
2932 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2933 a directory entry by the name of the lookup key. Previously, if a
2934 symlink pointed to a non-existing file or a file in a directory that
2935 Exim lacked permissions to read, a lookup for a key matching that
2936 symlink would fail. Now it is enough that a matching directory entry
2937 exists, symlink or not. (Bugzilla 503.)
2939 PH/03 The body_linecount and body_zerocount variables are now exported in the
2942 PH/04 Added the $dnslist_matched variable.
2944 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2945 This means they are set thereafter only if the connection becomes
2948 PH/06 Added the client_condition to authenticators so that some can be skipped
2949 by clients under certain conditions.
2951 PH/07 The error message for a badly-placed control=no_multiline_responses left
2952 "_responses" off the end of the name.
2954 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2956 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2957 (without spaces) instead of just copying the configuration text.
2959 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2961 PH/11 Added $max_received_linelength.
2963 PH/12 Added +ignore_defer and +include_defer to host lists.
2965 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2966 way in which PCRE > 7.0 is built.
2968 PH/14 Implemented queue_only_load_latch.
2970 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2971 MAIL command. The effect was to mangle the value on 64-bit systems.
2973 PH/16 Another patch from the Sieve maintainer.
2975 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2977 PH/18 If a system quota error occurred while trying to create the file for
2978 a maildir delivery, the message "Mailbox is full" was not appended to the
2979 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2980 only to a quota excession during the actual writing of the file.
2982 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2983 characters?) which causes problems in log lines. The DN values are now
2984 passed through string_printing() before being added to log lines.
2986 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2987 and InterBase are left for another time.)
2989 PH/21 Added message_body_newlines option.
2991 PH/22 Guard against possible overflow in moan_check_errorcopy().
2993 PH/23 POSIX allows open() to be a macro; guard against that.
2995 PH/24 If the recipient of an error message contained an @ in the local part
2996 (suitably quoted, of course), incorrect values were put in $domain and
2997 $local_part during the evaluation of errors_copy.
3003 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3004 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3007 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3008 issue a MAIL command.
3010 PH/02 In an ACL statement such as
3012 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3014 if a client was not listed at all, or was listed with a value other than
3015 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3016 the condition was not true (as it should be), so access was not denied.
3017 The bug was that the ! inversion was incorrectly passed on to the second
3018 item. This has been fixed.
3020 PH/03 Added additional dnslists conditions == and =& which are different from
3021 = and & when the dns lookup returns more than one IP address.
3023 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3024 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3026 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3027 FSYNC, which compiles an option called disable_fsync that allows for
3028 bypassing fsync(). The documentation is heavily laced with warnings.
3030 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3032 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3033 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3034 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3035 including adding "make clean"; (3) Added -fPIC when compiling the test
3036 dynamically loaded module, to get rid of a warning.
3038 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3039 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3040 = 0s. The bug is otherwise harmless.
3042 PH/07 There was a bug in the dovecot authenticator such that the value of
3043 $auth1 could be overwritten, and so not correctly preserved, after a
3044 successful authentication. This usually meant that the value preserved by
3045 the server_setid option was incorrect.
3047 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3049 PH/09 Installed PCRE release 7.0.
3051 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3052 run for batched SMTP input. It is now run at the start of every message
3053 in the batch. While fixing this I discovered that the process information
3054 (output by running exiwhat) was not always getting set for -bs and -bS
3055 input. This is fixed, and it now also says "batched" for BSMTP.
3057 PH/11 Added control=no_pipelining.
3059 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3060 patch, slightly modified), and move the expansion of helo_data till after
3061 the connection is made in the smtp transport (so it can use these
3064 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3066 PH/14 Added log_selector = +pid.
3068 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3070 PH/16 Add ${if forany and ${if forall.
3072 PH/17 Added dsn_from option to vary the From: line in DSNs.
3074 PH/18 Flush SMTP output before performing a callout, unless control =
3075 no_callout_flush is set.
3077 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3078 was true (the default) a successful delivery failed to delete the retry
3079 item, thus causing premature timeout of the address. The bug is now
3082 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3084 PH/21 Long custom messages for fakedefer and fakereject are now split up
3085 into multiline responses in the same way that messages for "deny" and
3086 other ACL rejections are.
3088 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3089 with slight modification.
3091 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3092 draft, changing the syntax and factoring some duplicate code".
3094 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3095 for deliveries of the second and subsequent messages over the same SMTP
3098 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3099 ${reduce, with only minor "tidies".
3101 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3103 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3104 expansion side effects.
3106 PH/27 When a message times out after an over-quota error from an Exim-imposed
3107 quota, the bounce message says "mailbox is full". This message was not
3108 being given when it was a system quota that was exceeded. It now should
3111 MH/03 Made $recipients available in local_scan(). local_scan() already has
3112 better access to the recipient list through recipients_list[], but
3113 $recipients can be useful in postmaster-provided expansion strings.
3115 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3116 in the case of a MAIL command with additional options following the
3117 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3118 were accidentally chopped off.
3120 PH/29 SMTP synchronization checks are implemented when a command is read -
3121 there is a check that no more input is waiting when there shouldn't be
3122 any. However, for some commands, a delay in an ACL can mean that it is
3123 some time before the response is written. In this time, more input might
3124 arrive, invalidly. So now there are extra checks after an ACL has run for
3125 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3126 pipelining has not been advertised.
3128 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3130 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3131 returned in the SMTP response when smtp_return_error_details was false.
3132 This has been fixed.
3134 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3135 instead of the C I/O that was originally supplied, because problems were
3136 reported on Solaris.
3138 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3139 Exim which did not show up earlier: it was assuming that a call to
3140 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3141 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3142 was a macro that became an assignment, so it seemed to work. This has
3143 changed to a proper function call with a void return, hence the compile
3144 error. Exim's code has been fixed.
3146 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3149 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3151 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3153 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3154 because it turns out to be too big or there is a timeout) there is no
3155 "Completed" line in the log. When some messages of this type were
3156 selected by exigrep, they were listed as "not completed". Others were
3157 picked up by some special patterns. I have improved the selection
3158 criteria to be more general.
3160 PH/38 The host_find_failed option in the manualroute router can now be set
3161 to "ignore", to completely ignore a host whose IP address cannot be
3162 found. If all hosts are ignored, the behaviour is controlled by the new
3163 host_all_ignored option.
3165 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3166 homing or because of multiple MX records with /mx) generated more than
3167 one IP address, and the following item turned out to be the local host,
3168 all the secondary addresses of the first item were incorrectly removed
3169 from the list, along with the local host and any following hosts (which
3170 is what is supposed to happen).
3172 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3173 whoever called Exim into the -H file. In the case of the daemon it was
3174 behaving confusingly. When first started, it used values for whoever
3175 started the daemon, but after a SIGHUP it used the Exim user (because it
3176 calls itself on a restart). I have changed the code so that it now always
3179 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3180 message are rejected with the same error (e.g. no authentication or bad
3181 sender address), and a DATA command is nevertheless sent (as can happen
3182 with PIPELINING or a stupid MUA), the error message that was given to the
3183 RCPT commands is included in the rejection of the DATA command. This is
3184 intended to be helpful for MUAs that show only the final error to their
3187 PH/42 Another patch from the Sieve maintainer.
3189 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3190 Eximstats - Fixed some broken HTML links and added missing column headers
3192 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3193 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3195 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3197 PH/43 Yet another patch from the Sieve maintainer.
3199 PH/44 I found a way to check for a TCP/IP connection going away before sending
3200 the response to the final '.' that terminates a message, but only in the
3201 case where the client has not sent further data following the '.'
3202 (unfortunately, this is allowed). However, in many cases there won't be
3203 any further data because there won't be any more messages to send. A call
3204 to select() can be used: if it shows that the input is "ready", there is
3205 either input waiting, or the socket has been closed. An attempt to read
3206 the next input character can distinguish the two cases. Previously, Exim
3207 would have sent an OK response which the client would never have see.
3208 This could lead to message repetition. This fix should cure that, at
3209 least in a lot of common cases.
3211 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3212 advertised in response to EHLO.
3218 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3219 fixed by 4.65/MH/01 (is this a record?) are fixed:
3221 (i) An empty string was always treated as zero by the numeric comparison
3222 operators. This behaviour has been restored.
3224 (ii) It is documented that the numeric comparison operators always treat
3225 their arguments as decimal numbers. This was broken in that numbers
3226 starting with 0 were being interpreted as octal.
3228 While fixing these problems I realized that there was another issue that
3229 hadn't been noticed. Values of message_size_limit (both the global option
3230 and the transport option) were treated as octal if they started with 0.
3231 The documentation was vague. These values are now always treated as
3232 decimal, and I will make that clear in the documentation.
3238 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3239 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3242 MH/01 Don't check that the operands of numeric comparison operators are
3243 integers when their expansion is in "skipping" mode (fixes bug
3244 introduced by 4.64-PH/07).
3246 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3247 child addresses, Exim now panics and dies. Previously, because the count
3248 is held in a short int, deliveries were likely to be lost. As such a
3249 large number of recipients for a single message is ridiculous
3250 (performance will be very, very poor), I have chosen to impose a limit
3251 rather than extend the field.
3257 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3258 leftover -K file (the existence of which was triggered by #402).
3259 While we were at it, introduced process PID as part of the -K
3260 filename. This should rule out race conditions when creating
3263 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3264 processing considerably. Previous code took too long for large mails,
3265 triggering a timeout which in turn triggers #401.
3267 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3268 in the DK code in transports.c. sendfile() is not really portable,
3269 hence the _LINUX specificness.
3271 TF/01 In the add_headers option to the mail command in an Exim filter,
3272 there was a bug that Exim would claim a syntax error in any
3273 header after the first one which had an odd number of characters
3276 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3277 callout verification, Exim cached a "reject" for the entire domain. This
3278 is correct for most verifications, but it is not correct for a recipient
3279 verification with use_sender or use_postmaster set, because in that case
3280 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3281 case of MAIL FROM:<> rejection from other early rejections (e.g.
3282 rejection of HELO). When verifying a recipient using a non-null MAIL
3283 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3284 Whatever the result of the callout, the value of the domain cache is
3285 left unchanged (for any other kind of callout, getting as far as trying
3286 RCPT means that the domain itself is ok).
3288 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3291 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3292 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3293 session, the daemon ignores ECONNECT errors and logs others; it now
3294 ignores EPIPE as well.
3296 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3297 (quoted-printable decoding).
3299 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3300 later the small subsequent patch to fix an introduced bug.
3302 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3304 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3306 PH/08 An error is now given if message_size_limit is specified negative.
3308 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3309 to be given (somewhat) arbitrary names.
3311 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3314 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3315 miscellaneous code fixes
3317 PH/10 Added the log_reject_target ACL modifier to specify where to log
3320 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3321 hostname. This is wrong, because it relates to the incoming message (and
3322 probably the interface on which it is arriving) and not to the outgoing
3323 callout (which could be using a different interface). This has been
3324 changed to use the value of the helo_data option from the smtp transport
3325 instead - this is what is used when a message is actually being sent. If
3326 there is no remote transport (possible with a router that sets up host
3327 addresses), $smtp_active_hostname is used.
3329 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3330 tweaks were necessary in order to get it to work (see also 21 below):
3331 (a) The code assumed that strncpy() returns a negative number on buffer
3332 overflow, which isn't the case. Replaced with Exim's string_format()
3334 (b) There were several signed/unsigned issues. I just did the minimum
3335 hacking in of casts. There is scope for a larger refactoring.
3336 (c) The code used strcasecmp() which is not a standard C function.
3337 Replaced with Exim's strcmpic() function.
3338 (d) The code set only $1; it now sets $auth1 as well.
3339 (e) A simple test gave the error "authentication client didn't specify
3340 service in request". It would seem that Dovecot has changed its
3341 interface. Fortunately there's a specification; I followed it and
3342 changed what the client sends and it appears to be working now.
3344 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3347 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3348 address A is aliased to B and C, where B exists and C does not. Without
3349 -v the output is "A verified" because verification stops after a
3350 successful redirection if more than one address is generated. However,
3351 with -v the child addresses are also verified. Exim was outputting "A
3352 failed to verify" and then showing the successful verification for C,
3353 with its parentage. It now outputs "B failed to verify", showing B's
3354 parentage before showing the successful verification of C.
3356 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3357 look up a TXT record in a specific list after matching in a combined
3360 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3361 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3362 they consult the DNS. I had assumed they would set it the way they
3363 wanted; and indeed my experiments on Linux seem to show that in some
3364 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3365 To be on the safe side, however, I have now made the interface to
3366 host_find_byname() similar to host_find_bydns(), with an argument
3367 containing the DNS resolver options. The host_find_byname() function now
3368 sets these options at its start, just as host_find_bydns() does. The smtp
3369 transport options dns_qualify_single and dns_search_parents are passed to
3370 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3371 of host_find_byname() use the default settings of RES_DEFNAMES
3372 (qualify_single) but not RES_DNSRCH (search_parents).
3374 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3375 spool_read_header() do less string testing, by means of a preliminary
3376 switch on the second character of optional "-foo" lines. (This is
3377 overdue, caused by the large number of possibilities that now exist.
3378 Originally there were few.) While I was there, I also converted the
3379 str(n)cmp tests so they don't re-test the leading "-" and the first
3380 character, in the hope this might squeeze out yet more improvement.
3382 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3383 flag allowing group syntax was set by the header_syntax check but not
3384 turned off, possible causing trouble later; (2) The flag was not being
3385 set at all for the header_verify test, causing "group"-style headers to
3386 be rejected. I have now set it in this case, and also caused header_
3387 verify to ignore an empty address taken from a group. While doing this, I
3388 came across some other cases where the code for allowing group syntax
3389 while scanning a header line wasn't quite right (mostly, not resetting
3390 the flag correctly in the right place). These bugs could have caused
3391 trouble for malformed header lines. I hope it is now all correct.
3393 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3394 with the "reply" argument non-NULL. The code, however (which originally
3395 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3396 but it didn't always do it. This confused somebody who was copying the
3397 code for some other use. I have removed all the tests.
3399 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3400 feature that was used to support insecure browsers during the U.S. crypto
3401 embargo. It requires special client support, and Exim is probably the
3402 only MTA that supported it -- and would never use it because real RSA is
3403 always available. This code has been removed, because it had the bad
3404 effect of slowing Exim down by computing (never used) parameters for the
3405 RSA_EXPORT functionality.
3407 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3408 authenticator to fail if there's a tab character in the incoming data
3409 (there should never be unless someone is messing about, as it's supposed
3410 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3411 if the connection is using TLS or if the remote IP is the same as the
3412 local IP, and the "valid-client-cert option" if a client certificate has
3415 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3416 authenticators. This can be used for authorization after authentication
3417 succeeds. (In the case of plaintext, it servers for both authentication
3420 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3421 if any retry times were supplied.
3423 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3424 connection, where there is no client IP address to check. In this
3425 situation, the verify now always succeeds.
3427 PH/25 Applied John Jetmore's -Mset patch.
3429 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3431 PH/27 In a string expansion for a processed (not raw) header when multiple
3432 headers of the same name were present, leading whitespace was being
3433 removed from all of them, but trailing whitespace was being removed only
3434 from the last one. Now trailing whitespace is removed from each header
3435 before concatenation. Completely empty headers in a concatenation (as
3436 before) are ignored.
3438 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3439 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3441 PH/29 [Removed. This was a change that I later backed out, and forgot to
3442 correct the ChangeLog entry (that I had efficiently created) before
3443 committing the later change.]
3445 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3446 address errors (4xx response to RCPT) over the same connection as other
3447 messages routed to the same hosts. Such deliveries are always "forced",
3448 so retry times are not inspected. This resulted in far too many retries
3449 for the affected addresses. The effect occurred only when there were more
3450 hosts than the hosts_max_try setting in the smtp transport when it had
3451 the 4xx errors. Those hosts that it had tried were not added to the list
3452 of hosts for which the message was waiting, so if all were tried, there
3453 was no problem. Two fixes have been applied:
3455 (i) If there are any address or message errors in an SMTP delivery, none
3456 of the hosts (tried or untried) are now added to the list of hosts
3457 for which the message is waiting, so the message should not be a
3458 candidate for sending over the same connection that was used for a
3459 successful delivery of some other message. This seems entirely
3460 reasonable: after all the message is NOT "waiting for some host".
3461 This is so "obvious" that I'm not sure why it wasn't done
3462 previously. Hope I haven't missed anything, but it can't do any
3463 harm, as the worst effect is to miss an optimization.
3465 (ii) If, despite (i), such a delivery is accidentally attempted, the
3466 routing retry time is respected, so at least it doesn't keep
3467 hammering the server.
3469 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3470 in ${readsocket because some servers need this prod.
3472 PH/32 Added some extra debug output when updating a wait-xxx database.
3474 PH/33 The hint "could be header name not terminated by colon", which has been
3475 given for certain expansion errors for a long time, was not being given
3476 for the ${if def:h_colon_omitted{... case.
3478 PH/34 The spec says: "With one important exception, whenever a domain list is
3479 being scanned, $domain contains the subject domain." There was at least
3480 one case where this was not true.
3482 PH/35 The error "getsockname() failed: connection reset by peer" was being
3483 written to the panic log as well as the main log, but it isn't really
3484 panic-worthy as it just means the connection died rather early on. I have
3485 removed the panic log writing for the ECONNRESET error when getsockname()
3488 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3489 runs only) independently of the message's sender address. This meant
3490 that, if the 4xx error was in fact related to the sender, a different
3491 message to the same recipient with a different sender could confuse
3492 things. In particular, this can happen when sending to a greylisting
3493 server, but other circumstances could also provoke similar problems.
3494 I have changed the default so that the retry time for these errors is now
3495 based a combination of the sender and recipient addresses. This change
3496 can be overridden by setting address_retry_include_sender=false in the
3499 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3500 remote server are returned as part of bounce messages. This was not
3501 happening for LMTP over a pipe (the lmtp transport), but now it is the
3502 same for both kinds of LMTP.
3504 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3505 in header lines that were in fact CNAMEs. This is no longer the case.
3507 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3508 and queue runs started by the daemon processed all messages. This has
3509 been fixed so that -R and -S can now usefully be given with -q<time>.
3511 PH/40 Import PCRE release 6.7 (fixes some bugs).
3513 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3515 PH/42 Give an error if -q is specified more than once.
3517 PH/43 Renamed the variables $interface_address and $interface_port as
3518 $received_ip_address and $received_port, to make it clear that these
3519 values apply to message reception, and not to the outgoing interface when
3520 a message is delivered. (The old names remain recognized, of course.)
3522 PH/44 There was no timeout on the connect() call when using a Unix domain
3523 socket in the ${readsocket expansion. There now is.
3525 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3526 be meaningful with "accept".
3528 SC/01 Eximstats V1.43
3529 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3531 SC/02 Eximstats V1.44
3532 Use a glob alias rather than an array ref in the generated
3533 parser. This improves both readability and performance.
3535 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3536 Collect SpamAssassin and rejection statistics.
3537 Don't display local sender or destination tables unless
3538 there is data to show.
3539 Added average volumes into the top table text output.
3541 SC/04 Eximstats V1.46
3542 Collect data on the number of addresses (recipients)
3543 as well as the number of messages.
3545 SC/05 Eximstats V1.47
3546 Added 'Message too big' to the list of mail rejection
3547 reasons (thanks to Marco Gaiarin).
3549 SC/06 Eximstats V1.48
3550 Mainlog lines which have GMT offsets and are too short to
3551 have a flag are now skipped.
3553 SC/07 Eximstats V1.49 (Alain Williams)
3554 Added the -emptyok flag.
3556 SC/08 Eximstats V1.50
3557 Fixes for obtaining the IP address from reject messages.
3559 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3560 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3561 whitespace changes from 4.64-PH/27
3563 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3566 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3567 are found, allow negative numbers in numeric criteria)
3569 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3571 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3572 to match changes made in 4.64-PH/43
3574 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3576 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3577 "domain" is actually an IP address, and force a failure. This locks out
3578 those revolvers/nameservers that support "A-for-A" lookups, in
3579 contravention of the specifications.
3581 PH/48 When a host name was looked up from an IP address, and the subsequent
3582 forward lookup of the name timed out, the host name was left in
3583 $sender_host_name, contrary to the specification.
3585 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3586 restricted to single-key lookups, Exim was not diagnosing an error if
3587 * or *@ was used with a query-style lookup.
3589 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3591 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3592 long ago, but noone interested enough thought of it. Let's just say that
3593 the "1.1" means that there are some new functions that weren't there at
3594 some point in the past.
3596 PH/51 Error processing for expansion failure of helo_data from an smtp
3597 transport during callout processing was broken.
3599 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3600 tested/used via the -bh/-bhc/-bs options.
3602 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3603 bug, fixed in subsequent PCRE releases).
3605 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3606 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3608 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3614 SC/01 Use a glob alias rather than an array ref in eximstats generated
3615 parser. This improves both readability and performance.
3617 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3618 Don't display local sender or destination tables in eximstats unless
3619 there is data to show.
3620 Added average volumes into the eximstats top table text output.
3622 SC/03 Collect data on the number of addresses (recipients) as well
3623 as the number of messages in eximstats.
3625 TF/01 Correct an error in the documentation for the redirect router. Exim
3626 does (usually) call initgroups() when daemonizing.
3628 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3629 with consistent privilege compared to when running as a daemon.
3631 TF/03 Note in the spec that $authenticated_id is not set for local
3632 submissions from trusted users.
3634 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3635 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3637 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3638 by adding some example configuration directives to the default
3639 configuration file. A little bit of work is required to uncomment the
3640 directives and define how usernames and passwords are checked, but
3641 there is now a framework to start from.
3643 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3644 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3645 without this. I don't know how relevant this is to other LDAP libraries.
3647 PH/02 Add the verb name to the "unknown ACL verb" error.
3649 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3651 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3653 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3654 directory not expanded when it should be if an expanded home directory
3655 was set for the address (which is overridden by the transport).
3657 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3660 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3661 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3662 because it is too late at that time, and has no effect.
3664 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3665 security issue with \' (bugzilla #107). I could not use the
3666 PQescapeStringConn() function, because it needs a PGconn value as one of
3669 PH/08 When testing addresses using -bt, indicate those final addresses that
3670 are duplicates that would not cause an additional delivery. At least one
3671 person was confused, thinking that -bt output corresponded to deliveries.
3672 (Suppressing duplicates isn't a good idea as you lose the information
3673 about possibly different redirections that led to the duplicates.)
3675 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3676 systems where poll() doesn't work, in particular OS X.
3678 PH/10 Added more information to debugging output for retry time not reached.
3680 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3681 operations in malware.c.
3683 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3686 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3687 syslog_duplication set false, log lines that would normally be written
3688 both the the main log and to the reject log were not written to syslog at
3691 PH/14 In the default configuration, change the use of "message" in ACL warn
3692 statements to "add_header".
3694 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3695 not followed by a command (e.g. "seen endif").
3697 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3698 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3701 PH/17 Added extra conditions to the default value of delay_warning_condition
3705 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3706 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3707 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3710 The Auto-Submitted: and various List- headers are standardised, whereas I
3711 don't think Precedence: ever was.
3713 PH/18 Refactored debugging code in route_finduser() to show more information,
3714 in particular, the error code if getpwnam() issues one.
3716 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3717 This is apparently needed in addition to the PH/07 change above to avoid
3718 any possible encoding problems.
3720 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3721 but not after initializing Perl.
3723 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3724 output them only if debugging. By default they are written stderr,
3725 apparently, which is not desirable.
3727 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3730 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3733 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3735 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3736 authenticated or an ident call has been made. Suppress the default
3737 values for $authenticated_id and $authenticated_sender (but permit -oMai
3738 and -oMas) when testing with -bh.
3740 PH/24 Re-jigged the order of the tests in the default configuration so that the
3741 tests for valid domains and recipients precede the DNS black list and CSA
3742 tests, on the grounds that those ones are more expensive.
3744 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3745 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3746 This bug exists in every version of Exim that I still have, right back to
3749 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3750 However, an attempt to turn on case-sensitivity in a regex key by
3751 including (?-i) didn't work because the subject string was already
3752 lowercased, and the effects were non-intuitive. It turns out that a
3753 one-line patch can be used to allow (?-i) to work as expected.
3759 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3760 other effects) broke the use of negated acl sub-conditions.
3762 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3765 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3766 "Deny" causes Exim to reject the incoming connection with a 554 error.
3767 Unfortunately, if there is a major crisis, such as a disk failure,
3768 tcp-wrappers gives "deny", whereas what one would like would be some
3769 kind of temporary error. A kludge has been added to help with this.
3770 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3771 554 error is used if errno is still zero or contains ENOENT (which occurs
3772 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3775 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3777 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3778 errors. Otherwise a message that provokes a temporary error (when other
3779 messages do not) can cause a whole host to time out.
3781 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3782 addresses were routed directly to files or pipes from a redirect router.
3783 File deliveries just didn't batch; pipe deliveries might have suffered
3786 PH/06 A failure to get a lock for a hints database would erroneously always say
3787 "Failed to get write lock", even when it was really a read lock.
3789 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3790 of 0600. This has been changed to use the value of the lockfile_mode
3791 option (which defaults to 0600).
3793 PH/08 Applied small patch from the Sieve maintainer.
3795 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3796 folder from quota calculations, a direct delivery into this folder messed
3797 up the contents of the maildirsize file. This was because the regex was
3798 used only to exclude .Trash (or whatever) when the size of the mailbox
3799 was calculated. There was no check that a delivery was happening into an
3800 excluded directory. This bug has been fixed by ignoring all quota
3801 processing for deliveries into excluded directories.
3803 PH/10 Added the maildirfolder_create_regex option to appendfile.
3809 PH/01 The code for finding all the local interface addresses on a FreeBSD
3810 system running IPv6 was broken. This may well have applied to all BSD
3811 systems, as well as to others that have similar system calls. The broken
3812 code found IPv4 interfaces correctly, but gave incorrect values for the
3813 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3814 that it would not match correctly against @[] and not recognize the IPv6
3817 PH/02 The ipliteral router was not recognizing addresses of the form user@
3818 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3820 PH/03 Added disable_ipv6.
3822 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3823 decoded form, because it is most often used to construct To: headers
3824 lines in autoreplies, and the decoded form may well be syntactically
3825 invalid. However, $reply_address has leading white space removed, and all
3826 newlines turned into spaces so that the autoreply transport does not
3829 PH/05 If group was specified without a user on a router, and no group or user
3830 was specified on a transport, the group from the router was ignored.
3832 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3833 for visible compile-time settings that can be used to change these
3834 numbers, for those that want even more. Backwards compatibility with old
3835 spool files has been maintained. However, going back to a previous Exim
3836 release will lost any variables that are in spool files.
3838 PH/07 Two small changes when running in the test harness: increase delay when
3839 passing a TCP/IP connection to a new process, in case the original
3840 process has to generate a bounce, and remove special handling of
3841 127.0.0.2 (sic), which is no longer necessary.
3843 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3844 be the same on different OS.
3846 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3849 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3850 whether --show-vars was specified or not
3852 JJ/02 exipick: Added support for new ACL variable spool format introduced
3855 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3856 syntactically invalid From: or Reply-to: line, and a filter used this to
3857 generate an autoreply, and therefore failed to obtain an address for the
3858 autoreply, Exim could try to deliver to a non-existent relative file
3859 name, causing unrelated and misleading errors. What now happens is that
3860 it logs this as a hard delivery error, but does not attempt to create a
3863 PH/11 The exinext utility has a -C option for testing purposes, but although
3864 the given file was scanned by exinext itself; it wasn't being passed on
3865 when Exim was called.
3867 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3868 an end-of-file indication when reading a command response.
3870 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3871 compiled. In many other places in Exim, IPv6 addresses are always
3872 recognized, so I have changed this. It also means that IPv4 domain
3873 literals of the form [IPV4:n.n.n.n] are now always recognized.
3875 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3876 used if the router is not running as root, for example, when verifying at
3877 ACL time, or when using -bh. The debugging output from this situation was
3878 non-existent - all you got was a failure to exec. I have made two
3881 (a) Failures to set uid/gid, the current directory, or a process leader
3882 in a subprocess such as that created by queryprogram now generate
3883 suitable debugging output when -d is set.
3885 (b) The queryprogram router detects when it is not running as root,
3886 outputs suitable debugging information if -d is set, and then runs
3887 the subprocess without attempting to change uid/gid.
3889 PH/15 Minor change to Makefile for building test_host (undocumented testing
3892 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3893 additional section of a DNS packet that returns MX or SRV records.
3894 Instead, it always explicitly searches for A/AAAA records. This avoids
3895 major problems that occur when a DNS server includes only records of one
3896 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3897 fixed another bug: if SRV records were looked up and the corresponding
3898 address records were *not* found in the additional section, the port
3899 values from the SRV records were lost.
3901 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3902 using the correct key (the original address) when searching the retry
3903 rules in order to find which one to use for generating the retry hint.
3905 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3906 adding the default one. Similarly, if it contains a Reply-To: header, the
3907 errors_reply_to option, if set, is not used.
3909 PH/19 When calculating a retry time, Exim used to measure the "time since
3910 failure" by looking at the "first failed" field in the retry record. Now
3911 it does not use this if it is later than than the arrival time of the
3912 message. Instead it uses the arrival time. This makes for better
3913 behaviour in cases where some deliveries succeed, thus re-setting the
3914 "first failed" field. An example is a quota failure for a huge message
3915 when small messages continue to be delivered. Without this change, the
3916 "time since failure" will always be short, possible causing more frequent
3917 delivery attempts for the huge message than are intended.
3918 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3920 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3921 $1, $2, $3) because the numerical variables can be reset during some
3922 expansion items (e.g. "match"), thereby losing the authentication data.
3924 PH/21 Make -bV show the size of off_t variables so that the test suite can
3925 decide whether to run tests for quotas > 2G.
3927 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3928 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3929 filecount value is greater than 2G or if a quota value is greater than 2G
3930 on a system where the size of off_t is not greater than 4, a panic error
3933 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3934 never match. The debug and -bh output now contains an explicit error
3935 message indicating a malformed IPv4 address or mask.
3937 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3938 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3939 PH/23 above applies.
3941 PH/25 Do not write to syslog when running in the test harness. The only
3942 occasion when this arises is a failure to open the main or panic logs
3943 (for which there is an explicit test).
3945 PH/26 Added the /no_tell option to "control=freeze".
3947 PH/27 If a host name lookup failed very early in a connection, for example, if
3948 the IP address matched host_lookup and the reverse lookup yielded a name
3949 that did not have a forward lookup, an error message of the form "no IP
3950 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3951 could be logged. Now it outputs the IP address instead of "NULL".
3953 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3954 allows the sender and the authenticated sender to be set when
3955 submitting a message from within Exim. Since child_open_exim() is
3956 documented for local_scan(), the new function should be too.
3958 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3959 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3960 results in an empty string is now treated as unset.
3962 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3964 PH/31 Added sender_verify_fail logging option.
3966 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3967 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3968 tidied the source and removed it altogether.
3970 PH/33 When a queue run was abandoned because the load average was too high, a
3971 log line was always written; now it is written only if the queue_run log
3972 selector is set. In addition, the log line for abandonment now contains
3973 information about the queue run such as the pid. This is always present
3974 in "start" and "stop" lines but was omitted from the "abandon" line.
3976 PH/34 Omit spaces between a header name and the colon in the error message that
3977 is given when verify = headers_syntax fails (if there are lots of them,
3978 the message gets confusing).
3980 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3981 names, as there are now some PTR records that contain slashes. This check
3982 is only to protect against broken name servers that fall over on strange
3983 characters, so the fact that it applies to all lookups doesn't matter.
3985 PH/36 Now that the new test suite is complete, we can remove some of the
3986 special code in Exim that was needed for the old test suite. For example,
3987 sorting DNS records because real resolvers return them in an arbitrary
3988 order. The new test suite's fake resolver always returns records in the
3991 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3992 bounces) except when queue_only is set, to avoid logging races between
3993 the different processes.
3995 PH/38 Panic-die if .include specifies a non-absolute path.
3997 PH/39 A tweak to the "H" retry rule from its user.
3999 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4000 a label. They prevented compilation on older perls.
4002 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4003 a warning to be raised on newish perls.
4005 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4006 on queue. Changes to match documented behaviour of showing count of
4007 messages matching specified criteria.
4009 PH/40 Changed the default ident timeout from 30s to 5s.
4011 PH/41 Added support for the use of login_cap features, on those BSD systems
4012 that have them, for controlling the resources used by pipe deliveries.
4014 PH/42 The content-scanning code uses fopen() to create files in which to put
4015 message data. Previously it was not paying any attention to the mode of
4016 the files. Exim runs with umask(0) because the rest of the code creates
4017 files with open(), and sets the required mode explicitly. Thus, these
4018 files were ending up world-writeable. This was not a big issue, because,
4019 being within the spool directory, they were not world-accessible. I have
4020 created a function called modefopen, which takes an additional mode
4021 argument. It sets umask(777), creates the file, chmods it to the required
4022 mode, then resets the umask. All the relevant calls to fopen() in the
4023 content scanning code have been changed to use this function.
4025 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4026 to 24 hours. This avoids potential overflow problems when processing G
4027 and H retry rules. I suspect nobody ever tinkers with this value.
4029 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4031 PH/45 When the plaintext authenticator is running as a client, the server's
4032 challenges are checked to ensure they are valid base64 strings. By
4033 default, the authentication attempt is cancelled if an invalid string is
4034 received. Setting client_ignore_invalid_base64 true ignores these errors.
4035 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4036 they are received. Thus, the responses can be made to depend on the
4037 challenges. If an invalid string is ignored, an empty string is placed in
4040 PH/46 Messages that are created by the autoreply transport now contains a
4041 References: header, in accordance with RFCs 2822 and 3834.
4043 PH/47 Added authenticated_sender_force to the smtp transport.
4045 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4047 PH/49 Installed latest patch from the Sieve maintainer.
4049 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4050 was also set, the appendfile transport was unnecessarily scanning a
4051 directory of message files (e.g. for maildir delivery) to find the count
4052 of files (along with the size), even though it did not need this
4053 information. It now does the scan only if it needs to find either the
4054 size of the count of files.
4056 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4058 PH/52 Two bugs concerned with error handling when the smtp transport is
4061 (i) Exim was not creating retry information for temporary errors given
4062 for individual recipients after the DATA command when the smtp transport
4063 was used in LMTP mode. This meant that they could be retried too
4064 frequently, and not timed out correctly.
4066 (ii) Exim was setting the flag that allows error details to be returned
4067 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4068 recipients that were returned after the DATA command.
4070 PH/53 This is related to PH/52, but is more general: for any failing address,
4071 when detailed error information was permitted to be returned to the
4072 sender, but the error was temporary, then after the final timeout, only
4073 "retry timeout exceeded" was returned. Now it returns the full error as
4074 well as "retry timeout exceeded".
4076 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4077 do this, and (what is worse) MTAs that accept it.
4079 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4080 will now be deprecated.
4082 PH/56 New os.c-cygwin from the Cygwin maintainer.
4084 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4085 formats (previously only available in exim formats via -bpr, -bpru,
4086 and -bpra. Now also available in native and exiqgrep formats)
4088 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4089 with very large, slow to parse queues
4091 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4093 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4095 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4096 responses to authentication challenges, though it was showing the
4097 challenges; (ii) I've removed the CR characters from the debug output for
4100 PH/58 Allow for the insertion of a newline as well as a space when a string
4101 is turned into more than one encoded-word during RFC 2047 encoding. The
4102 Sieve code now uses this.
4104 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4105 data_4xx, lost_connection, tls_required.
4107 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4108 message was being sent as an SMTP response.
4110 PH/61 Add -l and -k options to exicyclog.
4112 PH/62 When verifying, if an address was redirected to one new address, so that
4113 verification continued, and the new address failed or deferred after
4114 having set something in $address_data, the value of $address_data was not
4115 passed back to the ACL. This was different to the case when no
4116 redirection occurred. The value is now passed back in both cases.
4118 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4119 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4120 use login_cap.h, so on its own it isn't the distinguishing feature. The
4121 new name refers directly to the setclassresources() function.
4123 PH/65 Added configuration files for NetBSD3.
4125 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4127 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4128 is preferred over IPv4.
4130 PH/68 The bounce_return_message and bounce_return_body options were not being
4131 honoured for bounces generated during the reception of non-SMTP messages.
4132 In particular, this applied to messages rejected by the ACL. This bug has
4133 been fixed. However, if bounce_return_message is true and bounce_return_
4134 body is false, the headers that are returned for a non-SMTP message
4135 include only those that have been read before the error was detected.
4136 (In the case of an ACL rejection, they have all been read.)
4138 PH/69 The HTML version of the specification is now built in a directory called
4139 spec_html instead of spec.html, because the latter looks like a path with
4140 a MIME-type, and this confuses some software.
4142 PH/70 Catch two compiler warnings in sieve.c.
4144 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4145 function verify_get_ident() calls ip_connect() to connect a socket, but
4146 if the "connect()" function timed out, ip_connect() used to close the
4147 socket. However, verify_get_ident() also closes the socket later, and in
4148 between Exim writes to the log, which may get opened at this point. When
4149 the socket was closed in ip_connect(), the log could get the same file
4150 descriptor number as the socket. This naturally causes chaos. The fix is
4151 not to close the socket in ip_connect(); the socket should be closed by
4152 the function that creates it. There was only one place in the code where
4153 this was missing, in the iplookup router, which I don't think anybody now
4154 uses, but I've fixed it anyway.
4156 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4157 well as to direct DNS lookups. Otherwise the handling of names in host
4158 lists is inconsistent and therefore confusing.
4164 PH/01 Two changes to the default runtime configuration:
4166 (1) Move the checks for relay_from_hosts and authenticated clients from
4167 after to before the (commented out) DNS black list checks.
4169 (2) Add control=submission to the relay_from_hosts and authenticated
4170 clients checks, on the grounds that messages accepted by these
4171 statements are most likely to be submissions.
4173 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4175 (1) Generate an error if the third argument for the ${prvs expansion is
4178 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4181 (3) Reset the variables that are obtained from the first argument of
4182 ${prvscheck and used in the second argument before leaving the code,
4183 because their memory is reclaimed, so using them afterwards may do
4186 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4187 one (it's much easier than Tom thought :-).
4189 (5) Because of (4), we can now allow for the use of $prvscheck_result
4190 inside the third argument.
4192 PH/03 For some reason, the default setting of PATH when running a command from
4193 a pipe transport was just "/usr/bin". I have changed it to
4196 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4197 anything to be listed in the output from -bV.
4199 PH/05 When a filter generated an autoreply, the entire To: header line was
4200 quoted in the delivery log line, like this:
4202 => >A.N.Other <ano@some.domain> <original@ddress> ...
4204 This has been changed so that it extracts the operative address. There
4205 may be more than one such address. If so, they are comma-separated, like
4208 => >ano@some.domain,ona@other.domain <original@ddress> ...
4210 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4211 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4212 not being looked up in the rDNS to get a host name, Exim was showing the
4213 IP address twice in Received: lines, even though the IP addresses were
4214 identical. For example:
4216 Received: from [1.2.3.4] (helo=[1.2.3.4])
4218 However, if the real host name was known, it was omitting the HELO data
4219 if it matched the actual IP address. This has been tidied up so that it
4220 doesn't show the same IP address twice.
4222 PH/07 When both +timestamp and +memory debugging was on, the value given by
4223 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4224 called by the debug printing, thereby overwriting the timestamp buffer.
4225 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4227 PH/08 When the original message was included in an autoreply transport, it
4228 always said "this is a copy of the message, including all the headers",
4229 even if body_only or headers_only was set. It now gives an appropriate
4232 PH/09 Applied a patch from the Sieve maintainer which:
4234 o fixes some comments
4235 o adds the (disabled) notify extension core
4236 o adds some debug output for the result of if/elsif tests
4237 o points to the current vacation draft in the documentation
4238 and documents the missing references header update
4242 o fixes a bug in processing the envelope test (when testing
4243 multiple envelope elements, the last element determined the
4246 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4247 Electronic Mail") by including:
4249 Auto-submitted: auto-generated
4251 in the messages that it generates (bounce messages and others, such as
4252 warnings). In the case of bounce messages for non-SMTP messages, there was
4253 also a typo: it was using "Auto_submitted" (underscore instead of
4254 hyphen). Since every message generated by Exim is necessarily in response
4255 to another message, thes have all been changed to:
4257 Auto-Submitted: auto-replied
4259 in accordance with these statements in the RFC:
4261 The auto-replied keyword:
4263 - SHOULD be used on messages sent in direct response to another
4264 message by an automatic process,
4266 - MUST NOT be used on manually-generated messages,
4268 - MAY be used on Delivery Status Notifications (DSNs) and Message
4269 Disposition Notifications (MDNs),
4271 - MUST NOT be used on messages generated by automatic or periodic
4272 processes, except for messages which are automatic responses to
4275 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4276 to the default Received: header definition.
4278 PH/12 Added log selector acl_warn_skipped (default on).
4280 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4281 variables because (a) they are in the wrong storage pool and (b) even if
4282 they were copied, it wouldn't work properly because of the caching.
4284 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4285 checking when decoding. Apparently there are clients that generate
4286 overlong encoded strings. Why am I not surprised?
4288 PH/15 If the first argument of "${if match_address" was not empty, but did not
4289 contain an "@" character, Exim crashed. Now it writes a panic log message
4290 and treats the condition as false.
4292 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4294 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4295 extension "envelope-auth". The code is finished and in agreement with
4296 other implementations, but there is no documentation so far and in fact,
4297 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4298 not changing the active code.
4300 Print executed "if" and "elsif" statements when debugging is used. This
4301 helps a great deal to understand what a filter does.
4303 Document more things not specified clearly in RFC3028. I had all this
4304 sorted out, when out of a sudden new issues came to my mind. Oops."
4306 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4309 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4310 canonical form (as documented). However, after a host name lookup from
4311 the IP address, check_host() was doing a simple string comparison with
4312 addresses acquired from the DNS when checking that the found name did
4313 have the original IP as one of its addresses. Since any found IPv6
4314 addresses are likely to be in abbreviated form, the comparison could
4315 fail. Luckily, there already exists a function for doing the comparison
4316 by converting both addresses to binary, so now that is used instead of
4317 the text comparison.
4319 PH/20 There was another similar case to PH/19, when a complete host name was
4320 given in a host list; looking up its IP address could give an abbreviated
4321 form, whereas the current host's name might or might not be abbreviated.
4322 The same fix has been applied.
4328 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4329 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4332 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4333 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4335 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4337 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4338 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4339 Also, for "Auto-Submitted", treat anything other than "no" as
4340 non-personal, in accordance with RFC 3834. (Previously it treated
4341 anything starting "auto-" as non-personal.)
4343 TF/01 The control=submission/name=... option had a problem with syntax
4344 errors if the name included a slash character. The /name= option
4345 now slurps the rest of the string, so it can include any characters
4346 but it must come last in the list of options (after /sender_retain
4349 PH/05 Some modifications to the interface to the fake nameserver for the new
4357 TK/01 Added the "success_on_redirect" address verification option. See
4358 NewStuff for rationale and an example.
4360 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4362 PH/02 Patch to exigrep to allow it to work on syslog lines.
4364 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4365 fread() to skip over the body file's header line, because in Cygwin the
4366 header line is locked and is inaccessible.
4368 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4369 co-exist for some time) to make it clear that it is the Exim ID that is
4370 referenced, not the Message-ID: header line.
4372 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4373 string_format() function, because snprintf() does not exist on all
4376 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4377 ${stat: expansion item.
4379 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4380 protocol synchronization error", to keep the pedants happy.
4382 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4383 well as for IRIX systems, when gcc is being used. See the host.c source
4386 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4388 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4391 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4392 transport and to the smtp transport in LMTP mode.
4394 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4396 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4397 run a filter in a subprocess. This could lead to confusion in subsequent
4398 lookups in the parent process. There should also be a search_tidyup() at
4399 the end of the subprocess.
4401 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4402 only if the host matched helo_try_verify_hosts, which caused the
4403 verification to occur when the EHLO/HELO command was issued. The ACL just
4404 tested the remembered result. Now, if a previous verification attempt has
4405 not happened, "verify = helo" does it there and then.
4407 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4409 TK/03 Fix log output including CR from clamd.
4411 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4412 exist provoked a memory error which could cause a segfault.
4414 PH/15 Installed PCRE 6.2
4416 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4418 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4419 of the problem. Specifically, suggested +O2 rather than +O1 for the
4422 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4424 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4425 record, the port was not correctly logged when the outgoing_port log
4426 selector was set (it logged the transort's default port).
4428 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4429 fallback_hosts, and "hosts" in the smtp transport.
4431 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4432 host errors such as "Connection refused".
4434 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4435 authentication with radiusclient 0.4.9:
4437 - Error returned from rc_read_config was caught wrongly
4438 - Username/password not passed on to radius server due to wrong length.
4440 The presumption is that some radiusclient API changes for 4.51/PH/17
4441 were not taken care of correctly. The code is still untested by me (my
4442 Linux distribution still has 0.3.2 of radiusclient), but it was
4443 contributed by a Radius user.
4445 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4446 expanding the "port" option of the smtp transport.
4448 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4449 while reading a MIME header. Thanks to Tom Hughes for a patch.
4451 PH/24 Include config.h inside local_scan.h so that configuration settings are
4454 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4455 that in an ACL for RCPT (for example), you can examine exactly what was
4458 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4459 commands, but it was not correctly comparing the address with the actual
4460 client host address. Thus, it would show the EHLO address in Received:
4461 header lines when this was not necessary.
4463 PH/27 Added the % operator to ${eval:}.
4465 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4466 it should be ignoring failures (because with -C, for example, it has lost
4467 privilege). It wasn't ignoring creation failures other than "already
4470 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4473 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4474 because an input line was too long, either on its own, or by virtue of
4475 too many continuations, the temporary file was not being removed, and the
4476 return code was incorrect.
4478 PH/31 Missing "BOOL" in function definition in filtertest.c.
4480 PH/32 Applied Sieve patches from the maintainer.
4482 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4484 PH/33 Added "verify = not_blind".
4486 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4487 Local/Makefile (with some defaults set). These are used in built scripts
4488 such as exicyclog, but they have never been used in the exim_install
4489 script (though there are many overriding facilities there). I have
4490 arranged that the exim_install script now takes note of these two
4493 PH/35 Installed configuration files for Dragonfly.
4495 PH/36 When a locally submitted message by a trusted user did not contain a
4496 From: header, and the sender address was obtained from -f or from an SMTP
4497 MAIL command, and the trusted user did not use -F to supply a sender
4498 name, $originator_name was incorrectly used when constructing a From:
4499 header. Furthermore, $originator_name was used for submission mode
4500 messages from external hosts without From: headers in a similar way,
4501 which is clearly wrong.
4503 PH/37 Added control=suppress_local_fixups.
4505 PH/38 When log_selector = +received_sender was set, and the addition of the
4506 sender made the log line's construction buffer exactly full, or one byte
4507 less than full, an overflow happened when the terminating "\n" was
4510 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4511 when the result of a list match is failure because a DNS lookup failed.
4513 PH/40 RM_COMMAND is now used in the building process.
4515 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4516 the "build-* directories that it finds.
4518 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4519 address in a domain literal was a prefix of an interface address.
4521 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4522 when verifying a sender address, unless rewrite_headers is false.
4524 PH/44 Wrote a long comment about why errors_to addresses are verified as
4525 recipients, not senders.
4527 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4528 the ratelimit ACL was added.
4530 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4532 PH/46 Added extra information about PostgreSQL errors to the error string.
4534 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4535 avoiding the need to install special zones in a real server. This is
4536 backwards compatible; if it can't find the fake resolver, it drops back.
4537 Thus, both old and new test suites can be run.
4539 TF/02 Added util/ratelimit.pl
4541 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4542 clock is set back in time.
4544 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4545 Candler <B.Candler@pobox.com>.
4547 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4548 OK for addresses that are the result of redirections.
4550 PH/48 A number of further additions for the benefit of the new test suite,
4551 including a fake gethostbyname() that interfaces to the fake DNS resolver
4554 TF/06 The fix for widen_domains has also been applied to qualify_single and
4555 search_parents which are the other dnslookup options that can cause
4558 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4561 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4563 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4564 service" if it encountered an MX record with an empty target hostname.
4565 The message is now "an MX or SRV record indicated no SMTP service".
4567 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4568 if the DNS of the sending site is misconfigured. This is quite a
4569 common situation. This change restores the behaviour of treating a
4570 helo verification defer as a failure.
4572 PH/51 If self=fail was set on a router, the bounce message did not include the
4573 actual error message.
4579 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4581 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4582 command itself ended in error, the underlying message about the transport
4583 filter timeout was being overwritten with the pipe command error. Now the
4584 underlying error message should be appended to the second error message.
4586 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4588 PH/02 Reduce the amount of output that "make" produces by default. Full output
4589 can still be requested.
4591 PH/03 The warning log line about a condition test deferring for a "warn" verb
4592 was being output only once per connection, rather than after each
4593 occurrence (because it was using the same function as for successful
4594 "warn" verbs). This seems wrong, so I have changed it.
4596 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4597 it should not have, which might have caused a crash in the right
4598 circumstances, but probably never did.
4600 PH/04 Installed a modified version of Tony Finch's patch to make submission
4601 mode fix the return path as well as the Sender: header line, and to
4602 add a /name= option so that you can make the user's friendly name appear
4605 TF/03 Added the control = fakedefer ACL modifier.
4607 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4608 Mark Lowes for thorough testing.
4610 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4612 TK/03 Merged latest SRS patch from Miles Wilton.
4614 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4615 with the definition in sysexits.h (which is #included earlier).
4616 Fortunately, Exim does not actually use EX_OK. The code used to try to
4617 preserve the sysexits.h value, by assuming that macro definitions were
4618 scanned for macro replacements. I have been disabused of this notion,
4619 so now the code just undefines EX_OK before #including unistd.h.
4621 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4622 in the smtp transport. When a block could not be written in a single
4623 write() function, the timeout was being re-applied to each part-write.
4624 This seems wrong - if the receiver was accepting one byte at a time it
4625 would take for ever. The timeout is now adjusted when this happens. It
4626 doesn't have to be particularly precise.
4628 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4629 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4631 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4632 without a domain if the check to <postmaster@domain> fails.
4634 SC/01 Eximstats: added -xls and the ability to specify output files
4635 (patch written by Frank Heydlauf).
4637 SC/02 Eximstats: use FileHandles for outputting results.
4639 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4641 SC/04 Eximstats: fixed display of large numbers with -nvr option
4643 SC/05 Eximstats: fixed merging of reports with empty tables.
4645 SC/06 Eximstats: added the -include_original_destination flag
4647 SC/07 Eximstats: removed tabs and trailing whitespace.
4649 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4651 TK/06 MBOX spool code: Add real "From " MBOX separator line
4652 so the .eml file is really in mbox format (even though
4653 most programs do not really care). Patch from Alex Miller.
4655 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4656 The latter is generated from $received_to and is only set if the
4657 message has one envelope recipient. SA can use these headers,
4658 obviously out-of-the-box. Patch from Alex Miller.
4660 PH/08 The ${def test on a variable was returning false if the variable's
4661 value was "0", contrary to what the specification has always said!
4662 The result should be true unless the variable is empty.
4664 PH/09 The syntax error of a character other than { following "${if
4665 def:variable_name" (after optional whitespace) was not being diagnosed.
4666 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4667 accidental colon was present, for example, could give incorrect results.
4669 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4670 result is used (not including appendfile, where other changes are about
4673 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4674 This involved changing a lot of size variables from int to off_t. It
4675 should work with maildirs and everything.
4677 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4678 spamd dying while we are connected to it.
4680 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4683 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4684 function for BDB 4.3.
4686 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4688 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4689 the sources for PCRE have been re-arranged and more files are now
4692 PH/15 The code I had for printing potentially long long variables in PH/11
4693 above was not the best (it lost precision). The length of off_t variables
4694 is now inspected at build time, and an appropriate printing format (%ld
4695 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4696 to be "long long int" or "long int". This is needed for the internal
4697 formatting function string_vformat().
4699 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4700 the configuration file to be ":syslog", then the script "guesses" where
4701 the logs files are, rather than using the compiled in default. In our
4702 case the guess is not the same as the compiled default, so the script
4703 suddenly stopped working when I started to use syslog. The patch checks
4704 to see if log_file_path is "". If so, it attempts to read it from exim
4705 with no configuration file to get the compiled in version, before it
4706 falls back to the previous guessing code."
4708 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4709 implementing BATV in an Exim configuration. See NewStuff for the gory
4712 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4713 Makefile that are specific to HP-UX.
4715 PH/18 If the "use_postmaster" option was set for a recipient callout together
4716 with the "random" option, the postmaster address was used as the MAIL
4717 FROM address for the random test, but not for the subsequent recipient
4718 test. It is now used for both.
4720 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4721 patch removes a few documentation additions to RFC 3028, because the
4722 latest draft now contains them. It adds the new en;ascii-case comparator
4723 and a new error check for 8bit text in MIME parts. Comparator and
4724 require names are now matched exactly. I enabled the subaddress
4725 extension, but it is not well tested yet (read: it works for me)."
4727 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4728 rework some of the code of TK/09 above to avoid the hardwired use of
4729 "%lld" and "long long". Replaced the call to snprintf() with a call to
4732 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4733 records point to non-existent hosts", "retry timeout exceeded", and
4734 "retry time not reached for any host after a long failure period".
4736 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4737 experimental DomainKeys support:
4739 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4740 (2) On an error such as an illegally used "control", the wrong name for
4741 the control was given.
4743 These problems did NOT occur unless DomainKeys support was compiled.
4745 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4747 PH/24 Added ${if match_ip condition.
4749 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4750 databases so that it will be absolutely obvious if a crash occurs in the
4751 DB library. This is a regular occurrence (often caused by mis-matched
4754 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4755 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4756 picked up on a user's system that detects such things. There doesn't seem
4757 to be a gcc warning option for this - only an attribute that has to be
4758 put on the function's prototype. It seems that in Fedora Core 4 they have
4759 set this on a number of new functions. No doubt there will be more in due
4762 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4763 specify a transport. However, if an address that was verified by such a
4764 router was the subject of a callout, Exim crashed because it tried to
4765 read the rcpt_include_affixes from the non-existent transport. Now it
4766 just assumes that the setting of that option is false. This bug was
4767 introduced by 4.51/PH/31.
4769 PH/28 Changed -d+all to exclude +memory, because that information is very
4770 rarely of interest, but it makes the output a lot bigger. People tend to
4771 do -d+all out of habit.
4773 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4774 code in os-type was giving problems when libc.so lives in lib64, like on
4777 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4778 aren't the modern standard, and it seems that some systems' include files
4779 don't always have them. Exim was already checking for some of the newer
4780 ones like T_AAAA, and defining it itself. I've added checks for all the
4781 record types that Exim uses.
4783 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4784 not automatically generating a new one, as it is supposed to. This
4785 prevented TLS from working. If the file did exist, but contained invalid
4786 data, a new version was generated, as expected. It was only the case of a
4787 non-existent file that was broken.
4789 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4790 with a change in libdomainkeys > 0.64.
4792 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4793 from DNS. If the selector record carries the flag, it now has
4794 precedence over the domain-wide flag.
4796 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4798 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4799 the use of a port name that isn't defined in /etc/services) occurred, the
4800 message was deferred as in a normal delivery, and thus remained on the
4801 spool, instead of being failed because of the mua_wrapper setting. This
4802 is now fixed, and I tidied up some of the mua_wrapper messages at the
4805 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4806 the messages in a hash of arrays rather than using individual hashes.
4807 This is a bit cleaner and results in dramatic memory savings, albeit
4808 at a slight CPU cost.
4810 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4811 as requested by Marc Sherman.
4813 SC/10 Eximstats: added histograms for user specified patterns as requested
4816 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4818 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4819 fopen() in the content-scanning modules that did not already have it.
4825 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4826 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4828 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4830 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4832 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4833 to test Sieve filters that use "vacation".
4835 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4836 that changes the way the GnuTLS parameters are stored in the cache file.
4837 The new format can be generated externally. For backward compatibility,
4838 if the data in the cache doesn't make sense, Exim assumes it has read an
4839 old-format file, and it generates new data and writes a new file. This
4840 means that you can't go back to an older release without removing the
4843 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4844 work if there are any delivery delays because "one_time" forces the
4845 parent to be marked "delivered", so its unseen clone is never tried
4846 again. For this reason, Exim now forbids the simultaneous setting of
4849 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4850 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4851 entry if you want to know the details. The fix, however, neglected to
4852 consider the case when local delivery batching is involved. The test for
4853 "previously delivered" was not happening when checking to see if an
4854 address could be batched with a previous (undelivered) one; under
4855 certain circumstances this could lead to multiple deliveries to the same
4858 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4859 in its include files, and this causes problems building Exim.
4861 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4862 header_syntax) but Exim was just ignoring anything given after a slash.
4863 In particular, this caused confusion with an attempt to use "verify =
4864 reverse_host_lookup/defer_ok". An error is now given when options are
4865 supplied for verify items that do not have them. (Maybe reverse_host_
4866 lookup should have a defer_ok option, but that's a different point.)
4868 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4869 defined by RFC 821) to 2048, because there were problems with some AUTH
4870 commands, and RFC 1869 says the size should be increased for extended
4871 SMTP commands that take arguments.
4873 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4876 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4877 "unknown" error; now it says that the functionality isn't in the binary.
4879 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4880 an address' error message when a string expansion fails (syntax or
4881 whatever). Otherwise the password may appear in the log. Following change
4882 PH/42 below, there is no longer a chance of it appearing in a bounce
4885 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4887 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4888 of its addresses was ever tried. (Bugzilla bug #2.)
4890 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4891 the result incorrectly in the debug output. (It correctly added a newline
4892 to what was transported.)
4894 TF/01 Added $received_time.
4896 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4897 commented out examples of how to interface to a virus scanner and to
4898 SpamAssassin. Also added commented examples of av_scanner and
4899 spamd_address settings.
4901 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4902 and controls are allowed in which ACLs. There were a couple of minor
4903 errors. Some of the entries in the conditions table (which is a table of
4904 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4905 negation of where the condition IS allowed.
4907 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4909 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4910 header file does not have a version number, so I've had to invent a new
4911 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4912 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4913 radiusclient), but it was contributed by a Radius user.
4915 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4916 files or directories, for OpenSSL.
4918 PH/19 When an Exim process that is running as root has to create an Exim log
4919 file, it does so in a subprocess that runs as exim:exim so as to get the
4920 ownership right at creation (otherwise, other Exim processes might see
4921 the file with the wrong ownership). There was no test for failure of this
4922 fork() call, which would lead to the process getting stuck as it waited
4923 for a non-existent subprocess. Forks do occasionally fail when resources
4924 run out. I reviewed all the other calls to fork(); they all seem to check
4927 PH/20 When checking for unexpected SMTP input at connect time (before writing
4928 the banner), Exim was not dealing correctly with a non-positive return
4929 from the read() function. If the client had disconnected by this time,
4930 the result was a log entry for a synchronization error with an empty
4931 string after "input=" when read() returned zero. If read() returned -1
4932 (an event I could not check), uninitialized data bytes were printed.
4933 There were reports of junk text (parts of files, etc) appearing after
4936 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4938 PH/22 Added support for macro redefinition, and (re)definition in between
4939 driver and ACL definitions.
4941 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4942 forgetting to use the resulting value; it was using the unexpanded value.
4944 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4945 hadn't been configured. The fix is from Juergen Kreileder, who
4946 understands it better than I do:
4948 "Here's what I see happening with three configured cyrus_sasl
4949 authenticators configured (plain, login, cram-md5):
4951 On startup auth_cyrus_sasl_init() gets called for each of these.
4952 This means three calls to sasl_listmech() without a specified mech_list.
4953 => SASL tests which mechs of all available mechs actually work
4954 => three warnings about OTP not working
4955 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4957 With the patch, sasl_listmech() also gets called three times. But now
4958 SASL's mech_list option is set to the server_mech specified in the the
4959 authenticator. Or in other words, the answer from sasl_listmech()
4960 gets limited to just the mech you're testing for (which is different
4962 => the return list contains just 'plain' or 'login', 'cram-md5' or
4963 nothing depending on the value of ob->server_mech.
4965 I've just tested the patch: Authentication still works fine,
4966 unavailable mechs specified in the exim configuration are still
4967 caught, and the auth.log warnings about OTP are gone."
4969 PH/25 When debugging is enabled, the contents of the command line are added
4970 to the debugging output, even when log_selector=+arguments is not
4973 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4974 answer is "GNU", and only if the return is "GNU/something" is the answer
4977 PH/27 $acl_verify_message is now set immediately after the failure of a
4978 verification in an ACL, and so is available in subsequent modifiers. In
4979 particular, the message can be preserved by coding like this:
4981 warn !verify = sender
4982 set acl_m0 = $acl_verify_message
4984 Previously, $acl_verify_message was set only while expanding "message"
4985 and "log_message" when a very denied access.
4987 PH/28 Modified OS/os.c-Linux with
4989 -#ifndef OS_LOAD_AVERAGE
4990 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4992 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4993 nomenclature these days.)
4995 PH/29 Installed patch from the Sieve maintainer that adds the options
4996 sieve_useraddress and sieve_subaddress to the redirect router.
4998 PH/30 In these circumstances:
4999 . Two addresses routed to the same list of hosts;
5000 . First host does not offer TLS;
5001 . First host accepts first address;
5002 . First host gives temporary error to second address;
5003 . Second host offers TLS and a TLS session is established;
5004 . Second host accepts second address.
5005 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5006 and peerdn, if requested) that were in fact used only for the second
5009 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5010 attention to any local part prefix or suffix that was matched by the
5011 router that accepted the address. It now behaves in the same way as it
5012 does for delivery: the affixes are removed from the local part unless
5013 rcpt_include_affixes is set on the transport.
5015 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5016 timeout during the DATA phase of an incoming message.
5018 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5019 applied a patch sanctioned by the Sieve maintainer.
5021 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5022 the uid or gid is negative. A case of a negative gid caused this to be
5023 noticed. The fix allows for either to be negative.
5025 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5026 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5029 PH/36 The change PH/12 above was broken. Fixed it.
5031 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5032 the grounds that routing the same address twice would always produce the
5033 same answer. This might have been true once, but it is certainly no
5034 longer true now. Routing a child address may depend on the previous
5035 routing that produced that child. Some complicated redirection strategies
5036 went wrong when messages had multiple recipients, and made Exim's
5037 behaviour dependent on the order in which the addresses were given.
5039 I have moved the duplicate checking until after the routing is complete.
5040 Exim scans the addresses that are assigned to local and remote
5041 transports, and removes any duplicates. This means that more work will be
5042 done, as duplicates will always all be routed, but duplicates are
5043 presumably rare, so I don't expect this is of any significance.
5045 For deliveries to pipes, files, and autoreplies, the duplicate checking
5046 still happens during the routing process, since they are not going to be
5049 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5050 It corrects a timeout issue with spamd. This is Ian's comment: "The
5051 background is that sometimes spamd either never reads data from a
5052 connection it has accepted, or it never writes response data. The exiscan
5053 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5054 blindly assumes that writes won't block so it may never time out."
5056 PH/39 Allow G after quota size as well as K and M.
5058 PH/40 The value set for $authenticated_id in an authenticator may not contain
5059 binary zeroes or newlines because the value is written to log lines and
5060 to spool files. There was no check on this. Now the value is run through
5061 the string_printing() function so that such characters are converted to
5062 printable escape sequences.
5064 PH/41 $message_linecount is a new variable that contains the total number of
5065 lines in the message. Compare $body_linecount, which is the count for the
5068 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5069 bounce and delay warning messages, except in certain special cases, which
5072 (a) An SMTP error message from a remote host;
5073 (b) A message specified in a :fail: redirection;
5074 (c) A message specified in a "fail" command in a system filter;
5075 (d) A message specified in a FAIL return from the queryprogram router;
5076 (e) A message specified by the cannot_route_message router option.
5078 In these cases only, Exim does include the error details in bounce and
5079 warning messages. There are also a few cases where bland messages such
5080 as "unrouteable address" or "local delivery error" are given.
5082 PH/43 $value is now also set for the "else" part of a ${run expansion.
5084 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5085 being worked on, but at least Exim now implements the latest version to
5088 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5089 process to complete was treated as a delivery failure, a timeout while
5090 writing the message to the pipe was logged, but erroneously treated as a
5091 successful delivery. Such timeouts include transport filter timeouts. For
5092 consistency with the overall process timeout, these timeouts are now
5093 treated as errors, giving rise to delivery failures by default. However,
5094 there is now a new Boolean option for the pipe transport called
5095 timeout_defer, which, if set TRUE, converts the failures into defers for
5096 both kinds of timeout. A transport filter timeout is now identified in
5099 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5100 systems where "make" and "gmake" are different, calling "gmake" at top
5101 level broke things. I've arranged for the value of $(MAKE) to be passed
5102 from the Makefile to this script so that it can call the same version of
5106 A note about Exim versions 4.44 and 4.50
5107 ----------------------------------------
5109 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5110 changes of various kinds. As a consequence, a big documentation update was
5111 needed. This delayed the release for rather longer than seemed good, especially
5112 in the light of a couple of (minor) security issues. Therefore, the changes
5113 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5114 release. So 4.44 and 4.50 are in effect two different branches that both start
5117 I have left the 4.50 change log unchanged; it contains all the changes since
5118 4.43. The change log for 4.44 is below; many of its items are identical to
5119 those for 4.50. This seems to be the most sensible way to preserve the
5120 historical information.
5126 1. Minor wording change to the doc/README.SIEVE file.
5128 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5129 computation of the current number of files was incorrect.
5131 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5132 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5135 4. Give more explanation in the error message when the command for a transport
5136 filter fails to execute.
5138 5. There are several places where Exim runs a non-Exim command in a
5139 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5140 was being done only for the command run by the queryprogram router. It is
5141 now done for all such subprocesses. The other cases are: ${run, transport
5142 filters, and the commands run by the lmtp and pipe transports.
5144 6. Added CONFIGURE_GROUP build-time option.
5146 7. Some older OS have a limit of 256 on the maximum number of file
5147 descriptors. Exim was using setrlimit() to set 1000 as a large value
5148 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5149 systems. I've change it so that if it can't get 1000, it tries for 256.
5151 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5152 was an oversight, and furthermore, ever since the addition of extra
5153 controls (e.g. 4.43/32), the checks on when to allow different forms of
5154 "control" were broken. There should now be diagnostics for all cases when a
5155 control that does not make sense is encountered.
5157 9. Added the /retain_sender option to "control=submission".
5159 10. $recipients is now available in the predata ACL (oversight).
5161 11. Tidy the search cache before the fork to do a delivery from a message
5162 received from the command line. Otherwise the child will trigger a lookup
5163 failure and thereby defer the delivery if it tries to use (for example) a
5164 cached ldap connection that the parent has called unbind on.
5166 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5167 of $address_data from the recipient verification was clobbered by the
5168 sender verification.
5170 13. The value of address_data from a sender verification is now available in
5171 $sender_address_data in subsequent conditions in the ACL statement.
5173 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5175 15. Added a new option "connect=<time>" to callout options, to set a different
5178 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5179 was its contents. (It was OK if the option was not defined at all.)
5181 17. A "Completed" log line is now written for messages that are removed from
5182 the spool by the -Mrm option.
5184 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5185 information about exactly what failed.
5187 19. Added -dd to debug only the daemon process.
5189 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5190 handles timeouts, both on the server side and network timeouts. Renamed the
5191 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5193 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5194 It is now set to "smtps".
5196 22. $host_address is now set to the target address during the checking of
5197 ignore_target_hosts.
5199 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5200 being passed; this would have caused $sender_host_name to have been used if
5201 matching the list had actually called for a host name (not very likely,
5202 since this list is usually IP addresses). A host name is now passed as
5205 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5206 code that specifies a non-restarting handler (typically sigaction() in
5207 modern systems) in an attempt to fix a rare and obscure crash bug.
5209 25. Narrowed the window for a race in the daemon that could cause it to ignore
5210 SIGCHLD signals. This is not a major problem, because they are used only to
5211 wake it up if nothing else does.
5213 26. A malformed maildirsize file could cause Exim to calculate negative values
5214 for the mailbox size or file count. Odd effects could occur as a result.
5215 The maildirsize information is now recalculated if the size or filecount
5218 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5219 support for a long time. Removed HAVE_SYS_VFS_H.
5221 28. Installed the latest version of exipick from John Jetmore.
5223 29. In an address list, if the pattern was not a regular expression, an empty
5224 subject address (from a bounce message) matched only if the pattern was an
5225 empty string. Non-empty patterns were not even tested. This was the wrong
5226 because it is perfectly reasonable to use an empty address as part of a
5227 database query. An empty address is now tested by patterns that are
5228 lookups. However, all the other forms of pattern expect the subject to
5229 contain a local part and a domain, and therefore, for them, an empty
5230 address still always fails if the pattern is not itself empty.
5232 30. Exim went into a mad DNS loop when attempting to do a callout where the
5233 host was specified on an smtp transport, and looking it up yielded more
5234 than one IP address.
5236 31. Re-factored the code for checking spool and log partition space into a
5237 function that finds that data and another that does the check. The former
5238 is then used to implement four new variables: $spool_space, $log_space,
5239 $spool_inodes, and $log_inodes.
5241 32. The RFC2047 encoding function was originally intended for short strings
5242 such as real names; it was not keeping to the 75-character limit for
5243 encoded words that the RFC imposes. It now respects the limit, and
5244 generates multiple encoded words if necessary. To be on the safe side, I
5245 have increased the buffer size for the ${rfc2047: expansion operator from
5248 33. It is now permitted to omit both strings after an "if" condition; if the
5249 condition is true, the result is "true". As before, when the second string
5250 is omitted, a false condition yields an empty string. This makes it less
5251 cumbersome to write custom ACL and router conditions.
5253 34. Failure to deliver a bounce message always caused it to be frozen, even if
5254 there was an errors_to setting on the router. The errors_to setting is now
5257 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5258 canonical form (fully expanded) before being placed in
5259 $sender_host_address.
5261 36. The table in the code that translates DNS record types into text (T_A to
5262 "A" for instance) was missing entries for NS and CNAME. It is just possible
5263 that this could have caused confusion if both these types were looked up
5264 for the same domain, because the text type is used as part of Exim's
5265 per-process caching. But the chance of anyone hitting this buglet seems
5268 37. The dnsdb lookup has been extended in a number of ways.
5270 (1) There is a new type, "zns", which walks up the domain tree until it
5271 finds some nameserver records. It should be used with care.
5273 (2) There is a new type, "mxh", which is like "mx" except that it returns
5274 just the host names, not the priorities.
5276 (3) It is now possible to give a list of domains (or IP addresses) to be
5277 looked up. The behaviour when one of the lookups defers can be
5278 controlled by a keyword.
5280 (4) It is now possible to specify the separator character for use when
5281 multiple records are returned.
5283 38. The dnslists ACL condition has been extended: it is now possible to supply
5284 a list of IP addresses and/or domains to be looked up in a particular DNS
5287 39. Added log_selector=+queue_time_overall.
5289 40. When running the queue in the test harness, wait just a tad after forking a
5290 delivery process, to get repeatability of debugging output.
5292 41. Include certificate and key file names in error message when GnuTLS fails
5293 to set them up, because the GnuTLS error message doesn't include the name
5294 of the failing file when there is a problem reading it.
5296 42. Allow both -bf and -bF in the same test run.
5298 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5300 44. The "Exiscan patch" is now merged into the mainline Exim source.
5302 45. Sometimes the final signoff response after QUIT could fail to get
5303 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5304 < 0 before doing a fflush(). This bug looks as though it goes back to the
5305 introduction of TLS in release 3.20, but "sometimes" must have been rare
5306 because the tests only now provoked it.
5308 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5309 (this can affect the format of dates).
5311 47. exim_tidydb, when checking for the continued existence of a message for
5312 which it has found a message-specific retry record, was not finding
5313 messages that were in split spool directories. Consequently, it was
5314 deleting retry records that should have stayed in existence.
5316 48. Steve fixed some bugs in eximstats.
5318 49. The SPA authentication driver was not abandoning authentication and moving
5319 on to the next authenticator when an expansion was forced to fail,
5320 contradicting the general specification for all authenticators. Instead it
5321 was generating a temporary error. It now behaves as specified.
5323 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5324 (the order specifies the preference for clients). The order is now AES256,
5325 AES128, 3DES, ARCFOUR128.
5327 51. Small patch to Sieve code - explicitly set From: when generating an
5330 52. Exim crashed if a remote delivery caused a very long error message to be
5331 recorded - for instance if somebody sent an entire SpamAssassin report back
5332 as a large number of 550 error lines. This bug was coincidentally fixed by
5333 increasing the size of one of Exim's internal buffers (big_buffer) that
5334 happened as part of the Exiscan merge. However, to be on the safe side, I
5335 have made the code more robust (and fixed the comments that describe what
5338 53. Now that there can be additional text after "Completed" in log lines (if
5339 the queue_time_overall log selector is set), a one-byte patch to exigrep
5340 was needed to allow it to recognize "Completed" as not the last thing in
5343 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5344 patch that reportedly fixes this has been added. I am not expert enough to
5345 create a test for it. This is what the patch creator wrote:
5347 "I found a little strange behaviour of ldap code when working with
5348 Windows 2003 AD Domain, where users was placed in more than one
5349 Organization Units. When I tried to give exim partial DN, the exit code
5350 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5351 But simultaneously result of request was absolutely normal ldap result,
5352 so I produce this patch..."
5354 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5355 REFERENCE, so I have modified the code to exclude the patch when that macro
5358 55. Some experimental protocols are using DNS PTR records for new purposes. The
5359 keys for these records are domain names, not reversed IP addresses. The
5360 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5361 leaves it alone. Component reversal etc. now happens only for IP addresses.
5364 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5366 57. Double the size of the debug message buffer (to 2048) so that more of very
5367 long debug lines gets shown.
5369 58. The exicyclog utility now does better if the number of log files to keep
5370 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5372 59. Two changes related to the smtp_active_hostname option:
5374 (1) $smtp_active_hostname is now available as a variable.
5375 (2) The default for smtp_banner uses $smtp_active_hostname instead
5376 of $primary_hostname.
5378 60. The host_aton() function is supposed to be passed a string that is known
5379 to be a valid IP address. However, in the case of IPv6 addresses, it was
5380 not checking this. This is a hostage to fortune. Exim now panics and dies
5381 if the condition is not met. A case was found where this could be provoked
5382 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5383 components; fortuitously, this particular loophole had already been fixed
5384 by change 4.50/55 above.
5386 If there are any other similar loopholes, the new check in host_aton()
5387 itself should stop them being exploited. The report I received stated that
5388 data on the command line could provoke the exploit when Exim was running as
5389 exim, but did not say which command line option was involved. All I could
5390 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5391 running as the user.
5394 61. There was a buffer overflow vulnerability in the SPA authentication code
5395 (which came originally from the Samba project). I have added a test to the
5396 spa_base64_to_bits() function which I hope fixes it.
5399 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5400 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5402 63. The daemon start-up calls getloadavg() while still root for those OS that
5403 need the first call to be done as root, but it missed one case: when
5404 deliver_queue_load_max is set with deliver_drop_privilege. This is
5405 necessary for the benefit of the queue runner, because there is no re-exec
5406 when deliver_drop_privilege is set.
5408 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5409 This has been fixed.
5411 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5412 was in use, was not putting the data itself into the right store pool;
5413 consequently, it could be overwritten for a subsequent message in the same
5414 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5417 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5419 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5420 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5421 the function were treating the return as a boolean value, which happened to
5422 work because 0=false and not-0=true, but is not correct code.
5424 68. The host_aton() function was not handling scoped IPv6 addresses (those
5425 with, for example, "%eth0" on the end) correctly.
5427 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5428 negated items (that is, ~something) in unsigned ints. Some compilers
5429 apparently mutter when there is no cast.
5431 70. If an address verification called from an ACL failed, and did not produce a
5432 user-specific message (i.e. there was only a "system" message), nothing was
5433 put in $acl_verify_message. In this situation, it now puts the system
5436 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5437 session; change 4.31/43 added the unwanted input to the log line - except
5438 that it did not do this in the start of session case. It now does.
5440 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5441 This is wrong and can cause the other end to generate a synchronization
5442 error if it is another Exim or anything else that does the synchronization
5443 check. A QUIT command is no longer sent after a timeout.
5445 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5446 during host lookups.
5448 74. The defer_ok option of callout verification was not working if it was used
5449 when verifying addresses in header lines, that is, for this case:
5451 verify = header_sender/callout=defer_ok
5453 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5454 those file descriptors could be used for SMTP connections. If anything
5455 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5456 could be sent to the SMTP client, causing chaos. The daemon now opens
5457 stdin, stdout, and stderr to /dev/null when it puts itself into the
5460 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5461 log by default. The user can override this with suitable Perl magic.
5463 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5464 the log message when discard triggers, was not working for the DATA ACL or
5465 for the non-SMTP ACL.
5467 78. Error message wording change in sieve.c.
5469 79. If smtp_accept_max_per_host was set, the number of connections could be
5470 restricted to fewer than expected, because the daemon was trying to set up
5471 a new connection before checking whether the processes handling previous
5472 connections had finished. The check for completed processes is now done
5473 earlier. On busy systems, this bug wouldn't be noticed because something
5474 else would have woken the daemon, and it would have reaped the completed
5477 80. If a message was submitted locally by a user whose login name contained one
5478 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5479 It caused a spool format error. I have fixed the spool reading code. A
5480 related problem was that the "from" clause in the Received: line became
5481 illegal because of the space(s). It is now covered by ${quote_local_part.
5483 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5486 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5488 83. Patch from Sieve maintainer to fix unterminated string problem in
5489 "vacation" handling.
5491 84. Some minor changes to the Linux configuration files to help with other
5492 OS variants using glibc.
5494 85. One more patch for Sieve to update vacation handling to latest spec.
5497 ----------------------------------------------------
5498 See the note above about the 4.44 and 4.50 releases.
5499 ----------------------------------------------------
5505 1. Change 4.43/35 introduced a bug that caused file counts to be
5506 incorrectly computed when quota_filecount was set in an appendfile
5509 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5510 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5513 3. Give more explanation in the error message when the command for a transport
5514 filter fails to execute.
5516 4. There are several places where Exim runs a non-Exim command in a
5517 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5518 was being done only for the command run by the queryprogram router. It is
5519 now done for all such subprocesses. The other cases are: ${run, transport
5520 filters, and the commands run by the lmtp and pipe transports.
5522 5. Some older OS have a limit of 256 on the maximum number of file
5523 descriptors. Exim was using setrlimit() to set 1000 as a large value
5524 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5525 systems. I've change it so that if it can't get 1000, it tries for 256.
5527 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5528 was an oversight, and furthermore, ever since the addition of extra
5529 controls (e.g. 4.43/32), the checks on when to allow different forms of
5530 "control" were broken. There should now be diagnostics for all cases when a
5531 control that does not make sense is encountered.
5533 7. $recipients is now available in the predata ACL (oversight).
5535 8. Tidy the search cache before the fork to do a delivery from a message
5536 received from the command line. Otherwise the child will trigger a lookup
5537 failure and thereby defer the delivery if it tries to use (for example) a
5538 cached ldap connection that the parent has called unbind on.
5540 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5541 of $address_data from the recipient verification was clobbered by the
5542 sender verification.
5544 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5545 was its contents. (It was OK if the option was not defined at all.)
5547 11. A "Completed" log line is now written for messages that are removed from
5548 the spool by the -Mrm option.
5550 12. $host_address is now set to the target address during the checking of
5551 ignore_target_hosts.
5553 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5554 being passed; this would have caused $sender_host_name to have been used if
5555 matching the list had actually called for a host name (not very likely,
5556 since this list is usually IP addresses). A host name is now passed as
5559 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5560 code that specifies a non-restarting handler (typically sigaction() in
5561 modern systems) in an attempt to fix a rare and obscure crash bug.
5563 15. Narrowed the window for a race in the daemon that could cause it to ignore
5564 SIGCHLD signals. This is not a major problem, because they are used only to
5565 wake it up if nothing else does.
5567 16. A malformed maildirsize file could cause Exim to calculate negative values
5568 for the mailbox size or file count. Odd effects could occur as a result.
5569 The maildirsize information is now recalculated if the size or filecount
5572 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5573 support for a long time. Removed HAVE_SYS_VFS_H.
5575 18. Updated exipick to current release from John Jetmore.
5577 19. Allow an empty sender to be matched against a lookup in an address list.
5578 Previously the only cases considered were a regular expression, or an
5581 20. Exim went into a mad DNS lookup loop when doing a callout where the
5582 host was specified on the transport, if the DNS lookup yielded more than
5585 21. The RFC2047 encoding function was originally intended for short strings
5586 such as real names; it was not keeping to the 75-character limit for
5587 encoded words that the RFC imposes. It now respects the limit, and
5588 generates multiple encoded words if necessary. To be on the safe side, I
5589 have increased the buffer size for the ${rfc2047: expansion operator from
5592 22. Failure to deliver a bounce message always caused it to be frozen, even if
5593 there was an errors_to setting on the router. The errors_to setting is now
5596 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5597 canonical form (fully expanded) before being placed in
5598 $sender_host_address.
5600 24. Updated eximstats to version 1.33
5602 25. Include certificate and key file names in error message when GnuTLS fails
5603 to set them up, because the GnuTLS error message doesn't include the name
5604 of the failing file when there is a problem reading it.
5606 26. Expand error message when OpenSSL has problems setting up cert/key files.
5609 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5610 (this can affect the format of dates).
5612 28. exim_tidydb, when checking for the continued existence of a message for
5613 which it has found a message-specific retry record, was not finding
5614 messages that were in split spool directories. Consequently, it was
5615 deleting retry records that should have stayed in existence.
5617 29. eximstats updated to version 1.35
5618 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5619 1.35 - bugfix such that pie charts by volume are generated correctly
5621 30. The SPA authentication driver was not abandoning authentication and moving
5622 on to the next authenticator when an expansion was forced to fail,
5623 contradicting the general specification for all authenticators. Instead it
5624 was generating a temporary error. It now behaves as specified.
5626 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5627 (the order specifies the preference for clients). The order is now AES256,
5628 AES128, 3DES, ARCFOUR128.
5630 31. Small patch to Sieve code - explicitly set From: when generating an
5633 32. Exim crashed if a remote delivery caused a very long error message to be
5634 recorded - for instance if somebody sent an entire SpamAssassin report back
5635 as a large number of 550 error lines. This bug was coincidentally fixed by
5636 increasing the size of one of Exim's internal buffers (big_buffer) that
5637 happened as part of the Exiscan merge. However, to be on the safe side, I
5638 have made the code more robust (and fixed the comments that describe what
5641 33. Some experimental protocols are using DNS PTR records for new purposes. The
5642 keys for these records are domain names, not reversed IP addresses. The
5643 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5644 leaves it alone. Component reversal etc. now happens only for IP addresses.
5647 34. The host_aton() function is supposed to be passed a string that is known
5648 to be a valid IP address. However, in the case of IPv6 addresses, it was
5649 not checking this. This is a hostage to fortune. Exim now panics and dies
5650 if the condition is not met. A case was found where this could be provoked
5651 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5652 components; fortuitously, this particular loophole had already been fixed
5653 by change 4.50/55 or 4.44/33 above.
5655 If there are any other similar loopholes, the new check in host_aton()
5656 itself should stop them being exploited. The report I received stated that
5657 data on the command line could provoke the exploit when Exim was running as
5658 exim, but did not say which command line option was involved. All I could
5659 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5660 running as the user.
5663 35. There was a buffer overflow vulnerability in the SPA authentication code
5664 (which came originally from the Samba project). I have added a test to the
5665 spa_base64_to_bits() function which I hope fixes it.
5668 36. The daemon start-up calls getloadavg() while still root for those OS that
5669 need the first call to be done as root, but it missed one case: when
5670 deliver_queue_load_max is set with deliver_drop_privilege. This is
5671 necessary for the benefit of the queue runner, because there is no re-exec
5672 when deliver_drop_privilege is set.
5674 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5675 was in use, was not putting the data itself into the right store pool;
5676 consequently, it could be overwritten for a subsequent message in the same
5677 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5680 38. Sometimes the final signoff response after QUIT could fail to get
5681 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5682 < 0 before doing a fflush(). This bug looks as though it goes back to the
5683 introduction of TLS in release 3.20, but "sometimes" must have been rare
5684 because the tests only now provoked it.
5690 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5691 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5692 Now it returns an integer. A number of calls were still expecting a T/F
5693 return. Fortuitously, in all cases, the tests worked in OK situations,
5694 which is the norm. However, things would have gone wrong on any write
5695 failures on the smtp file descriptor. This function is used when sending
5696 messages over SMTP and also when doing verify callouts.
5698 2. When Exim is called to do synchronous delivery of a locally submitted
5699 message (the -odf or -odi options), it no longer closes stderr before doing
5702 3. Implemented the mua_wrapper option.
5704 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5706 5. Implemented the functions header_remove(), header_testname(),
5707 header_add_at_position(), and receive_remove_recipient(), and exported them
5710 6. If an ACL "warn" statement specified the addition of headers, Exim already
5711 inserted X-ACL-Warn: at the start if there was no header name. However, it
5712 was not making this test for the second and subsequent header lines if
5713 there were newlines in the string. This meant that an invalid header could
5714 be inserted if Exim was badly configured.
5716 7. Allow an ACL "warn" statement to add header lines at the start or after all
5717 the Received: headers, as well as at the end.
5719 8. Added the rcpt_4xx retry error code.
5721 9. Added postmaster_mailfrom=xxx to callout verification option.
5723 10. Added mailfrom=xxxx to the callout verification option, for verify=
5726 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5727 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5729 12. Inserted some casts to stop certain compilers warning when using pointer
5730 differences as field lengths or precisions in printf-type calls (mostly
5731 affecting debugging statements).
5733 13. Added optional readline() support for -be (dynamically loaded).
5735 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5736 same clock tick as a message's arrival, so that its received time was the
5737 same as the "first fail" time on the retry record, and that message
5738 remained on the queue past the ultimate address timeout, every queue runner
5739 would try a delivery (because it was past the ultimate address timeout) but
5740 after another failure, the ultimate address timeout, which should have then
5741 bounced the address, did not kick in. This was a "< instead of <=" error;
5742 in most cases the first failure would have been in the next clock tick
5743 after the received time, and all would be well.
5745 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5746 being recognized when the domain list was tested by the match_domain
5747 condition in an expansion string.
5749 16. Added the ${str2b64: operator.
5751 17. Exim was always calling setrlimit() to set a large limit for the number of
5752 processes, without checking whether the existing limit was already
5753 adequate. (It did check for the limit on file descriptors.) Furthermore,
5754 errors from getrlimit() and setrlimit() were being ignored. Now they are
5755 logged to the main and panic logs, but Exim does carry on, to try to do its
5756 job under whatever limits there are.
5758 18. Imported PCRE 5.0.
5760 19. Trivial typo in log message " temporarily refused connection" (the leading
5763 20. If the log selector return_path_on_delivery was set and an address was
5764 redirected to /dev/null, the delivery process crashed because it assumed
5765 that a return path would always be set for a "successful" delivery. In this
5766 case, the whole delivery is bypassed as an optimization, and therefore no
5769 21. Internal re-arrangement: the function for sending a challenge and reading
5770 a response while authentication was assuming a zero-terminated challenge
5771 string. It's now changed to take a pointer and a length, to allow for
5772 binary data in such strings.
5774 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5776 23. Exim was not respecting finduser_retries when seeking the login of the
5777 uid under which it was called; it was always trying 10 times. (The default
5778 setting of finduser_retries is zero.) Also, it was sleeping after the final
5779 failure, which is pointless.
5781 24. Implemented tls_on_connect_ports.
5783 25. Implemented acl_smtp_predata.
5785 26. If the domain in control=submission is set empty, Exim assumes that the
5786 authenticated id is a complete email address when it generates From: or
5787 Sender: header lines.
5789 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5790 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5791 chown and chgrp in /bin and hostname in /usr/bin.
5793 28. Exim was keeping the "process log" file open after each use, just as it
5794 does for the main log. This opens the possibility of it remaining open for
5795 long periods when the USR1 signal hits a daemon. Occasional processlog
5796 errors were reported, that could have been caused by this. Anyway, it seems
5797 much more sensible not to leave this file open at all, so that is what now
5800 29. The long-running daemon process does not normally write to the log once it
5801 has entered its main loop, and it closes the log before doing so. This is
5802 so that log files can straightforwardly be renamed and moved. However,
5803 there are a couple of unusual error situations where the daemon does write
5804 log entries, and I had neglected to close the log afterwards.
5806 30. The text of an SMTP error response that was received during a remote
5807 delivery was being truncated at 512 bytes. This is too short for some of
5808 the long messages that one sometimes sees. I've increased the limit to
5811 31. It is now possible to make retry rules that apply only when a message has a
5812 specific sender, in particular, an empty sender.
5814 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5815 it possible to be selective about when SMTP synchronization is enforced.
5817 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5819 32. Implemented hosts_connection_nolog.
5821 33. Added an ACL for QUIT.
5823 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5826 35. Added mailbox_size and mailbox_filecount to appendfile.
5828 36. Added control = no_multiline_responses to ACLs.
5830 37. There was a bug in the logic of the code that waits for the clock to tick
5831 in the case where the clock went backwards by a substantial amount such
5832 that the microsecond fraction of "now" was more than the microsecond
5833 fraction of "then" (but the whole seconds number was less).
5835 38. Added support for the libradius Radius client library this is found on
5836 FreeBSD (previously only the radiusclient library was supported).
5842 1. When certain lookups returned multiple values in the form name=value, the
5843 quoting of the values was not always being done properly. Specifically:
5844 (a) If the value started with a double quote, but contained no whitespace,
5846 (b) If the value contained whitespace other than a space character (i.e.
5847 tabs or newlines or carriage returns) it was not quoted.
5848 This fix has been applied to the mysql and pgsql lookups by writing a
5849 separate quoting function and calling it from the lookup code. The fix
5850 should probably also be applied to nisplus, ibase and oracle lookups, but
5851 since I cannot test any of those, I have not disturbed their existing code.
5853 2. A hit in the callout cache for a specific address caused a log line with no
5854 reason for rejecting RCPT. Now it says "Previous (cached) callout
5855 verification failure".
5857 3. There was an off-by-one bug in the queryprogram router. An over-long
5858 return line was truncated at 256 instead of 255 characters, thereby
5859 overflowing its buffer with the terminating zero. As well as fixing this, I
5860 have increased the buffer size to 1024 (and made a note to document this).
5862 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5863 when Exim is waiting for an SMTP response from a remote server, Exim
5864 restarts its select() call on the socket, thereby resetting its timeout.
5865 This is not a problem when such interrupts are rare. Somebody set up a cron
5866 job to run exiwhat every 2 minutes, which is less than the normal select()
5867 timeout (5 or 10 minutes). This meant that the select() timeout never
5868 kicked in because it was always reset. I have fixed this by comparing the
5869 time when an interrupt arrives with the time at the start of the first call
5870 to select(). If more time than the timeout has elapsed, the interrupt is
5871 treated as a timeout.
5873 5. Some internal re-factoring in preparation for the addition of Sieve
5874 extensions (by MH). In particular, the "personal" test is moved to a
5875 separate function, and given an option for scanning Cc: and Bcc: (which is
5876 not set for Exim filters).
5878 6. When Exim created an email address using the login of the caller as the
5879 local part (e.g. when creating a From: or Sender: header line), it was not
5880 quoting the local part when it contained special characters such as @.
5882 7. Installed new OpenBSD configuration files.
5884 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5885 try to make them clearer.
5887 9. Callout options, other than the timeout value, were being ignored when
5888 verifying sender addresses in header lines. For example, when using
5890 verify = header_sender/callout=no_cache
5892 the cache was (incorrectly) being used.
5894 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5895 only the Cygwin environment.
5897 11. When return_path_on_delivery was set as a log selector, if different remote
5898 addresses in the same message used different return paths and parallel
5899 remote delivery occurred, the wrong values would sometimes be logged.
5900 (Whenever a remote delivery process finished, the return path value from
5901 the most recently started remote delivery process was logged.)
5903 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5904 lines when AUTH and/or TLS are in use. This is the "received protocol"
5905 field. Exim used to use "asmtp" for authenticated SMTP, without any
5906 indication (in the protocol name) for TLS use. Now it follows the RFC and
5907 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5908 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5909 names appear in log lines as well as in Received: header lines.
5911 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5912 extensions, and comparison tests, and to fix some bugs.
5914 14. Changes to the "personal" filter test:
5916 (1) The test was buggy in that it was just doing the equivalent of
5917 "contains" tests on header lines. For example, if a user's address was
5918 anne@some.where, the "personal" test would incorrectly be true for
5920 To: susanne@some.where
5922 This test is now done by extracting each address from the header in turn,
5923 and checking the entire address. Other tests that are part of "personal"
5924 are now done using regular expressions (for example, to check local parts
5925 of addresses in From: header lines).
5927 (2) The list of non-personal local parts in From: addresses has been
5928 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5929 taken from the Sieve specification recommendations.
5931 (3) If the message contains any header line starting with "List-" it is
5932 treated as non-personal.
5934 (4) The test for "circular" in the Subject: header line has been removed
5935 because it now seems ill-conceived.
5937 15. Minor typos in src/EDITME comments corrected.
5939 16. Installed latest exipick from John Jetmore.
5941 17. If headers_add on a router specified a text string that was too long for
5942 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5943 of string_sprintf() is now avoided.
5945 18. $message_body_size was not set (it was always zero) when running the DATA
5946 ACL and the local_scan() function.
5948 19. For the "mail" command in an Exim filter, no default was being set for
5949 the once_repeat time, causing a random time value to be used if "once" was
5950 specified. (If the value happened to be <= 0, no repeat happened.) The
5951 default is now 0s, meaning "never repeat". The "vacation" command was OK
5952 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5953 (I found it when inspecting the code).
5955 20. There is now an overall timeout for performing a callout verification. It
5956 defaults to 4 times the callout timeout, which applies to individual SMTP
5957 commands during the callout. The overall timeout applies when there is more
5958 than one host that can be tried. The timeout is checked before trying the
5959 next host. This prevents very long delays if there are a large number of
5960 hosts and all are timing out (e.g. when the network connections are timing
5961 out). The value of the overall timeout can be changed by specifying an
5962 additional sub-option for "callout", called "maxwait". For example:
5964 verify = sender/callout=5s,maxwait=20s
5966 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5967 to the end before writing, but this should make it even safer).
5969 22. Exim was forgetting that it had advertised PIPELINING for the second and
5970 subsequent messages on an SMTP connection. It was also not resetting its
5971 memory on STARTTLS and an internal HELO.
5973 23. When Exim logs an SMTP synchronization error within a session, it now
5974 records whether PIPELINING has been advertised or not.
5976 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5977 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5978 rather than long int.
5980 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5982 26. Added the never_mail option to autoreply.
5988 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5989 crash if the getsockname() call failed; this can happen if a connection is
5990 closed very soon after it is established. The problem was simply in the
5991 order in which certain operations were done, causing Exim to try to write
5992 to the SMTP stream before it had set up the file descriptor. The bug has
5993 been fixed by making things happen in the correct order.
5999 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6000 before the connection was closed, thus losing the rejection response.
6002 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6003 some early Solaris releases, but causes trouble in current releases where
6004 socklen_t is defined.
6006 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6009 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6012 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6013 that releasing the top of it at the end releases what was used for sub-
6014 expansions (unless the block got too big). However, discard this block if
6015 the first thing is a variable or header, so that we can use its block when
6016 it is dynamic (useful for very large $message_headers, for example).
6018 6. Lookups now cache *every* query, not just the most recent. A new, separate
6019 store pool is used for this. It can be recovered when all lookup caches are
6020 flushed. Lookups now release memory at the end of their result strings.
6021 This has involved some general refactoring of the lookup sources.
6023 7. Some code has been added to the store_xxx() functions to reduce the amount
6024 of flapping under certain conditions.
6026 8. log_incoming_interface used to affect only the <= reception log lines. Now
6027 it causes the local interface and port to be added to several more SMTP log
6028 lines, for example "SMTP connection from", and rejection lines.
6030 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6032 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6034 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6035 could be overwritten at the end of the current message (or the start of a
6036 new message if it was set in a HELO ACL). The value is now preserved for
6037 the duration of the SMTP connection.
6039 12. If a transport had a headers_rewrite setting, and a matching header line
6040 contained an unqualified address, that address was qualified, even if it
6041 did not match any rewriting rules. The underlying bug was that the values
6042 of the flags that permit the existence of unqualified sender and recipient
6043 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6044 non-local messages, and by -bnq for local messages) were not being
6045 preserved with the message after it was received.
6047 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6048 "next input=" as part of the text comprising the host identity instead of
6049 the correct text. The code was using the same buffer for two different
6050 strings. However, depending on which order the printing function evaluated
6051 its arguments, the bug did not always show up. Under Linux, for example, my
6052 test suite worked just fine.
6054 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6055 doesn't work with some older versions of Perl. It has been changed to "my",
6056 which in any case is probably the better facility to use.
6058 15. A really picky compiler found some instances of statements for creating
6059 error messages that either had too many or two few arguments for the format
6062 16. The size of the buffer for calls to the DNS resolver has been increased
6063 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6064 for addresses that have a lot of PTR records. This alleviates a problem; it
6065 does not fully solve it.
6067 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6068 buffer now truncates the list and logs the incident, which is the same
6069 action as happens when Exim is looking up a host name and its aliases.
6070 Previously in this situation something unpredictable would happen;
6071 sometimes it was "internal error: store_reset failed".
6073 18. If a server dropped the connection unexpectedly when an Exim client was
6074 using GnuTLS and trying to read a response, the client delivery process
6075 crashed while trying to generate an error log message.
6077 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6078 string, for example:
6080 warn message = H1: something\nH2: something
6082 the text was added as a single header line from Exim's point of view
6083 though it ended up OK in the delivered message. However, searching for the
6084 second and subsequent header lines using $h_h2: did not work. This has been
6085 fixed. Similarly, if a system filter added multiple headers in this way,
6086 the routers could not see them.
6088 20. Expanded the error message when iplsearch is called with an invalid key to
6089 suggest using net-iplsearch in a host list.
6091 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6092 ACLs are no longer actually imposed (and a message to that effect is
6095 22. If a "gecos" field in a passwd entry contained escaped characters, in
6096 particular, if it contained a \" sequence, Exim got it wrong when building
6097 a From: or a Sender: header from that name. A second bug also caused
6098 incorrect handling when an unquoted " was present following a character
6099 that needed quoting.
6101 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6102 was not being matched caselessly.
6104 24. Arranged for all hyphens in the exim.8 source to be escaped with
6107 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6108 properly. Recipient callout cache records were still being keyed to include
6109 the sender, even when use_sender was set false. This led to far more
6110 callouts that were necessary. The sender is no longer included in the key
6111 when use_sender is false.
6113 26. Added "control = submission" modifier to ACLs.
6115 27. Added the ${base62d: operator to decode base 62 numbers.
6117 28. dnsdb lookups can now access SRV records.
6119 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6120 the configuration file.
6122 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6123 (-v) mode. This makes the output for a verbose queue run more intelligible.
6125 31. Added a use_postmaster feature to recipient callouts.
6127 32. Added the $body_zerocount variable, containing the number of binary zero
6128 bytes in the message body.
6130 33. The time of last modification of the "new" subdirectory is now used as the
6131 "mailbox time last read" when there is a quota error for a maildir
6134 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6136 35. Added +ignore_unknown as a special item in host lists.
6138 36. Code for decoding IPv6 addresses in host lists is now included, even if
6139 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6140 address was recognized as an IP address, but was then not correctly decoded
6141 into binary, causing unexpected and incorrect effects when compared with
6148 1. Very minor rewording of debugging text in manualroute to say "list of
6149 hosts" instead of "hostlist".
6151 2. If verify=header_syntax was set, and a header line with an unqualified
6152 address (no domain) and a large number of spaces between the end of the
6153 name and the colon was received, the reception process suffered a buffer
6154 overflow, and (when I tested it) crashed. This was caused by some obsolete
6155 code that should have been removed. The fix is to remove it!
6157 3. When running in the test harness, delay a bit after writing a bounce
6158 message to get a bit more predictability in the log output.
6160 4. Added a call to search_tidyup() just before forking a reception process. In
6161 theory, someone could use a lookup in the expansion of smtp_accept_max_
6162 per_host which, without the tidyup, could leave open a database connection.
6164 5. Added the variables $recipient_data and $sender_data which get set from a
6165 lookup success in an ACL "recipients" or "senders" condition, or a router
6166 "senders" option, similar to $domain_data and $local_part_data.
6168 6. Moved the writing of debug_print from before to after the "senders" test
6171 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6172 problems for message scanning, either using a data ACL, or using
6173 local_scan() because the Received: header was not generated till after they
6174 were called (in order to set the time as the time of reception completion).
6175 I have revised the way this works. The header is now generated after the
6176 body is received, but before the ACL or local_scan() are called. After they
6177 are run, the timestamp in the header is updated.
6183 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6184 before starting a queue runner without re-exec. This happened only when
6185 deliver_drop_privilege was set or when the Exim user was set to root. The
6186 effect of the bug was that timeouts during subsequent deliveries caused
6187 crashes instead of being properly handled. The handler is now left at its
6188 default (and expected) setting.
6190 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6191 message, again when deliver_drop_privilege is set or Exim is run as root.
6192 The bug described in (1) was not present in this case, but the tidying up
6193 of the other signals was missing. I have made the two cases consistent.
6195 3. The ignore_target_hosts setting on a manualroute router was being ignored
6196 for hosts that were looked up using the /MX notation.
6198 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6201 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6202 operated on the sender address. After changing the $sender_address to <>
6203 for the sender address verify, Exim was re-instated it as the original
6204 (before rewriting) address, but remembering that it had rewritten it, so it
6205 wasn't rewriting it again. This bug also had the effect of breaking the
6206 sender address verification caching when the sender address was rewritten.
6208 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6209 This has been changed so that if the ip literal address matches
6210 ignore_target_hosts, the router declines.
6212 7. Added expansion conditions match_domain, match_address, and match_local_
6213 part (NOT match_host).
6215 8. The placeholder for the Received: header didn't have a length field set.
6217 9. Added code to Exim itself and to exim_lock to test for a specific race
6218 condition that could lead to file corruption when using MBX delivery. The
6219 issue is with the lockfile that is created in /tmp. If this file is removed
6220 after a process has opened it but before that process has acquired a lock,
6221 there is the potential for a second process to recreate the file and also
6222 acquire a lock. This could lead to two Exim processes writing to the file
6223 at the same time. The added code performs the same test as UW imapd; it
6224 checks after acquiring the lock that its file descriptor still refers to
6225 the same named file.
6227 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6228 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6231 11. Added the smtp_active_hostname option. If used, this will typically be made
6232 to depend on the incoming interface address. Because $interface_address is
6233 not set up until the daemon has forked a reception process, error responses
6234 that can happen earlier (such as "too many connections") no longer contain
6237 12. If an expansion in a condition on a "warn" statement fails because a lookup
6238 defers, the "warn" statement is abandoned, and the next ACL statement is
6239 processed. Previously this caused the whole ACL to be aborted.
6241 13. Added the iplsearch lookup type.
6243 14. Added ident_timeout as a log selector.
6245 15. Added tls_certificate_verified as a log selector.
6247 16. Added a global option tls_require_ciphers (compare the smtp transport
6248 option of the same name). This controls incoming TLS connections.
6250 17. I finally figured out how to make tls_require_ciphers do a similar thing
6251 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6252 before starting the TLS session.
6254 18. Tabs are now shown as \t in -bP output.
6256 19. If the log selector return_path_on_delivery was set, Exim crashed when
6257 bouncing a message because it had too many Received: header lines.
6259 20. If two routers both had headers_remove settings, and the first one included
6260 a superfluous trailing colon, the final name in the first list and the
6261 first name in the second list were incorrectly joined into one item (with a
6262 colon in the middle).
6268 1. Added -C and -D options to the exinext utility, mainly to make it easier
6269 to include in the automated testing, but these could be helpful when
6270 multiple configurations are in use.
6272 2. The exinext utility was not formatting the output nicely when there was
6273 an alternate port involved in the retry record key, nor when there was a
6274 message id as well (for retries that were specific to a specific message
6275 and a specific host). It was also confused by IPv6 addresses, because of
6276 the additional colons they contain. I have fixed the IPv4 problem, and
6277 patched it up to do a reasonable job for IPv6.
6279 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6280 delivery, the log line now contains "pipelined" if PIPELINING was used.
6282 4. An SMTP transport process used to panic and die if the bind() call to set
6283 an explicit outgoing interface failed. This has been changed; it is now
6284 treated in the same way as a connect() failure.
6286 5. A reference to $sender_host_name in the part of a conditional expansion
6287 that was being skipped was still causing a DNS lookup. This no longer
6290 6. The def: expansion condition was not recognizing references to header lines
6291 that used bh_ and bheader_.
6293 7. Added the _cache feature to named lists.
6295 8. The code for checking quota_filecount in the appendfile transport was
6296 allowing one more file than it should have been.
6298 9. For compatibility with Sendmail, the command line option
6306 and sets the incoming protocol and host name (for trusted callers). The
6307 host name and its colon can be omitted when only the protocol is to be set.
6308 Note the Exim already has two private options, -pd and -ps, that refer to
6309 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6310 "s", but I don't think that's a major issue.
6312 10. A number of refactoring changes to the code, none of which should affect
6315 (a) The number of logging options was getting close to filling up the
6316 32-bit word that was used as a bit map. I have split them into two classes:
6317 those that are passed in the argument to log_write(), and those that are
6318 only ever tested independently outside of that function. These are now in
6319 separate 32-bit words, so there is plenty of room for expansion again.
6320 There is no change in the user interface or the logging behaviour.
6322 (b) When building, for example, log lines, the code previously used a
6323 macro that called string_cat() twice, in order to add two strings. This is
6324 not really sufficiently general. Furthermore, there was one instance where
6325 it was actually wrong because one of the argument was used twice, and in
6326 one call a function was used. (As it happened, calling the function twice
6327 did not affect the overall behaviour.) The macro has been replaced by a
6328 function that can join an arbitrary number of extra strings onto a growing
6331 (c) The code for expansion conditions now uses a table and a binary chop
6332 instead of a serial search (which was left over from when there were very
6333 few conditions). Also, it now recognizes conditions like "pam" even when
6334 the relevant support is not compiled in: a suitably worded error message is
6335 given if an attempt is made to use such a condition.
6337 11. Added ${time_interval:xxxxx}.
6339 12. A bug was causing one of the ddress fields not to be passed back correctly
6340 from remote delivery subprocesses. The field in question was not being
6341 subsequently used, so this caused to problems in practice.
6343 13. Added new log selectors queue_time and deliver_time.
6345 14. Might have fixed a bug in maildirsizefile handling that threw up
6346 "unexpected character" debug warnings, and recalculated the data
6347 unnecessarily. In any case, I expanded the warning message to give more
6350 15. Added the message "Restricted characters in address" to the statements in
6351 the default ACL that block characters like @ and % in local parts.
6353 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6354 Three changes have been made:
6356 (a) There was a serious bug; a negative response to MAIL caused the whole
6357 recipient domain to be cached as invalid, thereby blocking all messages
6358 to all local parts at the same domain, from all senders. This bug has
6359 been fixed. The domain is no longer cached after a negative response to
6360 MAIL if the sender used is not empty.
6362 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6365 (c) A new callout option, "use_sender" has been added for people who want
6366 the modified behaviour.
6372 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6375 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6376 indeed breaks things for older releases.
6378 3. Added additional logging to the case where there is a problem reading data
6379 from a filter that is running in a subprocess using a pipe, in order to
6380 try to track down a specific problem.
6382 4. Testing facility fudge: when running in the test harness and attempting
6383 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6384 getting "No route to host". Convert this to a timeout.
6386 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6389 6. Some OS don't have socklen_t but use size_t instead. This affects the
6390 fifth argument of getsockopt() amongst other things. This is now
6391 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6392 can be set for individual OS. I have set it for SunOS5, OSF1, and
6393 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6394 some earlier ones do not.
6396 7. Change 4.30/15 was not doing the test caselessly.
6398 8. The standard form for an IPv6 address literal was being rejected by address
6399 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6400 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6401 this, as well as the form without the "IPv6" on the front (but only when
6402 address literals are enabled, of course).
6404 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6406 10. Exim crashed if a message with an empty sender address specified by -f
6407 encountered a router with an errors_to setting. This could be provoked only
6408 by a command such as
6412 where an empty string was supplied; "<>" did not hit this bug.
6414 11. Installed PCRE release 4.5.
6416 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6417 remained set. It is now erased.
6419 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6420 times from message ids (which are base 36 rather than the normal 62).
6422 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6423 were being counted as actual protocol errors, and logged if the log
6424 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6425 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6426 and DATA following a set of rejected RCPTs do not count as protocol errors.
6427 In other words, Exim assumes they were pipelined, though this may not
6428 actually be the case. Of course, in all cases the client gets an
6429 appropriate error code.
6431 15. If a lookup fails in an ACL condition, a message about the failure may
6432 be available; it is used if testing the ACL cannot continue, because most
6433 such messages specify what the cause of the deferral is. However, some
6434 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6435 that caused an old message to be retained and used if a later statement
6436 caused a defer, replacing the real cause of the deferral.
6438 16. If an IP address had so many PTR records that the DNS lookup buffer
6439 was not large enough to hold them, Exim could crash while trying to process
6440 the truncated data. It now detects and logs this case.
6442 17. Further to 4.21/58, another change has been made: if (and only if) the
6443 first line of a message (the first header line) ends with CRLF, a bare LF
6444 in a subsequent header line has a space inserted after it, so as not to
6445 terminate the header.
6447 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6448 unnecessarily, used atoi() instead of strtol(), and didn't check the
6449 termination when getting file sizes from file names by regex.
6451 19. Completely re-implemented the support for maildirsize files, in the light
6452 of a number of problems with the previous contributed implementation
6453 (4.30/29). In particular:
6455 . If the quota is zero, the maildirsize file is maintained, but no quota is
6458 . If the maildir directory does not exist, it is created before any attempt
6459 to write a maildirsize file.
6461 . The quota value in the file is just a cache; if the quota is changed in
6462 the transport, the new value overrides.
6464 . A regular expression is available for excluding directories from the
6467 20. The autoreply transport checks the characters in options that define the
6468 message's headers; it allows continued headers, but it was checking with
6469 isspace() after an embedded newline instead of explicitly looking for a
6472 21. If all the "regular" hosts to which an address was routed had passed their
6473 expiry times, and had not reached their retry times, the address was
6474 bounced, even if fallback hosts were defined. Now Exim should go on to try
6477 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6478 equivalent code in the SMTP transport. Some hosts send humungous responses
6479 to HELO/EHLO, more than 1024 it seems.
6481 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6482 gives compiler warnings in some environments. I've now done it "properly",
6485 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6486 (because of problems with the built-in one) was declared to return uschar *
6487 instead of char *, causing compiler failure.
6489 25. Fixed a file descriptor leak when processing alias/forward files.
6491 26. Fixed a minor format string issue in dbfn.c.
6493 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6495 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6496 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6499 29. When a sender address is verified, it is cached, to save repeating the test
6500 when there is more than one recipient in a message. However, when the
6501 verification involves a callout, it is possible for different callout
6502 options to be set for different recipients. It is too complicated to keep
6503 track of this in the cache, so now Exim always runs a verification when a
6504 callout is required, relying on the callout cache for the optimization.
6505 The overhead is duplication of the address routing, but this should not be
6508 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6509 to be verified with callout=postmaster, and the main callout worked but the
6510 postmaster check failed, the verification correctly failed. However, if a
6511 subsequent RCPT command asked for sender verification *without* the
6512 postmaster check, incorrect caching caused this verification also to fail,
6515 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6516 it was not caching the DNS options (qualify_single, search_parents) that
6517 were used when the lookup failed. A subsequent lookup with different
6518 options therefore always gave the same answer, though there were cases
6519 where it should not have. (Example: a "domains = !$mx_any" option on a
6520 dnslookup router: the "domains" option is always processed without any
6521 widening, but the router might have qualify_single set.) Now Exim uses the
6522 cached value only when the same options are set.
6524 32. Added John Jetmore's "exipick" utility to the distribution.
6526 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6527 than a timeout (e.g. a certificate is required, and is not provided), an
6528 Exim server now closes the connection immediately. Previously it waited for
6529 the client to close - but if the client is SSL, it seems that they each
6530 wait for each other, leading to a delay before one of them times out.
6532 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6533 maintained 0.8.x compatibility because I don't think many are using it, and
6534 it is clearly obsolete.
6536 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6539 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6540 client certificate was expired. A simple patch fixes this, though I don't
6541 understand the full logic of why the verify callback is called multiple
6544 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6545 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6546 which causes problems with some clients (such as the Certicom SSL Plus
6547 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6548 disables the coutermeasure allowing Eudora to connect."
6550 38. Exim was not checking that a write() to a log file succeeded. This could
6551 lead to Bad Things if a log got too big, in particular if it hit a file
6552 size limit. Exim now panics and dies if it cannot write to a log file, just
6553 as it does if it cannot open a log file.
6555 39. Modified OS/Makefile-Linux so that it now contains
6557 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6559 The two -D definitions ensure that Exim is compiled with large file
6560 support, which makes it possible to handle log files that are bigger than
6563 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6564 instance) a domain was checked against a named list that involved a lookup,
6565 causing $domain_data to be set, then another domain was checked against the
6566 same list, then the first domain was re-checked, the value of $domain_data
6567 after the final check could be wrong. In particular, if the second check
6568 failed, it could be set empty. This bug probably also applied to
6571 41. The strip_trailing_dot option was not being applied to the address given
6572 with the -f command-line option.
6574 42. The code for reading a message's header from the spool was incrementing
6575 $received_count, but never initializing it. This meant that the value was
6576 incorrect (doubled) while delivering a message in the same process in which
6577 it was received. In the most common configuration of Exim, this never
6578 happens - a fresh exec is done - but it can happen when
6579 deliver_drop_privilege is set.
6581 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6582 it now includes up to 150 characters of the unexpected data in the log
6585 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6586 and building data strings. The size of both of these buffers was 10 000
6587 bytes - far larger than anybody would *ever* want, thought I. Needless to
6588 say, somebody hit the limit. I have increased the maximum line length to
6589 20 000 and the maximum data length of concatenated lines to 100 000. I have
6590 also fixed two bugs, because there was no checking on these buffers. Tsk,
6591 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6592 buffer is too small.
6594 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6595 lsearch lookups. Now it does.
6597 46. When parsing a route_list item in a manualroute router, a fixed-length
6598 buffer was used for the list of hosts. I made this 1024 bytes long,
6599 thinking that nobody would ever have a list of hosts that long. Wrong.
6600 Somebody had a whole pile of complicated expansion conditions, and the
6601 string was silently truncated, leading to an expansion error. It turns out
6602 that it is easier to change to an unlimited length (owing to other changes
6603 that have happened since this code was originally written) than to build
6604 structure for giving a limitation error. The length of the item that
6605 expands into the list of hosts is now unlimited.
6607 47. The lsearch lookup could not handle data where the length of text line was
6608 more than 4095 characters. Such lines were truncated, leading to shortened
6609 data being returned. It should now handle lines of any length.
6611 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6612 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6615 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6616 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6617 no longer happens in comment lines. A list of replacements is now placed
6618 at the head of all of the source files, except those whose only change is
6619 to replace PERL_COMMAND in the very first #! line.
6621 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6622 messages on the queue, with a bottom-up merge sort, using code contributed
6623 by Michael Haardt. This should make operations like -bp somewhat faster on
6624 large queues. It won't affect queue runners, except when queue_run_in_order
6627 51. Installed eximstats 1.31 in the distribution.
6629 52. Added support for SRV lookups to the dnslookup router.
6631 53. If an ACL referred to $message_body or $message_body_end, the value was not
6632 reset for any messages that followed in the same SMTP session.
6634 54. The store-handling optimization for building very long strings was not
6635 differentiating between the different store pools. I don't think this
6636 actually made any difference in practice, but I've tidied it.
6638 55. While running the routers to verify a sender address, $sender_address
6639 was still set to the sender address. This is wrong, because when routing to
6640 send a bounce to the sender, it would be empty. Therefore, I have changed
6641 it so that, while verifying a sender address, $sender_address is set to <>.
6642 (There is no change to what happens when verifying a recipient address.)
6644 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6645 target A or AAAA records (if not already returned) without resetting the
6646 qualify_single or search_parents options of the DNS resolver. These are
6647 inappropriate in this case because the targets of MX and SRV records must
6648 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6649 target that, when qualified, matched something in the local domain. These
6650 two options are now turned off when doing these lookups.
6652 57. It seems that at least some releases of Reiserfs (which does not have the
6653 concept of a fixed number of inodes) returns zero and not -1 for the
6654 number of available inodes. This interacted badly with check_spool_inodes,
6655 which assumed that -1 was the "no such thing" setting. What I have done is
6656 to check that the total number of inodes is greater than zero before doing
6657 the test of how many are available.
6659 58. When a "warn" ACL statement has a log_message modifier, the message is
6660 remembered, and not repeated. This is to avoid a lot of repetition when a
6661 message has many recipients that cause the same warning to be written.
6662 However, Exim was preserving the list of already written lines for an
6663 entire SMTP session, which doesn't seem right. The memory is now reset if a
6664 new message is started.
6666 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6667 debugging output unless log_rewrite was also set.
6669 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6670 of (int)(handle) when we know that handle contains (void *)(-1).
6672 61. The Exim daemon panic-logs an error return when it closes the incoming
6673 connection. However "connection reset by peer" seems to be common, and
6674 isn't really an error worthy of noting specially, so that particular error
6677 62. When Exim is trying to find all the local interfaces, it used to panic and
6678 die if the ioctl to get the interface flags failed. However, it seems that
6679 on at least one OS (Solaris 9) it is possible to have an interface that is
6680 included in the list of interfaces, but for which you get a failure error
6681 for this call. This happens when the interface is not "plumbed" into a
6682 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6683 failure of the "get flags" call assumes that the interface is down.
6685 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6686 makes life easier for people who are doing arithmetic on fields extracted
6687 from dates, where you often get leading zeros that should not be
6688 interpreted as octal.
6690 64. Added qualify_domain to the redirect router, to override the global
6693 65. If a pathologically long header line contained very many addresses (the
6694 report of this problem mentioned 10 000) and each of them was rewritten,
6695 Exim could use up a very large amount of memory. (It kept on making new
6696 copies of the header line as it rewrote, and never released the old ones.)
6697 At the expense of a bit more processing, the header rewriting function has
6698 been changed so that it no longer eats memory in this way.
6700 66. The generation of the Received: header has been moved from the time that a
6701 message starts to be received, to the time that it finishes. The timestamp
6702 in the Received: header should now be very close to that of the <= log
6703 line. There are two side-effects of this change:
6705 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6706 logged header lines no longer include the local Received: line, because
6707 it has not yet been created. The same applies to a copy of the message
6708 that is returned to a non-SMTP sender when a message is rejected.
6710 (b) When a filter file is tested using -bf, no additional Received: header
6711 is added to the test message. After some thought, I decided that this
6714 This change does not affect the value of $received_for. It is still set
6715 after address rewriting, but before local_scan() is called.
6717 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6719 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6720 gave an unhelpful panic error message, and a defer error. I have managed to
6721 change this behaviour so that it now rejects any supplied certificate,
6722 which seems right, as the list of acceptable certificates is empty.
6724 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6725 gave an unhelpful defer error. I have not managed to make this reject any
6726 supplied certificates, but the error message it gives is "no certificate
6727 supplied", which is not helpful.
6729 70. exigrep's output now also includes lines that are not associated with any
6730 message, but which match the given pattern. Implemented by a patch from
6731 Martin Sluka, which also tidied up the Perl a bit.
6733 71. Recipient callout verification, like sender verification, was using <> in
6734 the MAIL FROM command. This isn't really the right thing, since the actual
6735 sender may affect whether the remote host accepts the recipient or not. I
6736 have changed it to use the actual sender in the callout; this means that
6737 the cache record is now keyed on a recipient/sender pair, not just the
6738 recipient address. There doesn't seem to be a real danger of callout loops,
6739 since a callout by the remote host to check the sender would use <>.
6740 [SEE ABOVE: changed after hitting problems.]
6742 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6743 temporary errors. However, in the case of such a code being given after
6744 the end of a data transmission (i.e. after ".") Exim was failing to write
6745 a retry record for the message. (Yes, there was some broken host that was
6746 actually sending 8xx at this point.)
6748 73. An unknown lookup type in a host list could cause Exim to panic-die when
6749 the list was checked. (An example that provoked this was putting <; in the
6750 middle of a list instead of at the start.) If this happened during a DATA
6751 ACL check, a -D file could be left lying around. This kind of configuration
6752 error no longer causes Exim to die; instead it causes a defer error. The
6753 incident is still logged to the main and panic logs.
6755 74. Buglet left over from Exim 3 conversion. The message "too many messages
6756 in one connection" was written to the rejectlog but not the mainlog, except
6757 when address rewriting (yes!) was being logged.
6759 75. Added write_rejectlog option.
6761 76. When a system filter was run not as root (that is, when system_filter_user
6762 was set), the values of the $n variables were not being returned to the
6763 main process; thus, they were not subsequently available in the $sn
6766 77. Added +return_path_on_delivery log selector.
6768 78. A connection timeout was being treated differently from recipients deferred
6769 when testing hosts_max_try with a message that was older than the host's
6770 retry timeout. (The host should not be counted, thus allowing all hosts to
6771 be tried at least once before bouncing.) This may have been the cause of an
6772 occasionally reported bug whereby a message would remain on the queue
6773 longer than the retry timeout, but would be bounced if a delivery was
6774 forced. I say "may" because I never totally pinned down the problem;
6775 setting up timeout/retry tests is difficult. See also the next item.
6777 79. The ultimate address timeout was not being applied to errors that involved
6778 a combination of host plus message (for example, a timeout on a MAIL
6779 command). When an address resolved to a number of possible hosts, and they
6780 were not all tried for each delivery (e.g. because of hosts_max_try), a
6781 message could remain on the queue longer than the retry timeout.
6783 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6786 81. Fixed an obscure SMTP outgoing bug which required at least the following
6787 conditions: (a) there was another message waiting for the same server;
6788 (b) the server returned 5xx to all RCPT commands in the first message so
6789 that the message was not completed; (c) the server dropped the connection
6790 or gave a negative response to the RSET that Exim sends to abort the
6791 transaction. The observed case was a dropped connection after DATA that had
6792 been sent in pipelining mode. That is, the server had advertised PIPELINING
6793 but was not implementing it correctly. The effect of the bug was incorrect
6794 behaviour, such as trying another host, and this could lead to a crash.
6800 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6801 and daemon.c were passed as pointers to ints; they should have been
6802 pointers to socklen_t variables (which are typically unsigned ints).
6804 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6807 3. Fixed a really odd bug that affected only the testing scheme; patching a
6808 certain fixed string in the binary changed the value of another string that
6809 happened to be identical to the end of the original first string.
6811 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6812 name", it returns that address as the IP address. On some operating
6813 systems (e.g. Solaris), it also passes back the IP address string as the
6814 "host name". However, on others (e.g. Linux), it passes back an empty
6815 string. Exim wasn't checking for this, and was changing the host name to an
6816 empty string, assuming it had been canonicalized.
6818 5. Although rare, it is permitted to have more than one PTR record for a given
6819 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6820 all the names associated with an address, because they do in Solaris.
6821 However, it seems that they do not in Linux for data that comes from the
6822 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6823 I found this out when I moved to a new Linux workstation and tried to run
6824 the Exim test suite.
6826 To get round this problem I have changed the code so that it now does its
6827 own call to the DNS to look up PTR records when searching for a host name.
6828 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6829 addresses that are only in /etc/hosts are still found.
6831 This behaviour is, however, controlled by an option called host_lookup_
6832 order, which defaults to "bydns:byaddr". If people want to use the other
6833 order, or indeed, just use one or the other means of lookup, they can
6834 specify it in this variable.
6836 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6837 some operating systems, this comes back from gethostbyaddr() as an empty
6838 string, and this is what Exim used to test for. However, it seems that in
6839 other systems, "." is yielded. Exim now tests for this case too.
6841 7. The values of check_spool_space and check_log_space are now held internally
6842 as a number of kilobytes instead of an absolute number of bytes. If a
6843 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6844 kilobyte. This means that much larger values can be stored.
6846 8. Exim monitor: an attempt to get the action menu when not actually pointing
6847 at a message produces an empty menu entitled "No message selected". This
6848 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6849 no entries in it ("Shell widget menu has zero width and/or height"). So I
6850 have added a single, blank menu entry in this case.
6852 9. Added ${quote_local_part.
6854 10. MIME decoding is now applied to the contents of Subject: header lines when
6857 11. Now that a reference to $sender_host_address automatically causes a reverse
6858 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6859 host lookup before query-style lookups in lists that might use this
6860 variable. This has therefore been abolished, and the "net-" prefix is no
6861 longer necessary for query-style lookups.
6863 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6864 have been a typo for LFLAGS, so it has been changed.
6866 13. The install script calls Exim with "-C /dev/null" in order to find the
6867 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6868 to be output. However, since Exim outputs its version number before the
6869 error, it didn't break the script. It just looked ugly. I fixed this by
6870 always allowing "-C /dev/null" if the caller is root.
6872 14. Ignore overlarge ACL variable number when reading spool file - insurance
6873 against a later release with more variables having written the file.
6875 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6876 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6877 this, as well as the form without the "IPv6" on the front.
6879 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6880 OS/Makefile-Darwin file.
6882 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6883 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6885 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6886 variable. However, some OS use ints for inodes. I've added cast to long int
6887 to get rid of the compiler warning.
6889 19. I had forgotten to lock out "/../" in configuration file names when
6890 ALT_CONFIG_PREFIX was set.
6892 20. Routers used for verification do not need to specify transports. However,
6893 if such a router generated a host list, and callout was configured, Exim
6894 crashed, because it could not find a port number from the (non-existent)
6895 transport. It now assumes port 25 in this circumstance.
6897 21. Added the -t option to exigrep.
6899 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6900 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6901 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6902 them is set, LOOKUP_LSEARCH is forced.
6904 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6906 24. Added sender and host information to the "rejected by local_scan()" log
6907 line; previously there was no indication of these.
6909 25. Added .include_if_exists.
6911 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6912 Linux. It turns out that not all file systems support this. Apparently some
6913 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6914 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6915 not supported on the file descriptor, is now ignored when Exim is trying to
6916 sync a directory. This applies only to Linux.
6918 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6920 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6921 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6924 29. Added support for maildirsize files from supplied patch (modified a bit).
6926 30. The use of :fail: followed by an empty string could lead Exim to respond to
6927 sender verification failures with (e.g.):
6929 550 Verification failed for <xxx>
6930 550 Sender verify failed
6932 where the first response line was missing the '-' that indicates it is not
6933 the final line of the response.
6935 31. The loop for finding the name of the user that called Exim had a hardwired
6936 limit of 10; it now uses the value of finduser_retries, which is used for
6937 all other user lookups.
6939 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6942 33. Exim was neglecting to zero errno before one call of strtol() when
6943 expanding a string and expecting an integer value. On some systems this
6944 resulted in spurious "integer overflow" errors. Also, it was casting the
6945 result into an int without checking.
6947 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6948 did not work. The code looks as if it has *never* worked, though it appears
6949 to have been documented since at least release 1.62. I have made it work.
6951 35. The "timeout_DNS" error in retry rules, also documented since at least
6952 1.62, also never worked. As it isn't clear exactly what this means, and
6953 clearly it isn't a major issue, I have abolished the feature by treating it
6954 as "timeout", and writing a warning to the main and panic logs.
6956 36. The display of retry rules for -brt wasn't always showing the error code
6959 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6960 timeout_connect_A, timeout_connect_MX.
6962 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6963 to the empty sender.
6965 39. The daemon was not analysing the content of -oX till after it had closed
6966 stderr and disconnected from the controlling terminal. This meant that any
6967 syntax errors were only noted on the panic log, and the return code from
6968 the command was 0. By re-arranging the code a little, I've made the
6969 decoding happen first, so such errors now appear on stderr, and the return
6970 code is 1. However, the actual setting up of the sockets still happens in
6971 the disconnected process, so errors there are still only recorded on the
6974 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6975 connections (as happens on some IP stacks) was logged at start up time as
6976 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6977 it from "IPv6 and IPv4", which means that two separate sockets are being
6980 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6981 says whether AF_INET or AF_INET6 was passed as an argument.
6983 42. Exiwhat output was messed up when time zones were included in log
6986 43. Exiwhat now gives more information about the daemon's listening ports,
6987 and whether -tls-on-connect was used.
6989 44. The "port" option of the smtp transport is now expanded.
6991 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6992 silently ignored. Now an error message is written to the main and panic
6995 46. There's a new ACL modifier called "logwrite" which writes to a log file
6996 as soon as it is encountered.
6998 47. Added $local_user_uid and $local_user_gid at routing time.
7000 48. Exim crashed when trying to verify a sender address that was being
7003 49. Exim was recognizing only a space character after ".include". It now also
7004 recognizes a tab character.
7006 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7007 extracting the relevant information from the specification. The man page no
7008 longer contains scrambled data for the -d option, and I've added a section
7009 at the front about calling Exim under different names.
7011 51. Added "extra_headers" argument to the "mail" command in filter files.
7013 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7016 53. Installed eximstats 1.29.
7018 54. Added transport_filter_timeout as a generic transport option.
7020 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7021 Cc: header lines. This was required by RFC 822, but it not required by RFC
7024 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7025 incoming messages that did not have them. Now it does so only if the
7026 message originates locally, that is, if there is no associated remote host
7027 address. When Resent- header lines are present, this applies to the Resent-
7028 lines rather than the non-Resent- lines.
7030 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7031 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7033 58. Messages for configuration errors now include the name of the main
7034 configuration file - useful now that there may be more than one file in a
7035 list (.included file names were always shown).
7037 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7038 for those rare installations that do not start the daemon as root or run it
7039 setuid root. I've cut out the call to initgroups() if the daemon is not
7042 60. The Exim user and group can now be bound into the binary as text strings
7043 that are looked up at the start of Exim's processing.
7045 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7047 62. Added $mailstore_basename variable.
7049 63. Installed patch to sieve.c from Michael Haardt.
7051 64. When Exim failed to open the panic log after failing to open the main log,
7052 the original message it was trying to log was written to stderr and debug
7053 output, but if they were not available (the usual case in production), it
7054 was lost. Now it is written to syslog before the two lines that record the
7055 failures to open the logs.
7057 65. Users' Exim filters run in subprocesses under the user's uid. It is
7058 possible for a "deliver" command or an alias in a "personal" command to
7059 provoke an address rewrite. If logging of address rewriting is configured,
7060 this fails because the process is not running as root or exim. There may be
7061 a better way of dealing with this, but for the moment (because 4.30 needs
7062 to be released), I have disabled address rewrite logging when running a
7063 filter in a non-root, non-exim process.
7069 1. The buildconfig auxiliary program wasn't quoting the value set for
7070 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7071 not defined. This bug was masked in 4.22 by the effect that was fixed in
7074 2. Some messages that were rejected after a message id was allocated were
7075 shown as "incomplete" by exigrep. It no longer does this for messages that
7076 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7078 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7079 have allow_domain_literals set, the ID did not get logged in the <= line.
7080 Domain literals are now always recognized in Message-ID: header lines.
7082 4. The first argument for a ${extract expansion item is the key name or field
7083 number. Leading and trailing spaces in this item were not being ignored,
7084 causing some misleading effects.
7086 5. When deliver_drop_privilege was set, single queue runner processes started
7087 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7088 same command in the process it spins off) were not dropping privilege.
7090 6. When the daemon running as "exim" started a queue runner, it always
7091 re-executed Exim in the spun-off process. This is a waste of effort when
7092 deliver_drop_privilege is set. The new process now just calls the
7093 queue-runner function directly.
7099 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7102 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7103 ignored. Though the use of -oP was forcing the writing of a pid file, it
7104 was always written to the default place.
7106 3. If the message "no IP address found for host xxxx" is generated during
7107 incoming verification, it is now followed by identification of the incoming
7108 connection (so you can more easily find what provoked it).
7110 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7112 5. Added some features to "harden" Exim a bit more against certain attacks:
7114 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7115 be put in Local/Makefile. This is like the never_users runtime option,
7116 but it cannot be overridden. The default setting is "root".
7118 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7119 prefix string with which any file named in a -C command line option
7122 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7123 is retained for -C and -D only if the caller of Exim is root. Without
7124 it, the exim user may also use -C and -D and retain privilege.
7126 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7127 command line option is disabled.
7129 6. Macro names set by the -D option must start with an upper case letter, just
7130 like macro names defined in the configuration file.
7132 7. Added "dereference=" facility to LDAP.
7134 8. Two instances of the typo "uknown" in the source files are fixed.
7136 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7137 the Configure-Makefile script screwed up while processing it.
7139 10. Incorporated PCRE 4.4.
7141 11. The SMTP synchronization check was not operating right at the start of an
7142 SMTP session. For example, it could not catch a HELO sent before the client
7143 waited for the greeting. There is now a check for outstanding input at the
7144 point when the greeting is written. Because of the duplex, asynchronous
7145 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7146 way, but not yet received, when the check is performed.
7148 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7149 on TCP/IP sockets, because this apparently causes some broken clients to
7152 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7153 unchanged) from the Cygwin maintainer.
7155 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7156 was supported instead of testing for mbx. Effectively a typo.
7158 15. The spa authenticator server code was not checking that the input it
7159 received was valid base64.
7161 16. The debug output line for the "set" modifier in ACLs was not showing the
7162 name of the variable that was being set.
7164 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7166 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7167 lookup on its own. Something else was needed to trigger the lookup. For
7168 example, a match in host_lookup or the need for a host name in a host list.
7169 Now, if $sender_host_name is referenced and the host name has not yet been
7170 looked up, a lookup is performed. If the lookup fails, the variable remains
7171 empty, and $host_lookup_failed is set to "1".
7173 19. Added "eqi" as a case-independent comparison operator.
7175 20. The saslauthd authentication condition could segfault if neither service
7176 nor realm was specified.
7178 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7179 error message that was logged was misleading, and incoming SMTP
7180 connections were dropped. The message is now more accurate, and temporary
7181 errors are given to SMTP connections.
7183 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7184 (except RSET) with a 421 error, until QUIT is received. However, it was
7185 failing to send a response to QUIT.
7187 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7188 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7189 ACL is now run after the helo_try_verify_hosts code.
7191 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7192 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7193 case-independent, so other case variants are also recognized). Apparently
7194 some systems use these upper case variants.
7196 25. If more than two messages were waiting for the same host, and a transport
7197 filter was specified for the transport, Exim sent two messages over the
7198 same TCP/IP connection, and then failed with "socket operation on non-
7199 socket" when it tried to send the third.
7201 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7203 27. The extern definition of crypt16() in expand.c was not being excluded when
7204 the OS had its own crypt16() function.
7206 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7207 as a preferred synonym for return_size_limit, both as an option and as an
7210 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7212 30. Changed the default configuration ACL to relax the local part checking rule
7213 for addresses that are not in any local domains. For these addresses,
7214 slashes and pipe symbols are allowed within local parts, but the sequence
7215 /../ is explicitly forbidden.
7217 31. SPA server authentication was not clearing the challenge buffer before
7220 32. log_message in a "warn" ACL statement was writing to the reject log as
7221 well as to the main log, which contradicts the documentation and doesn't
7222 seem right (because no rejection is happening). So I have stopped it.
7224 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7225 However, I am unable to do any testing of this.
7227 34. Fixed an infelicity in the appendfile transport. When checking directories
7228 for a mailbox, to see if any needed to be created, it was accidentally
7229 using path names with one or more superfluous leading slashes; tracing
7230 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7232 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7233 contents are added to the log line that is written for every discarded
7234 recipient. (Previously a log_message setting was ignored.)
7236 36. The ${quote: operator now quotes the string if it is empty.
7238 37. The install script runs exim in order to find its version number. If for
7239 some reason other than non-existence or emptiness, which it checks, it
7240 could not run './exim', it was installing it with an empty version number,
7241 i.e. as "exim-". This error state is now caught, and the installation is
7244 38. An argument was missing from the function that creates an error message
7245 when Exim fails to connect to the socket for saslauthd authentication.
7246 This could cause Exim to crash, or give a corrupted message.
7248 39. Added isip, isip4, and isip6 to ${if conditions.
7250 40. The ACL variables $acl_xx are now saved with the message, and can be
7251 accessed later in routers, transports, and filters.
7253 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7254 strings in the file are not string-expanded.
7256 42. If a MAIL command specified a SIZE value that was too large to fit into an
7257 int variable, the check against message_size_limit failed. Such values are
7258 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7259 day this will have to be increased, but I don't think I want to be around
7260 when emails are that large.
7267 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7268 iconv() is not standard in FreeBSD.
7270 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7271 IPv6 enabled. The observed symptom was a segmentation fault on return from
7272 the function os_common_find_running_interfaces() in src/os.c.
7274 3. In the check_special_case() function in daemon.c I had used "errno" as an
7275 argument name, which causes warnings on some systems. This was basically a
7276 typo, since it was named "eno" in the comments!
7278 4. The code that waits for the clock to tick (at a resolution of some fraction
7279 of a second) so as to ensure message-id uniqueness was always waiting for
7280 at least one whole tick, when it could have waited for less. [This is
7281 almost certainly not relevant at current processor speeds, where it is
7282 unlikely to ever wait at all. But we try to future-proof.]
7284 5. The function that sleeps for a time interval that includes fractions of a
7285 second contained a race. It did not block SIGALRM between setting the
7286 timer, and suspending (a couple of lines later). If the interval was short
7287 and the sigsuspend() was delayed until after it had expired, the suspension
7288 never ended. On busy systems this could lead to processes getting stuck for
7291 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7292 process, before it forks any delivery processes. The open lookup caching
7293 mechanism meant that the open file or database connection was passed into
7294 the delivery process. The problem was that delivery processes always tidy
7295 up cached lookup data. This could cause a problem for the next delivery
7296 process started by the queue runner, because the external queue runner
7297 process does not know about the closure. So the next delivery process
7298 still has data in the lookup cache. In the case of a file lookup, there was
7299 no problem because closing a file descriptor in a subprocess doesn't affect
7300 the parent. However, if the lookup was caching a connection to a database,
7301 the connection was closed, and the second delivery process was likely to
7302 see errors such as "PGSQL: query failed: server closed the connection
7303 unexpectedly". The problem has been fixed by closing all cached lookups
7304 in a queue runner before running a delivery process.
7306 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7307 seem to have the "const" qualifier which it has on other OS. I've
7310 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7311 *of the same type* (client or server) with the same public name that an
7312 error should be diagnosed.
7314 9. When Exim looked up a host name for an IP address, but failed to find the
7315 original IP address when looking up the host name (a safety check), it
7316 output the message "<ip address> does not match any IP for NULL", which was
7317 confusing, to say the least. The bug was that the host name should have
7318 appeared instead of "NULL".
7320 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7321 user that is built into the binary, and the -C or -D options is used, root
7322 privilege is dropped before the configuration file is read. In addition,
7323 logging is switched to stderr instead of the normal log files. If the
7324 configuration then re-defines the Exim user, the unprivileged environment
7325 is probably not what is expected, so Exim logs a panic warning message (but
7328 However, if deliver_drop_privilege is set, the unprivileged state may well
7329 be exactly what is intended, so the warning has been cut out in that case,
7330 and Exim is allowed to try to write to its normal log files.
7336 1. smtp_return_error_details was not giving details for temporary sender
7337 or receiver verification errors.
7339 2. Diagnose a configuration error if two authenticators have the same public
7342 3. Exim used not to create the message log file for a message until the first
7343 delivery attempt. This could be confusing when incoming messages were held
7344 for policy or load reasons. The message log file is now created at the time
7345 the message is received, and an initial "Received" line is written to it.
7347 4. The automatically generated man page for command line options had a minor
7348 bug that caused no ill effects; however, a more serious problem was that
7349 the procedure for building the man page automatically didn't always
7350 operate. Consequently, release 4.20 contains an out-of-date version. This
7351 shouldn't happen again.
7353 5. When building Exim with embedded Perl support, the script that builds the
7354 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7355 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7357 6. The freeze_tell option was not being used for messages that were frozen on
7358 arrival, either by an ACL or by local_scan().
7360 7. Added the smtp_incomplete_transaction log selector.
7362 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7363 was accepting AUTH without a new EHLO.
7365 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7366 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7369 10. Logging of TCP/IP connections (when configured) now happens in the main
7370 daemon process instead of the child process, so that the TCP/IP connection
7371 count is more accurate (but it can never be perfect).
7373 11. The use of "drop" in a nested ACL was not being handled correctly in the
7374 outer ACL. Now, if condition failure induced by the nested "drop" causes
7375 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7376 or "require"), the connection is dropped.
7378 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7379 that yield "discard" can now be used with an "accept" or a "discard" verb,
7380 but an error is generated for any others (because I can't see a useful way
7381 to define what should happen).
7383 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7384 are now processed in the same way as lines in the Exim configuration file.
7385 In particular, continuation lines are supported.
7387 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7389 15. Added -ti meaning -t -i.
7391 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7392 domains, and warn by logging if others are found.
7394 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7395 returned from GIFCONF when Exim is trying to find the list of interfaces on
7396 a host. The code in os.c has been modified to copy each ifreq to an aligned
7397 structure in all cases.
7399 Also, in some cases, the returned ifreq's were being copied to a 'struct
7400 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7401 means the last couple of bytes of an IPv6 address could be chopped if the
7402 ifreq contained only a normal sockaddr (14 bytes storage).
7404 18. Named domain lists were not supported in the hosts_treat_as_local option.
7405 An entry such as +xxxx was not recognized, and was treated as a literal
7408 19. Ensure that header lines added by a DATA ACL are included in the reject log
7409 if the ACL subsequently rejects the message.
7411 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7412 MD5 (which is deprecated).
7414 21. When testing a filter file using -bf, Exim was writing a message when it
7415 took the sender from a "From " line in the message, but it was not doing so
7416 when it took $return_path from a Return-Path: header line. It now does.
7418 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7419 with a valid header line field name (a series of printing characters
7420 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7422 23. Changed "disc" in the source to "disk" to conform to the documentation and
7423 the book and for uniformity.
7425 24. Ignore Sendmail's -Ooption=value command line item.
7427 25. When execve() failed while trying to run a command in a pipe transport,
7428 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7429 could be confused with a return value of 69 from the command itself. This
7430 has been changed to 127, the value the shell returns if it is asked to run
7431 a non-existent command. The wording for the related log line suggests a
7432 non-existent command as the problem.
7434 26. If received_header_text expands to an empty string, do not add a Received:
7435 header line to the message. (Well, it adds a token one on the spool, but
7436 marks it "old" so that it doesn't get used or transmitted.)
7438 27. Installed eximstats 1.28 (addition of -nt option).
7440 28. There was no check for failure on the call to getsockname() in the daemon
7441 code. This can fail if there is a shortage of resources on the system, with
7442 ENOMEM, for example. A temporary error is now given on failure.
7444 29. Contrary to the C standard, it seems that in some environments, the
7445 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7446 program. Exim now does this explicitly; it affects the formatting of
7447 timestamps using strftime().
7449 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7450 complaints. I've now initialized all the variables, to avoid this.
7452 32. Header lines added by a system filter were not being "seen" during
7453 transport-time rewrites.
7455 33. The info_callback() function passed to OpenSSL is set up with type void
7456 (*)(SSL *, int, int), as described somewhere. However, when calling the
7457 function (actually a macro) that sets it up, the type void(*)() is
7458 expected. I've put in a cast to prevent warnings from picky compilers.
7460 34. If a DNS black list lookup found a CNAME record, but there were no A
7461 records associated with the domain it pointed at, Exim crashed.
7463 35. If a DNS black list lookup returned more than one A record, Exim ignored
7464 all but the first. It now scans all returned addresses if a particular IP
7465 value is being sought. In this situation, the contents of the
7466 $dnslist_value variable are a list of all the addresses, separated by a
7469 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7470 to accept a host name and all its aliases if the forward lookup for any of
7471 them yielded the IP address of the incoming connection. Now it accepts only
7472 those names whose forward lookup yields the correct IP address. Any other
7473 names are discarded. This closes a loophole whereby a rogue DNS
7474 administrator could create reverse DNS records to break through a
7475 wildcarded host restriction in an ACL.
7477 37. If a user filter or a system filter that ran in a subprocess used any of
7478 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7479 the wrong values were passed to the pipe command ($thisaddress had the
7480 value of $0, $0 had the value of $1, etc). This bug was introduced by
7481 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7483 38. Improved the line breaking for long SMTP error messages from ACLs.
7484 Previously, if there was no break point between 40 and 75 characters, Exim
7485 left the rest of the message alone. Two changes have been made: (a) I've
7486 reduced the minimum length to 35 characters; (b) if it can't find a break
7487 point between 35 and 75 characters, it looks ahead and uses the first one
7488 that it finds. This may give the occasional overlong line, but at least the
7489 remaining text gets split now.
7491 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7492 file descriptors might be low, and that setting 1000 would always raise it.
7493 It turns out that in some environments, the limit is already over 1000 and
7494 that lowering it causes trouble. So now Exim takes care not to decrease it.
7496 40. When delivering a message, the value of $return_path is set to $sender_
7497 address at the start of routing (routers may change the value). By an
7498 oversight, this default was not being set up when an address was tested by
7499 -bt or -bv, which affected the outcome if any router or filter referred to
7502 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7503 log only when "message" or "log_message" are set. However, if one of the
7504 conditions was an address verification, or a call to a nested ACL, the
7505 messages generated by the underlying test were being passed through. This
7506 no longer happens. The underlying message is available in $acl_verify_
7507 message for both "message" and "log_message" expansions, so it can be
7508 passed through if needed.
7510 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7511 new expansion $bh_ to give the encoded byte string without charset
7512 translation. Translation happens only if iconv() is available; HAVE_ICONV
7513 indicates this at build time. HEADERS_CHARSET gives the charset to
7514 translate to; headers_charset can change it in the configuration, and
7515 "headers charset" can change it in an individual filter file.
7517 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7518 that creates RFC 2047 encoded "words" labels them as that charset instead
7519 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7520 expansion operator; (ii) when Exim creates a From: line for a local
7521 message; (iii) when a header line is rewritten to include a "phrase" part.
7523 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7524 buggy, causing it to skip the first lines of messages whose message ID
7525 ended in 'D'. This would not have bitten before Exim release 4.14, because
7526 message IDs were unlikely to end in 'D' before then. The effect was to have
7527 incorrect size information for certain domains.
7529 45. #include "config.h" was missing at the start of the crypt16.c module. This
7530 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7533 46. If there was a timeout during a "random" callout check, Exim treated it as
7534 a failure of the random address, and carried on sending RSET and the real
7535 address. If the delay was just some slowness somewhere, the response to the
7536 original RCPT would be taken as a response to RSET and so on, causing
7537 mayhem of various kinds.
7539 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7540 when I implemented it. It didn't allow for the fact that some option values
7541 may legitimately be negative (e.g. size_addition), and it didn't even do
7542 the right test for positive values.
7544 48. Domain names in DNS records are case-independent. Exim always looks them up
7545 in lower case. Some resolvers return domain names in exactly the case they
7546 appear in the zone file, that is, they may contain uppercase letters. Not
7547 all resolvers do this - some return always lower case. Exim was treating a
7548 change of case by a resolver as a change of domain, similar to a widening
7549 of a domain abbreviation. This triggered its re-routing code and so it was
7550 trying to route what was effectively the same domain again. This normally
7551 caused routing to fail (because the router wouldn't handle the domain
7552 twice). Now Exim checks for this case specially, and just changes the
7553 casing of the domain that it ultimately uses when it transmits the message
7556 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7559 50. If a filter generated a file delivery with a non-absolute name (possible if
7560 no home directory exists for the router), the forbid_file option was not
7563 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7564 the existing equality matching.
7566 52. Exim was using ints instead of ino_t variables in some places where it was
7567 dealing with inode numbers.
7569 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7570 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7571 called TMPDIR, and if it finds it is different, it changes its value.
7573 54. The smtp_printf() function is now made available to local_scan() so
7574 additional output lines can be written before returning. There is also an
7575 smtp_fflush() function to enable the detection of a dropped connection.
7576 The variables smtp_input and smtp_batched_input are exported to
7579 55. Changed the default runtime configuration: the message "Unknown user"
7580 has been removed from the ACL, and instead placed on the localuser router,
7581 using the cannot_route_message feature. This means that any verification
7582 failures that generate their own messages won't get overridden. Similarly,
7583 the "Unrouteable address" message that was in the ACL for unverifiable
7584 relay addresses has also been removed.
7586 56. Added hosts_avoid_esmtp to the smtp transport.
7588 57. The exicyclog script was not checking for the esoteric option
7589 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7590 will work only if exicyclog is run under the appropriate euid.
7592 58. Following a discussion on the list, the rules by which Exim recognises line
7593 endings on incoming messages have been changed. The -dropcr and drop_cr
7594 options are now no-ops, retained only for backwards compatibility. The
7595 following line terminators are recognized: LF CRLF CR. However, special
7596 processing applies to CR:
7598 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7599 nor a local message in the state where . is a terminator.
7601 (ii) If a bare CR is encountered in a header line, an extra space is added
7602 after the line terminator so as not to end the header. The reasoning
7603 behind this is that bare CRs in header lines are most likely either
7604 to be mistakes, or people trying to play silly games.
7606 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7607 was being incorrectly given as 18 bytes larger than it should have been.
7608 This is a VOB (very old bug).
7610 60. This may never have affected anything current, but just in case it has:
7611 When the local host is found other than at the start of a list of hosts,
7612 the local host, those with the same MX, and any that follow, are discarded.
7613 When the list in question was part of a longer list of hosts, the following
7614 hosts (not currently being processed) were also being discarded. This no
7615 longer happens. I'm not sure if this situation could ever has previously
7618 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7621 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7622 header. This is something that is recommended in a new Internet Draft, and
7623 is something that is documented as being done by Sendmail. There are two
7624 possible values. For messages generated by the autoreply transport, Exim
7627 Auto-Submitted: auto-replied
7629 whereas for all other generated messages (e.g. bounces) it adds
7631 Auto-Submitted: auto-generated
7633 63. The "personal" condition in filters now includes a test for the
7634 Auto-Submitted: header. If it contains the string "auto-" the message it
7635 not considered personal.
7637 64. Added rcpt_include_affixes as a generic transport option.
7639 65. Added queue_only_override (default true).
7641 66. Added the syslog_duplication option.
7643 67. If what should have been the first header line of a message consisted of
7644 a space followed by a colon, Exim was mis-interpreting it as a header line.
7645 It isn't of course - it is syntactically invalid and should therefore be
7646 treated as the start of the message body. The misbehaviour could have
7647 caused a number of strange effects, including loss of data in subsequent
7648 header lines, and spool format errors.
7650 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7651 client host had authenticated. This control can now be exercised by an ACL
7652 for more flexibility.
7654 69. By default, callouts do not happen when testing with -bh. There is now a
7655 variant, -bhc, which does actually run the callout code, including
7656 consulting and updating the callout cache.
7658 70. Added support for saslauthd authentication, courtesy of Alexander
7661 71. If statvfs() failed on the spool or log directories while checking their
7662 size for availability, Exim confusingly gave the error "space shortage".
7663 Furthermore, in debugging mode it crashed with a floating point exception.
7664 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7665 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7666 serious problem, Exim now writes to the main and panic logs when this
7667 happens, with details of the failure. It then refuses to accept the
7668 incoming message, giving the message "spool directory problem" or "log
7669 directory problem" with a 421 code for SMTP messages.
7671 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7672 0, 1, and 2 exist, because some OS complain for execs without them (see
7673 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7674 descriptors. However, the code omitted to check that the open succeeded,
7675 causing mysterious errors if for some reason the permissions on /dev/null
7676 got screwed. Now Exim writes a message to the main and panic logs, and
7677 bombs out if it can't open /dev/null.
7679 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7680 interact so that it is all more flexible. It is supposed to remain
7681 backwards compatible. Also added extra_local_interfaces.
7683 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7684 to bomb out with an assertion failure - to the client this appears as a
7685 connection drop. This problem occurs in the part of the code that was taken
7686 from the Samba project. Fortunately, the assertion is in a very simple
7687 function, so I have fixed this by reproducing the function inline in the
7688 one place where it is called, and arranging for authentication to fail
7689 instead of killing the process with assert().
7691 75. The SPA client code was not working when the server requested OEM rather
7692 than Unicode encoding.
7694 76. Added code to make require_files with a specific uid setting more usable in
7695 the case where statting the file as root fails - usually a non-root-mounted
7696 NFS file system. When this happens and the failure is EACCES, Exim now
7697 forks a subprocess and does the per-uid checking as the relevant uid.
7699 77. Added process_log_path.
7701 78. If log_file_path was not explicitly set, a setting of check_log_space or
7702 check_log_inodes was ignored.
7704 79. If a space check for the spool or log partitions fails, the incident is now
7705 logged. Of course, in the latter case the data may get lost...
7707 80. Added the %p formatting code to string_format() so that it can be used to
7708 print addresses in debug_print(). Adjusted all the address printing in the
7709 debugging in store.c to use %p rather than %d.
7711 81. There was a concern that a line of code in smtp_in.c could overflow a
7712 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7713 initially expressed, the concern was not well-founded, because trailing
7714 spaces are removed early. However, if the trailing spaces were followed by
7715 a NULL, they did not get removed, so the overflow was possible. Two fixes
7718 (a) I re-wrote the offending code in a cleaner fashion.
7719 (b) If an incoming SMTP command contains a NULL character, it is rejected
7722 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7723 runs initgroups(), so that if the Exim user is in any additional groups,
7724 they will be used during message reception.
7730 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob