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.
138 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
139 side, to decode a TLS 1.3 packet capture.
141 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
142 Previously the default library behaviour applied, sending two, each in
145 JH/03 Debug output for ACL now gives the config file name and line number for
148 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
150 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
152 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
153 buffer overrun for (non-chunking) other transports.
155 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
156 TLS1.3, means that a server rejecting a client certificate is not visible
157 to the client until the first read of encrypted data (typically the
158 response to EHLO). Add detection for that case and treat it as a failed
159 TLS connection attempt, so that the normal retry-in-clear can work (if
160 suitably configured).
162 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
163 and/or domain. Found and fixed by Jason Betts.
165 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
166 configuration). If a CNAME target was not a wellformed name pattern, a
169 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
170 the OS reports them interleaved with other addresses.
172 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
173 used both for input and for a verify callout, both encrypted, SMTP
174 responses being sent by the server could be lost. This resulted in
175 dropped connections and sometimes bounces generated by a peer sending
178 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
179 string. Previously it was possible to cause undefined behaviour in a
180 library routine (usually a crash). Found by "zerons".
182 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
185 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
186 API was removed, so update to use the newer ones.
188 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
189 any timeout set, is taking a long time. Previously we would hang on to a
190 rotated logfile "forever" if the input was arriving with long gaps
191 (a previous attempt to fix addressed lack, for a long time, of initial
194 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
195 shared (NFS) environment. The length of the tempfile name is now
196 4 + 16 ("hdr.$message_exim_id") which might break on file
197 systems which restrict the file name length to lower values.
200 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
201 shared (NFS) environment.
203 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
204 did for all versions <4.90). Notably -M, -m, --invert, -I may be
207 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
208 on some platforms for bit 31.
210 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
211 to changes apparently associated with TLS1.3 handling some of the APIs
212 previously used were either nonfunctional or inappropriate. Strings
213 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
214 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
215 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
216 This affects log line X= elements, the $tls_{in,out}_cipher variables,
217 and the use of specific cipher names in the encrypted= ACL condition.
219 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
221 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
222 verification result was not updated unless hosts_require_ocsp applied.
224 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
225 queue_list_requires_admin set to false, non-admin users were denied the
228 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
229 directory-of-certs mode. Previously they were advertised despite the
232 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
233 A single TCP connection by a client will now hold a TLS connection open
234 for multiple message deliveries, by default. Previoud the default was to
237 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
238 default. If built with the facility, DANE will be used. The facility
239 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
241 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
242 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
243 must be defined and you must still, unless you define DISABLE_TLS, manage
244 the the include-dir and library-file requirements that go with that
245 choice. Non-TLS builds are still supported.
247 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
250 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
251 default. If the platform supports and has the facility enabled, it will
252 be requested on all coneections.
254 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
255 controlled by the build-time option SUPPORT_PIPE_CONNECT.
257 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
259 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
260 success-DSN messages. Previously the From: header was always the default
261 one for these; the option was ignored.
263 JH/28 Fix the timeout on smtp response to apply to the whole response.
264 Previously it was reset for every read, so a teergrubing peer sending
265 single bytes within the time limit could extend the connection for a
266 long time. Credit to Qualsys Security Advisory Team for the discovery.
268 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
269 delivery address, which leaked information of the results of local
270 forwarding. Change to the original envelope recipient address, per
273 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
274 requested. Previously not bounce was generated and a log entry of
275 error ignored was made.
277 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
279 JH/32 Introduce a general tainting mechanism for values read from the input
280 channel, and values derived from them. Refuse to expand any tainted
281 values, to catch one form of exploit.
283 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
284 was unused and the unexpanded text used for the test. Found and
285 fixed by Ruben Jenster.
287 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
288 an attempt to use a TLS library read routine dereffed a nul pointer,
291 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
292 them in our smtp response.
294 JH/36 Have the general router option retry_use_local_part default to true when
295 any of the restrictive preconditions are set (to anything). Previously it
296 was only for check_local user. The change removes one item of manual
297 configuration which is required for proper retries when a remote router
298 handles a subset of addresses for a domain.
300 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
301 link count into consideration.
303 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
304 caused the extension of big_buffer, the following lines were ignored.
306 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
307 accordance with RFC 2308. Previously there was no expiry, so a longlived
308 receive process (eg. due to ACL delays) versus a short SOA value could
311 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
313 JH/39 Promote DMARC support to mainline.
315 JH/40 Bug 2452: Add a References: header to DSNs.
317 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
318 parameters. The relevant library call is documented as "Deprecated: This
319 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
320 3.6.0, DH parameters are negotiated following RFC7919."
322 HS/06 Change the default of dnssec_request_domains to "*"
324 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
325 carried on and emitted a BDAT command, even when PIPELINING was not
328 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
329 buffer was used for the filename, resulting in a trap when tainted
330 arguments (eg. $domain) were used.
332 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
333 recommended to avoid a possible server-load attack. The feature can be
334 re-enabled via the openssl_options main cofiguration option.
336 JH/45 local_scan API: documented the current smtp_printf() call. This changed
337 for version 4.90 - adding a "more data" boolean to the arguments.
338 Bumped the ABI version number also, this having been missed previously;
339 release versions 4.90 to 4.92.3 inclusive were effectively broken in
340 respect of usage of smtp_printf() by either local_scan code or libraries
341 accessed via the ${dlfunc } expansion item. Both will need coding
342 adjustment for any calls to smtp_printf() to match the new function
343 signature; a FALSE value for the new argument is always safe.
345 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
346 the file-offset (which the Linux syscall does, and exim expects); this
347 resulted in an indefinite loop.
349 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
350 to do ARC verification. The Authentication-Results: header line added
351 by the configuration then had no ARC item.
357 JH/01 Remove code calling the customisable local_scan function, unless a new
358 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
360 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
361 non-signal-safe functions being used.
363 JH/03 Bug 2269: When presented with a received message having a stupidly large
364 number of DKIM-Signature headers, disable DKIM verification to avoid
365 a resource-consumption attack. The limit is set at twenty.
367 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
368 report of oldest_pass in ${authres } in consequence, and separate out
369 some descriptions of reasons for verification fail.
371 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
372 files in the spool were present and unlocked. A queue-runner could spot
373 them, resulting in a duplicate delivery. Fix that by doing the unlock
374 after the unlink. Investigation by Tim Stewart. Take the opportunity to
375 add more error-checking on spoolfile handling while that code is being
378 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
379 No known attacks, no CVE, this is defensive hardening.
381 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
382 a queue-runner could start a delivery while other operations were ongoing.
383 Cutthrough delivery was a common victim, resulting in duplicate delivery.
384 Found and investigated by Tim Stewart. Fix by using the open message data
385 file handle rather than opening another, and not locally closing it (which
386 releases a lock) for that case, while creating the temporary .eml format
387 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
389 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
390 $sender_verify_failure/$recipient_verify_failure to "random".
392 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
395 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
396 Previously this would segfault.
398 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
401 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
402 like zero, since the resolver should be doing this for us, But we need one
403 as a CNAME but no MX presence gets the CNAME returned; we need to check
404 that doesn't point to an MX to declare it "no MX returned" rather than
405 "error, loop". A new main option is added so the older capability of
406 following some limited number of chain links is maintained.
408 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
410 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
411 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
412 now-deprecated earlier definitions used only specified the range up to TLS
413 1.2 (in the older-version library docs).
415 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
417 JH/15 Rework TLS client-side context management. Stop using a global, and
418 explicitly pass a context around. This enables future use of TLS for
419 connections to service-daemons (eg. malware scanning) while a client smtp
420 connection is using TLS; with cutthrough connections this is quite likely.
422 JH/16 Fix ARC verification to do AS checks in reverse order.
424 JH/17 Support a "tls" option on the ${readsocket } expansion item.
426 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
427 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
428 Previously the "utf8" would be re-prepended for every additional message.
430 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
431 Previously thery were accepted, resulting in issues when attempting to
432 forward messages to a non-supporting MTA.
434 PP/02 Let -n work with printing macros too, not just options.
436 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
437 one parent address was copied, and bogus data was used at delivery-logging
438 time. Either a crash (after delivery) or bogus log data could result.
439 Discovery and analysis by Tim Stewart.
441 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
442 Previously if the string ended mid-character, we did not insert the
443 promised '?' replacement.
445 PP/04 Documentation: current string operators work on bytes, not codepoints.
447 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
448 should pack well giving a smaller memory footprint so better caching and
449 therefore performance. Group the declarations where this can't be done so
450 that the byte-sized flag variables are not interspersed among pointer
451 variables, giving a better chance of good packing by the compiler.
453 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
454 non-null, to avoid issues with sites running BATV. Previously reports were
455 sent with an empty envelope sender so looked like bounces.
457 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
458 The ignore_error flag wasn't being returned from the filter subprocess so
459 was not set for later routers. Investigation and fix by Matthias Kurz.
461 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
462 and a msg:complete for the whole, when a message is manually removed using
463 -Mrm. Developement by Matthias Kurz, hacked on by JH.
465 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
466 a "Gnu special" function, asprintf() in the DB utility binary builds; I
467 hope that is portable enough.
469 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
470 requiring a known-CA anchor certificate; make it now rely entirely on the
471 TLSA as an anchor. Checking the name on the leaf cert against the name
472 on the A-record for the host is still done for TA (but not for EE mode).
474 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
475 included in delivery lines for non-proxied connections, when compiled with
476 SUPPORT_SOCKS and running with proxy logging enabled.
478 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
479 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
480 move the existing event to fire before the normal logging of message
481 failure so that custom logging is bracketed by normal logging.
483 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
484 msg:fail:internal event. Developement by Matthias Kurz.
486 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
487 far too small for todays use of crypto signatures stored there. Go all
488 the way to the max DNS message size of 64kB, even though this might be
489 overmuch for IOT constrained device use.
491 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
492 copy a string over itself. The library routine is documented as not
493 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
495 JH/32 For main options check_spool_space and check_inode_space, where the
496 platform supports 64b integers, support more than the previous 2^31 kB
497 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
498 the previous G, M, k.
500 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
501 $authenticated_fail_id variable on authentication failure. Previously
504 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
505 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
506 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
507 GNUTLS_SEC_PARAM_MEDIUM.
509 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
510 side. Previously we would continue as if no SNI had been received.
512 JH/36 Harden the handling of string-lists. When a list consisted of a sole
513 "<" character, which should be a list-separator specification, we walked
514 off past the nul-terimation.
516 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
517 causes) even when the retry time is not yet met. Previously they were
518 not, meaning that when (say) an account was over-quota and temp-rejecting,
519 and multiple senders' messages were queued, only one sender would get
520 notified on each configured delay_warning cycle.
522 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
524 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
525 cipher-suites, an error can be left on the stack even for a succeeding
526 accept; this results in impossible error messages when a later operation
529 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
530 return error codes indicating retry. Under TLS1.3 this becomes required.
532 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
533 it only wrote the new authenticators, resulting in a lack of tracking of
534 peer changes of ESMTP extensions until the next cache flush.
536 JH/41 Fix the loop reading a message header line to check for integer overflow,
537 and more-often against header_maxsize. Previously a crafted message could
538 induce a crash of the recive process; now the message is cleanly rejected.
540 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
541 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
547 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
548 When redis_servers is set to a list of > 1 element, and the Redis servers
549 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
550 case of MOVED into a DEFER case instead, thus moving the query onto the
551 next server in the list. For a cluster of N elements, all N servers must
552 be defined in redis_servers.
554 GF/02 Catch and remove uninitialized value warning in exiqsumm
555 Check for existence of @ARGV before looking at $ARGV[0]
557 JH/01 Replace the store_release() internal interface with store_newblock(),
558 which internalises the check required to safely use the old one, plus
559 the allocate and data copy operations duplicated in both (!) of the
560 extant use locations.
562 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
563 modifier. This matches the restriction on the commandline.
565 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
566 Previously only the last row was returned.
568 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
569 we assumed that tags in the header were well-formed, and parsed the
570 element content after inspecting only the first char of the tag.
571 Assumptions at that stage could crash the receive process on malformed
574 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
575 While running the DKIM ACL we operate on the Permanent memory pool so that
576 variables created with "set" persist to the DATA ACL. Also (at any time)
577 DNS lookups that fail create cache records using the Permanent pool. But
578 expansions release any allocations made on the current pool - so a dnsdb
579 lookup expansion done in the DKIM ACL releases the memory used for the
580 DNS negative-cache, and bad things result. Solution is to switch to the
581 Main pool for expansions.
582 While we're in that code, add checks on the DNS cache during store_reset,
583 active in the testsuite.
584 Problem spotted, and debugging aided, by Wolfgang Breyha.
586 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
587 When none of the hosts presented to a transport match an already-open
588 connection, close it and proceed with the list. Previously we would
589 queue the message. Spotted by Lena with Yahoo, probably involving
592 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
593 Previously a spurious "250 OK id=" response was appended to the proper
596 JH/08 The "support for" informational output now, which built with Content
597 Scanning support, has a line for the malware scanner interfaces compiled
598 in. Interface can be individually included or not at build time.
600 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
601 by the template makefile "src/EDITME". The "STREAM" support for an older
602 ClamAV interface method is removed.
604 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
605 rows affected is given instead).
607 JH/11 The runtime Berkeley DB library version is now additionally output by
608 "exim -d -bV". Previously only the compile-time version was shown.
610 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
611 SMTP connection. Previously, when one had more receipients than the
612 first, an abortive onward connection was made. Move to full support for
613 multiple onward connections in sequence, handling cutthrough connection
614 for all multi-message initiating connections.
616 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
617 routers. Previously, a multi-recipient message would fail to match the
618 onward-connection opened for the first recipient, and cause its closure.
620 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
621 a timeout on read on a GnuTLS initiating connection, resulting in the
622 initiating connection being dropped. This mattered most when the callout
623 was marked defer_ok. Fix to keep the two timeout-detection methods
626 JH/15 Relax results from ACL control request to enable cutthrough, in
627 unsupported situations, from error to silently (except under debug)
628 ignoring. This covers use with PRDR, frozen messages, queue-only and
631 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
633 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
634 metadata, resulting in a crash in free().
636 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
637 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
638 Broken also in d185889f4, with init system revamp.
640 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
641 Previously we abruptly closed the connection after reading a malware-
642 found indication; now we go on to read the "scan ok" response line,
645 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
646 ACL. Previously, a crash would result.
648 JH/19 Speed up macro lookups during configuration file read, by skipping non-
649 macro text after a replacement (previously it was only once per line) and
650 by skipping builtin macros when searching for an uppercase lead character.
652 JH/20 DANE support moved from Experimental to mainline. The Makefile control
653 for the build is renamed.
655 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
656 was allocated for every new TLS startup, meaning one per message. Fix
657 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
659 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
660 reported the original. Fix to report (as far as possible) the ACL
661 result replacing the original.
663 JH/23 Fix memory leak during multi-message connections using STARTTLS under
664 OpenSSL. Certificate information is loaded for every new TLS startup,
665 and the resources needed to be freed.
667 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
669 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
672 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
673 DATA response info to the (existing) per-recipient response info for
674 the "C=" log element. It can have useful tracking info from the
675 destination system. Patch from Simon Arlott.
677 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
678 length value. Previously this would segfault.
680 HS/02 Support Avast multiline protoocol, this allows passing flags to
681 newer versions of the scanner.
683 JH/28 Ensure that variables possibly set during message acceptance are marked
684 dead before release of memory in the daemon loop. This stops complaints
685 about them when the debug_store option is enabled. Discovered specifically
686 for sender_rate_period, but applies to a whole set of variables.
687 Do the same for the queue-runner and queue-list loops, for variables set
688 from spool message files. Do the same for the SMTP per-message loop, for
689 certain variables indirectly set in ACL operations.
691 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
692 as a multi-recipient message from a mailinglist manager). The coding had
693 an arbitrary cutoff number of characters while checking for more input;
694 enforced by writing a NUL into the buffer. This corrupted long / fast
695 input. The problem was exposed more widely when more pipelineing of SMTP
696 responses was introduced, and one Exim system was feeding another.
697 The symptom is log complaints of SMTP syntax error (NUL chars) on the
698 receiving system, and refused recipients seen by the sending system
699 (propating to people being dropped from mailing lists).
700 Discovered and pinpointed by David Carter.
702 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
703 replaced by the ${authresults } expansion.
705 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
707 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
708 allows proper process termination in container environments.
710 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
711 Previously the "final dot" had a newline after it; ensure it is CR,LF.
713 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
714 and "err_perm", deprecated since 4.83 when the RFC-defined words
715 "temperror" and "permerror" were introduced.
717 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
718 transport-filters or DKIM-signing. The restriction was lost in the
719 consolidation of verify-callout and delivery SMTP handling.
720 Extend the restriction to also cover ARC-signing.
722 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
723 in defer=pass mode supply a 450 to the initiator. Previously the message
726 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
727 tls_require_ciphers is used as before.
729 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
730 "pass_unscanned". Only tmpfails from the scanner are written to
731 the paniclog, as they may require admin intervention (permission
732 denied, license issues). Other scanner errors (like decompression
733 bombs) do not cause a paniclog entry.
735 JH/36 Fix reinitialisation of DKIM logging variable between messages.
736 Previously it was possible to log spurious information in receive log
739 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
740 triggered odd behaviour from Outlook Express clients.
742 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
745 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
746 since the IETF WG has not yet settled on that versus the original
747 "bare" representation.
749 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
750 Previously the millisecond value corrupted the output.
751 Fix also for syslog_pid=no and log_selector +pid, for which the pid
752 corrupted the output.
758 JH/01 Rework error string handling in TLS interface so that the caller in
759 more cases is responsible for logging. This permits library-sourced
760 string to be attached to addresses during delivery, and collapses
761 pairs of long lines into single ones.
763 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
764 during configuration. Wildcards are allowed and expanded.
766 JH/02 Rework error string handling in DKIM to pass more info back to callers.
767 This permits better logging.
769 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
770 do not close it down and have the child transport start it up again on
771 the passed-on TCP connection. Instead, proxy the child (and any
772 subsequent ones) for TLS via a unix-domain socket channel. Logging is
773 affected: the continued delivery log lines do not have any DNSSEC, TLS
774 Certificate or OCSP information. TLS cipher information is still logged.
776 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
777 identical IP addresses on different listening ports. Will also affect
780 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
781 add noisy ifdef guards to special-case this sillyness.
782 Patch from Bernd Kuhls.
784 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
785 than 255 are no longer allowed.
787 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
788 Disable the session-cache too, which might reduce our load. Since we
789 currrectly use a new context for every connection, both as server and
790 client, there is no benefit for these.
791 GnuTLS appears to not support tickets server-side by default (we don't
792 call gnutls_session_ticket_enable_server()) but client side is enabled
793 by default on recent versions (3.1.3 +) unless the PFS priority string
796 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
797 <https://reproducible-builds.org/specs/source-date-epoch/>.
799 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
800 the check for any unsuccessful recipients did not notice the limit, and
801 erroneously found still-pending ones.
803 JH/08 Pipeline CHUNKING command and data together, on kernels that support
804 MSG_MORE. Only in-clear (not on TLS connections).
806 JH/09 Avoid using a temporary file during transport using dkim. Unless a
807 transport-filter is involved we can buffer the headers in memory for
808 creating the signature, and read the spool data file once for the
809 signature and again for transmission.
811 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
812 4.77 as the kernel support then wasn't solid, having issues in 64bit
813 mode. Now, it's been long enough. Add support for FreeBSD also.
815 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
816 case where the routing stage had gathered several addresses to send to
817 a host before calling the transport for the first, we previously failed
818 to close down TLS in the old transport process before passing the TCP
819 connection to the new process. The new one sent a STARTTLS command
820 which naturally failed, giving a failed delivery and bloating the retry
821 database. Investigation and fix prototype from Wolfgang Breyha.
823 JH/12 Fix check on SMTP command input synchronisation. Previously there were
824 false-negatives in the check that the sender had not preempted a response
825 or prompt from Exim (running as a server), due to that code's lack of
826 awareness of the SMTP input buffering.
828 PP/04 Add commandline_checks_require_admin option.
829 Exim drops privileges sanely, various checks such as -be aren't a
830 security problem, as long as you trust local users with access to their
831 own account. When invoked by services which pass untrusted data to
832 Exim, this might be an issue. Set this option in main configuration
833 AND make fixes to the calling application, such as using `--` to stop
836 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
837 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
838 responses to those, into a single TLS record each way (this usually means
839 a single packet). As a side issue, smtp_enforce_sync now works on TLS
842 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
843 affects you only if you're dancing at the edge of the param size limits.
844 If you are, and this message makes sense to you, then: raise the
845 configured limit or use OpenSSL 1.1. Nothing we can do for older
848 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
849 cmdline element to get the documented default one. Previously it was
852 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
853 are unsuable for non-message-specific reasons - eg. connection timeout,
854 banner-time rejection.
856 JH/16 Fix logging of delivery remote port, when specified by router, under
859 PP/06 Repair manualroute's ability to take options in any order, even if one
860 is the name of a transport.
863 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
865 JH/17 Change the list-building routines interface to use the expanding-string
866 triplet model, for better allocation and copying behaviour.
868 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
869 Previously it was constructed the first time a possibly-matching string
870 was met in the configuration file input during startup; now it is done
873 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
874 compatible one, to avoid the (poorly documented) possibility of a config
875 file in the working directory redirecting the DB files, possibly correpting
876 some existing file. CVE-2017-10140 assigned for BDB.
878 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
879 cache-hot. Previously, although the result was properly cached, the
880 initial verify call returned a defer.
882 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
883 the main verify for receipient in uncached-mode.
885 JH/22 Retire historical build files to an "unsupported" subdir. These are
886 defined as "ones for which we have no current evidence of testing".
888 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
889 if present. Previously it was ignored.
891 JH/24 Start using specified-initialisers in C structure init coding. This is
892 a C99 feature (it's 2017, so now considered safe).
894 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
895 if was a fixed-sized field and bitmask ops via macros; it is now more
898 PP/07 GitHub PR 56: Apply MariaDB build fix.
899 Patch provided by Jaroslav Škarvada.
901 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
902 during Coverity cleanups [4.87 JH/47]
903 Diagnosis and fix provided by Michael Fischer v. Mollard.
905 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
906 the right size to place the terminating semicolon on its own folded
907 line, the header hash was calculated to an incorrect value thanks to
908 the (relaxed) space the fold became.
910 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
911 and confused the parent.
913 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
914 which could crash as a result. This could lead to undeliverable messages.
916 JH/28 Logging: "next input sent too soon" now shows where input was truncated
919 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
920 matters on fast-turnover and PID-randomising systems, which were getting
921 out-of-order delivery.
923 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
924 a possibly-overlapping copy. The symptom was that "Remote host closed
925 connection in response to HELO" was logged instead of the actual 4xx
928 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
929 Previously only that bufferd was discarded, resulting in SYMTP command
932 JH/32 DKIM: when a message has multiple signatures matching an identity given
933 in dkim_verify_signers, run the dkim acl once for each. Previously only
934 one run was done. Bug 2189.
936 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
937 "panic the current process" to "deliberately defer". The panic log is
938 still written with the problem list name; the mail and reject logs now
939 get a temp-reject line for the message that was being handled, saying
940 something like "domains check lookup or other defer". The SMTP 451
941 message is still "Temporary local problem".
943 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
944 A crafted sequence of BDAT commands could result in in-use memory beeing
945 freed. CVE-2017-16943.
947 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
948 from SMTP input. Previously it was always done; now only done for DATA
949 and not BDAT commands. CVE-2017-16944.
951 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
952 to the message (such as an overlong header line). Previously this was
953 not done and we did not exit BDAT mode. Followon from the previous item
954 though a different problem.
960 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
961 than -2003 did; needs libidn2 in addition to libidn.
963 JH/02 The path option on a pipe transport is now expanded before use.
965 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
966 Patch provided by "Björn", documentation fix added too.
968 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
969 missing a wire-to-host endian conversion.
971 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
972 close after a BDAT command line could be taken as a following command,
973 giving a synch failure. Fix by only checking for synch immediately
974 before acknowledging the chunk.
976 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
977 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
978 macro. Patches provided by Josh Soref.
980 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
981 Previously we did not; the RFC seems ambiguous and VRFY is not listed
982 by IANA as a service extension. However, John Klensin suggests that we
985 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
986 the dkim code may be unix-mode line endings rather than smtp wire-format
987 CRLF, so prepend a CR to any bare LF.
989 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
990 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
992 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
993 the first were themselves being wrongly included in the feed into dkim
994 processing; with most chunk sizes in use this resulted in an incorrect
995 body hash calculated value.
997 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
998 DKIM signature block, for verification. Although advised against by
999 standards it is specifically not ruled illegal.
1001 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1003 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1004 missing a body hash (the bh= tag).
1006 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1007 It seems that HAProxy sends the Proxy Protocol information in clear and
1008 only then does a TLS startup, so do the same.
1010 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1011 TCP connections (such as for Spamd) unless the daemon successfully set
1012 Fast Open mode on its listening sockets. This fixes breakage seen on
1013 too-old kernels or those not configured for Fast Open, at the cost of
1014 requiring both directions being enabled for TFO, and TFO never being used
1015 by non-daemon-related Exim processes.
1017 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1018 endings, at least on the first header line. Try to canonify any that get
1019 past that check, despite the cost.
1021 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1022 now limited to an arbitrary five deep, while parsing addresses with the
1023 strip_excess_angle_brackets option enabled.
1025 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1026 instead leave the unprompted TLS handshake in socket buffer for the
1027 TLS library to consume.
1029 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1031 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1033 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1034 region in one operation, for speed, and this leaves assigned pointers
1035 dangling. Add checks run only under the testsuite which checks all
1036 variables at a store-reset and panics on a dangling pointer; add code
1037 explicitly nulling out all the variables discovered. Fixes one known
1038 bug: a transport crash, where a dangling pointer for $sending_ip_address
1039 originally assigned in a verify callout, is re-used.
1041 PP/06 Drop '.' from @INC in various Perl scripts.
1043 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1045 PP/08 Reduce a number of compilation warnings under clang; building with
1046 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1047 should be warning-free.
1049 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1051 HS/01 Fix portability problems introduced by PP/08 for platforms where
1052 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1054 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1055 chunk. This allows us to accept broken chunked messages. We need a more
1056 general solution here.
1058 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1059 already-broken messages in the queue.
1061 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1063 JH/19 Fix reference counting bug in routing-generated-address tracking.
1069 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1070 supports it and a size is available (ie. the sending peer gave us one).
1072 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1073 years of being deprecated). The replacements are the ACLs
1074 acl_smtp_mime and acl_not_smtp_mime.
1076 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1077 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1078 or even an in-clear connection were permitted. Now, if the host lookup
1079 was dnssec and dane was requested then the host is only used if the
1080 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1081 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1082 if one fails this test.
1083 This means that a poorly-configured remote DNS will make it incommunicado;
1084 but it protects against a DNS-interception attack on it.
1086 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1087 non-noisy when a race steals the message being considered.
1089 JH/05 If main configuration option tls_certificate is unset, generate a
1090 self-signed certificate for inbound TLS connections.
1092 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1093 in rewrites and routers.
1095 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1096 and logged a warning sing 4.83; now they are a configuration file error.
1098 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1099 (lacking @domain). Apply the same qualification processing as RCPT.
1101 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1103 JH/10 Support ${sha256:} applied to a string (as well as the previous
1106 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1107 a cutthrough deliver is pending, as we always want to make a connection.
1108 This also avoids re-routing the message when later placing the cutthrough
1109 connection after a verify cache hit.
1110 Do not update it with the verify result either.
1112 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1113 when routing results in more than one destination address.
1115 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1116 signing (which inhibits the cutthrough capability). Previously only
1117 the presence of an option was tested; now an expansion evaluating as
1118 empty is permissible (obviously it should depend only on data available
1119 when the cutthrough connection is made).
1121 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1122 the relevant preceding SMTP command did not note the pipelining mode.
1124 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1125 Previously they were not counted.
1127 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1128 as one having no matching records. Previously we deferred the message
1129 that needed the lookup.
1131 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1132 distinguished as "(=".
1134 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1135 for missing MX records. Previously it only worked for missing A records.
1137 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1139 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1140 after the data-go-ahead and data-ack. Patch from Jason Betts.
1142 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1143 even for a "none" policy. Patch from Tony Meyer.
1145 JH/22 Fix continued use of a connection for further deliveries. If a port was
1146 specified by a router, it must also match for the delivery to be
1149 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1150 When one of the recipients of a message was unsuitable for the connection
1151 (has no matching addresses), we lost track of needing to mark it
1152 deferred. As a result mail would be lost.
1154 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1156 JH/25 Decoding ACL controls is now done using a binary search; the source code
1157 takes up less space and should be simpler to maintain. Merge the ACL
1158 condition decode tables also, with similar effect.
1160 JH/26 Fix problem with one_time used on a redirect router which returned the
1161 parent address unchanged. A retry would see the parent address marked as
1162 delivered, so not attempt the (identical) child. As a result mail would
1165 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1166 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1167 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1170 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1171 Main config options check_{log,spool}_{inodes,space} are now
1172 100 inodes, 10MB unless set otherwise in the configuration.
1174 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1175 Previously it only applied to the main-section connection policy
1178 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1180 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1181 by me. Added RFC7919 DH primes as an alternative.
1183 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1184 are not in the system include path.
1186 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1187 GnuTLS, when a session startup failed (eg because the client disconnected)
1188 Exim did stdio operations after fclose. This was exposed by a recent
1189 change which nulled out the file handle after the fclose.
1191 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1192 signed directly by the cert-signing cert, rather than an intermediate
1193 OCSP-signing cert. This is the model used by LetsEncrypt.
1195 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1197 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1198 an incoming connection.
1200 HS/02 Bug 1802: Do not half-close the connection after sending a request
1203 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1204 fallback to "prime256v1".
1206 JH/34 SECURITY: Use proper copy of DATA command in error message.
1207 Could leak key material. Remotely exploitable. CVE-2016-9963.
1213 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1214 and 3.4.4 - once the server is enabled to respond to an OCSP request
1215 it does even when not requested, resulting in a stapling non-aware
1216 client dropping the TLS connection.
1218 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1219 support variable-length bit vectors. No functional change.
1221 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1222 The I= interface field on outgoing lines is now after the H= remote
1223 host field, same as incoming lines. There is a separate
1224 outgoing_interface log selector which allows you to disable the
1227 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1228 If not running log_selector +smtp_connection the mainlog would be held
1229 open indefinitely after a "too many connections" event, including to a
1230 deleted file after a log rotate. Leave the per net connection logging
1231 leaving it open for efficiency as that will be quickly detected by the
1232 check on the next write.
1234 HS/01 Bug 1671: Fix post transport crash.
1235 Processing the wait-<transport> messages could crash the delivery
1236 process if the message IDs didn't exist for some reason. When
1237 using 'split_spool_directory=yes' the construction of the spool
1238 file name failed already, exposing the same netto behaviour.
1240 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1241 mime_regex ACL conditions.
1243 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1244 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1245 to HELO, local diagnostic string.
1247 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1248 log line to debug. Even when configured with a tls authenticator many
1249 client connections are expected to not authenticate in this way, so
1250 an authenticate fail is not an error.
1252 HS/02 Add the Exim version string to the process info. This way exiwhat
1253 gives some more detail about the running daemon.
1255 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1256 matter for fast-change records such as DNSBLs.
1258 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1259 retry record, even if constant. There may be multiple transports with
1260 different interface settings and the retry behaviour needs to be kept
1263 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1265 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1267 JH/10 Bug 840: fix log_defer_output option of pipe transport
1269 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1270 hosts_randomize. This matters a lot when combined with mua_wrapper.
1272 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1273 ${quote_pgsql:<string>} operator.
1275 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1277 JH/14 Change header file name for freeradius-client. Relevant if compiling
1278 with Radius support; from the Gentoo tree and checked under Fedora.
1280 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1282 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1283 option transport_filter, meaning no filtering.
1285 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1287 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1288 defaults to "*" (all hosts). The variable is now available when not built
1289 with TLS, default unset, mainly to enable keeping the testsuite sane.
1290 If a server certificate is not supplied (via tls_certificate) an error is
1291 logged, and clients will find TLS connections fail on startup. Presumably
1292 they will retry in-clear.
1293 Packagers of Exim are strongly encouraged to create a server certificate
1294 at installation time.
1296 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1297 with the $config_file variable.
1299 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1300 in transport context, after the attempt, and per-recipient. The latter type
1301 is per host attempted. The event data is the error message, and the errno
1302 information encodes the lookup type (A vs. MX) used for the (first) host,
1303 and the trailing two digits of the smtp 4xx response.
1305 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1306 to write to mainlog (or rejectlog, paniclog) in the window between file
1307 creation and permissions/ownership being changed. Particularly affects
1308 installations where exicyclog is run as root, rather than exim user;
1309 result is that the running daemon panics and dies.
1311 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1313 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1314 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1315 "pri" and "weight". Note that the previous implicit priority given by the
1316 list order is no longer honoured.
1318 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1319 for DKIM processing.
1321 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1322 by defining SUPPORT_SOCKS.
1324 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1325 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1326 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1327 variables are renamed to proxy_{local,external}_{address,port}.
1329 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1330 for a build by defining SUPPORT_I18N
1332 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1333 of the query string, and make ${quote_redis:} do that quoting.
1335 JH/29 Move Events support from Experimental to mainline, enabled by default
1336 and removable for a build by defining DISABLE_EVENT.
1338 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1340 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1341 cached by the daemon.
1343 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1344 by defining LOOKUP_REDIS. The libhiredis library is required.
1346 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1347 keys are given for lookup.
1349 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1350 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1351 only supported when built with TLS support. The PolarSSL SHA routines
1352 are still used when the TLS library is too old for convenient support.
1354 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1355 openssl_options), for security. OpenSSL forces this from version 1.1.0
1356 server-side so match that on older versions.
1358 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1359 allocation for $value could be released as the expansion processing
1360 concluded, but leaving the global pointer active for it.
1362 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1363 and to use the domains and local_parts ACL conditions.
1365 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1366 incorrectly not doubled on cutthrough transmission, hence seen as a
1367 body-termination at the receiving system - resulting in truncated mails.
1368 Commonly the sender saw a TCP-level error, and retransmitted the message
1369 via the normal store-and-forward channel. This could result in duplicates
1370 received - but deduplicating mailstores were liable to retain only the
1371 initial truncated version.
1373 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1375 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1377 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1378 we're in there, support oversigning also; bug 1309.
1380 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1382 HS/04 Add support for keep_environment and add_environment options.
1384 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1385 either intentional arithmetic overflow during PRNG, or testing config-
1388 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1389 delivery resulted in actual delivery. Cancel cutthrough before DATA
1392 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1393 rejects a recipient: pass the reject to the originator.
1395 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1396 Many were false-positives and ignorable, but it's worth fixing the
1399 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1400 for the new environment-manipulation done at startup. Move the routines
1401 from being local to tls.c to being global via the os.c file.
1403 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1404 an extract embedded as result-arg for a map, the first arg for extract
1405 is unavailable so we cannot tell if this is a numbered or keyed
1406 extraction. Accept either.
1412 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1415 JH/02 The smtp transport option "multi_domain" is now expanded.
1417 JH/03 The smtp transport now requests PRDR by default, if the server offers
1420 JH/04 Certificate name checking on server certificates, when exim is a client,
1421 is now done by default. The transport option tls_verify_cert_hostnames
1422 can be used to disable this per-host. The build option
1423 EXPERIMENTAL_CERTNAMES is withdrawn.
1425 JH/05 The value of the tls_verify_certificates smtp transport and main options
1426 default to the word "system" to access the system default CA bundle.
1427 For GnuTLS, only version 3.0.20 or later.
1429 JH/06 Verification of the server certificate for a TLS connection is now tried
1430 (but not required) by default. The verification status is now logged by
1431 default, for both outbound TLS and client-certificate supplying inbound
1434 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1437 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1438 Status Notification (bounce) messages are now MIME format per RFC 3464.
1439 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1440 under the control of the dsn_advertise_hosts option, and routers may
1441 have a dsn_lasthop option.
1443 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1444 default, modifiable by a malware= option. The list separator for
1445 the options can now be changed in the usual way. Bug 68.
1447 JH/10 The smtp_receive_timeout main option is now expanded before use.
1449 JH/11 The incoming_interface log option now also enables logging of the
1450 local interface on delivery outgoing connections.
1452 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1453 if the interface and destination host and port all match.
1455 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1458 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1459 Patch from Andrew Lewis.
1461 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1462 now supports optional time-restrictions, weighting, and priority
1463 modifiers per server. Patch originally by <rommer@active.by>.
1465 JH/16 The spamd_address main option now supports a mixed list of local
1466 and remote servers. Remote servers can be IPv6 addresses, and
1467 specify a port-range.
1469 JH/17 Bug 68: The spamd_address main option now supports an optional
1470 timeout value per server.
1472 JH/18 Bug 1581: Router and transport options headers_add/remove can
1473 now have the list separator specified.
1475 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1478 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1481 JH/21 Support for the A6 type of dns record is withdrawn.
1483 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1484 rather than the verbs used.
1486 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1487 from 255 to 1024 chars.
1489 JH/24 Verification callouts now attempt to use TLS by default.
1491 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1492 are generic router options now. The defaults didn't change.
1494 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1495 Original patch from Alexander Shikoff, worked over by JH.
1497 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1498 files if ZCAT_COMMAND is not executable.
1500 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1502 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1504 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1505 Normally benign, it bites when the pair was led to by a CNAME;
1506 modern usage is to not canonicalize the domain to a CNAME target
1507 (and we were inconsistent anyway for A-only vs AAAA+A).
1509 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1511 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1512 when evaluating $sender_host_dnssec.
1514 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1515 $sender_helo_dnssec variable.
1517 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1519 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1521 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1523 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1524 documented as working, but never had. Support all but $spam_report.
1526 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1527 added for tls authenticator.
1529 HS/03 Add perl_taintmode main config option
1535 TL/01 When running the test suite, the README says that variables such as
1536 no_msglog_check are global and can be placed anywhere in a specific
1537 test's script, however it was observed that placement needed to be near
1538 the beginning for it to behave that way. Changed the runtest perl
1539 script to read through the entire script once to detect and set these
1540 variables, reset to the beginning of the script, and then run through
1541 the script parsing/test process like normal.
1543 TL/02 The BSD's have an arc4random API. One of the functions to induce
1544 adding randomness was arc4random_stir(), but it has been removed in
1545 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1546 function when detected.
1548 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1549 cause callback expansion.
1551 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1552 syntax errors in an expansion can be treated as a string instead of
1553 logging or causing an error, due to the internal use of bool_lax
1554 instead of bool when processing it.
1556 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1557 server certificates when making smtp deliveries.
1559 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1561 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1563 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1565 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1566 Merged patch from Sebastian Wiedenroth.
1568 JH/05 Fix results-pipe from transport process. Several recipients, combined
1569 with certificate use, exposed issues where response data items split
1570 over buffer boundaries were not parsed properly. This eventually
1571 resulted in duplicates being sent. This issue only became common enough
1572 to notice due to the introduction of connection certificate information,
1573 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1575 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1576 size buffer was used, resulting in syntax errors when an expansion
1579 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1580 version 3.3.6 or later.
1582 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1583 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1584 both become "event_action", the variables become $event_name, $event_data
1585 and $event_defer_errno. There is a new variable $verify_mode, usable in
1586 routers, transports and related events. The tls:cert event is now also
1587 raised for inbound connections, if the main configuration event_action
1590 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1591 early OCSP support, but no stapling (appears to be less than 1.0.0).
1593 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1594 server certificate names available under the smtp transport option
1595 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1598 JH/10 Time-related extraction expansions from certificates now use the main
1599 option "timezone" setting for output formatting, and are consistent
1600 between OpenSSL and GnuTLS compilations. Bug 1541.
1602 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1603 encoded parameter in the incoming message. Bug 1558.
1605 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1606 include certificate info, eximon was claiming there were spoolfile
1609 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1611 JH/14 Log delivery-related information more consistently, using the sequence
1612 "H=<name> [<ip>]" wherever possible.
1614 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1615 are problematic for Debian distribution, omit them from the release
1618 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1620 JH/16 Fix string representation of time values on 64bit time_t architectures.
1623 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1624 output list separator was used.
1629 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1630 checkers that were complaining about end of non-void function with no
1633 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1634 This was a regression introduced in 4.83 by another bugfix.
1636 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1638 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1639 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1645 TF/01 Correctly close the server side of TLS when forking for delivery.
1647 When a message was received over SMTP with TLS, Exim failed to clear up
1648 the incoming connection properly after forking off the child process to
1649 deliver the message. In some situations the subsequent outgoing
1650 delivery connection happened to have the same fd number as the incoming
1651 connection previously had. Exim would try to use TLS and fail, logging
1652 a "Bad file descriptor" error.
1654 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1655 utilities have not been installed.
1657 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1658 temporary space as the ACL may create new global variables.
1660 TL/01 LDAP support uses per connection or global context settings, depending
1661 upon the detected version of the libraries at build time.
1663 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1664 to extract and use the src ip:port in logging and expansions as if it
1665 were a direct connection from the outside internet. PPv2 support was
1666 updated based on HAProxy spec change in May 2014.
1668 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1670 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1671 Properly escape header and check for NULL return.
1673 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1676 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1678 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1679 characters in header names, implemented as a verify condition.
1680 Contributed by Michael Fischer v. Mollard.
1682 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1683 results permerror and temperror. Previous values are deprecated but
1684 still accepted. In a future release, err_perm and err_temp will be
1685 completely removed, which will be a backward incompatibility if the
1686 ACL tests for either of these two old results. Patch contributed by
1687 user bes-internal on the mailing list.
1689 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1691 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1692 selectors, in both main and reject logs.
1694 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1697 JH/07 Add malware type "sock" for talking to simple daemon.
1699 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1701 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1702 routers/transports under cutthrough routing.
1704 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1705 numbers. Touch up "bool" conditional to keep the same definition.
1707 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1709 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1711 JH/12 Expand items in router/transport headers_add or headers_remove lists
1712 individually rather than the list as a whole. Bug 1452.
1714 Required for reasonable handling of multiple headers_ options when
1715 they may be empty; requires that headers_remove items with embedded
1716 colons must have them doubled (or the list-separator changed).
1718 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1719 view the policy declared in the DMARC record. Currently, $dmarc_status
1720 is a combined value of both the record presence and the result of the
1723 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1725 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1726 dnslookup router and the smtp transport (applying to the forward
1729 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1730 of ldap servers used for a specific lookup. Patch provided by Heiko
1733 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1734 New variable $lookup_dnssec_authenticated for observability.
1736 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1737 Patch submitted by Lars Timman.
1739 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1741 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1742 Requires trusted mode and valid format message id, aborts otherwise.
1743 Patch contributed by Heiko Schlichting.
1745 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1746 certextract with support for various fields. Bug 1358.
1748 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1749 is requested by default, modifiable by smtp transport option
1752 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1753 operate on certificate variables to give certificate fingerprints
1754 Also new ${sha256:cert_variable}.
1756 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1758 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1759 Christian Aistleitner.
1761 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1763 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1764 file. Patch from Wolfgang Breyha.
1766 JH/25 Expand the coverage of the delivery $host and $host_address to
1767 client authenticators run in verify callout. Bug 1476.
1769 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1770 align with daemon_smtp_ports. Bug 72.
1772 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1773 support and error reporting did not work properly.
1775 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1776 and is readable. Patch from Andrew Colin Kissa.
1778 TL/14 Enhance documentation of ${run expansion and how it parses the
1779 commandline after expansion, particularly in the case when an
1780 unquoted variable expansion results in an empty value.
1782 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1784 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1785 by renaming away from it. Addresses GH issue 15, reported by
1788 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1790 TL/15 SECURITY: prevent double expansion in math comparison functions
1791 (can expand unsanitized data). Not remotely exploitable.
1798 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1800 PP/02 Make -n do something, by making it not do something.
1801 When combined with -bP, the name of an option is not output.
1803 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1806 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1807 $sender_host_name and config options to manage this, and basic check
1810 PP/05 DSCP support for outbound connections and control modifier for inbound.
1812 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1813 (Only plugin which currently uses this is kerberos4, which nobody should
1814 be using, but we should make it available and other future plugins might
1815 conceivably use it, even though it would break NAT; stuff *should* be
1816 using channel bindings instead).
1818 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1819 name; added for Sendmail compatibility; requires admin caller.
1820 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1821 just ignore it); requires trusted caller.
1822 Also parse but ignore: -Ac -Am -X<logfile>
1825 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1827 TL/02 Add +smtp_confirmation as a default logging option.
1829 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1830 Patch by Magnus Holmgren from 2007-02-20.
1832 TL/04 Bugzilla 1281 - Spec typo.
1833 Bugzilla 1283 - Spec typo.
1834 Bugzilla 1290 - Spec grammar fixes.
1836 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1838 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1840 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1841 mailing list by by Dmitry Isaikin.
1843 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1845 JH/02 Support "G" suffix to numbers in ${if comparisons.
1847 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1849 NM/01 Bugzilla 1197 - Spec typo
1850 Bugzilla 1196 - Spec examples corrections
1852 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1854 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1855 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1858 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1859 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1861 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1862 "acl {{name}{arg}...}", and optional args on acl condition
1865 JH/05 Permit multiple router/transport headers_add/remove lines.
1867 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1869 JH/07 Avoid using a waiting database for a single-message-only transport.
1870 Performance patch from Paul Fisher. Bugzilla 1262.
1872 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1875 JH/09 Add $headers_added variable, with content from use of ACL modifier
1876 add_header (but not yet added to the message). Bugzilla 199.
1878 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1879 Pulled from Bugzilla 817 by Wolfgang Breyha.
1881 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1883 (nb: this is the same fix as in Exim 4.80.1)
1885 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1888 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1891 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1893 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1894 advertise SMTP AUTH mechanism to us, instead of a generic
1895 protocol violation error. Also, make Exim more robust to bad
1896 data from the Dovecot auth socket.
1898 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1900 When a queue runner is handling a message, Exim first routes the
1901 recipient addresses, during which it prunes them based on the retry
1902 hints database. After that it attempts to deliver the message to
1903 any remaining recipients. It then updates the hints database using
1906 So if a recipient address works intermittently, it can get repeatedly
1907 deferred at routing time. The retry hints record remains fresh so the
1908 address never reaches the final cutoff time.
1910 This is a fairly common occurrence when a user is bumping up against
1911 their storage quota. Exim had some logic in its local delivery code
1912 to deal with this. However it did not apply to per-recipient defers
1913 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1915 This change adds a proper retry rule check during routing so that the
1916 final cutoff time is checked against the message's age. We only do
1917 this check if there is an address retry record and there is not a
1918 domain retry record; this implies that previous attempts to handle
1919 the address had the retry_use_local_parts option turned on. We use
1920 this as an approximation for the destination being like a local
1921 delivery, as in LMTP.
1923 I suspect this new check makes the old local delivery cutoff check
1924 redundant, but I have not verified this so I left the code in place.
1926 TF/02 Correct gecos expansion when From: is a prefix of the username.
1928 Test 0254 submits a message to Exim with the header
1932 When I ran the test suite under the user fanf2, Exim expanded
1933 the header to contain my full name, whereas it should have added
1934 a Resent-Sender: header. It erroneously treats any prefix of the
1935 username as equal to the username.
1937 This change corrects that bug.
1939 GF/01 DCC debug and logging tidyup
1940 Error conditions log to paniclog rather than rejectlog.
1941 Debug lines prefixed by "DCC: " to remove any ambiguity.
1943 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1945 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1946 Bug spotted by Jeremy Harris; was flawed since initial commit.
1947 Would have resulted in OCSP responses post-SNI triggering an Exim
1948 NULL dereference and crash.
1950 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1952 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1953 Bug detection, analysis and fix by Samuel Thibault.
1954 Bugzilla 1331, Debian bug #698092.
1956 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1958 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1959 Server implementation by Todd Lyons, client by JH.
1960 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1961 config variable "prdr_enable" controls whether the server
1962 advertises the facility. If the client requests PRDR a new
1963 acl_data_smtp_prdr ACL is called once for each recipient, after
1964 the body content is received and before the acl_smtp_data ACL.
1965 The client is controlled by both of: a hosts_try_prdr option
1966 on the smtp transport, and the server advertisement.
1967 Default client logging of deliveries and rejections involving
1968 PRDR are flagged with the string "PRDR".
1970 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1971 fclose(). Diagnosis by Todd Lyons.
1973 PP/17 Update configure.default to handle IPv6 localhost better.
1974 Patch by Alain Williams (plus minor tweaks).
1977 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1978 This is now consistent with GnuTLS, and is now documented: the
1979 previous undocumented portable approach to treating the option as
1980 unset was to force an expansion failure. That still works, and
1981 an empty string is now equivalent.
1983 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1984 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1985 not performing validation itself.
1987 PP/20 Added force_command boolean option to pipe transport.
1988 Patch from Nick Koston, of cPanel Inc.
1990 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1993 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1995 PP/21 Fix eximon continuous updating with timestamped log-files.
1996 Broken in a format-string cleanup in 4.80, missed when I repaired the
1997 other false fix of the same issue.
1998 Report and fix from Heiko Schlichting.
2001 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2002 Report from Prashanth Katuri.
2004 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2005 It's SecureTransport, so affects any MacOS clients which use the
2006 system-integrated TLS libraries, including email clients.
2008 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2009 using a MIME ACL for non-SMTP local injection.
2010 Report and assistance in diagnosis by Warren Baker.
2012 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2014 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2016 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2017 last id that failed so it may be referenced in subsequent ACL's.
2019 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2022 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2023 ldap library initialization, allowing self-signed CA's to be
2024 used. Also properly sets require_cert option later in code by
2025 using NULL (global ldap config) instead of ldap handle (per
2026 session). Bug diagnosis and testing by alxgomz.
2028 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2029 the src/util/ subdirectory.
2031 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2032 renamed to Transport Post Delivery Action by Jeremy Harris, as
2035 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2036 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2037 redis_servers = needs to be configured which will be used by the redis
2038 lookup. Patch from Warren Baker, of The Packet Hub.
2040 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2042 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2043 hostname or reverse DNS when processing a host list. Used suggestions
2044 from multiple comments on this bug.
2046 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2048 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2049 Missed a few lines, added it to make the runtest require no keyboard
2052 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2053 contains upper case chars. Make router use caseful_local_part.
2055 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2056 support when GnuTLS has been built with p11-kit.
2062 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2064 This, or similar/improved, will also be change PP/11 of 4.82.
2070 PP/01 Handle short writes when writing local log-files.
2071 In practice, only affects FreeBSD (8 onwards).
2072 Bugzilla 1053, with thanks to Dmitry Isaikin.
2074 NM/01 Bugzilla 949 - Documentation tweak
2076 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2079 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2081 PP/02 Implemented gsasl authenticator.
2083 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2085 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2086 `pkg-config foo` for cflags/libs.
2088 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2089 with rest of GSASL and with heimdal_gssapi.
2091 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2092 `pkg-config foo` for cflags/libs for the TLS implementation.
2094 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2095 properties get this fed in as external SSF. A number of robustness
2096 and debugging improvements to the cyrus_sasl authenticator.
2098 PP/08 cyrus_sasl server now expands the server_realm option.
2100 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2101 Patch by Jeremy Harris.
2103 PP/10 Added dbmjz lookup type.
2105 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2107 PP/12 MAIL args handles TAB as well as SP, for better interop with
2108 non-compliant senders.
2109 Analysis and variant patch by Todd Lyons.
2111 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2112 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2113 Patch from Dirk Mueller <dmueller@suse.com>
2115 PP/13 tls_peerdn now print-escaped for spool files.
2116 Observed some $tls_peerdn in wild which contained \n, which resulted
2117 in spool file corruption.
2119 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2120 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2121 or write after TLS renegotiation, which otherwise led to messages
2124 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2125 as a tracking header (ie: a signed header comes before the signature).
2126 Patch from Wolfgang Breyha.
2128 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2129 comma-sep list; embedded commas doubled.
2131 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2133 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2135 Report and patch from Dmitry Banschikov.
2137 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2138 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2139 failures. We appear to now support TLS1.1+ with Exim.
2141 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2142 lets Exim select keys and certificates based upon TLS SNI from client.
2143 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2144 before an outbound SMTP session. New log_selector, +tls_sni.
2146 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2147 NULL dereference. Report and patch from Alun Jones.
2149 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2150 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2151 Not seeing resolver debug output on NetBSD, but suspect this is a
2152 resolver implementation change.
2154 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2155 Left warnings. Added "eximon gdb" invocation mode.
2157 PP/21 Defaulting "accept_8bitmime" to true, not false.
2159 PP/22 Added -bw for inetd wait mode support.
2161 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2162 locate the relevant includes and libraries. Made this the default.
2164 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2165 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2167 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2168 This may cause build issues on older platforms.
2170 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2171 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2172 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2173 Added SNI support via GnuTLS too.
2174 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2176 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2178 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2179 Applied second patch from Janne, implementing suggestion to default
2180 multiple-strings-in-record handling to match SPF spec.
2182 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2184 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2185 read-only, out of scope).
2186 Patch from Wolfgang Breyha, report from Stuart Northfield.
2188 PP/29 Fix three issues highlighted by clang analyser static analysis.
2189 Only crash-plausible issue would require the Cambridge-specific
2190 iplookup router and a misconfiguration.
2191 Report from Marcin Mirosław.
2193 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2195 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2196 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2197 As part of this, removing so much warning spew let me fix some minor
2198 real issues in debug logging.
2200 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2201 assignment on my part. Fixed.
2203 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2204 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2205 Janne Snabb (who went above and beyond: thank you).
2207 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2208 string otherwise requires a connection and a bunch more work and it's
2209 relatively easy to get wrong. Should also expose TLS library linkage
2212 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2213 64-bit ${eval} (JH/03).
2215 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2216 GNU libc to support some of the 64-bit stuff, should not lead to
2217 conflicts. Defined before os.h is pulled in, so if a given platform
2218 needs to override this, it can.
2220 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2221 protection layer was required, which is not implemented.
2222 Bugzilla 1254, patch from Wolfgang Breyha.
2224 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2225 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2226 tls_dhparam take prime identifiers. Also unbreak combination of
2227 OpenSSL+DH_params+TLSSNI.
2229 PP/39 Disable SSLv2 by default in OpenSSL support.
2235 PP/01 Solaris build fix for Oracle's LDAP libraries.
2236 Bugzilla 1109, patch from Stephen Usher.
2238 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2240 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2243 TF/02 Fix a couple more cases where we did not log the error message
2244 when unlink() failed. See also change 4.74-TF/03.
2246 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2247 lock up or crash if it happened to be inside a call to libc when it
2248 got a SIGUSR1 from exiwhat.
2250 The SIGUSR1 handler appends the current process status to the process
2251 log which is later printed by exiwhat. It used to use the general
2252 purpose logging code to do this, but several functions it calls are
2253 not safe for signals.
2255 The new output code in the SIGUSR1 handler is specific to the process
2256 log, and simple enough that it's easy to inspect for signal safety.
2257 Removing some special cases also simplifies the general logging code.
2258 Removing the spurious timestamps from the process log simplifies
2261 TF/04 Improved ratelimit ACL condition.
2263 The /noupdate option has been deprecated in favour of /readonly which
2264 has clearer semantics. The /leaky, /strict, and /readonly update modes
2265 are mutually exclusive. The update mode is no longer included in the
2266 database key; it just determines when the database is updated. (This
2267 means that when you upgrade Exim will forget old rate measurements.)
2269 Exim now checks that the per_* options are used with an update mode that
2270 makes sense for the current ACL. For example, when Exim is processing a
2271 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2272 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2273 must specify per_mail/readonly. If you omit the update mode it defaults to
2274 /leaky where that makes sense (as before) or /readonly where required.
2276 The /noupdate option is now undocumented but still supported for
2277 backwards compatibility. It is equivalent to /readonly except that in
2278 ACLs where /readonly is required you may specify /leaky/noupdate or
2279 /strict/noupdate which are treated the same as /readonly.
2281 A useful new feature is the /count= option. This is a generalization
2282 of the per_byte option, so that you can measure the throughput of other
2283 aggregate values. For example, the per_byte option is now equivalent
2284 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2286 The per_rcpt option has been generalized using the /count= mechanism
2287 (though it's more complicated than the per_byte equivalence). When it is
2288 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2289 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2290 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2291 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2292 Note that using per_rcpt with a non-readonly update mode in more than
2293 one ACL will cause the recipients to be double-counted. (The per_mail
2294 and per_byte options don't have this problem.)
2296 The handling of very low rates has changed slightly. If the computed rate
2297 is less than the event's count (usually one) then this event is the first
2298 after a long gap. In this case the rate is set to the same as this event's
2299 count, so that the first message of a spam run is counted properly.
2301 The major new feature is a mechanism for counting the rate of unique
2302 events. The new per_addr option counts the number of different
2303 recipients that someone has sent messages to in the last time period. It
2304 behaves like per_rcpt if all the recipient addresses are different, but
2305 duplicate recipient addresses do not increase the measured rate. Like
2306 the /count= option this is a general mechanism, so the per_addr option
2307 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2308 example, measure the rate that a client uses different sender addresses
2309 with the options per_mail/unique=$sender_address. There are further
2310 details in the main documentation.
2312 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2314 TF/06 Removed a few PCRE remnants.
2316 TF/07 Automatically extract Exim's version number from tags in the git
2317 repository when doing development or release builds.
2319 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2320 Bugzilla 879. Patch from Paul Fisher.
2322 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2323 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2326 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2328 PP/05 Variable $av_failed, true if the AV scanner deferred.
2329 Bugzilla 1078. Patch from John Horne.
2331 PP/06 Stop make process more reliably on build failure.
2332 Bugzilla 1087. Patch from Heiko Schlittermann.
2334 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2335 Bugzilla 1089. Patch from Heiko Schlittermann.
2337 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2338 Bugzilla 1131. Patch from Holger Weiß.
2340 PP/09 Handle IPv6 addresses with SPF.
2341 Bugzilla 860. Patch from Wolfgang Breyha.
2343 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2345 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2348 PP/11 match_* no longer expand right-hand-side by default.
2349 New compile-time build option, EXPAND_LISTMATCH_RHS.
2350 New expansion conditions, "inlist", "inlisti".
2352 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2354 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2356 PP/14 fix log_write() format string regression from TF/03.
2357 Bugzilla 1152. Patch from Dmitry Isaikin.
2363 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2365 PP/02 Harmonised TLS library version reporting; only show if debugging.
2366 Layout now matches that introduced for other libraries in 4.74 PP/03.
2368 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2370 PP/04 New "dns_use_edns0" global option.
2372 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2375 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2376 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2378 TK/01 Updated PolarSSL code to 0.14.2.
2379 Bugzilla 1097. Patch from Andreas Metzler.
2381 PP/07 Catch divide-by-zero in ${eval:...}.
2382 Fixes bugzilla 1102.
2384 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2387 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2388 format-string attack -- SECURITY: remote arbitrary code execution.
2390 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2391 time unintentionally subject to list matching rules, letting the header
2392 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2393 arbitrary string expansion). This allowed for information disclosure.
2395 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2396 INT_MIN/-1 -- value coerced to INT_MAX.
2402 NM/01 Workaround for PCRE version dependency in version reporting
2405 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2406 This fixes portability to compilers other than gcc, notably
2407 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2409 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2410 makefiles for portability to HP-UX and POSIX correctness.
2412 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2413 Also via indented variable definition in the Makefile.
2414 (Debugging by Oliver Heesakkers).
2416 PP/02 Restore caching of spamd results with expanded spamd_address.
2417 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2419 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2420 Improves build reliability. Fix from: Frank Elsner
2422 NM/02 Fix wide character breakage in the rfc2047 coding
2423 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2425 NM/03 Allow underscore in dnslist lookups
2426 Fixes bug 1026. Patch from Graeme Fowler
2428 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2429 Code patches from Adam Ciarcinski of NetBSD.
2431 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2434 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2435 is logged, to avoid truncation. Patch from John Horne.
2437 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2438 Patch from Jakob Hirsch.
2440 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2441 SQL string expansion failure details.
2442 Patch from Andrey Oktyabrski.
2444 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2445 Patch from Simon Arlott.
2447 PP/09 New lookups functionality failed to compile on old gcc which rejects
2448 extern declarations in function scope.
2449 Patch from Oliver Fleischmann
2451 PP/10 Use sig_atomic_t for flags set from signal handlers.
2452 Check getgroups() return and improve debugging.
2453 Fixed developed for diagnosis in bug 927 (which turned out to be
2456 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2457 Patch from Mark Zealey.
2459 PP/12 Bugzilla 1056: Improved spamd server selection.
2460 Patch from Mark Zealey.
2462 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2463 Based on patch from Heiko Schlittermann.
2465 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2466 Patch from Uwe Doering, sign-off by Michael Haardt.
2468 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2469 variable declaration deep within a block. Bug and patch from
2472 PP/15 lookups-Makefile IRIX compatibility coercion.
2474 PP/16 Make DISABLE_DKIM build knob functional.
2476 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2477 Patch by Simon Arlott
2479 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2480 variable argument macros. Our copy now differs from upstream.
2486 TF/01 Failure to get a lock on a hints database can have serious
2487 consequences so log it to the panic log.
2489 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2490 controlled using the smtp_confirmation log selector.
2492 TF/03 Include the error message when we fail to unlink a spool file.
2494 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2495 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2496 for maintaining out-of-tree patches for some time.
2498 PP/01 Bugzilla 139: Documentation and portability issues.
2499 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2500 Handle per-OS dynamic-module compilation flags.
2502 PP/02 Let /dev/null have normal permissions.
2503 The 4.73 fixes were a little too stringent and complained about the
2504 permissions on /dev/null. Exempt it from some checks.
2505 Reported by Andreas M. Kirchwitz.
2507 PP/03 Report version information for many libraries, including
2508 Exim version information for dynamically loaded libraries. Created
2509 version.h, now support a version extension string for distributors
2510 who patch heavily. Dynamic module ABI change.
2512 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2513 privilege escalation vulnerability whereby the Exim run-time user
2514 can cause root to append content of the attacker's choosing to
2517 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2520 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2521 If dropping privileges for untrusted macros, we disabled normal logging
2522 on the basis that it would fail; for the Exim run-time user, this is not
2523 the case, and it resulted in successful deliveries going unlogged.
2524 Fixed. Reported by Andreas Metzler.
2530 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2531 only prepend for the Resent-* case. Fixes regression introduced in
2532 Exim 4.70 by NM/22 for Bugzilla 607.
2534 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2535 increasing numbers of administrators be bitten by this.
2537 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2539 PP/03 Bugzilla 994: added openssl_options main configuration option.
2541 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2543 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2545 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2547 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2548 without a peer certificate, leading to a segfault because of an
2549 assumption that peers always have certificates. Be a little more
2550 paranoid. Problem reported by Martin Tscholak.
2552 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2553 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2554 NB: ClamAV planning to remove STREAM in "middle of 2010".
2555 CL also introduces -bmalware, various -d+acl logging additions and
2556 more caution in buffer sizes.
2558 PP/09 Implemented reverse_ip expansion operator.
2560 PP/10 Bugzilla 937: provide a "debug" ACL control.
2562 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2564 PP/12 Bugzilla 973: Implement --version.
2566 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2568 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2570 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2572 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2573 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2574 ignore trailing whitespace.
2576 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2578 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2581 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2582 Notification from Dr Andrew Aitchison.
2584 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2585 ExtendedDetectionInfo response format.
2586 Notification from John Horne.
2588 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2591 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2592 XSL and documented dependency on system catalogs, with examples of how
2595 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2598 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2599 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2600 configuration file which is writeable by the Exim user or group.
2602 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2603 of configuration files to cover files specified with the -C option if
2604 they are going to be used with root privileges, not just the default
2607 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2608 option (effectively making it always true).
2610 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2611 files to be used while preserving root privileges.
2613 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2614 that rogue child processes cannot use them.
2616 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2617 run-time user, instead of root.
2619 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2620 Exim run-time user without dropping privileges.
2622 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2623 result string, instead of calling string_vformat() twice with the same
2626 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2627 for other users. Others should always drop root privileges if they use
2628 -C on the command line, even for a whitelisted configure file.
2630 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2632 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2638 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2639 $data_path, and $header_path variables; fixed documentation bugs and
2642 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2643 exipick to access non-standard spools, including the "frozen" queue
2646 NM/01 Bugzilla 965: Support mysql stored procedures.
2647 Patch from Alain Williams
2649 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2651 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2652 Patch from Andreas Metzler
2654 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2655 Patch from Kirill Miazine
2657 NM/05 Bugzilla 671: Added umask to procmail example.
2659 JJ/03 installed exipick 20100323.0, fixing doc bug
2661 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2662 directory. Notification and patch from Dan Rosenberg.
2664 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2666 TK/02 Improve log output when DKIM signing operation fails.
2668 MH/01 Treat the transport option dkim_domain as a colon separated
2669 list, not as a single string, and sign the message with each element,
2670 omitting multiple occurences of the same signer.
2672 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2673 Bugzilla 985, 986. Patch by Simon Arlott
2675 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2676 Patch by Simon Arlott
2678 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2679 MBX locking. Notification from Dan Rosenberg.
2685 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2687 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2689 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2691 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2693 NM/04 Bugzilla 915: Flag broken perl installation during build.
2699 TK/01 Added patch by Johannes Berg that expands the main option
2700 "spamd_address" if it starts with a dollar sign.
2702 TK/02 Write list of recipients to X-Envelope-Sender header when building
2703 the mbox-format spool file for content scanning (suggested by Jakob
2706 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2707 (http://www.dcc-servers.net/) support via dccifd. Activated by
2708 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2710 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2711 by Mark Daniel Reidel <mr@df.eu>.
2713 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2714 When building exim an external PCRE library is now needed -
2715 PCRE is a system library on the majority of modern systems.
2716 See entry on PCRE_LIBS in EDITME file.
2718 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2719 conversation. Added nologin parameter to request.
2720 Patch contributed by Kirill Miazine.
2722 TF/01 Do not log submission mode rewrites if they do not change the address.
2724 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2726 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2727 log files in place. Contributed by Roberto Lima.
2729 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2731 TF/03 Bugzilla 615: When checking the local_parts router precondition
2732 after a local_part_suffix or local_part_prefix option, Exim now
2733 does not use the address's named list lookup cache, since this
2734 contains cached lookups for the whole local part.
2736 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2737 Robert Millan. Documentation is in experimental-spec.txt.
2739 TF/04 Bugzilla 668: Fix parallel build (make -j).
2741 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2743 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2744 Patch provided by Jan Srzednicki.
2746 TF/05 Leading white space used to be stripped from $spam_report which
2747 wrecked the formatting. Now it is preserved.
2749 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2750 that they are available at delivery time.
2752 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2754 TF/08 TLS error reporting now respects the incoming_interface and
2755 incoming_port log selectors.
2757 TF/09 Produce a more useful error message if an SMTP transport's hosts
2758 setting expands to an empty string.
2760 NM/06 Bugzilla 744: EXPN did not work under TLS.
2761 Patch provided by Phil Pennock.
2763 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2764 Patch provided by Richard Godbee.
2766 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2767 acl_smtp_notquit, added index entry.
2769 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2770 Patch provided by Eugene Bujak.
2772 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2773 accept(). Patch provided by Maxim Dounin.
2775 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2776 Patch provided by Phil Pennock.
2778 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2780 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2781 Patch provided by Brad "anomie" Jorsch.
2783 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2784 Patch provided by Dean Brooks.
2786 TK/05 Add native DKIM support (does not depend on external libraries).
2788 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2789 Patch provided by Graeme Fowler.
2791 NM/16 Bugzilla 851: Documentation example syntax fix.
2793 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2795 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2798 NM/19 Bugzilla 745: TLS version reporting.
2799 Patch provided by Phil Pennock.
2801 NM/20 Bugzilla 167: bool: condition support.
2802 Patch provided by Phil Pennock.
2804 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2805 clients. Patch provided by Phil Pennock.
2807 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2808 Patch provided by Brad "anomie" Jorsch.
2810 NM/23 Bugzilla 687: Fix misparses in eximstats.
2811 Patch provided by Heiko Schlittermann.
2813 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2814 Patch provided by Heiko Schlittermann.
2816 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2817 plus update to original patch.
2819 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2821 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2822 Patch provided by David Brownlee.
2824 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2826 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2828 NM/30 Bugzilla 888: TLS documentation bugfixes.
2830 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2832 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2833 Unlike the original bugzilla I have changed all shell scripts in src tree.
2835 NM/33 Bugzilla 898: Transport filter timeout fix.
2836 Patch by Todd Rinaldo.
2838 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2839 Patch by Serge Demonchaux.
2841 NM/35 Bugzilla 39: Base64 decode bug fixes.
2842 Patch by Jakob Hirsch.
2844 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2846 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2848 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2850 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2856 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2857 ALT-N's libdkim that I have put here:
2858 http://duncanthrax.net/exim-experimental/
2860 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2861 were called 'true' and it seems that C99 defines that as a reserved
2862 keyword to be used with 'bool' variable types. That means you could
2863 not include C99-style headers which use bools without triggering
2864 build errors in sieve.c.
2866 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2867 as mailq or other aliases. Changed the --help handling significantly
2868 to do whats expected. exim_usage() emits usage/help information.
2870 SC/01 Added the -bylocaldomain option to eximstats.
2872 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2874 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2876 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2882 PH/01 Another patch from the Sieve maintainer.
2884 PH/02 When an IPv6 address is converted to a string for single-key lookup
2885 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2886 dots are used instead of colons so that keys in lsearch files need not
2887 contain colons. This was done some time before quoting was made available
2888 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2889 (notated using the quote facility) so as to distinguish them from IPv4
2890 keys. This meant that lookups for IP addresses in host lists did not work
2891 for iplsearch lookups.
2893 This has been fixed by arranging for IPv6 addresses to be expressed with
2894 colons if the lookup type is iplsearch. This is not incompatible, because
2895 previously such lookups could never work.
2897 The situation is now rather anomalous, since one *can* have colons in
2898 ordinary lsearch keys. However, making the change in all cases is
2899 incompatible and would probably break a number of configurations.
2901 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2904 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2905 conversion specification without a maximum field width, thereby enabling
2906 a rogue spamd server to cause a buffer overflow. While nobody in their
2907 right mind would setup Exim to query an untrusted spamd server, an
2908 attacker that gains access to a server running spamd could potentially
2909 exploit this vulnerability to run arbitrary code as the Exim user.
2911 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2912 $primary_hostname instead of what libspf2 thinks the hosts name is.
2914 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2915 a directory entry by the name of the lookup key. Previously, if a
2916 symlink pointed to a non-existing file or a file in a directory that
2917 Exim lacked permissions to read, a lookup for a key matching that
2918 symlink would fail. Now it is enough that a matching directory entry
2919 exists, symlink or not. (Bugzilla 503.)
2921 PH/03 The body_linecount and body_zerocount variables are now exported in the
2924 PH/04 Added the $dnslist_matched variable.
2926 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2927 This means they are set thereafter only if the connection becomes
2930 PH/06 Added the client_condition to authenticators so that some can be skipped
2931 by clients under certain conditions.
2933 PH/07 The error message for a badly-placed control=no_multiline_responses left
2934 "_responses" off the end of the name.
2936 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2938 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2939 (without spaces) instead of just copying the configuration text.
2941 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2943 PH/11 Added $max_received_linelength.
2945 PH/12 Added +ignore_defer and +include_defer to host lists.
2947 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2948 way in which PCRE > 7.0 is built.
2950 PH/14 Implemented queue_only_load_latch.
2952 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2953 MAIL command. The effect was to mangle the value on 64-bit systems.
2955 PH/16 Another patch from the Sieve maintainer.
2957 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2959 PH/18 If a system quota error occurred while trying to create the file for
2960 a maildir delivery, the message "Mailbox is full" was not appended to the
2961 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2962 only to a quota excession during the actual writing of the file.
2964 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2965 characters?) which causes problems in log lines. The DN values are now
2966 passed through string_printing() before being added to log lines.
2968 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2969 and InterBase are left for another time.)
2971 PH/21 Added message_body_newlines option.
2973 PH/22 Guard against possible overflow in moan_check_errorcopy().
2975 PH/23 POSIX allows open() to be a macro; guard against that.
2977 PH/24 If the recipient of an error message contained an @ in the local part
2978 (suitably quoted, of course), incorrect values were put in $domain and
2979 $local_part during the evaluation of errors_copy.
2985 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2986 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2989 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2990 issue a MAIL command.
2992 PH/02 In an ACL statement such as
2994 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2996 if a client was not listed at all, or was listed with a value other than
2997 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2998 the condition was not true (as it should be), so access was not denied.
2999 The bug was that the ! inversion was incorrectly passed on to the second
3000 item. This has been fixed.
3002 PH/03 Added additional dnslists conditions == and =& which are different from
3003 = and & when the dns lookup returns more than one IP address.
3005 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3006 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3008 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3009 FSYNC, which compiles an option called disable_fsync that allows for
3010 bypassing fsync(). The documentation is heavily laced with warnings.
3012 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3014 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3015 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3016 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3017 including adding "make clean"; (3) Added -fPIC when compiling the test
3018 dynamically loaded module, to get rid of a warning.
3020 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3021 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3022 = 0s. The bug is otherwise harmless.
3024 PH/07 There was a bug in the dovecot authenticator such that the value of
3025 $auth1 could be overwritten, and so not correctly preserved, after a
3026 successful authentication. This usually meant that the value preserved by
3027 the server_setid option was incorrect.
3029 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3031 PH/09 Installed PCRE release 7.0.
3033 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3034 run for batched SMTP input. It is now run at the start of every message
3035 in the batch. While fixing this I discovered that the process information
3036 (output by running exiwhat) was not always getting set for -bs and -bS
3037 input. This is fixed, and it now also says "batched" for BSMTP.
3039 PH/11 Added control=no_pipelining.
3041 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3042 patch, slightly modified), and move the expansion of helo_data till after
3043 the connection is made in the smtp transport (so it can use these
3046 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3048 PH/14 Added log_selector = +pid.
3050 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3052 PH/16 Add ${if forany and ${if forall.
3054 PH/17 Added dsn_from option to vary the From: line in DSNs.
3056 PH/18 Flush SMTP output before performing a callout, unless control =
3057 no_callout_flush is set.
3059 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3060 was true (the default) a successful delivery failed to delete the retry
3061 item, thus causing premature timeout of the address. The bug is now
3064 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3066 PH/21 Long custom messages for fakedefer and fakereject are now split up
3067 into multiline responses in the same way that messages for "deny" and
3068 other ACL rejections are.
3070 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3071 with slight modification.
3073 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3074 draft, changing the syntax and factoring some duplicate code".
3076 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3077 for deliveries of the second and subsequent messages over the same SMTP
3080 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3081 ${reduce, with only minor "tidies".
3083 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3085 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3086 expansion side effects.
3088 PH/27 When a message times out after an over-quota error from an Exim-imposed
3089 quota, the bounce message says "mailbox is full". This message was not
3090 being given when it was a system quota that was exceeded. It now should
3093 MH/03 Made $recipients available in local_scan(). local_scan() already has
3094 better access to the recipient list through recipients_list[], but
3095 $recipients can be useful in postmaster-provided expansion strings.
3097 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3098 in the case of a MAIL command with additional options following the
3099 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3100 were accidentally chopped off.
3102 PH/29 SMTP synchronization checks are implemented when a command is read -
3103 there is a check that no more input is waiting when there shouldn't be
3104 any. However, for some commands, a delay in an ACL can mean that it is
3105 some time before the response is written. In this time, more input might
3106 arrive, invalidly. So now there are extra checks after an ACL has run for
3107 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3108 pipelining has not been advertised.
3110 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3112 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3113 returned in the SMTP response when smtp_return_error_details was false.
3114 This has been fixed.
3116 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3117 instead of the C I/O that was originally supplied, because problems were
3118 reported on Solaris.
3120 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3121 Exim which did not show up earlier: it was assuming that a call to
3122 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3123 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3124 was a macro that became an assignment, so it seemed to work. This has
3125 changed to a proper function call with a void return, hence the compile
3126 error. Exim's code has been fixed.
3128 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3131 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3133 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3135 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3136 because it turns out to be too big or there is a timeout) there is no
3137 "Completed" line in the log. When some messages of this type were
3138 selected by exigrep, they were listed as "not completed". Others were
3139 picked up by some special patterns. I have improved the selection
3140 criteria to be more general.
3142 PH/38 The host_find_failed option in the manualroute router can now be set
3143 to "ignore", to completely ignore a host whose IP address cannot be
3144 found. If all hosts are ignored, the behaviour is controlled by the new
3145 host_all_ignored option.
3147 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3148 homing or because of multiple MX records with /mx) generated more than
3149 one IP address, and the following item turned out to be the local host,
3150 all the secondary addresses of the first item were incorrectly removed
3151 from the list, along with the local host and any following hosts (which
3152 is what is supposed to happen).
3154 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3155 whoever called Exim into the -H file. In the case of the daemon it was
3156 behaving confusingly. When first started, it used values for whoever
3157 started the daemon, but after a SIGHUP it used the Exim user (because it
3158 calls itself on a restart). I have changed the code so that it now always
3161 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3162 message are rejected with the same error (e.g. no authentication or bad
3163 sender address), and a DATA command is nevertheless sent (as can happen
3164 with PIPELINING or a stupid MUA), the error message that was given to the
3165 RCPT commands is included in the rejection of the DATA command. This is
3166 intended to be helpful for MUAs that show only the final error to their
3169 PH/42 Another patch from the Sieve maintainer.
3171 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3172 Eximstats - Fixed some broken HTML links and added missing column headers
3174 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3175 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3177 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3179 PH/43 Yet another patch from the Sieve maintainer.
3181 PH/44 I found a way to check for a TCP/IP connection going away before sending
3182 the response to the final '.' that terminates a message, but only in the
3183 case where the client has not sent further data following the '.'
3184 (unfortunately, this is allowed). However, in many cases there won't be
3185 any further data because there won't be any more messages to send. A call
3186 to select() can be used: if it shows that the input is "ready", there is
3187 either input waiting, or the socket has been closed. An attempt to read
3188 the next input character can distinguish the two cases. Previously, Exim
3189 would have sent an OK response which the client would never have see.
3190 This could lead to message repetition. This fix should cure that, at
3191 least in a lot of common cases.
3193 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3194 advertised in response to EHLO.
3200 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3201 fixed by 4.65/MH/01 (is this a record?) are fixed:
3203 (i) An empty string was always treated as zero by the numeric comparison
3204 operators. This behaviour has been restored.
3206 (ii) It is documented that the numeric comparison operators always treat
3207 their arguments as decimal numbers. This was broken in that numbers
3208 starting with 0 were being interpreted as octal.
3210 While fixing these problems I realized that there was another issue that
3211 hadn't been noticed. Values of message_size_limit (both the global option
3212 and the transport option) were treated as octal if they started with 0.
3213 The documentation was vague. These values are now always treated as
3214 decimal, and I will make that clear in the documentation.
3220 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3221 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3224 MH/01 Don't check that the operands of numeric comparison operators are
3225 integers when their expansion is in "skipping" mode (fixes bug
3226 introduced by 4.64-PH/07).
3228 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3229 child addresses, Exim now panics and dies. Previously, because the count
3230 is held in a short int, deliveries were likely to be lost. As such a
3231 large number of recipients for a single message is ridiculous
3232 (performance will be very, very poor), I have chosen to impose a limit
3233 rather than extend the field.
3239 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3240 leftover -K file (the existence of which was triggered by #402).
3241 While we were at it, introduced process PID as part of the -K
3242 filename. This should rule out race conditions when creating
3245 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3246 processing considerably. Previous code took too long for large mails,
3247 triggering a timeout which in turn triggers #401.
3249 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3250 in the DK code in transports.c. sendfile() is not really portable,
3251 hence the _LINUX specificness.
3253 TF/01 In the add_headers option to the mail command in an Exim filter,
3254 there was a bug that Exim would claim a syntax error in any
3255 header after the first one which had an odd number of characters
3258 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3259 callout verification, Exim cached a "reject" for the entire domain. This
3260 is correct for most verifications, but it is not correct for a recipient
3261 verification with use_sender or use_postmaster set, because in that case
3262 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3263 case of MAIL FROM:<> rejection from other early rejections (e.g.
3264 rejection of HELO). When verifying a recipient using a non-null MAIL
3265 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3266 Whatever the result of the callout, the value of the domain cache is
3267 left unchanged (for any other kind of callout, getting as far as trying
3268 RCPT means that the domain itself is ok).
3270 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3273 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3274 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3275 session, the daemon ignores ECONNECT errors and logs others; it now
3276 ignores EPIPE as well.
3278 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3279 (quoted-printable decoding).
3281 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3282 later the small subsequent patch to fix an introduced bug.
3284 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3286 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3288 PH/08 An error is now given if message_size_limit is specified negative.
3290 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3291 to be given (somewhat) arbitrary names.
3293 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3296 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3297 miscellaneous code fixes
3299 PH/10 Added the log_reject_target ACL modifier to specify where to log
3302 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3303 hostname. This is wrong, because it relates to the incoming message (and
3304 probably the interface on which it is arriving) and not to the outgoing
3305 callout (which could be using a different interface). This has been
3306 changed to use the value of the helo_data option from the smtp transport
3307 instead - this is what is used when a message is actually being sent. If
3308 there is no remote transport (possible with a router that sets up host
3309 addresses), $smtp_active_hostname is used.
3311 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3312 tweaks were necessary in order to get it to work (see also 21 below):
3313 (a) The code assumed that strncpy() returns a negative number on buffer
3314 overflow, which isn't the case. Replaced with Exim's string_format()
3316 (b) There were several signed/unsigned issues. I just did the minimum
3317 hacking in of casts. There is scope for a larger refactoring.
3318 (c) The code used strcasecmp() which is not a standard C function.
3319 Replaced with Exim's strcmpic() function.
3320 (d) The code set only $1; it now sets $auth1 as well.
3321 (e) A simple test gave the error "authentication client didn't specify
3322 service in request". It would seem that Dovecot has changed its
3323 interface. Fortunately there's a specification; I followed it and
3324 changed what the client sends and it appears to be working now.
3326 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3329 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3330 address A is aliased to B and C, where B exists and C does not. Without
3331 -v the output is "A verified" because verification stops after a
3332 successful redirection if more than one address is generated. However,
3333 with -v the child addresses are also verified. Exim was outputting "A
3334 failed to verify" and then showing the successful verification for C,
3335 with its parentage. It now outputs "B failed to verify", showing B's
3336 parentage before showing the successful verification of C.
3338 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3339 look up a TXT record in a specific list after matching in a combined
3342 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3343 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3344 they consult the DNS. I had assumed they would set it the way they
3345 wanted; and indeed my experiments on Linux seem to show that in some
3346 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3347 To be on the safe side, however, I have now made the interface to
3348 host_find_byname() similar to host_find_bydns(), with an argument
3349 containing the DNS resolver options. The host_find_byname() function now
3350 sets these options at its start, just as host_find_bydns() does. The smtp
3351 transport options dns_qualify_single and dns_search_parents are passed to
3352 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3353 of host_find_byname() use the default settings of RES_DEFNAMES
3354 (qualify_single) but not RES_DNSRCH (search_parents).
3356 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3357 spool_read_header() do less string testing, by means of a preliminary
3358 switch on the second character of optional "-foo" lines. (This is
3359 overdue, caused by the large number of possibilities that now exist.
3360 Originally there were few.) While I was there, I also converted the
3361 str(n)cmp tests so they don't re-test the leading "-" and the first
3362 character, in the hope this might squeeze out yet more improvement.
3364 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3365 flag allowing group syntax was set by the header_syntax check but not
3366 turned off, possible causing trouble later; (2) The flag was not being
3367 set at all for the header_verify test, causing "group"-style headers to
3368 be rejected. I have now set it in this case, and also caused header_
3369 verify to ignore an empty address taken from a group. While doing this, I
3370 came across some other cases where the code for allowing group syntax
3371 while scanning a header line wasn't quite right (mostly, not resetting
3372 the flag correctly in the right place). These bugs could have caused
3373 trouble for malformed header lines. I hope it is now all correct.
3375 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3376 with the "reply" argument non-NULL. The code, however (which originally
3377 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3378 but it didn't always do it. This confused somebody who was copying the
3379 code for some other use. I have removed all the tests.
3381 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3382 feature that was used to support insecure browsers during the U.S. crypto
3383 embargo. It requires special client support, and Exim is probably the
3384 only MTA that supported it -- and would never use it because real RSA is
3385 always available. This code has been removed, because it had the bad
3386 effect of slowing Exim down by computing (never used) parameters for the
3387 RSA_EXPORT functionality.
3389 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3390 authenticator to fail if there's a tab character in the incoming data
3391 (there should never be unless someone is messing about, as it's supposed
3392 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3393 if the connection is using TLS or if the remote IP is the same as the
3394 local IP, and the "valid-client-cert option" if a client certificate has
3397 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3398 authenticators. This can be used for authorization after authentication
3399 succeeds. (In the case of plaintext, it servers for both authentication
3402 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3403 if any retry times were supplied.
3405 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3406 connection, where there is no client IP address to check. In this
3407 situation, the verify now always succeeds.
3409 PH/25 Applied John Jetmore's -Mset patch.
3411 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3413 PH/27 In a string expansion for a processed (not raw) header when multiple
3414 headers of the same name were present, leading whitespace was being
3415 removed from all of them, but trailing whitespace was being removed only
3416 from the last one. Now trailing whitespace is removed from each header
3417 before concatenation. Completely empty headers in a concatenation (as
3418 before) are ignored.
3420 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3421 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3423 PH/29 [Removed. This was a change that I later backed out, and forgot to
3424 correct the ChangeLog entry (that I had efficiently created) before
3425 committing the later change.]
3427 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3428 address errors (4xx response to RCPT) over the same connection as other
3429 messages routed to the same hosts. Such deliveries are always "forced",
3430 so retry times are not inspected. This resulted in far too many retries
3431 for the affected addresses. The effect occurred only when there were more
3432 hosts than the hosts_max_try setting in the smtp transport when it had
3433 the 4xx errors. Those hosts that it had tried were not added to the list
3434 of hosts for which the message was waiting, so if all were tried, there
3435 was no problem. Two fixes have been applied:
3437 (i) If there are any address or message errors in an SMTP delivery, none
3438 of the hosts (tried or untried) are now added to the list of hosts
3439 for which the message is waiting, so the message should not be a
3440 candidate for sending over the same connection that was used for a
3441 successful delivery of some other message. This seems entirely
3442 reasonable: after all the message is NOT "waiting for some host".
3443 This is so "obvious" that I'm not sure why it wasn't done
3444 previously. Hope I haven't missed anything, but it can't do any
3445 harm, as the worst effect is to miss an optimization.
3447 (ii) If, despite (i), such a delivery is accidentally attempted, the
3448 routing retry time is respected, so at least it doesn't keep
3449 hammering the server.
3451 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3452 in ${readsocket because some servers need this prod.
3454 PH/32 Added some extra debug output when updating a wait-xxx database.
3456 PH/33 The hint "could be header name not terminated by colon", which has been
3457 given for certain expansion errors for a long time, was not being given
3458 for the ${if def:h_colon_omitted{... case.
3460 PH/34 The spec says: "With one important exception, whenever a domain list is
3461 being scanned, $domain contains the subject domain." There was at least
3462 one case where this was not true.
3464 PH/35 The error "getsockname() failed: connection reset by peer" was being
3465 written to the panic log as well as the main log, but it isn't really
3466 panic-worthy as it just means the connection died rather early on. I have
3467 removed the panic log writing for the ECONNRESET error when getsockname()
3470 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3471 runs only) independently of the message's sender address. This meant
3472 that, if the 4xx error was in fact related to the sender, a different
3473 message to the same recipient with a different sender could confuse
3474 things. In particular, this can happen when sending to a greylisting
3475 server, but other circumstances could also provoke similar problems.
3476 I have changed the default so that the retry time for these errors is now
3477 based a combination of the sender and recipient addresses. This change
3478 can be overridden by setting address_retry_include_sender=false in the
3481 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3482 remote server are returned as part of bounce messages. This was not
3483 happening for LMTP over a pipe (the lmtp transport), but now it is the
3484 same for both kinds of LMTP.
3486 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3487 in header lines that were in fact CNAMEs. This is no longer the case.
3489 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3490 and queue runs started by the daemon processed all messages. This has
3491 been fixed so that -R and -S can now usefully be given with -q<time>.
3493 PH/40 Import PCRE release 6.7 (fixes some bugs).
3495 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3497 PH/42 Give an error if -q is specified more than once.
3499 PH/43 Renamed the variables $interface_address and $interface_port as
3500 $received_ip_address and $received_port, to make it clear that these
3501 values apply to message reception, and not to the outgoing interface when
3502 a message is delivered. (The old names remain recognized, of course.)
3504 PH/44 There was no timeout on the connect() call when using a Unix domain
3505 socket in the ${readsocket expansion. There now is.
3507 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3508 be meaningful with "accept".
3510 SC/01 Eximstats V1.43
3511 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3513 SC/02 Eximstats V1.44
3514 Use a glob alias rather than an array ref in the generated
3515 parser. This improves both readability and performance.
3517 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3518 Collect SpamAssassin and rejection statistics.
3519 Don't display local sender or destination tables unless
3520 there is data to show.
3521 Added average volumes into the top table text output.
3523 SC/04 Eximstats V1.46
3524 Collect data on the number of addresses (recipients)
3525 as well as the number of messages.
3527 SC/05 Eximstats V1.47
3528 Added 'Message too big' to the list of mail rejection
3529 reasons (thanks to Marco Gaiarin).
3531 SC/06 Eximstats V1.48
3532 Mainlog lines which have GMT offsets and are too short to
3533 have a flag are now skipped.
3535 SC/07 Eximstats V1.49 (Alain Williams)
3536 Added the -emptyok flag.
3538 SC/08 Eximstats V1.50
3539 Fixes for obtaining the IP address from reject messages.
3541 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3542 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3543 whitespace changes from 4.64-PH/27
3545 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3548 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3549 are found, allow negative numbers in numeric criteria)
3551 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3553 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3554 to match changes made in 4.64-PH/43
3556 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3558 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3559 "domain" is actually an IP address, and force a failure. This locks out
3560 those revolvers/nameservers that support "A-for-A" lookups, in
3561 contravention of the specifications.
3563 PH/48 When a host name was looked up from an IP address, and the subsequent
3564 forward lookup of the name timed out, the host name was left in
3565 $sender_host_name, contrary to the specification.
3567 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3568 restricted to single-key lookups, Exim was not diagnosing an error if
3569 * or *@ was used with a query-style lookup.
3571 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3573 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3574 long ago, but noone interested enough thought of it. Let's just say that
3575 the "1.1" means that there are some new functions that weren't there at
3576 some point in the past.
3578 PH/51 Error processing for expansion failure of helo_data from an smtp
3579 transport during callout processing was broken.
3581 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3582 tested/used via the -bh/-bhc/-bs options.
3584 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3585 bug, fixed in subsequent PCRE releases).
3587 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3588 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3590 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3596 SC/01 Use a glob alias rather than an array ref in eximstats generated
3597 parser. This improves both readability and performance.
3599 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3600 Don't display local sender or destination tables in eximstats unless
3601 there is data to show.
3602 Added average volumes into the eximstats top table text output.
3604 SC/03 Collect data on the number of addresses (recipients) as well
3605 as the number of messages in eximstats.
3607 TF/01 Correct an error in the documentation for the redirect router. Exim
3608 does (usually) call initgroups() when daemonizing.
3610 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3611 with consistent privilege compared to when running as a daemon.
3613 TF/03 Note in the spec that $authenticated_id is not set for local
3614 submissions from trusted users.
3616 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3617 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3619 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3620 by adding some example configuration directives to the default
3621 configuration file. A little bit of work is required to uncomment the
3622 directives and define how usernames and passwords are checked, but
3623 there is now a framework to start from.
3625 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3626 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3627 without this. I don't know how relevant this is to other LDAP libraries.
3629 PH/02 Add the verb name to the "unknown ACL verb" error.
3631 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3633 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3635 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3636 directory not expanded when it should be if an expanded home directory
3637 was set for the address (which is overridden by the transport).
3639 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3642 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3643 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3644 because it is too late at that time, and has no effect.
3646 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3647 security issue with \' (bugzilla #107). I could not use the
3648 PQescapeStringConn() function, because it needs a PGconn value as one of
3651 PH/08 When testing addresses using -bt, indicate those final addresses that
3652 are duplicates that would not cause an additional delivery. At least one
3653 person was confused, thinking that -bt output corresponded to deliveries.
3654 (Suppressing duplicates isn't a good idea as you lose the information
3655 about possibly different redirections that led to the duplicates.)
3657 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3658 systems where poll() doesn't work, in particular OS X.
3660 PH/10 Added more information to debugging output for retry time not reached.
3662 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3663 operations in malware.c.
3665 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3668 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3669 syslog_duplication set false, log lines that would normally be written
3670 both the the main log and to the reject log were not written to syslog at
3673 PH/14 In the default configuration, change the use of "message" in ACL warn
3674 statements to "add_header".
3676 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3677 not followed by a command (e.g. "seen endif").
3679 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3680 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3683 PH/17 Added extra conditions to the default value of delay_warning_condition
3687 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3688 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3689 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3692 The Auto-Submitted: and various List- headers are standardised, whereas I
3693 don't think Precedence: ever was.
3695 PH/18 Refactored debugging code in route_finduser() to show more information,
3696 in particular, the error code if getpwnam() issues one.
3698 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3699 This is apparently needed in addition to the PH/07 change above to avoid
3700 any possible encoding problems.
3702 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3703 but not after initializing Perl.
3705 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3706 output them only if debugging. By default they are written stderr,
3707 apparently, which is not desirable.
3709 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3712 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3715 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3717 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3718 authenticated or an ident call has been made. Suppress the default
3719 values for $authenticated_id and $authenticated_sender (but permit -oMai
3720 and -oMas) when testing with -bh.
3722 PH/24 Re-jigged the order of the tests in the default configuration so that the
3723 tests for valid domains and recipients precede the DNS black list and CSA
3724 tests, on the grounds that those ones are more expensive.
3726 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3727 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3728 This bug exists in every version of Exim that I still have, right back to
3731 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3732 However, an attempt to turn on case-sensitivity in a regex key by
3733 including (?-i) didn't work because the subject string was already
3734 lowercased, and the effects were non-intuitive. It turns out that a
3735 one-line patch can be used to allow (?-i) to work as expected.
3741 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3742 other effects) broke the use of negated acl sub-conditions.
3744 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3747 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3748 "Deny" causes Exim to reject the incoming connection with a 554 error.
3749 Unfortunately, if there is a major crisis, such as a disk failure,
3750 tcp-wrappers gives "deny", whereas what one would like would be some
3751 kind of temporary error. A kludge has been added to help with this.
3752 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3753 554 error is used if errno is still zero or contains ENOENT (which occurs
3754 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3757 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3759 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3760 errors. Otherwise a message that provokes a temporary error (when other
3761 messages do not) can cause a whole host to time out.
3763 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3764 addresses were routed directly to files or pipes from a redirect router.
3765 File deliveries just didn't batch; pipe deliveries might have suffered
3768 PH/06 A failure to get a lock for a hints database would erroneously always say
3769 "Failed to get write lock", even when it was really a read lock.
3771 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3772 of 0600. This has been changed to use the value of the lockfile_mode
3773 option (which defaults to 0600).
3775 PH/08 Applied small patch from the Sieve maintainer.
3777 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3778 folder from quota calculations, a direct delivery into this folder messed
3779 up the contents of the maildirsize file. This was because the regex was
3780 used only to exclude .Trash (or whatever) when the size of the mailbox
3781 was calculated. There was no check that a delivery was happening into an
3782 excluded directory. This bug has been fixed by ignoring all quota
3783 processing for deliveries into excluded directories.
3785 PH/10 Added the maildirfolder_create_regex option to appendfile.
3791 PH/01 The code for finding all the local interface addresses on a FreeBSD
3792 system running IPv6 was broken. This may well have applied to all BSD
3793 systems, as well as to others that have similar system calls. The broken
3794 code found IPv4 interfaces correctly, but gave incorrect values for the
3795 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3796 that it would not match correctly against @[] and not recognize the IPv6
3799 PH/02 The ipliteral router was not recognizing addresses of the form user@
3800 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3802 PH/03 Added disable_ipv6.
3804 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3805 decoded form, because it is most often used to construct To: headers
3806 lines in autoreplies, and the decoded form may well be syntactically
3807 invalid. However, $reply_address has leading white space removed, and all
3808 newlines turned into spaces so that the autoreply transport does not
3811 PH/05 If group was specified without a user on a router, and no group or user
3812 was specified on a transport, the group from the router was ignored.
3814 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3815 for visible compile-time settings that can be used to change these
3816 numbers, for those that want even more. Backwards compatibility with old
3817 spool files has been maintained. However, going back to a previous Exim
3818 release will lost any variables that are in spool files.
3820 PH/07 Two small changes when running in the test harness: increase delay when
3821 passing a TCP/IP connection to a new process, in case the original
3822 process has to generate a bounce, and remove special handling of
3823 127.0.0.2 (sic), which is no longer necessary.
3825 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3826 be the same on different OS.
3828 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3831 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3832 whether --show-vars was specified or not
3834 JJ/02 exipick: Added support for new ACL variable spool format introduced
3837 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3838 syntactically invalid From: or Reply-to: line, and a filter used this to
3839 generate an autoreply, and therefore failed to obtain an address for the
3840 autoreply, Exim could try to deliver to a non-existent relative file
3841 name, causing unrelated and misleading errors. What now happens is that
3842 it logs this as a hard delivery error, but does not attempt to create a
3845 PH/11 The exinext utility has a -C option for testing purposes, but although
3846 the given file was scanned by exinext itself; it wasn't being passed on
3847 when Exim was called.
3849 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3850 an end-of-file indication when reading a command response.
3852 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3853 compiled. In many other places in Exim, IPv6 addresses are always
3854 recognized, so I have changed this. It also means that IPv4 domain
3855 literals of the form [IPV4:n.n.n.n] are now always recognized.
3857 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3858 used if the router is not running as root, for example, when verifying at
3859 ACL time, or when using -bh. The debugging output from this situation was
3860 non-existent - all you got was a failure to exec. I have made two
3863 (a) Failures to set uid/gid, the current directory, or a process leader
3864 in a subprocess such as that created by queryprogram now generate
3865 suitable debugging output when -d is set.
3867 (b) The queryprogram router detects when it is not running as root,
3868 outputs suitable debugging information if -d is set, and then runs
3869 the subprocess without attempting to change uid/gid.
3871 PH/15 Minor change to Makefile for building test_host (undocumented testing
3874 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3875 additional section of a DNS packet that returns MX or SRV records.
3876 Instead, it always explicitly searches for A/AAAA records. This avoids
3877 major problems that occur when a DNS server includes only records of one
3878 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3879 fixed another bug: if SRV records were looked up and the corresponding
3880 address records were *not* found in the additional section, the port
3881 values from the SRV records were lost.
3883 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3884 using the correct key (the original address) when searching the retry
3885 rules in order to find which one to use for generating the retry hint.
3887 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3888 adding the default one. Similarly, if it contains a Reply-To: header, the
3889 errors_reply_to option, if set, is not used.
3891 PH/19 When calculating a retry time, Exim used to measure the "time since
3892 failure" by looking at the "first failed" field in the retry record. Now
3893 it does not use this if it is later than than the arrival time of the
3894 message. Instead it uses the arrival time. This makes for better
3895 behaviour in cases where some deliveries succeed, thus re-setting the
3896 "first failed" field. An example is a quota failure for a huge message
3897 when small messages continue to be delivered. Without this change, the
3898 "time since failure" will always be short, possible causing more frequent
3899 delivery attempts for the huge message than are intended.
3900 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3902 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3903 $1, $2, $3) because the numerical variables can be reset during some
3904 expansion items (e.g. "match"), thereby losing the authentication data.
3906 PH/21 Make -bV show the size of off_t variables so that the test suite can
3907 decide whether to run tests for quotas > 2G.
3909 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3910 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3911 filecount value is greater than 2G or if a quota value is greater than 2G
3912 on a system where the size of off_t is not greater than 4, a panic error
3915 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3916 never match. The debug and -bh output now contains an explicit error
3917 message indicating a malformed IPv4 address or mask.
3919 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3920 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3921 PH/23 above applies.
3923 PH/25 Do not write to syslog when running in the test harness. The only
3924 occasion when this arises is a failure to open the main or panic logs
3925 (for which there is an explicit test).
3927 PH/26 Added the /no_tell option to "control=freeze".
3929 PH/27 If a host name lookup failed very early in a connection, for example, if
3930 the IP address matched host_lookup and the reverse lookup yielded a name
3931 that did not have a forward lookup, an error message of the form "no IP
3932 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3933 could be logged. Now it outputs the IP address instead of "NULL".
3935 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3936 allows the sender and the authenticated sender to be set when
3937 submitting a message from within Exim. Since child_open_exim() is
3938 documented for local_scan(), the new function should be too.
3940 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3941 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3942 results in an empty string is now treated as unset.
3944 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3946 PH/31 Added sender_verify_fail logging option.
3948 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3949 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3950 tidied the source and removed it altogether.
3952 PH/33 When a queue run was abandoned because the load average was too high, a
3953 log line was always written; now it is written only if the queue_run log
3954 selector is set. In addition, the log line for abandonment now contains
3955 information about the queue run such as the pid. This is always present
3956 in "start" and "stop" lines but was omitted from the "abandon" line.
3958 PH/34 Omit spaces between a header name and the colon in the error message that
3959 is given when verify = headers_syntax fails (if there are lots of them,
3960 the message gets confusing).
3962 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3963 names, as there are now some PTR records that contain slashes. This check
3964 is only to protect against broken name servers that fall over on strange
3965 characters, so the fact that it applies to all lookups doesn't matter.
3967 PH/36 Now that the new test suite is complete, we can remove some of the
3968 special code in Exim that was needed for the old test suite. For example,
3969 sorting DNS records because real resolvers return them in an arbitrary
3970 order. The new test suite's fake resolver always returns records in the
3973 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3974 bounces) except when queue_only is set, to avoid logging races between
3975 the different processes.
3977 PH/38 Panic-die if .include specifies a non-absolute path.
3979 PH/39 A tweak to the "H" retry rule from its user.
3981 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3982 a label. They prevented compilation on older perls.
3984 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3985 a warning to be raised on newish perls.
3987 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3988 on queue. Changes to match documented behaviour of showing count of
3989 messages matching specified criteria.
3991 PH/40 Changed the default ident timeout from 30s to 5s.
3993 PH/41 Added support for the use of login_cap features, on those BSD systems
3994 that have them, for controlling the resources used by pipe deliveries.
3996 PH/42 The content-scanning code uses fopen() to create files in which to put
3997 message data. Previously it was not paying any attention to the mode of
3998 the files. Exim runs with umask(0) because the rest of the code creates
3999 files with open(), and sets the required mode explicitly. Thus, these
4000 files were ending up world-writeable. This was not a big issue, because,
4001 being within the spool directory, they were not world-accessible. I have
4002 created a function called modefopen, which takes an additional mode
4003 argument. It sets umask(777), creates the file, chmods it to the required
4004 mode, then resets the umask. All the relevant calls to fopen() in the
4005 content scanning code have been changed to use this function.
4007 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4008 to 24 hours. This avoids potential overflow problems when processing G
4009 and H retry rules. I suspect nobody ever tinkers with this value.
4011 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4013 PH/45 When the plaintext authenticator is running as a client, the server's
4014 challenges are checked to ensure they are valid base64 strings. By
4015 default, the authentication attempt is cancelled if an invalid string is
4016 received. Setting client_ignore_invalid_base64 true ignores these errors.
4017 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4018 they are received. Thus, the responses can be made to depend on the
4019 challenges. If an invalid string is ignored, an empty string is placed in
4022 PH/46 Messages that are created by the autoreply transport now contains a
4023 References: header, in accordance with RFCs 2822 and 3834.
4025 PH/47 Added authenticated_sender_force to the smtp transport.
4027 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4029 PH/49 Installed latest patch from the Sieve maintainer.
4031 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4032 was also set, the appendfile transport was unnecessarily scanning a
4033 directory of message files (e.g. for maildir delivery) to find the count
4034 of files (along with the size), even though it did not need this
4035 information. It now does the scan only if it needs to find either the
4036 size of the count of files.
4038 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4040 PH/52 Two bugs concerned with error handling when the smtp transport is
4043 (i) Exim was not creating retry information for temporary errors given
4044 for individual recipients after the DATA command when the smtp transport
4045 was used in LMTP mode. This meant that they could be retried too
4046 frequently, and not timed out correctly.
4048 (ii) Exim was setting the flag that allows error details to be returned
4049 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4050 recipients that were returned after the DATA command.
4052 PH/53 This is related to PH/52, but is more general: for any failing address,
4053 when detailed error information was permitted to be returned to the
4054 sender, but the error was temporary, then after the final timeout, only
4055 "retry timeout exceeded" was returned. Now it returns the full error as
4056 well as "retry timeout exceeded".
4058 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4059 do this, and (what is worse) MTAs that accept it.
4061 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4062 will now be deprecated.
4064 PH/56 New os.c-cygwin from the Cygwin maintainer.
4066 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4067 formats (previously only available in exim formats via -bpr, -bpru,
4068 and -bpra. Now also available in native and exiqgrep formats)
4070 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4071 with very large, slow to parse queues
4073 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4075 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4077 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4078 responses to authentication challenges, though it was showing the
4079 challenges; (ii) I've removed the CR characters from the debug output for
4082 PH/58 Allow for the insertion of a newline as well as a space when a string
4083 is turned into more than one encoded-word during RFC 2047 encoding. The
4084 Sieve code now uses this.
4086 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4087 data_4xx, lost_connection, tls_required.
4089 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4090 message was being sent as an SMTP response.
4092 PH/61 Add -l and -k options to exicyclog.
4094 PH/62 When verifying, if an address was redirected to one new address, so that
4095 verification continued, and the new address failed or deferred after
4096 having set something in $address_data, the value of $address_data was not
4097 passed back to the ACL. This was different to the case when no
4098 redirection occurred. The value is now passed back in both cases.
4100 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4101 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4102 use login_cap.h, so on its own it isn't the distinguishing feature. The
4103 new name refers directly to the setclassresources() function.
4105 PH/65 Added configuration files for NetBSD3.
4107 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4109 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4110 is preferred over IPv4.
4112 PH/68 The bounce_return_message and bounce_return_body options were not being
4113 honoured for bounces generated during the reception of non-SMTP messages.
4114 In particular, this applied to messages rejected by the ACL. This bug has
4115 been fixed. However, if bounce_return_message is true and bounce_return_
4116 body is false, the headers that are returned for a non-SMTP message
4117 include only those that have been read before the error was detected.
4118 (In the case of an ACL rejection, they have all been read.)
4120 PH/69 The HTML version of the specification is now built in a directory called
4121 spec_html instead of spec.html, because the latter looks like a path with
4122 a MIME-type, and this confuses some software.
4124 PH/70 Catch two compiler warnings in sieve.c.
4126 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4127 function verify_get_ident() calls ip_connect() to connect a socket, but
4128 if the "connect()" function timed out, ip_connect() used to close the
4129 socket. However, verify_get_ident() also closes the socket later, and in
4130 between Exim writes to the log, which may get opened at this point. When
4131 the socket was closed in ip_connect(), the log could get the same file
4132 descriptor number as the socket. This naturally causes chaos. The fix is
4133 not to close the socket in ip_connect(); the socket should be closed by
4134 the function that creates it. There was only one place in the code where
4135 this was missing, in the iplookup router, which I don't think anybody now
4136 uses, but I've fixed it anyway.
4138 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4139 well as to direct DNS lookups. Otherwise the handling of names in host
4140 lists is inconsistent and therefore confusing.
4146 PH/01 Two changes to the default runtime configuration:
4148 (1) Move the checks for relay_from_hosts and authenticated clients from
4149 after to before the (commented out) DNS black list checks.
4151 (2) Add control=submission to the relay_from_hosts and authenticated
4152 clients checks, on the grounds that messages accepted by these
4153 statements are most likely to be submissions.
4155 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4157 (1) Generate an error if the third argument for the ${prvs expansion is
4160 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4163 (3) Reset the variables that are obtained from the first argument of
4164 ${prvscheck and used in the second argument before leaving the code,
4165 because their memory is reclaimed, so using them afterwards may do
4168 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4169 one (it's much easier than Tom thought :-).
4171 (5) Because of (4), we can now allow for the use of $prvscheck_result
4172 inside the third argument.
4174 PH/03 For some reason, the default setting of PATH when running a command from
4175 a pipe transport was just "/usr/bin". I have changed it to
4178 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4179 anything to be listed in the output from -bV.
4181 PH/05 When a filter generated an autoreply, the entire To: header line was
4182 quoted in the delivery log line, like this:
4184 => >A.N.Other <ano@some.domain> <original@ddress> ...
4186 This has been changed so that it extracts the operative address. There
4187 may be more than one such address. If so, they are comma-separated, like
4190 => >ano@some.domain,ona@other.domain <original@ddress> ...
4192 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4193 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4194 not being looked up in the rDNS to get a host name, Exim was showing the
4195 IP address twice in Received: lines, even though the IP addresses were
4196 identical. For example:
4198 Received: from [1.2.3.4] (helo=[1.2.3.4])
4200 However, if the real host name was known, it was omitting the HELO data
4201 if it matched the actual IP address. This has been tidied up so that it
4202 doesn't show the same IP address twice.
4204 PH/07 When both +timestamp and +memory debugging was on, the value given by
4205 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4206 called by the debug printing, thereby overwriting the timestamp buffer.
4207 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4209 PH/08 When the original message was included in an autoreply transport, it
4210 always said "this is a copy of the message, including all the headers",
4211 even if body_only or headers_only was set. It now gives an appropriate
4214 PH/09 Applied a patch from the Sieve maintainer which:
4216 o fixes some comments
4217 o adds the (disabled) notify extension core
4218 o adds some debug output for the result of if/elsif tests
4219 o points to the current vacation draft in the documentation
4220 and documents the missing references header update
4224 o fixes a bug in processing the envelope test (when testing
4225 multiple envelope elements, the last element determined the
4228 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4229 Electronic Mail") by including:
4231 Auto-submitted: auto-generated
4233 in the messages that it generates (bounce messages and others, such as
4234 warnings). In the case of bounce messages for non-SMTP messages, there was
4235 also a typo: it was using "Auto_submitted" (underscore instead of
4236 hyphen). Since every message generated by Exim is necessarily in response
4237 to another message, thes have all been changed to:
4239 Auto-Submitted: auto-replied
4241 in accordance with these statements in the RFC:
4243 The auto-replied keyword:
4245 - SHOULD be used on messages sent in direct response to another
4246 message by an automatic process,
4248 - MUST NOT be used on manually-generated messages,
4250 - MAY be used on Delivery Status Notifications (DSNs) and Message
4251 Disposition Notifications (MDNs),
4253 - MUST NOT be used on messages generated by automatic or periodic
4254 processes, except for messages which are automatic responses to
4257 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4258 to the default Received: header definition.
4260 PH/12 Added log selector acl_warn_skipped (default on).
4262 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4263 variables because (a) they are in the wrong storage pool and (b) even if
4264 they were copied, it wouldn't work properly because of the caching.
4266 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4267 checking when decoding. Apparently there are clients that generate
4268 overlong encoded strings. Why am I not surprised?
4270 PH/15 If the first argument of "${if match_address" was not empty, but did not
4271 contain an "@" character, Exim crashed. Now it writes a panic log message
4272 and treats the condition as false.
4274 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4276 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4277 extension "envelope-auth". The code is finished and in agreement with
4278 other implementations, but there is no documentation so far and in fact,
4279 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4280 not changing the active code.
4282 Print executed "if" and "elsif" statements when debugging is used. This
4283 helps a great deal to understand what a filter does.
4285 Document more things not specified clearly in RFC3028. I had all this
4286 sorted out, when out of a sudden new issues came to my mind. Oops."
4288 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4291 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4292 canonical form (as documented). However, after a host name lookup from
4293 the IP address, check_host() was doing a simple string comparison with
4294 addresses acquired from the DNS when checking that the found name did
4295 have the original IP as one of its addresses. Since any found IPv6
4296 addresses are likely to be in abbreviated form, the comparison could
4297 fail. Luckily, there already exists a function for doing the comparison
4298 by converting both addresses to binary, so now that is used instead of
4299 the text comparison.
4301 PH/20 There was another similar case to PH/19, when a complete host name was
4302 given in a host list; looking up its IP address could give an abbreviated
4303 form, whereas the current host's name might or might not be abbreviated.
4304 The same fix has been applied.
4310 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4311 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4314 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4315 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4317 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4319 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4320 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4321 Also, for "Auto-Submitted", treat anything other than "no" as
4322 non-personal, in accordance with RFC 3834. (Previously it treated
4323 anything starting "auto-" as non-personal.)
4325 TF/01 The control=submission/name=... option had a problem with syntax
4326 errors if the name included a slash character. The /name= option
4327 now slurps the rest of the string, so it can include any characters
4328 but it must come last in the list of options (after /sender_retain
4331 PH/05 Some modifications to the interface to the fake nameserver for the new
4339 TK/01 Added the "success_on_redirect" address verification option. See
4340 NewStuff for rationale and an example.
4342 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4344 PH/02 Patch to exigrep to allow it to work on syslog lines.
4346 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4347 fread() to skip over the body file's header line, because in Cygwin the
4348 header line is locked and is inaccessible.
4350 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4351 co-exist for some time) to make it clear that it is the Exim ID that is
4352 referenced, not the Message-ID: header line.
4354 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4355 string_format() function, because snprintf() does not exist on all
4358 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4359 ${stat: expansion item.
4361 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4362 protocol synchronization error", to keep the pedants happy.
4364 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4365 well as for IRIX systems, when gcc is being used. See the host.c source
4368 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4370 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4373 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4374 transport and to the smtp transport in LMTP mode.
4376 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4378 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4379 run a filter in a subprocess. This could lead to confusion in subsequent
4380 lookups in the parent process. There should also be a search_tidyup() at
4381 the end of the subprocess.
4383 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4384 only if the host matched helo_try_verify_hosts, which caused the
4385 verification to occur when the EHLO/HELO command was issued. The ACL just
4386 tested the remembered result. Now, if a previous verification attempt has
4387 not happened, "verify = helo" does it there and then.
4389 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4391 TK/03 Fix log output including CR from clamd.
4393 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4394 exist provoked a memory error which could cause a segfault.
4396 PH/15 Installed PCRE 6.2
4398 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4400 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4401 of the problem. Specifically, suggested +O2 rather than +O1 for the
4404 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4406 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4407 record, the port was not correctly logged when the outgoing_port log
4408 selector was set (it logged the transort's default port).
4410 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4411 fallback_hosts, and "hosts" in the smtp transport.
4413 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4414 host errors such as "Connection refused".
4416 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4417 authentication with radiusclient 0.4.9:
4419 - Error returned from rc_read_config was caught wrongly
4420 - Username/password not passed on to radius server due to wrong length.
4422 The presumption is that some radiusclient API changes for 4.51/PH/17
4423 were not taken care of correctly. The code is still untested by me (my
4424 Linux distribution still has 0.3.2 of radiusclient), but it was
4425 contributed by a Radius user.
4427 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4428 expanding the "port" option of the smtp transport.
4430 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4431 while reading a MIME header. Thanks to Tom Hughes for a patch.
4433 PH/24 Include config.h inside local_scan.h so that configuration settings are
4436 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4437 that in an ACL for RCPT (for example), you can examine exactly what was
4440 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4441 commands, but it was not correctly comparing the address with the actual
4442 client host address. Thus, it would show the EHLO address in Received:
4443 header lines when this was not necessary.
4445 PH/27 Added the % operator to ${eval:}.
4447 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4448 it should be ignoring failures (because with -C, for example, it has lost
4449 privilege). It wasn't ignoring creation failures other than "already
4452 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4455 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4456 because an input line was too long, either on its own, or by virtue of
4457 too many continuations, the temporary file was not being removed, and the
4458 return code was incorrect.
4460 PH/31 Missing "BOOL" in function definition in filtertest.c.
4462 PH/32 Applied Sieve patches from the maintainer.
4464 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4466 PH/33 Added "verify = not_blind".
4468 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4469 Local/Makefile (with some defaults set). These are used in built scripts
4470 such as exicyclog, but they have never been used in the exim_install
4471 script (though there are many overriding facilities there). I have
4472 arranged that the exim_install script now takes note of these two
4475 PH/35 Installed configuration files for Dragonfly.
4477 PH/36 When a locally submitted message by a trusted user did not contain a
4478 From: header, and the sender address was obtained from -f or from an SMTP
4479 MAIL command, and the trusted user did not use -F to supply a sender
4480 name, $originator_name was incorrectly used when constructing a From:
4481 header. Furthermore, $originator_name was used for submission mode
4482 messages from external hosts without From: headers in a similar way,
4483 which is clearly wrong.
4485 PH/37 Added control=suppress_local_fixups.
4487 PH/38 When log_selector = +received_sender was set, and the addition of the
4488 sender made the log line's construction buffer exactly full, or one byte
4489 less than full, an overflow happened when the terminating "\n" was
4492 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4493 when the result of a list match is failure because a DNS lookup failed.
4495 PH/40 RM_COMMAND is now used in the building process.
4497 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4498 the "build-* directories that it finds.
4500 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4501 address in a domain literal was a prefix of an interface address.
4503 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4504 when verifying a sender address, unless rewrite_headers is false.
4506 PH/44 Wrote a long comment about why errors_to addresses are verified as
4507 recipients, not senders.
4509 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4510 the ratelimit ACL was added.
4512 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4514 PH/46 Added extra information about PostgreSQL errors to the error string.
4516 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4517 avoiding the need to install special zones in a real server. This is
4518 backwards compatible; if it can't find the fake resolver, it drops back.
4519 Thus, both old and new test suites can be run.
4521 TF/02 Added util/ratelimit.pl
4523 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4524 clock is set back in time.
4526 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4527 Candler <B.Candler@pobox.com>.
4529 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4530 OK for addresses that are the result of redirections.
4532 PH/48 A number of further additions for the benefit of the new test suite,
4533 including a fake gethostbyname() that interfaces to the fake DNS resolver
4536 TF/06 The fix for widen_domains has also been applied to qualify_single and
4537 search_parents which are the other dnslookup options that can cause
4540 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4543 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4545 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4546 service" if it encountered an MX record with an empty target hostname.
4547 The message is now "an MX or SRV record indicated no SMTP service".
4549 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4550 if the DNS of the sending site is misconfigured. This is quite a
4551 common situation. This change restores the behaviour of treating a
4552 helo verification defer as a failure.
4554 PH/51 If self=fail was set on a router, the bounce message did not include the
4555 actual error message.
4561 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4563 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4564 command itself ended in error, the underlying message about the transport
4565 filter timeout was being overwritten with the pipe command error. Now the
4566 underlying error message should be appended to the second error message.
4568 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4570 PH/02 Reduce the amount of output that "make" produces by default. Full output
4571 can still be requested.
4573 PH/03 The warning log line about a condition test deferring for a "warn" verb
4574 was being output only once per connection, rather than after each
4575 occurrence (because it was using the same function as for successful
4576 "warn" verbs). This seems wrong, so I have changed it.
4578 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4579 it should not have, which might have caused a crash in the right
4580 circumstances, but probably never did.
4582 PH/04 Installed a modified version of Tony Finch's patch to make submission
4583 mode fix the return path as well as the Sender: header line, and to
4584 add a /name= option so that you can make the user's friendly name appear
4587 TF/03 Added the control = fakedefer ACL modifier.
4589 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4590 Mark Lowes for thorough testing.
4592 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4594 TK/03 Merged latest SRS patch from Miles Wilton.
4596 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4597 with the definition in sysexits.h (which is #included earlier).
4598 Fortunately, Exim does not actually use EX_OK. The code used to try to
4599 preserve the sysexits.h value, by assuming that macro definitions were
4600 scanned for macro replacements. I have been disabused of this notion,
4601 so now the code just undefines EX_OK before #including unistd.h.
4603 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4604 in the smtp transport. When a block could not be written in a single
4605 write() function, the timeout was being re-applied to each part-write.
4606 This seems wrong - if the receiver was accepting one byte at a time it
4607 would take for ever. The timeout is now adjusted when this happens. It
4608 doesn't have to be particularly precise.
4610 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4611 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4613 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4614 without a domain if the check to <postmaster@domain> fails.
4616 SC/01 Eximstats: added -xls and the ability to specify output files
4617 (patch written by Frank Heydlauf).
4619 SC/02 Eximstats: use FileHandles for outputting results.
4621 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4623 SC/04 Eximstats: fixed display of large numbers with -nvr option
4625 SC/05 Eximstats: fixed merging of reports with empty tables.
4627 SC/06 Eximstats: added the -include_original_destination flag
4629 SC/07 Eximstats: removed tabs and trailing whitespace.
4631 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4633 TK/06 MBOX spool code: Add real "From " MBOX separator line
4634 so the .eml file is really in mbox format (even though
4635 most programs do not really care). Patch from Alex Miller.
4637 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4638 The latter is generated from $received_to and is only set if the
4639 message has one envelope recipient. SA can use these headers,
4640 obviously out-of-the-box. Patch from Alex Miller.
4642 PH/08 The ${def test on a variable was returning false if the variable's
4643 value was "0", contrary to what the specification has always said!
4644 The result should be true unless the variable is empty.
4646 PH/09 The syntax error of a character other than { following "${if
4647 def:variable_name" (after optional whitespace) was not being diagnosed.
4648 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4649 accidental colon was present, for example, could give incorrect results.
4651 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4652 result is used (not including appendfile, where other changes are about
4655 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4656 This involved changing a lot of size variables from int to off_t. It
4657 should work with maildirs and everything.
4659 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4660 spamd dying while we are connected to it.
4662 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4665 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4666 function for BDB 4.3.
4668 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4670 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4671 the sources for PCRE have been re-arranged and more files are now
4674 PH/15 The code I had for printing potentially long long variables in PH/11
4675 above was not the best (it lost precision). The length of off_t variables
4676 is now inspected at build time, and an appropriate printing format (%ld
4677 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4678 to be "long long int" or "long int". This is needed for the internal
4679 formatting function string_vformat().
4681 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4682 the configuration file to be ":syslog", then the script "guesses" where
4683 the logs files are, rather than using the compiled in default. In our
4684 case the guess is not the same as the compiled default, so the script
4685 suddenly stopped working when I started to use syslog. The patch checks
4686 to see if log_file_path is "". If so, it attempts to read it from exim
4687 with no configuration file to get the compiled in version, before it
4688 falls back to the previous guessing code."
4690 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4691 implementing BATV in an Exim configuration. See NewStuff for the gory
4694 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4695 Makefile that are specific to HP-UX.
4697 PH/18 If the "use_postmaster" option was set for a recipient callout together
4698 with the "random" option, the postmaster address was used as the MAIL
4699 FROM address for the random test, but not for the subsequent recipient
4700 test. It is now used for both.
4702 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4703 patch removes a few documentation additions to RFC 3028, because the
4704 latest draft now contains them. It adds the new en;ascii-case comparator
4705 and a new error check for 8bit text in MIME parts. Comparator and
4706 require names are now matched exactly. I enabled the subaddress
4707 extension, but it is not well tested yet (read: it works for me)."
4709 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4710 rework some of the code of TK/09 above to avoid the hardwired use of
4711 "%lld" and "long long". Replaced the call to snprintf() with a call to
4714 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4715 records point to non-existent hosts", "retry timeout exceeded", and
4716 "retry time not reached for any host after a long failure period".
4718 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4719 experimental DomainKeys support:
4721 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4722 (2) On an error such as an illegally used "control", the wrong name for
4723 the control was given.
4725 These problems did NOT occur unless DomainKeys support was compiled.
4727 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4729 PH/24 Added ${if match_ip condition.
4731 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4732 databases so that it will be absolutely obvious if a crash occurs in the
4733 DB library. This is a regular occurrence (often caused by mis-matched
4736 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4737 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4738 picked up on a user's system that detects such things. There doesn't seem
4739 to be a gcc warning option for this - only an attribute that has to be
4740 put on the function's prototype. It seems that in Fedora Core 4 they have
4741 set this on a number of new functions. No doubt there will be more in due
4744 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4745 specify a transport. However, if an address that was verified by such a
4746 router was the subject of a callout, Exim crashed because it tried to
4747 read the rcpt_include_affixes from the non-existent transport. Now it
4748 just assumes that the setting of that option is false. This bug was
4749 introduced by 4.51/PH/31.
4751 PH/28 Changed -d+all to exclude +memory, because that information is very
4752 rarely of interest, but it makes the output a lot bigger. People tend to
4753 do -d+all out of habit.
4755 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4756 code in os-type was giving problems when libc.so lives in lib64, like on
4759 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4760 aren't the modern standard, and it seems that some systems' include files
4761 don't always have them. Exim was already checking for some of the newer
4762 ones like T_AAAA, and defining it itself. I've added checks for all the
4763 record types that Exim uses.
4765 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4766 not automatically generating a new one, as it is supposed to. This
4767 prevented TLS from working. If the file did exist, but contained invalid
4768 data, a new version was generated, as expected. It was only the case of a
4769 non-existent file that was broken.
4771 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4772 with a change in libdomainkeys > 0.64.
4774 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4775 from DNS. If the selector record carries the flag, it now has
4776 precedence over the domain-wide flag.
4778 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4780 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4781 the use of a port name that isn't defined in /etc/services) occurred, the
4782 message was deferred as in a normal delivery, and thus remained on the
4783 spool, instead of being failed because of the mua_wrapper setting. This
4784 is now fixed, and I tidied up some of the mua_wrapper messages at the
4787 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4788 the messages in a hash of arrays rather than using individual hashes.
4789 This is a bit cleaner and results in dramatic memory savings, albeit
4790 at a slight CPU cost.
4792 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4793 as requested by Marc Sherman.
4795 SC/10 Eximstats: added histograms for user specified patterns as requested
4798 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4800 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4801 fopen() in the content-scanning modules that did not already have it.
4807 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4808 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4810 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4812 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4814 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4815 to test Sieve filters that use "vacation".
4817 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4818 that changes the way the GnuTLS parameters are stored in the cache file.
4819 The new format can be generated externally. For backward compatibility,
4820 if the data in the cache doesn't make sense, Exim assumes it has read an
4821 old-format file, and it generates new data and writes a new file. This
4822 means that you can't go back to an older release without removing the
4825 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4826 work if there are any delivery delays because "one_time" forces the
4827 parent to be marked "delivered", so its unseen clone is never tried
4828 again. For this reason, Exim now forbids the simultaneous setting of
4831 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4832 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4833 entry if you want to know the details. The fix, however, neglected to
4834 consider the case when local delivery batching is involved. The test for
4835 "previously delivered" was not happening when checking to see if an
4836 address could be batched with a previous (undelivered) one; under
4837 certain circumstances this could lead to multiple deliveries to the same
4840 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4841 in its include files, and this causes problems building Exim.
4843 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4844 header_syntax) but Exim was just ignoring anything given after a slash.
4845 In particular, this caused confusion with an attempt to use "verify =
4846 reverse_host_lookup/defer_ok". An error is now given when options are
4847 supplied for verify items that do not have them. (Maybe reverse_host_
4848 lookup should have a defer_ok option, but that's a different point.)
4850 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4851 defined by RFC 821) to 2048, because there were problems with some AUTH
4852 commands, and RFC 1869 says the size should be increased for extended
4853 SMTP commands that take arguments.
4855 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4858 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4859 "unknown" error; now it says that the functionality isn't in the binary.
4861 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4862 an address' error message when a string expansion fails (syntax or
4863 whatever). Otherwise the password may appear in the log. Following change
4864 PH/42 below, there is no longer a chance of it appearing in a bounce
4867 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4869 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4870 of its addresses was ever tried. (Bugzilla bug #2.)
4872 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4873 the result incorrectly in the debug output. (It correctly added a newline
4874 to what was transported.)
4876 TF/01 Added $received_time.
4878 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4879 commented out examples of how to interface to a virus scanner and to
4880 SpamAssassin. Also added commented examples of av_scanner and
4881 spamd_address settings.
4883 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4884 and controls are allowed in which ACLs. There were a couple of minor
4885 errors. Some of the entries in the conditions table (which is a table of
4886 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4887 negation of where the condition IS allowed.
4889 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4891 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4892 header file does not have a version number, so I've had to invent a new
4893 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4894 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4895 radiusclient), but it was contributed by a Radius user.
4897 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4898 files or directories, for OpenSSL.
4900 PH/19 When an Exim process that is running as root has to create an Exim log
4901 file, it does so in a subprocess that runs as exim:exim so as to get the
4902 ownership right at creation (otherwise, other Exim processes might see
4903 the file with the wrong ownership). There was no test for failure of this
4904 fork() call, which would lead to the process getting stuck as it waited
4905 for a non-existent subprocess. Forks do occasionally fail when resources
4906 run out. I reviewed all the other calls to fork(); they all seem to check
4909 PH/20 When checking for unexpected SMTP input at connect time (before writing
4910 the banner), Exim was not dealing correctly with a non-positive return
4911 from the read() function. If the client had disconnected by this time,
4912 the result was a log entry for a synchronization error with an empty
4913 string after "input=" when read() returned zero. If read() returned -1
4914 (an event I could not check), uninitialized data bytes were printed.
4915 There were reports of junk text (parts of files, etc) appearing after
4918 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4920 PH/22 Added support for macro redefinition, and (re)definition in between
4921 driver and ACL definitions.
4923 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4924 forgetting to use the resulting value; it was using the unexpanded value.
4926 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4927 hadn't been configured. The fix is from Juergen Kreileder, who
4928 understands it better than I do:
4930 "Here's what I see happening with three configured cyrus_sasl
4931 authenticators configured (plain, login, cram-md5):
4933 On startup auth_cyrus_sasl_init() gets called for each of these.
4934 This means three calls to sasl_listmech() without a specified mech_list.
4935 => SASL tests which mechs of all available mechs actually work
4936 => three warnings about OTP not working
4937 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4939 With the patch, sasl_listmech() also gets called three times. But now
4940 SASL's mech_list option is set to the server_mech specified in the the
4941 authenticator. Or in other words, the answer from sasl_listmech()
4942 gets limited to just the mech you're testing for (which is different
4944 => the return list contains just 'plain' or 'login', 'cram-md5' or
4945 nothing depending on the value of ob->server_mech.
4947 I've just tested the patch: Authentication still works fine,
4948 unavailable mechs specified in the exim configuration are still
4949 caught, and the auth.log warnings about OTP are gone."
4951 PH/25 When debugging is enabled, the contents of the command line are added
4952 to the debugging output, even when log_selector=+arguments is not
4955 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4956 answer is "GNU", and only if the return is "GNU/something" is the answer
4959 PH/27 $acl_verify_message is now set immediately after the failure of a
4960 verification in an ACL, and so is available in subsequent modifiers. In
4961 particular, the message can be preserved by coding like this:
4963 warn !verify = sender
4964 set acl_m0 = $acl_verify_message
4966 Previously, $acl_verify_message was set only while expanding "message"
4967 and "log_message" when a very denied access.
4969 PH/28 Modified OS/os.c-Linux with
4971 -#ifndef OS_LOAD_AVERAGE
4972 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4974 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4975 nomenclature these days.)
4977 PH/29 Installed patch from the Sieve maintainer that adds the options
4978 sieve_useraddress and sieve_subaddress to the redirect router.
4980 PH/30 In these circumstances:
4981 . Two addresses routed to the same list of hosts;
4982 . First host does not offer TLS;
4983 . First host accepts first address;
4984 . First host gives temporary error to second address;
4985 . Second host offers TLS and a TLS session is established;
4986 . Second host accepts second address.
4987 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4988 and peerdn, if requested) that were in fact used only for the second
4991 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4992 attention to any local part prefix or suffix that was matched by the
4993 router that accepted the address. It now behaves in the same way as it
4994 does for delivery: the affixes are removed from the local part unless
4995 rcpt_include_affixes is set on the transport.
4997 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4998 timeout during the DATA phase of an incoming message.
5000 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5001 applied a patch sanctioned by the Sieve maintainer.
5003 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5004 the uid or gid is negative. A case of a negative gid caused this to be
5005 noticed. The fix allows for either to be negative.
5007 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5008 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5011 PH/36 The change PH/12 above was broken. Fixed it.
5013 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5014 the grounds that routing the same address twice would always produce the
5015 same answer. This might have been true once, but it is certainly no
5016 longer true now. Routing a child address may depend on the previous
5017 routing that produced that child. Some complicated redirection strategies
5018 went wrong when messages had multiple recipients, and made Exim's
5019 behaviour dependent on the order in which the addresses were given.
5021 I have moved the duplicate checking until after the routing is complete.
5022 Exim scans the addresses that are assigned to local and remote
5023 transports, and removes any duplicates. This means that more work will be
5024 done, as duplicates will always all be routed, but duplicates are
5025 presumably rare, so I don't expect this is of any significance.
5027 For deliveries to pipes, files, and autoreplies, the duplicate checking
5028 still happens during the routing process, since they are not going to be
5031 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5032 It corrects a timeout issue with spamd. This is Ian's comment: "The
5033 background is that sometimes spamd either never reads data from a
5034 connection it has accepted, or it never writes response data. The exiscan
5035 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5036 blindly assumes that writes won't block so it may never time out."
5038 PH/39 Allow G after quota size as well as K and M.
5040 PH/40 The value set for $authenticated_id in an authenticator may not contain
5041 binary zeroes or newlines because the value is written to log lines and
5042 to spool files. There was no check on this. Now the value is run through
5043 the string_printing() function so that such characters are converted to
5044 printable escape sequences.
5046 PH/41 $message_linecount is a new variable that contains the total number of
5047 lines in the message. Compare $body_linecount, which is the count for the
5050 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5051 bounce and delay warning messages, except in certain special cases, which
5054 (a) An SMTP error message from a remote host;
5055 (b) A message specified in a :fail: redirection;
5056 (c) A message specified in a "fail" command in a system filter;
5057 (d) A message specified in a FAIL return from the queryprogram router;
5058 (e) A message specified by the cannot_route_message router option.
5060 In these cases only, Exim does include the error details in bounce and
5061 warning messages. There are also a few cases where bland messages such
5062 as "unrouteable address" or "local delivery error" are given.
5064 PH/43 $value is now also set for the "else" part of a ${run expansion.
5066 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5067 being worked on, but at least Exim now implements the latest version to
5070 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5071 process to complete was treated as a delivery failure, a timeout while
5072 writing the message to the pipe was logged, but erroneously treated as a
5073 successful delivery. Such timeouts include transport filter timeouts. For
5074 consistency with the overall process timeout, these timeouts are now
5075 treated as errors, giving rise to delivery failures by default. However,
5076 there is now a new Boolean option for the pipe transport called
5077 timeout_defer, which, if set TRUE, converts the failures into defers for
5078 both kinds of timeout. A transport filter timeout is now identified in
5081 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5082 systems where "make" and "gmake" are different, calling "gmake" at top
5083 level broke things. I've arranged for the value of $(MAKE) to be passed
5084 from the Makefile to this script so that it can call the same version of
5088 A note about Exim versions 4.44 and 4.50
5089 ----------------------------------------
5091 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5092 changes of various kinds. As a consequence, a big documentation update was
5093 needed. This delayed the release for rather longer than seemed good, especially
5094 in the light of a couple of (minor) security issues. Therefore, the changes
5095 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5096 release. So 4.44 and 4.50 are in effect two different branches that both start
5099 I have left the 4.50 change log unchanged; it contains all the changes since
5100 4.43. The change log for 4.44 is below; many of its items are identical to
5101 those for 4.50. This seems to be the most sensible way to preserve the
5102 historical information.
5108 1. Minor wording change to the doc/README.SIEVE file.
5110 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5111 computation of the current number of files was incorrect.
5113 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5114 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5117 4. Give more explanation in the error message when the command for a transport
5118 filter fails to execute.
5120 5. There are several places where Exim runs a non-Exim command in a
5121 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5122 was being done only for the command run by the queryprogram router. It is
5123 now done for all such subprocesses. The other cases are: ${run, transport
5124 filters, and the commands run by the lmtp and pipe transports.
5126 6. Added CONFIGURE_GROUP build-time option.
5128 7. Some older OS have a limit of 256 on the maximum number of file
5129 descriptors. Exim was using setrlimit() to set 1000 as a large value
5130 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5131 systems. I've change it so that if it can't get 1000, it tries for 256.
5133 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5134 was an oversight, and furthermore, ever since the addition of extra
5135 controls (e.g. 4.43/32), the checks on when to allow different forms of
5136 "control" were broken. There should now be diagnostics for all cases when a
5137 control that does not make sense is encountered.
5139 9. Added the /retain_sender option to "control=submission".
5141 10. $recipients is now available in the predata ACL (oversight).
5143 11. Tidy the search cache before the fork to do a delivery from a message
5144 received from the command line. Otherwise the child will trigger a lookup
5145 failure and thereby defer the delivery if it tries to use (for example) a
5146 cached ldap connection that the parent has called unbind on.
5148 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5149 of $address_data from the recipient verification was clobbered by the
5150 sender verification.
5152 13. The value of address_data from a sender verification is now available in
5153 $sender_address_data in subsequent conditions in the ACL statement.
5155 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5157 15. Added a new option "connect=<time>" to callout options, to set a different
5160 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5161 was its contents. (It was OK if the option was not defined at all.)
5163 17. A "Completed" log line is now written for messages that are removed from
5164 the spool by the -Mrm option.
5166 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5167 information about exactly what failed.
5169 19. Added -dd to debug only the daemon process.
5171 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5172 handles timeouts, both on the server side and network timeouts. Renamed the
5173 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5175 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5176 It is now set to "smtps".
5178 22. $host_address is now set to the target address during the checking of
5179 ignore_target_hosts.
5181 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5182 being passed; this would have caused $sender_host_name to have been used if
5183 matching the list had actually called for a host name (not very likely,
5184 since this list is usually IP addresses). A host name is now passed as
5187 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5188 code that specifies a non-restarting handler (typically sigaction() in
5189 modern systems) in an attempt to fix a rare and obscure crash bug.
5191 25. Narrowed the window for a race in the daemon that could cause it to ignore
5192 SIGCHLD signals. This is not a major problem, because they are used only to
5193 wake it up if nothing else does.
5195 26. A malformed maildirsize file could cause Exim to calculate negative values
5196 for the mailbox size or file count. Odd effects could occur as a result.
5197 The maildirsize information is now recalculated if the size or filecount
5200 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5201 support for a long time. Removed HAVE_SYS_VFS_H.
5203 28. Installed the latest version of exipick from John Jetmore.
5205 29. In an address list, if the pattern was not a regular expression, an empty
5206 subject address (from a bounce message) matched only if the pattern was an
5207 empty string. Non-empty patterns were not even tested. This was the wrong
5208 because it is perfectly reasonable to use an empty address as part of a
5209 database query. An empty address is now tested by patterns that are
5210 lookups. However, all the other forms of pattern expect the subject to
5211 contain a local part and a domain, and therefore, for them, an empty
5212 address still always fails if the pattern is not itself empty.
5214 30. Exim went into a mad DNS loop when attempting to do a callout where the
5215 host was specified on an smtp transport, and looking it up yielded more
5216 than one IP address.
5218 31. Re-factored the code for checking spool and log partition space into a
5219 function that finds that data and another that does the check. The former
5220 is then used to implement four new variables: $spool_space, $log_space,
5221 $spool_inodes, and $log_inodes.
5223 32. The RFC2047 encoding function was originally intended for short strings
5224 such as real names; it was not keeping to the 75-character limit for
5225 encoded words that the RFC imposes. It now respects the limit, and
5226 generates multiple encoded words if necessary. To be on the safe side, I
5227 have increased the buffer size for the ${rfc2047: expansion operator from
5230 33. It is now permitted to omit both strings after an "if" condition; if the
5231 condition is true, the result is "true". As before, when the second string
5232 is omitted, a false condition yields an empty string. This makes it less
5233 cumbersome to write custom ACL and router conditions.
5235 34. Failure to deliver a bounce message always caused it to be frozen, even if
5236 there was an errors_to setting on the router. The errors_to setting is now
5239 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5240 canonical form (fully expanded) before being placed in
5241 $sender_host_address.
5243 36. The table in the code that translates DNS record types into text (T_A to
5244 "A" for instance) was missing entries for NS and CNAME. It is just possible
5245 that this could have caused confusion if both these types were looked up
5246 for the same domain, because the text type is used as part of Exim's
5247 per-process caching. But the chance of anyone hitting this buglet seems
5250 37. The dnsdb lookup has been extended in a number of ways.
5252 (1) There is a new type, "zns", which walks up the domain tree until it
5253 finds some nameserver records. It should be used with care.
5255 (2) There is a new type, "mxh", which is like "mx" except that it returns
5256 just the host names, not the priorities.
5258 (3) It is now possible to give a list of domains (or IP addresses) to be
5259 looked up. The behaviour when one of the lookups defers can be
5260 controlled by a keyword.
5262 (4) It is now possible to specify the separator character for use when
5263 multiple records are returned.
5265 38. The dnslists ACL condition has been extended: it is now possible to supply
5266 a list of IP addresses and/or domains to be looked up in a particular DNS
5269 39. Added log_selector=+queue_time_overall.
5271 40. When running the queue in the test harness, wait just a tad after forking a
5272 delivery process, to get repeatability of debugging output.
5274 41. Include certificate and key file names in error message when GnuTLS fails
5275 to set them up, because the GnuTLS error message doesn't include the name
5276 of the failing file when there is a problem reading it.
5278 42. Allow both -bf and -bF in the same test run.
5280 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5282 44. The "Exiscan patch" is now merged into the mainline Exim source.
5284 45. Sometimes the final signoff response after QUIT could fail to get
5285 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5286 < 0 before doing a fflush(). This bug looks as though it goes back to the
5287 introduction of TLS in release 3.20, but "sometimes" must have been rare
5288 because the tests only now provoked it.
5290 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5291 (this can affect the format of dates).
5293 47. exim_tidydb, when checking for the continued existence of a message for
5294 which it has found a message-specific retry record, was not finding
5295 messages that were in split spool directories. Consequently, it was
5296 deleting retry records that should have stayed in existence.
5298 48. Steve fixed some bugs in eximstats.
5300 49. The SPA authentication driver was not abandoning authentication and moving
5301 on to the next authenticator when an expansion was forced to fail,
5302 contradicting the general specification for all authenticators. Instead it
5303 was generating a temporary error. It now behaves as specified.
5305 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5306 (the order specifies the preference for clients). The order is now AES256,
5307 AES128, 3DES, ARCFOUR128.
5309 51. Small patch to Sieve code - explicitly set From: when generating an
5312 52. Exim crashed if a remote delivery caused a very long error message to be
5313 recorded - for instance if somebody sent an entire SpamAssassin report back
5314 as a large number of 550 error lines. This bug was coincidentally fixed by
5315 increasing the size of one of Exim's internal buffers (big_buffer) that
5316 happened as part of the Exiscan merge. However, to be on the safe side, I
5317 have made the code more robust (and fixed the comments that describe what
5320 53. Now that there can be additional text after "Completed" in log lines (if
5321 the queue_time_overall log selector is set), a one-byte patch to exigrep
5322 was needed to allow it to recognize "Completed" as not the last thing in
5325 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5326 patch that reportedly fixes this has been added. I am not expert enough to
5327 create a test for it. This is what the patch creator wrote:
5329 "I found a little strange behaviour of ldap code when working with
5330 Windows 2003 AD Domain, where users was placed in more than one
5331 Organization Units. When I tried to give exim partial DN, the exit code
5332 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5333 But simultaneously result of request was absolutely normal ldap result,
5334 so I produce this patch..."
5336 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5337 REFERENCE, so I have modified the code to exclude the patch when that macro
5340 55. Some experimental protocols are using DNS PTR records for new purposes. The
5341 keys for these records are domain names, not reversed IP addresses. The
5342 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5343 leaves it alone. Component reversal etc. now happens only for IP addresses.
5346 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5348 57. Double the size of the debug message buffer (to 2048) so that more of very
5349 long debug lines gets shown.
5351 58. The exicyclog utility now does better if the number of log files to keep
5352 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5354 59. Two changes related to the smtp_active_hostname option:
5356 (1) $smtp_active_hostname is now available as a variable.
5357 (2) The default for smtp_banner uses $smtp_active_hostname instead
5358 of $primary_hostname.
5360 60. The host_aton() function is supposed to be passed a string that is known
5361 to be a valid IP address. However, in the case of IPv6 addresses, it was
5362 not checking this. This is a hostage to fortune. Exim now panics and dies
5363 if the condition is not met. A case was found where this could be provoked
5364 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5365 components; fortuitously, this particular loophole had already been fixed
5366 by change 4.50/55 above.
5368 If there are any other similar loopholes, the new check in host_aton()
5369 itself should stop them being exploited. The report I received stated that
5370 data on the command line could provoke the exploit when Exim was running as
5371 exim, but did not say which command line option was involved. All I could
5372 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5373 running as the user.
5376 61. There was a buffer overflow vulnerability in the SPA authentication code
5377 (which came originally from the Samba project). I have added a test to the
5378 spa_base64_to_bits() function which I hope fixes it.
5381 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5382 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5384 63. The daemon start-up calls getloadavg() while still root for those OS that
5385 need the first call to be done as root, but it missed one case: when
5386 deliver_queue_load_max is set with deliver_drop_privilege. This is
5387 necessary for the benefit of the queue runner, because there is no re-exec
5388 when deliver_drop_privilege is set.
5390 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5391 This has been fixed.
5393 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5394 was in use, was not putting the data itself into the right store pool;
5395 consequently, it could be overwritten for a subsequent message in the same
5396 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5399 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5401 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5402 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5403 the function were treating the return as a boolean value, which happened to
5404 work because 0=false and not-0=true, but is not correct code.
5406 68. The host_aton() function was not handling scoped IPv6 addresses (those
5407 with, for example, "%eth0" on the end) correctly.
5409 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5410 negated items (that is, ~something) in unsigned ints. Some compilers
5411 apparently mutter when there is no cast.
5413 70. If an address verification called from an ACL failed, and did not produce a
5414 user-specific message (i.e. there was only a "system" message), nothing was
5415 put in $acl_verify_message. In this situation, it now puts the system
5418 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5419 session; change 4.31/43 added the unwanted input to the log line - except
5420 that it did not do this in the start of session case. It now does.
5422 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5423 This is wrong and can cause the other end to generate a synchronization
5424 error if it is another Exim or anything else that does the synchronization
5425 check. A QUIT command is no longer sent after a timeout.
5427 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5428 during host lookups.
5430 74. The defer_ok option of callout verification was not working if it was used
5431 when verifying addresses in header lines, that is, for this case:
5433 verify = header_sender/callout=defer_ok
5435 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5436 those file descriptors could be used for SMTP connections. If anything
5437 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5438 could be sent to the SMTP client, causing chaos. The daemon now opens
5439 stdin, stdout, and stderr to /dev/null when it puts itself into the
5442 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5443 log by default. The user can override this with suitable Perl magic.
5445 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5446 the log message when discard triggers, was not working for the DATA ACL or
5447 for the non-SMTP ACL.
5449 78. Error message wording change in sieve.c.
5451 79. If smtp_accept_max_per_host was set, the number of connections could be
5452 restricted to fewer than expected, because the daemon was trying to set up
5453 a new connection before checking whether the processes handling previous
5454 connections had finished. The check for completed processes is now done
5455 earlier. On busy systems, this bug wouldn't be noticed because something
5456 else would have woken the daemon, and it would have reaped the completed
5459 80. If a message was submitted locally by a user whose login name contained one
5460 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5461 It caused a spool format error. I have fixed the spool reading code. A
5462 related problem was that the "from" clause in the Received: line became
5463 illegal because of the space(s). It is now covered by ${quote_local_part.
5465 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5468 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5470 83. Patch from Sieve maintainer to fix unterminated string problem in
5471 "vacation" handling.
5473 84. Some minor changes to the Linux configuration files to help with other
5474 OS variants using glibc.
5476 85. One more patch for Sieve to update vacation handling to latest spec.
5479 ----------------------------------------------------
5480 See the note above about the 4.44 and 4.50 releases.
5481 ----------------------------------------------------
5487 1. Change 4.43/35 introduced a bug that caused file counts to be
5488 incorrectly computed when quota_filecount was set in an appendfile
5491 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5492 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5495 3. Give more explanation in the error message when the command for a transport
5496 filter fails to execute.
5498 4. There are several places where Exim runs a non-Exim command in a
5499 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5500 was being done only for the command run by the queryprogram router. It is
5501 now done for all such subprocesses. The other cases are: ${run, transport
5502 filters, and the commands run by the lmtp and pipe transports.
5504 5. Some older OS have a limit of 256 on the maximum number of file
5505 descriptors. Exim was using setrlimit() to set 1000 as a large value
5506 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5507 systems. I've change it so that if it can't get 1000, it tries for 256.
5509 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5510 was an oversight, and furthermore, ever since the addition of extra
5511 controls (e.g. 4.43/32), the checks on when to allow different forms of
5512 "control" were broken. There should now be diagnostics for all cases when a
5513 control that does not make sense is encountered.
5515 7. $recipients is now available in the predata ACL (oversight).
5517 8. Tidy the search cache before the fork to do a delivery from a message
5518 received from the command line. Otherwise the child will trigger a lookup
5519 failure and thereby defer the delivery if it tries to use (for example) a
5520 cached ldap connection that the parent has called unbind on.
5522 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5523 of $address_data from the recipient verification was clobbered by the
5524 sender verification.
5526 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5527 was its contents. (It was OK if the option was not defined at all.)
5529 11. A "Completed" log line is now written for messages that are removed from
5530 the spool by the -Mrm option.
5532 12. $host_address is now set to the target address during the checking of
5533 ignore_target_hosts.
5535 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5536 being passed; this would have caused $sender_host_name to have been used if
5537 matching the list had actually called for a host name (not very likely,
5538 since this list is usually IP addresses). A host name is now passed as
5541 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5542 code that specifies a non-restarting handler (typically sigaction() in
5543 modern systems) in an attempt to fix a rare and obscure crash bug.
5545 15. Narrowed the window for a race in the daemon that could cause it to ignore
5546 SIGCHLD signals. This is not a major problem, because they are used only to
5547 wake it up if nothing else does.
5549 16. A malformed maildirsize file could cause Exim to calculate negative values
5550 for the mailbox size or file count. Odd effects could occur as a result.
5551 The maildirsize information is now recalculated if the size or filecount
5554 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5555 support for a long time. Removed HAVE_SYS_VFS_H.
5557 18. Updated exipick to current release from John Jetmore.
5559 19. Allow an empty sender to be matched against a lookup in an address list.
5560 Previously the only cases considered were a regular expression, or an
5563 20. Exim went into a mad DNS lookup loop when doing a callout where the
5564 host was specified on the transport, if the DNS lookup yielded more than
5567 21. The RFC2047 encoding function was originally intended for short strings
5568 such as real names; it was not keeping to the 75-character limit for
5569 encoded words that the RFC imposes. It now respects the limit, and
5570 generates multiple encoded words if necessary. To be on the safe side, I
5571 have increased the buffer size for the ${rfc2047: expansion operator from
5574 22. Failure to deliver a bounce message always caused it to be frozen, even if
5575 there was an errors_to setting on the router. The errors_to setting is now
5578 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5579 canonical form (fully expanded) before being placed in
5580 $sender_host_address.
5582 24. Updated eximstats to version 1.33
5584 25. Include certificate and key file names in error message when GnuTLS fails
5585 to set them up, because the GnuTLS error message doesn't include the name
5586 of the failing file when there is a problem reading it.
5588 26. Expand error message when OpenSSL has problems setting up cert/key files.
5591 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5592 (this can affect the format of dates).
5594 28. exim_tidydb, when checking for the continued existence of a message for
5595 which it has found a message-specific retry record, was not finding
5596 messages that were in split spool directories. Consequently, it was
5597 deleting retry records that should have stayed in existence.
5599 29. eximstats updated to version 1.35
5600 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5601 1.35 - bugfix such that pie charts by volume are generated correctly
5603 30. The SPA authentication driver was not abandoning authentication and moving
5604 on to the next authenticator when an expansion was forced to fail,
5605 contradicting the general specification for all authenticators. Instead it
5606 was generating a temporary error. It now behaves as specified.
5608 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5609 (the order specifies the preference for clients). The order is now AES256,
5610 AES128, 3DES, ARCFOUR128.
5612 31. Small patch to Sieve code - explicitly set From: when generating an
5615 32. Exim crashed if a remote delivery caused a very long error message to be
5616 recorded - for instance if somebody sent an entire SpamAssassin report back
5617 as a large number of 550 error lines. This bug was coincidentally fixed by
5618 increasing the size of one of Exim's internal buffers (big_buffer) that
5619 happened as part of the Exiscan merge. However, to be on the safe side, I
5620 have made the code more robust (and fixed the comments that describe what
5623 33. Some experimental protocols are using DNS PTR records for new purposes. The
5624 keys for these records are domain names, not reversed IP addresses. The
5625 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5626 leaves it alone. Component reversal etc. now happens only for IP addresses.
5629 34. The host_aton() function is supposed to be passed a string that is known
5630 to be a valid IP address. However, in the case of IPv6 addresses, it was
5631 not checking this. This is a hostage to fortune. Exim now panics and dies
5632 if the condition is not met. A case was found where this could be provoked
5633 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5634 components; fortuitously, this particular loophole had already been fixed
5635 by change 4.50/55 or 4.44/33 above.
5637 If there are any other similar loopholes, the new check in host_aton()
5638 itself should stop them being exploited. The report I received stated that
5639 data on the command line could provoke the exploit when Exim was running as
5640 exim, but did not say which command line option was involved. All I could
5641 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5642 running as the user.
5645 35. There was a buffer overflow vulnerability in the SPA authentication code
5646 (which came originally from the Samba project). I have added a test to the
5647 spa_base64_to_bits() function which I hope fixes it.
5650 36. The daemon start-up calls getloadavg() while still root for those OS that
5651 need the first call to be done as root, but it missed one case: when
5652 deliver_queue_load_max is set with deliver_drop_privilege. This is
5653 necessary for the benefit of the queue runner, because there is no re-exec
5654 when deliver_drop_privilege is set.
5656 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5657 was in use, was not putting the data itself into the right store pool;
5658 consequently, it could be overwritten for a subsequent message in the same
5659 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5662 38. Sometimes the final signoff response after QUIT could fail to get
5663 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5664 < 0 before doing a fflush(). This bug looks as though it goes back to the
5665 introduction of TLS in release 3.20, but "sometimes" must have been rare
5666 because the tests only now provoked it.
5672 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5673 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5674 Now it returns an integer. A number of calls were still expecting a T/F
5675 return. Fortuitously, in all cases, the tests worked in OK situations,
5676 which is the norm. However, things would have gone wrong on any write
5677 failures on the smtp file descriptor. This function is used when sending
5678 messages over SMTP and also when doing verify callouts.
5680 2. When Exim is called to do synchronous delivery of a locally submitted
5681 message (the -odf or -odi options), it no longer closes stderr before doing
5684 3. Implemented the mua_wrapper option.
5686 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5688 5. Implemented the functions header_remove(), header_testname(),
5689 header_add_at_position(), and receive_remove_recipient(), and exported them
5692 6. If an ACL "warn" statement specified the addition of headers, Exim already
5693 inserted X-ACL-Warn: at the start if there was no header name. However, it
5694 was not making this test for the second and subsequent header lines if
5695 there were newlines in the string. This meant that an invalid header could
5696 be inserted if Exim was badly configured.
5698 7. Allow an ACL "warn" statement to add header lines at the start or after all
5699 the Received: headers, as well as at the end.
5701 8. Added the rcpt_4xx retry error code.
5703 9. Added postmaster_mailfrom=xxx to callout verification option.
5705 10. Added mailfrom=xxxx to the callout verification option, for verify=
5708 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5709 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5711 12. Inserted some casts to stop certain compilers warning when using pointer
5712 differences as field lengths or precisions in printf-type calls (mostly
5713 affecting debugging statements).
5715 13. Added optional readline() support for -be (dynamically loaded).
5717 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5718 same clock tick as a message's arrival, so that its received time was the
5719 same as the "first fail" time on the retry record, and that message
5720 remained on the queue past the ultimate address timeout, every queue runner
5721 would try a delivery (because it was past the ultimate address timeout) but
5722 after another failure, the ultimate address timeout, which should have then
5723 bounced the address, did not kick in. This was a "< instead of <=" error;
5724 in most cases the first failure would have been in the next clock tick
5725 after the received time, and all would be well.
5727 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5728 being recognized when the domain list was tested by the match_domain
5729 condition in an expansion string.
5731 16. Added the ${str2b64: operator.
5733 17. Exim was always calling setrlimit() to set a large limit for the number of
5734 processes, without checking whether the existing limit was already
5735 adequate. (It did check for the limit on file descriptors.) Furthermore,
5736 errors from getrlimit() and setrlimit() were being ignored. Now they are
5737 logged to the main and panic logs, but Exim does carry on, to try to do its
5738 job under whatever limits there are.
5740 18. Imported PCRE 5.0.
5742 19. Trivial typo in log message " temporarily refused connection" (the leading
5745 20. If the log selector return_path_on_delivery was set and an address was
5746 redirected to /dev/null, the delivery process crashed because it assumed
5747 that a return path would always be set for a "successful" delivery. In this
5748 case, the whole delivery is bypassed as an optimization, and therefore no
5751 21. Internal re-arrangement: the function for sending a challenge and reading
5752 a response while authentication was assuming a zero-terminated challenge
5753 string. It's now changed to take a pointer and a length, to allow for
5754 binary data in such strings.
5756 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5758 23. Exim was not respecting finduser_retries when seeking the login of the
5759 uid under which it was called; it was always trying 10 times. (The default
5760 setting of finduser_retries is zero.) Also, it was sleeping after the final
5761 failure, which is pointless.
5763 24. Implemented tls_on_connect_ports.
5765 25. Implemented acl_smtp_predata.
5767 26. If the domain in control=submission is set empty, Exim assumes that the
5768 authenticated id is a complete email address when it generates From: or
5769 Sender: header lines.
5771 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5772 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5773 chown and chgrp in /bin and hostname in /usr/bin.
5775 28. Exim was keeping the "process log" file open after each use, just as it
5776 does for the main log. This opens the possibility of it remaining open for
5777 long periods when the USR1 signal hits a daemon. Occasional processlog
5778 errors were reported, that could have been caused by this. Anyway, it seems
5779 much more sensible not to leave this file open at all, so that is what now
5782 29. The long-running daemon process does not normally write to the log once it
5783 has entered its main loop, and it closes the log before doing so. This is
5784 so that log files can straightforwardly be renamed and moved. However,
5785 there are a couple of unusual error situations where the daemon does write
5786 log entries, and I had neglected to close the log afterwards.
5788 30. The text of an SMTP error response that was received during a remote
5789 delivery was being truncated at 512 bytes. This is too short for some of
5790 the long messages that one sometimes sees. I've increased the limit to
5793 31. It is now possible to make retry rules that apply only when a message has a
5794 specific sender, in particular, an empty sender.
5796 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5797 it possible to be selective about when SMTP synchronization is enforced.
5799 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5801 32. Implemented hosts_connection_nolog.
5803 33. Added an ACL for QUIT.
5805 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5808 35. Added mailbox_size and mailbox_filecount to appendfile.
5810 36. Added control = no_multiline_responses to ACLs.
5812 37. There was a bug in the logic of the code that waits for the clock to tick
5813 in the case where the clock went backwards by a substantial amount such
5814 that the microsecond fraction of "now" was more than the microsecond
5815 fraction of "then" (but the whole seconds number was less).
5817 38. Added support for the libradius Radius client library this is found on
5818 FreeBSD (previously only the radiusclient library was supported).
5824 1. When certain lookups returned multiple values in the form name=value, the
5825 quoting of the values was not always being done properly. Specifically:
5826 (a) If the value started with a double quote, but contained no whitespace,
5828 (b) If the value contained whitespace other than a space character (i.e.
5829 tabs or newlines or carriage returns) it was not quoted.
5830 This fix has been applied to the mysql and pgsql lookups by writing a
5831 separate quoting function and calling it from the lookup code. The fix
5832 should probably also be applied to nisplus, ibase and oracle lookups, but
5833 since I cannot test any of those, I have not disturbed their existing code.
5835 2. A hit in the callout cache for a specific address caused a log line with no
5836 reason for rejecting RCPT. Now it says "Previous (cached) callout
5837 verification failure".
5839 3. There was an off-by-one bug in the queryprogram router. An over-long
5840 return line was truncated at 256 instead of 255 characters, thereby
5841 overflowing its buffer with the terminating zero. As well as fixing this, I
5842 have increased the buffer size to 1024 (and made a note to document this).
5844 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5845 when Exim is waiting for an SMTP response from a remote server, Exim
5846 restarts its select() call on the socket, thereby resetting its timeout.
5847 This is not a problem when such interrupts are rare. Somebody set up a cron
5848 job to run exiwhat every 2 minutes, which is less than the normal select()
5849 timeout (5 or 10 minutes). This meant that the select() timeout never
5850 kicked in because it was always reset. I have fixed this by comparing the
5851 time when an interrupt arrives with the time at the start of the first call
5852 to select(). If more time than the timeout has elapsed, the interrupt is
5853 treated as a timeout.
5855 5. Some internal re-factoring in preparation for the addition of Sieve
5856 extensions (by MH). In particular, the "personal" test is moved to a
5857 separate function, and given an option for scanning Cc: and Bcc: (which is
5858 not set for Exim filters).
5860 6. When Exim created an email address using the login of the caller as the
5861 local part (e.g. when creating a From: or Sender: header line), it was not
5862 quoting the local part when it contained special characters such as @.
5864 7. Installed new OpenBSD configuration files.
5866 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5867 try to make them clearer.
5869 9. Callout options, other than the timeout value, were being ignored when
5870 verifying sender addresses in header lines. For example, when using
5872 verify = header_sender/callout=no_cache
5874 the cache was (incorrectly) being used.
5876 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5877 only the Cygwin environment.
5879 11. When return_path_on_delivery was set as a log selector, if different remote
5880 addresses in the same message used different return paths and parallel
5881 remote delivery occurred, the wrong values would sometimes be logged.
5882 (Whenever a remote delivery process finished, the return path value from
5883 the most recently started remote delivery process was logged.)
5885 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5886 lines when AUTH and/or TLS are in use. This is the "received protocol"
5887 field. Exim used to use "asmtp" for authenticated SMTP, without any
5888 indication (in the protocol name) for TLS use. Now it follows the RFC and
5889 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5890 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5891 names appear in log lines as well as in Received: header lines.
5893 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5894 extensions, and comparison tests, and to fix some bugs.
5896 14. Changes to the "personal" filter test:
5898 (1) The test was buggy in that it was just doing the equivalent of
5899 "contains" tests on header lines. For example, if a user's address was
5900 anne@some.where, the "personal" test would incorrectly be true for
5902 To: susanne@some.where
5904 This test is now done by extracting each address from the header in turn,
5905 and checking the entire address. Other tests that are part of "personal"
5906 are now done using regular expressions (for example, to check local parts
5907 of addresses in From: header lines).
5909 (2) The list of non-personal local parts in From: addresses has been
5910 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5911 taken from the Sieve specification recommendations.
5913 (3) If the message contains any header line starting with "List-" it is
5914 treated as non-personal.
5916 (4) The test for "circular" in the Subject: header line has been removed
5917 because it now seems ill-conceived.
5919 15. Minor typos in src/EDITME comments corrected.
5921 16. Installed latest exipick from John Jetmore.
5923 17. If headers_add on a router specified a text string that was too long for
5924 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5925 of string_sprintf() is now avoided.
5927 18. $message_body_size was not set (it was always zero) when running the DATA
5928 ACL and the local_scan() function.
5930 19. For the "mail" command in an Exim filter, no default was being set for
5931 the once_repeat time, causing a random time value to be used if "once" was
5932 specified. (If the value happened to be <= 0, no repeat happened.) The
5933 default is now 0s, meaning "never repeat". The "vacation" command was OK
5934 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5935 (I found it when inspecting the code).
5937 20. There is now an overall timeout for performing a callout verification. It
5938 defaults to 4 times the callout timeout, which applies to individual SMTP
5939 commands during the callout. The overall timeout applies when there is more
5940 than one host that can be tried. The timeout is checked before trying the
5941 next host. This prevents very long delays if there are a large number of
5942 hosts and all are timing out (e.g. when the network connections are timing
5943 out). The value of the overall timeout can be changed by specifying an
5944 additional sub-option for "callout", called "maxwait". For example:
5946 verify = sender/callout=5s,maxwait=20s
5948 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5949 to the end before writing, but this should make it even safer).
5951 22. Exim was forgetting that it had advertised PIPELINING for the second and
5952 subsequent messages on an SMTP connection. It was also not resetting its
5953 memory on STARTTLS and an internal HELO.
5955 23. When Exim logs an SMTP synchronization error within a session, it now
5956 records whether PIPELINING has been advertised or not.
5958 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5959 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5960 rather than long int.
5962 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5964 26. Added the never_mail option to autoreply.
5970 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5971 crash if the getsockname() call failed; this can happen if a connection is
5972 closed very soon after it is established. The problem was simply in the
5973 order in which certain operations were done, causing Exim to try to write
5974 to the SMTP stream before it had set up the file descriptor. The bug has
5975 been fixed by making things happen in the correct order.
5981 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5982 before the connection was closed, thus losing the rejection response.
5984 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5985 some early Solaris releases, but causes trouble in current releases where
5986 socklen_t is defined.
5988 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5991 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5994 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5995 that releasing the top of it at the end releases what was used for sub-
5996 expansions (unless the block got too big). However, discard this block if
5997 the first thing is a variable or header, so that we can use its block when
5998 it is dynamic (useful for very large $message_headers, for example).
6000 6. Lookups now cache *every* query, not just the most recent. A new, separate
6001 store pool is used for this. It can be recovered when all lookup caches are
6002 flushed. Lookups now release memory at the end of their result strings.
6003 This has involved some general refactoring of the lookup sources.
6005 7. Some code has been added to the store_xxx() functions to reduce the amount
6006 of flapping under certain conditions.
6008 8. log_incoming_interface used to affect only the <= reception log lines. Now
6009 it causes the local interface and port to be added to several more SMTP log
6010 lines, for example "SMTP connection from", and rejection lines.
6012 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6014 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6016 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6017 could be overwritten at the end of the current message (or the start of a
6018 new message if it was set in a HELO ACL). The value is now preserved for
6019 the duration of the SMTP connection.
6021 12. If a transport had a headers_rewrite setting, and a matching header line
6022 contained an unqualified address, that address was qualified, even if it
6023 did not match any rewriting rules. The underlying bug was that the values
6024 of the flags that permit the existence of unqualified sender and recipient
6025 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6026 non-local messages, and by -bnq for local messages) were not being
6027 preserved with the message after it was received.
6029 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6030 "next input=" as part of the text comprising the host identity instead of
6031 the correct text. The code was using the same buffer for two different
6032 strings. However, depending on which order the printing function evaluated
6033 its arguments, the bug did not always show up. Under Linux, for example, my
6034 test suite worked just fine.
6036 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6037 doesn't work with some older versions of Perl. It has been changed to "my",
6038 which in any case is probably the better facility to use.
6040 15. A really picky compiler found some instances of statements for creating
6041 error messages that either had too many or two few arguments for the format
6044 16. The size of the buffer for calls to the DNS resolver has been increased
6045 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6046 for addresses that have a lot of PTR records. This alleviates a problem; it
6047 does not fully solve it.
6049 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6050 buffer now truncates the list and logs the incident, which is the same
6051 action as happens when Exim is looking up a host name and its aliases.
6052 Previously in this situation something unpredictable would happen;
6053 sometimes it was "internal error: store_reset failed".
6055 18. If a server dropped the connection unexpectedly when an Exim client was
6056 using GnuTLS and trying to read a response, the client delivery process
6057 crashed while trying to generate an error log message.
6059 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6060 string, for example:
6062 warn message = H1: something\nH2: something
6064 the text was added as a single header line from Exim's point of view
6065 though it ended up OK in the delivered message. However, searching for the
6066 second and subsequent header lines using $h_h2: did not work. This has been
6067 fixed. Similarly, if a system filter added multiple headers in this way,
6068 the routers could not see them.
6070 20. Expanded the error message when iplsearch is called with an invalid key to
6071 suggest using net-iplsearch in a host list.
6073 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6074 ACLs are no longer actually imposed (and a message to that effect is
6077 22. If a "gecos" field in a passwd entry contained escaped characters, in
6078 particular, if it contained a \" sequence, Exim got it wrong when building
6079 a From: or a Sender: header from that name. A second bug also caused
6080 incorrect handling when an unquoted " was present following a character
6081 that needed quoting.
6083 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6084 was not being matched caselessly.
6086 24. Arranged for all hyphens in the exim.8 source to be escaped with
6089 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6090 properly. Recipient callout cache records were still being keyed to include
6091 the sender, even when use_sender was set false. This led to far more
6092 callouts that were necessary. The sender is no longer included in the key
6093 when use_sender is false.
6095 26. Added "control = submission" modifier to ACLs.
6097 27. Added the ${base62d: operator to decode base 62 numbers.
6099 28. dnsdb lookups can now access SRV records.
6101 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6102 the configuration file.
6104 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6105 (-v) mode. This makes the output for a verbose queue run more intelligible.
6107 31. Added a use_postmaster feature to recipient callouts.
6109 32. Added the $body_zerocount variable, containing the number of binary zero
6110 bytes in the message body.
6112 33. The time of last modification of the "new" subdirectory is now used as the
6113 "mailbox time last read" when there is a quota error for a maildir
6116 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6118 35. Added +ignore_unknown as a special item in host lists.
6120 36. Code for decoding IPv6 addresses in host lists is now included, even if
6121 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6122 address was recognized as an IP address, but was then not correctly decoded
6123 into binary, causing unexpected and incorrect effects when compared with
6130 1. Very minor rewording of debugging text in manualroute to say "list of
6131 hosts" instead of "hostlist".
6133 2. If verify=header_syntax was set, and a header line with an unqualified
6134 address (no domain) and a large number of spaces between the end of the
6135 name and the colon was received, the reception process suffered a buffer
6136 overflow, and (when I tested it) crashed. This was caused by some obsolete
6137 code that should have been removed. The fix is to remove it!
6139 3. When running in the test harness, delay a bit after writing a bounce
6140 message to get a bit more predictability in the log output.
6142 4. Added a call to search_tidyup() just before forking a reception process. In
6143 theory, someone could use a lookup in the expansion of smtp_accept_max_
6144 per_host which, without the tidyup, could leave open a database connection.
6146 5. Added the variables $recipient_data and $sender_data which get set from a
6147 lookup success in an ACL "recipients" or "senders" condition, or a router
6148 "senders" option, similar to $domain_data and $local_part_data.
6150 6. Moved the writing of debug_print from before to after the "senders" test
6153 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6154 problems for message scanning, either using a data ACL, or using
6155 local_scan() because the Received: header was not generated till after they
6156 were called (in order to set the time as the time of reception completion).
6157 I have revised the way this works. The header is now generated after the
6158 body is received, but before the ACL or local_scan() are called. After they
6159 are run, the timestamp in the header is updated.
6165 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6166 before starting a queue runner without re-exec. This happened only when
6167 deliver_drop_privilege was set or when the Exim user was set to root. The
6168 effect of the bug was that timeouts during subsequent deliveries caused
6169 crashes instead of being properly handled. The handler is now left at its
6170 default (and expected) setting.
6172 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6173 message, again when deliver_drop_privilege is set or Exim is run as root.
6174 The bug described in (1) was not present in this case, but the tidying up
6175 of the other signals was missing. I have made the two cases consistent.
6177 3. The ignore_target_hosts setting on a manualroute router was being ignored
6178 for hosts that were looked up using the /MX notation.
6180 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6183 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6184 operated on the sender address. After changing the $sender_address to <>
6185 for the sender address verify, Exim was re-instated it as the original
6186 (before rewriting) address, but remembering that it had rewritten it, so it
6187 wasn't rewriting it again. This bug also had the effect of breaking the
6188 sender address verification caching when the sender address was rewritten.
6190 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6191 This has been changed so that if the ip literal address matches
6192 ignore_target_hosts, the router declines.
6194 7. Added expansion conditions match_domain, match_address, and match_local_
6195 part (NOT match_host).
6197 8. The placeholder for the Received: header didn't have a length field set.
6199 9. Added code to Exim itself and to exim_lock to test for a specific race
6200 condition that could lead to file corruption when using MBX delivery. The
6201 issue is with the lockfile that is created in /tmp. If this file is removed
6202 after a process has opened it but before that process has acquired a lock,
6203 there is the potential for a second process to recreate the file and also
6204 acquire a lock. This could lead to two Exim processes writing to the file
6205 at the same time. The added code performs the same test as UW imapd; it
6206 checks after acquiring the lock that its file descriptor still refers to
6207 the same named file.
6209 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6210 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6213 11. Added the smtp_active_hostname option. If used, this will typically be made
6214 to depend on the incoming interface address. Because $interface_address is
6215 not set up until the daemon has forked a reception process, error responses
6216 that can happen earlier (such as "too many connections") no longer contain
6219 12. If an expansion in a condition on a "warn" statement fails because a lookup
6220 defers, the "warn" statement is abandoned, and the next ACL statement is
6221 processed. Previously this caused the whole ACL to be aborted.
6223 13. Added the iplsearch lookup type.
6225 14. Added ident_timeout as a log selector.
6227 15. Added tls_certificate_verified as a log selector.
6229 16. Added a global option tls_require_ciphers (compare the smtp transport
6230 option of the same name). This controls incoming TLS connections.
6232 17. I finally figured out how to make tls_require_ciphers do a similar thing
6233 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6234 before starting the TLS session.
6236 18. Tabs are now shown as \t in -bP output.
6238 19. If the log selector return_path_on_delivery was set, Exim crashed when
6239 bouncing a message because it had too many Received: header lines.
6241 20. If two routers both had headers_remove settings, and the first one included
6242 a superfluous trailing colon, the final name in the first list and the
6243 first name in the second list were incorrectly joined into one item (with a
6244 colon in the middle).
6250 1. Added -C and -D options to the exinext utility, mainly to make it easier
6251 to include in the automated testing, but these could be helpful when
6252 multiple configurations are in use.
6254 2. The exinext utility was not formatting the output nicely when there was
6255 an alternate port involved in the retry record key, nor when there was a
6256 message id as well (for retries that were specific to a specific message
6257 and a specific host). It was also confused by IPv6 addresses, because of
6258 the additional colons they contain. I have fixed the IPv4 problem, and
6259 patched it up to do a reasonable job for IPv6.
6261 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6262 delivery, the log line now contains "pipelined" if PIPELINING was used.
6264 4. An SMTP transport process used to panic and die if the bind() call to set
6265 an explicit outgoing interface failed. This has been changed; it is now
6266 treated in the same way as a connect() failure.
6268 5. A reference to $sender_host_name in the part of a conditional expansion
6269 that was being skipped was still causing a DNS lookup. This no longer
6272 6. The def: expansion condition was not recognizing references to header lines
6273 that used bh_ and bheader_.
6275 7. Added the _cache feature to named lists.
6277 8. The code for checking quota_filecount in the appendfile transport was
6278 allowing one more file than it should have been.
6280 9. For compatibility with Sendmail, the command line option
6288 and sets the incoming protocol and host name (for trusted callers). The
6289 host name and its colon can be omitted when only the protocol is to be set.
6290 Note the Exim already has two private options, -pd and -ps, that refer to
6291 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6292 "s", but I don't think that's a major issue.
6294 10. A number of refactoring changes to the code, none of which should affect
6297 (a) The number of logging options was getting close to filling up the
6298 32-bit word that was used as a bit map. I have split them into two classes:
6299 those that are passed in the argument to log_write(), and those that are
6300 only ever tested independently outside of that function. These are now in
6301 separate 32-bit words, so there is plenty of room for expansion again.
6302 There is no change in the user interface or the logging behaviour.
6304 (b) When building, for example, log lines, the code previously used a
6305 macro that called string_cat() twice, in order to add two strings. This is
6306 not really sufficiently general. Furthermore, there was one instance where
6307 it was actually wrong because one of the argument was used twice, and in
6308 one call a function was used. (As it happened, calling the function twice
6309 did not affect the overall behaviour.) The macro has been replaced by a
6310 function that can join an arbitrary number of extra strings onto a growing
6313 (c) The code for expansion conditions now uses a table and a binary chop
6314 instead of a serial search (which was left over from when there were very
6315 few conditions). Also, it now recognizes conditions like "pam" even when
6316 the relevant support is not compiled in: a suitably worded error message is
6317 given if an attempt is made to use such a condition.
6319 11. Added ${time_interval:xxxxx}.
6321 12. A bug was causing one of the ddress fields not to be passed back correctly
6322 from remote delivery subprocesses. The field in question was not being
6323 subsequently used, so this caused to problems in practice.
6325 13. Added new log selectors queue_time and deliver_time.
6327 14. Might have fixed a bug in maildirsizefile handling that threw up
6328 "unexpected character" debug warnings, and recalculated the data
6329 unnecessarily. In any case, I expanded the warning message to give more
6332 15. Added the message "Restricted characters in address" to the statements in
6333 the default ACL that block characters like @ and % in local parts.
6335 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6336 Three changes have been made:
6338 (a) There was a serious bug; a negative response to MAIL caused the whole
6339 recipient domain to be cached as invalid, thereby blocking all messages
6340 to all local parts at the same domain, from all senders. This bug has
6341 been fixed. The domain is no longer cached after a negative response to
6342 MAIL if the sender used is not empty.
6344 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6347 (c) A new callout option, "use_sender" has been added for people who want
6348 the modified behaviour.
6354 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6357 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6358 indeed breaks things for older releases.
6360 3. Added additional logging to the case where there is a problem reading data
6361 from a filter that is running in a subprocess using a pipe, in order to
6362 try to track down a specific problem.
6364 4. Testing facility fudge: when running in the test harness and attempting
6365 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6366 getting "No route to host". Convert this to a timeout.
6368 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6371 6. Some OS don't have socklen_t but use size_t instead. This affects the
6372 fifth argument of getsockopt() amongst other things. This is now
6373 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6374 can be set for individual OS. I have set it for SunOS5, OSF1, and
6375 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6376 some earlier ones do not.
6378 7. Change 4.30/15 was not doing the test caselessly.
6380 8. The standard form for an IPv6 address literal was being rejected by address
6381 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6382 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6383 this, as well as the form without the "IPv6" on the front (but only when
6384 address literals are enabled, of course).
6386 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6388 10. Exim crashed if a message with an empty sender address specified by -f
6389 encountered a router with an errors_to setting. This could be provoked only
6390 by a command such as
6394 where an empty string was supplied; "<>" did not hit this bug.
6396 11. Installed PCRE release 4.5.
6398 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6399 remained set. It is now erased.
6401 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6402 times from message ids (which are base 36 rather than the normal 62).
6404 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6405 were being counted as actual protocol errors, and logged if the log
6406 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6407 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6408 and DATA following a set of rejected RCPTs do not count as protocol errors.
6409 In other words, Exim assumes they were pipelined, though this may not
6410 actually be the case. Of course, in all cases the client gets an
6411 appropriate error code.
6413 15. If a lookup fails in an ACL condition, a message about the failure may
6414 be available; it is used if testing the ACL cannot continue, because most
6415 such messages specify what the cause of the deferral is. However, some
6416 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6417 that caused an old message to be retained and used if a later statement
6418 caused a defer, replacing the real cause of the deferral.
6420 16. If an IP address had so many PTR records that the DNS lookup buffer
6421 was not large enough to hold them, Exim could crash while trying to process
6422 the truncated data. It now detects and logs this case.
6424 17. Further to 4.21/58, another change has been made: if (and only if) the
6425 first line of a message (the first header line) ends with CRLF, a bare LF
6426 in a subsequent header line has a space inserted after it, so as not to
6427 terminate the header.
6429 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6430 unnecessarily, used atoi() instead of strtol(), and didn't check the
6431 termination when getting file sizes from file names by regex.
6433 19. Completely re-implemented the support for maildirsize files, in the light
6434 of a number of problems with the previous contributed implementation
6435 (4.30/29). In particular:
6437 . If the quota is zero, the maildirsize file is maintained, but no quota is
6440 . If the maildir directory does not exist, it is created before any attempt
6441 to write a maildirsize file.
6443 . The quota value in the file is just a cache; if the quota is changed in
6444 the transport, the new value overrides.
6446 . A regular expression is available for excluding directories from the
6449 20. The autoreply transport checks the characters in options that define the
6450 message's headers; it allows continued headers, but it was checking with
6451 isspace() after an embedded newline instead of explicitly looking for a
6454 21. If all the "regular" hosts to which an address was routed had passed their
6455 expiry times, and had not reached their retry times, the address was
6456 bounced, even if fallback hosts were defined. Now Exim should go on to try
6459 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6460 equivalent code in the SMTP transport. Some hosts send humungous responses
6461 to HELO/EHLO, more than 1024 it seems.
6463 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6464 gives compiler warnings in some environments. I've now done it "properly",
6467 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6468 (because of problems with the built-in one) was declared to return uschar *
6469 instead of char *, causing compiler failure.
6471 25. Fixed a file descriptor leak when processing alias/forward files.
6473 26. Fixed a minor format string issue in dbfn.c.
6475 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6477 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6478 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6481 29. When a sender address is verified, it is cached, to save repeating the test
6482 when there is more than one recipient in a message. However, when the
6483 verification involves a callout, it is possible for different callout
6484 options to be set for different recipients. It is too complicated to keep
6485 track of this in the cache, so now Exim always runs a verification when a
6486 callout is required, relying on the callout cache for the optimization.
6487 The overhead is duplication of the address routing, but this should not be
6490 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6491 to be verified with callout=postmaster, and the main callout worked but the
6492 postmaster check failed, the verification correctly failed. However, if a
6493 subsequent RCPT command asked for sender verification *without* the
6494 postmaster check, incorrect caching caused this verification also to fail,
6497 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6498 it was not caching the DNS options (qualify_single, search_parents) that
6499 were used when the lookup failed. A subsequent lookup with different
6500 options therefore always gave the same answer, though there were cases
6501 where it should not have. (Example: a "domains = !$mx_any" option on a
6502 dnslookup router: the "domains" option is always processed without any
6503 widening, but the router might have qualify_single set.) Now Exim uses the
6504 cached value only when the same options are set.
6506 32. Added John Jetmore's "exipick" utility to the distribution.
6508 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6509 than a timeout (e.g. a certificate is required, and is not provided), an
6510 Exim server now closes the connection immediately. Previously it waited for
6511 the client to close - but if the client is SSL, it seems that they each
6512 wait for each other, leading to a delay before one of them times out.
6514 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6515 maintained 0.8.x compatibility because I don't think many are using it, and
6516 it is clearly obsolete.
6518 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6521 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6522 client certificate was expired. A simple patch fixes this, though I don't
6523 understand the full logic of why the verify callback is called multiple
6526 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6527 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6528 which causes problems with some clients (such as the Certicom SSL Plus
6529 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6530 disables the coutermeasure allowing Eudora to connect."
6532 38. Exim was not checking that a write() to a log file succeeded. This could
6533 lead to Bad Things if a log got too big, in particular if it hit a file
6534 size limit. Exim now panics and dies if it cannot write to a log file, just
6535 as it does if it cannot open a log file.
6537 39. Modified OS/Makefile-Linux so that it now contains
6539 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6541 The two -D definitions ensure that Exim is compiled with large file
6542 support, which makes it possible to handle log files that are bigger than
6545 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6546 instance) a domain was checked against a named list that involved a lookup,
6547 causing $domain_data to be set, then another domain was checked against the
6548 same list, then the first domain was re-checked, the value of $domain_data
6549 after the final check could be wrong. In particular, if the second check
6550 failed, it could be set empty. This bug probably also applied to
6553 41. The strip_trailing_dot option was not being applied to the address given
6554 with the -f command-line option.
6556 42. The code for reading a message's header from the spool was incrementing
6557 $received_count, but never initializing it. This meant that the value was
6558 incorrect (doubled) while delivering a message in the same process in which
6559 it was received. In the most common configuration of Exim, this never
6560 happens - a fresh exec is done - but it can happen when
6561 deliver_drop_privilege is set.
6563 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6564 it now includes up to 150 characters of the unexpected data in the log
6567 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6568 and building data strings. The size of both of these buffers was 10 000
6569 bytes - far larger than anybody would *ever* want, thought I. Needless to
6570 say, somebody hit the limit. I have increased the maximum line length to
6571 20 000 and the maximum data length of concatenated lines to 100 000. I have
6572 also fixed two bugs, because there was no checking on these buffers. Tsk,
6573 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6574 buffer is too small.
6576 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6577 lsearch lookups. Now it does.
6579 46. When parsing a route_list item in a manualroute router, a fixed-length
6580 buffer was used for the list of hosts. I made this 1024 bytes long,
6581 thinking that nobody would ever have a list of hosts that long. Wrong.
6582 Somebody had a whole pile of complicated expansion conditions, and the
6583 string was silently truncated, leading to an expansion error. It turns out
6584 that it is easier to change to an unlimited length (owing to other changes
6585 that have happened since this code was originally written) than to build
6586 structure for giving a limitation error. The length of the item that
6587 expands into the list of hosts is now unlimited.
6589 47. The lsearch lookup could not handle data where the length of text line was
6590 more than 4095 characters. Such lines were truncated, leading to shortened
6591 data being returned. It should now handle lines of any length.
6593 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6594 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6597 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6598 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6599 no longer happens in comment lines. A list of replacements is now placed
6600 at the head of all of the source files, except those whose only change is
6601 to replace PERL_COMMAND in the very first #! line.
6603 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6604 messages on the queue, with a bottom-up merge sort, using code contributed
6605 by Michael Haardt. This should make operations like -bp somewhat faster on
6606 large queues. It won't affect queue runners, except when queue_run_in_order
6609 51. Installed eximstats 1.31 in the distribution.
6611 52. Added support for SRV lookups to the dnslookup router.
6613 53. If an ACL referred to $message_body or $message_body_end, the value was not
6614 reset for any messages that followed in the same SMTP session.
6616 54. The store-handling optimization for building very long strings was not
6617 differentiating between the different store pools. I don't think this
6618 actually made any difference in practice, but I've tidied it.
6620 55. While running the routers to verify a sender address, $sender_address
6621 was still set to the sender address. This is wrong, because when routing to
6622 send a bounce to the sender, it would be empty. Therefore, I have changed
6623 it so that, while verifying a sender address, $sender_address is set to <>.
6624 (There is no change to what happens when verifying a recipient address.)
6626 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6627 target A or AAAA records (if not already returned) without resetting the
6628 qualify_single or search_parents options of the DNS resolver. These are
6629 inappropriate in this case because the targets of MX and SRV records must
6630 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6631 target that, when qualified, matched something in the local domain. These
6632 two options are now turned off when doing these lookups.
6634 57. It seems that at least some releases of Reiserfs (which does not have the
6635 concept of a fixed number of inodes) returns zero and not -1 for the
6636 number of available inodes. This interacted badly with check_spool_inodes,
6637 which assumed that -1 was the "no such thing" setting. What I have done is
6638 to check that the total number of inodes is greater than zero before doing
6639 the test of how many are available.
6641 58. When a "warn" ACL statement has a log_message modifier, the message is
6642 remembered, and not repeated. This is to avoid a lot of repetition when a
6643 message has many recipients that cause the same warning to be written.
6644 However, Exim was preserving the list of already written lines for an
6645 entire SMTP session, which doesn't seem right. The memory is now reset if a
6646 new message is started.
6648 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6649 debugging output unless log_rewrite was also set.
6651 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6652 of (int)(handle) when we know that handle contains (void *)(-1).
6654 61. The Exim daemon panic-logs an error return when it closes the incoming
6655 connection. However "connection reset by peer" seems to be common, and
6656 isn't really an error worthy of noting specially, so that particular error
6659 62. When Exim is trying to find all the local interfaces, it used to panic and
6660 die if the ioctl to get the interface flags failed. However, it seems that
6661 on at least one OS (Solaris 9) it is possible to have an interface that is
6662 included in the list of interfaces, but for which you get a failure error
6663 for this call. This happens when the interface is not "plumbed" into a
6664 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6665 failure of the "get flags" call assumes that the interface is down.
6667 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6668 makes life easier for people who are doing arithmetic on fields extracted
6669 from dates, where you often get leading zeros that should not be
6670 interpreted as octal.
6672 64. Added qualify_domain to the redirect router, to override the global
6675 65. If a pathologically long header line contained very many addresses (the
6676 report of this problem mentioned 10 000) and each of them was rewritten,
6677 Exim could use up a very large amount of memory. (It kept on making new
6678 copies of the header line as it rewrote, and never released the old ones.)
6679 At the expense of a bit more processing, the header rewriting function has
6680 been changed so that it no longer eats memory in this way.
6682 66. The generation of the Received: header has been moved from the time that a
6683 message starts to be received, to the time that it finishes. The timestamp
6684 in the Received: header should now be very close to that of the <= log
6685 line. There are two side-effects of this change:
6687 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6688 logged header lines no longer include the local Received: line, because
6689 it has not yet been created. The same applies to a copy of the message
6690 that is returned to a non-SMTP sender when a message is rejected.
6692 (b) When a filter file is tested using -bf, no additional Received: header
6693 is added to the test message. After some thought, I decided that this
6696 This change does not affect the value of $received_for. It is still set
6697 after address rewriting, but before local_scan() is called.
6699 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6701 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6702 gave an unhelpful panic error message, and a defer error. I have managed to
6703 change this behaviour so that it now rejects any supplied certificate,
6704 which seems right, as the list of acceptable certificates is empty.
6706 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6707 gave an unhelpful defer error. I have not managed to make this reject any
6708 supplied certificates, but the error message it gives is "no certificate
6709 supplied", which is not helpful.
6711 70. exigrep's output now also includes lines that are not associated with any
6712 message, but which match the given pattern. Implemented by a patch from
6713 Martin Sluka, which also tidied up the Perl a bit.
6715 71. Recipient callout verification, like sender verification, was using <> in
6716 the MAIL FROM command. This isn't really the right thing, since the actual
6717 sender may affect whether the remote host accepts the recipient or not. I
6718 have changed it to use the actual sender in the callout; this means that
6719 the cache record is now keyed on a recipient/sender pair, not just the
6720 recipient address. There doesn't seem to be a real danger of callout loops,
6721 since a callout by the remote host to check the sender would use <>.
6722 [SEE ABOVE: changed after hitting problems.]
6724 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6725 temporary errors. However, in the case of such a code being given after
6726 the end of a data transmission (i.e. after ".") Exim was failing to write
6727 a retry record for the message. (Yes, there was some broken host that was
6728 actually sending 8xx at this point.)
6730 73. An unknown lookup type in a host list could cause Exim to panic-die when
6731 the list was checked. (An example that provoked this was putting <; in the
6732 middle of a list instead of at the start.) If this happened during a DATA
6733 ACL check, a -D file could be left lying around. This kind of configuration
6734 error no longer causes Exim to die; instead it causes a defer error. The
6735 incident is still logged to the main and panic logs.
6737 74. Buglet left over from Exim 3 conversion. The message "too many messages
6738 in one connection" was written to the rejectlog but not the mainlog, except
6739 when address rewriting (yes!) was being logged.
6741 75. Added write_rejectlog option.
6743 76. When a system filter was run not as root (that is, when system_filter_user
6744 was set), the values of the $n variables were not being returned to the
6745 main process; thus, they were not subsequently available in the $sn
6748 77. Added +return_path_on_delivery log selector.
6750 78. A connection timeout was being treated differently from recipients deferred
6751 when testing hosts_max_try with a message that was older than the host's
6752 retry timeout. (The host should not be counted, thus allowing all hosts to
6753 be tried at least once before bouncing.) This may have been the cause of an
6754 occasionally reported bug whereby a message would remain on the queue
6755 longer than the retry timeout, but would be bounced if a delivery was
6756 forced. I say "may" because I never totally pinned down the problem;
6757 setting up timeout/retry tests is difficult. See also the next item.
6759 79. The ultimate address timeout was not being applied to errors that involved
6760 a combination of host plus message (for example, a timeout on a MAIL
6761 command). When an address resolved to a number of possible hosts, and they
6762 were not all tried for each delivery (e.g. because of hosts_max_try), a
6763 message could remain on the queue longer than the retry timeout.
6765 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6768 81. Fixed an obscure SMTP outgoing bug which required at least the following
6769 conditions: (a) there was another message waiting for the same server;
6770 (b) the server returned 5xx to all RCPT commands in the first message so
6771 that the message was not completed; (c) the server dropped the connection
6772 or gave a negative response to the RSET that Exim sends to abort the
6773 transaction. The observed case was a dropped connection after DATA that had
6774 been sent in pipelining mode. That is, the server had advertised PIPELINING
6775 but was not implementing it correctly. The effect of the bug was incorrect
6776 behaviour, such as trying another host, and this could lead to a crash.
6782 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6783 and daemon.c were passed as pointers to ints; they should have been
6784 pointers to socklen_t variables (which are typically unsigned ints).
6786 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6789 3. Fixed a really odd bug that affected only the testing scheme; patching a
6790 certain fixed string in the binary changed the value of another string that
6791 happened to be identical to the end of the original first string.
6793 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6794 name", it returns that address as the IP address. On some operating
6795 systems (e.g. Solaris), it also passes back the IP address string as the
6796 "host name". However, on others (e.g. Linux), it passes back an empty
6797 string. Exim wasn't checking for this, and was changing the host name to an
6798 empty string, assuming it had been canonicalized.
6800 5. Although rare, it is permitted to have more than one PTR record for a given
6801 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6802 all the names associated with an address, because they do in Solaris.
6803 However, it seems that they do not in Linux for data that comes from the
6804 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6805 I found this out when I moved to a new Linux workstation and tried to run
6806 the Exim test suite.
6808 To get round this problem I have changed the code so that it now does its
6809 own call to the DNS to look up PTR records when searching for a host name.
6810 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6811 addresses that are only in /etc/hosts are still found.
6813 This behaviour is, however, controlled by an option called host_lookup_
6814 order, which defaults to "bydns:byaddr". If people want to use the other
6815 order, or indeed, just use one or the other means of lookup, they can
6816 specify it in this variable.
6818 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6819 some operating systems, this comes back from gethostbyaddr() as an empty
6820 string, and this is what Exim used to test for. However, it seems that in
6821 other systems, "." is yielded. Exim now tests for this case too.
6823 7. The values of check_spool_space and check_log_space are now held internally
6824 as a number of kilobytes instead of an absolute number of bytes. If a
6825 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6826 kilobyte. This means that much larger values can be stored.
6828 8. Exim monitor: an attempt to get the action menu when not actually pointing
6829 at a message produces an empty menu entitled "No message selected". This
6830 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6831 no entries in it ("Shell widget menu has zero width and/or height"). So I
6832 have added a single, blank menu entry in this case.
6834 9. Added ${quote_local_part.
6836 10. MIME decoding is now applied to the contents of Subject: header lines when
6839 11. Now that a reference to $sender_host_address automatically causes a reverse
6840 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6841 host lookup before query-style lookups in lists that might use this
6842 variable. This has therefore been abolished, and the "net-" prefix is no
6843 longer necessary for query-style lookups.
6845 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6846 have been a typo for LFLAGS, so it has been changed.
6848 13. The install script calls Exim with "-C /dev/null" in order to find the
6849 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6850 to be output. However, since Exim outputs its version number before the
6851 error, it didn't break the script. It just looked ugly. I fixed this by
6852 always allowing "-C /dev/null" if the caller is root.
6854 14. Ignore overlarge ACL variable number when reading spool file - insurance
6855 against a later release with more variables having written the file.
6857 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6858 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6859 this, as well as the form without the "IPv6" on the front.
6861 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6862 OS/Makefile-Darwin file.
6864 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6865 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6867 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6868 variable. However, some OS use ints for inodes. I've added cast to long int
6869 to get rid of the compiler warning.
6871 19. I had forgotten to lock out "/../" in configuration file names when
6872 ALT_CONFIG_PREFIX was set.
6874 20. Routers used for verification do not need to specify transports. However,
6875 if such a router generated a host list, and callout was configured, Exim
6876 crashed, because it could not find a port number from the (non-existent)
6877 transport. It now assumes port 25 in this circumstance.
6879 21. Added the -t option to exigrep.
6881 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6882 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6883 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6884 them is set, LOOKUP_LSEARCH is forced.
6886 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6888 24. Added sender and host information to the "rejected by local_scan()" log
6889 line; previously there was no indication of these.
6891 25. Added .include_if_exists.
6893 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6894 Linux. It turns out that not all file systems support this. Apparently some
6895 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6896 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6897 not supported on the file descriptor, is now ignored when Exim is trying to
6898 sync a directory. This applies only to Linux.
6900 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6902 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6903 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6906 29. Added support for maildirsize files from supplied patch (modified a bit).
6908 30. The use of :fail: followed by an empty string could lead Exim to respond to
6909 sender verification failures with (e.g.):
6911 550 Verification failed for <xxx>
6912 550 Sender verify failed
6914 where the first response line was missing the '-' that indicates it is not
6915 the final line of the response.
6917 31. The loop for finding the name of the user that called Exim had a hardwired
6918 limit of 10; it now uses the value of finduser_retries, which is used for
6919 all other user lookups.
6921 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6924 33. Exim was neglecting to zero errno before one call of strtol() when
6925 expanding a string and expecting an integer value. On some systems this
6926 resulted in spurious "integer overflow" errors. Also, it was casting the
6927 result into an int without checking.
6929 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6930 did not work. The code looks as if it has *never* worked, though it appears
6931 to have been documented since at least release 1.62. I have made it work.
6933 35. The "timeout_DNS" error in retry rules, also documented since at least
6934 1.62, also never worked. As it isn't clear exactly what this means, and
6935 clearly it isn't a major issue, I have abolished the feature by treating it
6936 as "timeout", and writing a warning to the main and panic logs.
6938 36. The display of retry rules for -brt wasn't always showing the error code
6941 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6942 timeout_connect_A, timeout_connect_MX.
6944 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6945 to the empty sender.
6947 39. The daemon was not analysing the content of -oX till after it had closed
6948 stderr and disconnected from the controlling terminal. This meant that any
6949 syntax errors were only noted on the panic log, and the return code from
6950 the command was 0. By re-arranging the code a little, I've made the
6951 decoding happen first, so such errors now appear on stderr, and the return
6952 code is 1. However, the actual setting up of the sockets still happens in
6953 the disconnected process, so errors there are still only recorded on the
6956 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6957 connections (as happens on some IP stacks) was logged at start up time as
6958 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6959 it from "IPv6 and IPv4", which means that two separate sockets are being
6962 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6963 says whether AF_INET or AF_INET6 was passed as an argument.
6965 42. Exiwhat output was messed up when time zones were included in log
6968 43. Exiwhat now gives more information about the daemon's listening ports,
6969 and whether -tls-on-connect was used.
6971 44. The "port" option of the smtp transport is now expanded.
6973 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6974 silently ignored. Now an error message is written to the main and panic
6977 46. There's a new ACL modifier called "logwrite" which writes to a log file
6978 as soon as it is encountered.
6980 47. Added $local_user_uid and $local_user_gid at routing time.
6982 48. Exim crashed when trying to verify a sender address that was being
6985 49. Exim was recognizing only a space character after ".include". It now also
6986 recognizes a tab character.
6988 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6989 extracting the relevant information from the specification. The man page no
6990 longer contains scrambled data for the -d option, and I've added a section
6991 at the front about calling Exim under different names.
6993 51. Added "extra_headers" argument to the "mail" command in filter files.
6995 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6998 53. Installed eximstats 1.29.
7000 54. Added transport_filter_timeout as a generic transport option.
7002 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7003 Cc: header lines. This was required by RFC 822, but it not required by RFC
7006 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7007 incoming messages that did not have them. Now it does so only if the
7008 message originates locally, that is, if there is no associated remote host
7009 address. When Resent- header lines are present, this applies to the Resent-
7010 lines rather than the non-Resent- lines.
7012 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7013 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7015 58. Messages for configuration errors now include the name of the main
7016 configuration file - useful now that there may be more than one file in a
7017 list (.included file names were always shown).
7019 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7020 for those rare installations that do not start the daemon as root or run it
7021 setuid root. I've cut out the call to initgroups() if the daemon is not
7024 60. The Exim user and group can now be bound into the binary as text strings
7025 that are looked up at the start of Exim's processing.
7027 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7029 62. Added $mailstore_basename variable.
7031 63. Installed patch to sieve.c from Michael Haardt.
7033 64. When Exim failed to open the panic log after failing to open the main log,
7034 the original message it was trying to log was written to stderr and debug
7035 output, but if they were not available (the usual case in production), it
7036 was lost. Now it is written to syslog before the two lines that record the
7037 failures to open the logs.
7039 65. Users' Exim filters run in subprocesses under the user's uid. It is
7040 possible for a "deliver" command or an alias in a "personal" command to
7041 provoke an address rewrite. If logging of address rewriting is configured,
7042 this fails because the process is not running as root or exim. There may be
7043 a better way of dealing with this, but for the moment (because 4.30 needs
7044 to be released), I have disabled address rewrite logging when running a
7045 filter in a non-root, non-exim process.
7051 1. The buildconfig auxiliary program wasn't quoting the value set for
7052 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7053 not defined. This bug was masked in 4.22 by the effect that was fixed in
7056 2. Some messages that were rejected after a message id was allocated were
7057 shown as "incomplete" by exigrep. It no longer does this for messages that
7058 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7060 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7061 have allow_domain_literals set, the ID did not get logged in the <= line.
7062 Domain literals are now always recognized in Message-ID: header lines.
7064 4. The first argument for a ${extract expansion item is the key name or field
7065 number. Leading and trailing spaces in this item were not being ignored,
7066 causing some misleading effects.
7068 5. When deliver_drop_privilege was set, single queue runner processes started
7069 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7070 same command in the process it spins off) were not dropping privilege.
7072 6. When the daemon running as "exim" started a queue runner, it always
7073 re-executed Exim in the spun-off process. This is a waste of effort when
7074 deliver_drop_privilege is set. The new process now just calls the
7075 queue-runner function directly.
7081 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7084 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7085 ignored. Though the use of -oP was forcing the writing of a pid file, it
7086 was always written to the default place.
7088 3. If the message "no IP address found for host xxxx" is generated during
7089 incoming verification, it is now followed by identification of the incoming
7090 connection (so you can more easily find what provoked it).
7092 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7094 5. Added some features to "harden" Exim a bit more against certain attacks:
7096 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7097 be put in Local/Makefile. This is like the never_users runtime option,
7098 but it cannot be overridden. The default setting is "root".
7100 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7101 prefix string with which any file named in a -C command line option
7104 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7105 is retained for -C and -D only if the caller of Exim is root. Without
7106 it, the exim user may also use -C and -D and retain privilege.
7108 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7109 command line option is disabled.
7111 6. Macro names set by the -D option must start with an upper case letter, just
7112 like macro names defined in the configuration file.
7114 7. Added "dereference=" facility to LDAP.
7116 8. Two instances of the typo "uknown" in the source files are fixed.
7118 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7119 the Configure-Makefile script screwed up while processing it.
7121 10. Incorporated PCRE 4.4.
7123 11. The SMTP synchronization check was not operating right at the start of an
7124 SMTP session. For example, it could not catch a HELO sent before the client
7125 waited for the greeting. There is now a check for outstanding input at the
7126 point when the greeting is written. Because of the duplex, asynchronous
7127 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7128 way, but not yet received, when the check is performed.
7130 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7131 on TCP/IP sockets, because this apparently causes some broken clients to
7134 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7135 unchanged) from the Cygwin maintainer.
7137 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7138 was supported instead of testing for mbx. Effectively a typo.
7140 15. The spa authenticator server code was not checking that the input it
7141 received was valid base64.
7143 16. The debug output line for the "set" modifier in ACLs was not showing the
7144 name of the variable that was being set.
7146 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7148 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7149 lookup on its own. Something else was needed to trigger the lookup. For
7150 example, a match in host_lookup or the need for a host name in a host list.
7151 Now, if $sender_host_name is referenced and the host name has not yet been
7152 looked up, a lookup is performed. If the lookup fails, the variable remains
7153 empty, and $host_lookup_failed is set to "1".
7155 19. Added "eqi" as a case-independent comparison operator.
7157 20. The saslauthd authentication condition could segfault if neither service
7158 nor realm was specified.
7160 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7161 error message that was logged was misleading, and incoming SMTP
7162 connections were dropped. The message is now more accurate, and temporary
7163 errors are given to SMTP connections.
7165 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7166 (except RSET) with a 421 error, until QUIT is received. However, it was
7167 failing to send a response to QUIT.
7169 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7170 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7171 ACL is now run after the helo_try_verify_hosts code.
7173 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7174 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7175 case-independent, so other case variants are also recognized). Apparently
7176 some systems use these upper case variants.
7178 25. If more than two messages were waiting for the same host, and a transport
7179 filter was specified for the transport, Exim sent two messages over the
7180 same TCP/IP connection, and then failed with "socket operation on non-
7181 socket" when it tried to send the third.
7183 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7185 27. The extern definition of crypt16() in expand.c was not being excluded when
7186 the OS had its own crypt16() function.
7188 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7189 as a preferred synonym for return_size_limit, both as an option and as an
7192 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7194 30. Changed the default configuration ACL to relax the local part checking rule
7195 for addresses that are not in any local domains. For these addresses,
7196 slashes and pipe symbols are allowed within local parts, but the sequence
7197 /../ is explicitly forbidden.
7199 31. SPA server authentication was not clearing the challenge buffer before
7202 32. log_message in a "warn" ACL statement was writing to the reject log as
7203 well as to the main log, which contradicts the documentation and doesn't
7204 seem right (because no rejection is happening). So I have stopped it.
7206 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7207 However, I am unable to do any testing of this.
7209 34. Fixed an infelicity in the appendfile transport. When checking directories
7210 for a mailbox, to see if any needed to be created, it was accidentally
7211 using path names with one or more superfluous leading slashes; tracing
7212 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7214 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7215 contents are added to the log line that is written for every discarded
7216 recipient. (Previously a log_message setting was ignored.)
7218 36. The ${quote: operator now quotes the string if it is empty.
7220 37. The install script runs exim in order to find its version number. If for
7221 some reason other than non-existence or emptiness, which it checks, it
7222 could not run './exim', it was installing it with an empty version number,
7223 i.e. as "exim-". This error state is now caught, and the installation is
7226 38. An argument was missing from the function that creates an error message
7227 when Exim fails to connect to the socket for saslauthd authentication.
7228 This could cause Exim to crash, or give a corrupted message.
7230 39. Added isip, isip4, and isip6 to ${if conditions.
7232 40. The ACL variables $acl_xx are now saved with the message, and can be
7233 accessed later in routers, transports, and filters.
7235 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7236 strings in the file are not string-expanded.
7238 42. If a MAIL command specified a SIZE value that was too large to fit into an
7239 int variable, the check against message_size_limit failed. Such values are
7240 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7241 day this will have to be increased, but I don't think I want to be around
7242 when emails are that large.
7249 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7250 iconv() is not standard in FreeBSD.
7252 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7253 IPv6 enabled. The observed symptom was a segmentation fault on return from
7254 the function os_common_find_running_interfaces() in src/os.c.
7256 3. In the check_special_case() function in daemon.c I had used "errno" as an
7257 argument name, which causes warnings on some systems. This was basically a
7258 typo, since it was named "eno" in the comments!
7260 4. The code that waits for the clock to tick (at a resolution of some fraction
7261 of a second) so as to ensure message-id uniqueness was always waiting for
7262 at least one whole tick, when it could have waited for less. [This is
7263 almost certainly not relevant at current processor speeds, where it is
7264 unlikely to ever wait at all. But we try to future-proof.]
7266 5. The function that sleeps for a time interval that includes fractions of a
7267 second contained a race. It did not block SIGALRM between setting the
7268 timer, and suspending (a couple of lines later). If the interval was short
7269 and the sigsuspend() was delayed until after it had expired, the suspension
7270 never ended. On busy systems this could lead to processes getting stuck for
7273 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7274 process, before it forks any delivery processes. The open lookup caching
7275 mechanism meant that the open file or database connection was passed into
7276 the delivery process. The problem was that delivery processes always tidy
7277 up cached lookup data. This could cause a problem for the next delivery
7278 process started by the queue runner, because the external queue runner
7279 process does not know about the closure. So the next delivery process
7280 still has data in the lookup cache. In the case of a file lookup, there was
7281 no problem because closing a file descriptor in a subprocess doesn't affect
7282 the parent. However, if the lookup was caching a connection to a database,
7283 the connection was closed, and the second delivery process was likely to
7284 see errors such as "PGSQL: query failed: server closed the connection
7285 unexpectedly". The problem has been fixed by closing all cached lookups
7286 in a queue runner before running a delivery process.
7288 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7289 seem to have the "const" qualifier which it has on other OS. I've
7292 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7293 *of the same type* (client or server) with the same public name that an
7294 error should be diagnosed.
7296 9. When Exim looked up a host name for an IP address, but failed to find the
7297 original IP address when looking up the host name (a safety check), it
7298 output the message "<ip address> does not match any IP for NULL", which was
7299 confusing, to say the least. The bug was that the host name should have
7300 appeared instead of "NULL".
7302 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7303 user that is built into the binary, and the -C or -D options is used, root
7304 privilege is dropped before the configuration file is read. In addition,
7305 logging is switched to stderr instead of the normal log files. If the
7306 configuration then re-defines the Exim user, the unprivileged environment
7307 is probably not what is expected, so Exim logs a panic warning message (but
7310 However, if deliver_drop_privilege is set, the unprivileged state may well
7311 be exactly what is intended, so the warning has been cut out in that case,
7312 and Exim is allowed to try to write to its normal log files.
7318 1. smtp_return_error_details was not giving details for temporary sender
7319 or receiver verification errors.
7321 2. Diagnose a configuration error if two authenticators have the same public
7324 3. Exim used not to create the message log file for a message until the first
7325 delivery attempt. This could be confusing when incoming messages were held
7326 for policy or load reasons. The message log file is now created at the time
7327 the message is received, and an initial "Received" line is written to it.
7329 4. The automatically generated man page for command line options had a minor
7330 bug that caused no ill effects; however, a more serious problem was that
7331 the procedure for building the man page automatically didn't always
7332 operate. Consequently, release 4.20 contains an out-of-date version. This
7333 shouldn't happen again.
7335 5. When building Exim with embedded Perl support, the script that builds the
7336 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7337 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7339 6. The freeze_tell option was not being used for messages that were frozen on
7340 arrival, either by an ACL or by local_scan().
7342 7. Added the smtp_incomplete_transaction log selector.
7344 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7345 was accepting AUTH without a new EHLO.
7347 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7348 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7351 10. Logging of TCP/IP connections (when configured) now happens in the main
7352 daemon process instead of the child process, so that the TCP/IP connection
7353 count is more accurate (but it can never be perfect).
7355 11. The use of "drop" in a nested ACL was not being handled correctly in the
7356 outer ACL. Now, if condition failure induced by the nested "drop" causes
7357 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7358 or "require"), the connection is dropped.
7360 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7361 that yield "discard" can now be used with an "accept" or a "discard" verb,
7362 but an error is generated for any others (because I can't see a useful way
7363 to define what should happen).
7365 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7366 are now processed in the same way as lines in the Exim configuration file.
7367 In particular, continuation lines are supported.
7369 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7371 15. Added -ti meaning -t -i.
7373 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7374 domains, and warn by logging if others are found.
7376 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7377 returned from GIFCONF when Exim is trying to find the list of interfaces on
7378 a host. The code in os.c has been modified to copy each ifreq to an aligned
7379 structure in all cases.
7381 Also, in some cases, the returned ifreq's were being copied to a 'struct
7382 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7383 means the last couple of bytes of an IPv6 address could be chopped if the
7384 ifreq contained only a normal sockaddr (14 bytes storage).
7386 18. Named domain lists were not supported in the hosts_treat_as_local option.
7387 An entry such as +xxxx was not recognized, and was treated as a literal
7390 19. Ensure that header lines added by a DATA ACL are included in the reject log
7391 if the ACL subsequently rejects the message.
7393 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7394 MD5 (which is deprecated).
7396 21. When testing a filter file using -bf, Exim was writing a message when it
7397 took the sender from a "From " line in the message, but it was not doing so
7398 when it took $return_path from a Return-Path: header line. It now does.
7400 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7401 with a valid header line field name (a series of printing characters
7402 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7404 23. Changed "disc" in the source to "disk" to conform to the documentation and
7405 the book and for uniformity.
7407 24. Ignore Sendmail's -Ooption=value command line item.
7409 25. When execve() failed while trying to run a command in a pipe transport,
7410 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7411 could be confused with a return value of 69 from the command itself. This
7412 has been changed to 127, the value the shell returns if it is asked to run
7413 a non-existent command. The wording for the related log line suggests a
7414 non-existent command as the problem.
7416 26. If received_header_text expands to an empty string, do not add a Received:
7417 header line to the message. (Well, it adds a token one on the spool, but
7418 marks it "old" so that it doesn't get used or transmitted.)
7420 27. Installed eximstats 1.28 (addition of -nt option).
7422 28. There was no check for failure on the call to getsockname() in the daemon
7423 code. This can fail if there is a shortage of resources on the system, with
7424 ENOMEM, for example. A temporary error is now given on failure.
7426 29. Contrary to the C standard, it seems that in some environments, the
7427 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7428 program. Exim now does this explicitly; it affects the formatting of
7429 timestamps using strftime().
7431 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7432 complaints. I've now initialized all the variables, to avoid this.
7434 32. Header lines added by a system filter were not being "seen" during
7435 transport-time rewrites.
7437 33. The info_callback() function passed to OpenSSL is set up with type void
7438 (*)(SSL *, int, int), as described somewhere. However, when calling the
7439 function (actually a macro) that sets it up, the type void(*)() is
7440 expected. I've put in a cast to prevent warnings from picky compilers.
7442 34. If a DNS black list lookup found a CNAME record, but there were no A
7443 records associated with the domain it pointed at, Exim crashed.
7445 35. If a DNS black list lookup returned more than one A record, Exim ignored
7446 all but the first. It now scans all returned addresses if a particular IP
7447 value is being sought. In this situation, the contents of the
7448 $dnslist_value variable are a list of all the addresses, separated by a
7451 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7452 to accept a host name and all its aliases if the forward lookup for any of
7453 them yielded the IP address of the incoming connection. Now it accepts only
7454 those names whose forward lookup yields the correct IP address. Any other
7455 names are discarded. This closes a loophole whereby a rogue DNS
7456 administrator could create reverse DNS records to break through a
7457 wildcarded host restriction in an ACL.
7459 37. If a user filter or a system filter that ran in a subprocess used any of
7460 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7461 the wrong values were passed to the pipe command ($thisaddress had the
7462 value of $0, $0 had the value of $1, etc). This bug was introduced by
7463 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7465 38. Improved the line breaking for long SMTP error messages from ACLs.
7466 Previously, if there was no break point between 40 and 75 characters, Exim
7467 left the rest of the message alone. Two changes have been made: (a) I've
7468 reduced the minimum length to 35 characters; (b) if it can't find a break
7469 point between 35 and 75 characters, it looks ahead and uses the first one
7470 that it finds. This may give the occasional overlong line, but at least the
7471 remaining text gets split now.
7473 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7474 file descriptors might be low, and that setting 1000 would always raise it.
7475 It turns out that in some environments, the limit is already over 1000 and
7476 that lowering it causes trouble. So now Exim takes care not to decrease it.
7478 40. When delivering a message, the value of $return_path is set to $sender_
7479 address at the start of routing (routers may change the value). By an
7480 oversight, this default was not being set up when an address was tested by
7481 -bt or -bv, which affected the outcome if any router or filter referred to
7484 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7485 log only when "message" or "log_message" are set. However, if one of the
7486 conditions was an address verification, or a call to a nested ACL, the
7487 messages generated by the underlying test were being passed through. This
7488 no longer happens. The underlying message is available in $acl_verify_
7489 message for both "message" and "log_message" expansions, so it can be
7490 passed through if needed.
7492 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7493 new expansion $bh_ to give the encoded byte string without charset
7494 translation. Translation happens only if iconv() is available; HAVE_ICONV
7495 indicates this at build time. HEADERS_CHARSET gives the charset to
7496 translate to; headers_charset can change it in the configuration, and
7497 "headers charset" can change it in an individual filter file.
7499 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7500 that creates RFC 2047 encoded "words" labels them as that charset instead
7501 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7502 expansion operator; (ii) when Exim creates a From: line for a local
7503 message; (iii) when a header line is rewritten to include a "phrase" part.
7505 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7506 buggy, causing it to skip the first lines of messages whose message ID
7507 ended in 'D'. This would not have bitten before Exim release 4.14, because
7508 message IDs were unlikely to end in 'D' before then. The effect was to have
7509 incorrect size information for certain domains.
7511 45. #include "config.h" was missing at the start of the crypt16.c module. This
7512 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7515 46. If there was a timeout during a "random" callout check, Exim treated it as
7516 a failure of the random address, and carried on sending RSET and the real
7517 address. If the delay was just some slowness somewhere, the response to the
7518 original RCPT would be taken as a response to RSET and so on, causing
7519 mayhem of various kinds.
7521 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7522 when I implemented it. It didn't allow for the fact that some option values
7523 may legitimately be negative (e.g. size_addition), and it didn't even do
7524 the right test for positive values.
7526 48. Domain names in DNS records are case-independent. Exim always looks them up
7527 in lower case. Some resolvers return domain names in exactly the case they
7528 appear in the zone file, that is, they may contain uppercase letters. Not
7529 all resolvers do this - some return always lower case. Exim was treating a
7530 change of case by a resolver as a change of domain, similar to a widening
7531 of a domain abbreviation. This triggered its re-routing code and so it was
7532 trying to route what was effectively the same domain again. This normally
7533 caused routing to fail (because the router wouldn't handle the domain
7534 twice). Now Exim checks for this case specially, and just changes the
7535 casing of the domain that it ultimately uses when it transmits the message
7538 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7541 50. If a filter generated a file delivery with a non-absolute name (possible if
7542 no home directory exists for the router), the forbid_file option was not
7545 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7546 the existing equality matching.
7548 52. Exim was using ints instead of ino_t variables in some places where it was
7549 dealing with inode numbers.
7551 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7552 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7553 called TMPDIR, and if it finds it is different, it changes its value.
7555 54. The smtp_printf() function is now made available to local_scan() so
7556 additional output lines can be written before returning. There is also an
7557 smtp_fflush() function to enable the detection of a dropped connection.
7558 The variables smtp_input and smtp_batched_input are exported to
7561 55. Changed the default runtime configuration: the message "Unknown user"
7562 has been removed from the ACL, and instead placed on the localuser router,
7563 using the cannot_route_message feature. This means that any verification
7564 failures that generate their own messages won't get overridden. Similarly,
7565 the "Unrouteable address" message that was in the ACL for unverifiable
7566 relay addresses has also been removed.
7568 56. Added hosts_avoid_esmtp to the smtp transport.
7570 57. The exicyclog script was not checking for the esoteric option
7571 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7572 will work only if exicyclog is run under the appropriate euid.
7574 58. Following a discussion on the list, the rules by which Exim recognises line
7575 endings on incoming messages have been changed. The -dropcr and drop_cr
7576 options are now no-ops, retained only for backwards compatibility. The
7577 following line terminators are recognized: LF CRLF CR. However, special
7578 processing applies to CR:
7580 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7581 nor a local message in the state where . is a terminator.
7583 (ii) If a bare CR is encountered in a header line, an extra space is added
7584 after the line terminator so as not to end the header. The reasoning
7585 behind this is that bare CRs in header lines are most likely either
7586 to be mistakes, or people trying to play silly games.
7588 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7589 was being incorrectly given as 18 bytes larger than it should have been.
7590 This is a VOB (very old bug).
7592 60. This may never have affected anything current, but just in case it has:
7593 When the local host is found other than at the start of a list of hosts,
7594 the local host, those with the same MX, and any that follow, are discarded.
7595 When the list in question was part of a longer list of hosts, the following
7596 hosts (not currently being processed) were also being discarded. This no
7597 longer happens. I'm not sure if this situation could ever has previously
7600 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7603 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7604 header. This is something that is recommended in a new Internet Draft, and
7605 is something that is documented as being done by Sendmail. There are two
7606 possible values. For messages generated by the autoreply transport, Exim
7609 Auto-Submitted: auto-replied
7611 whereas for all other generated messages (e.g. bounces) it adds
7613 Auto-Submitted: auto-generated
7615 63. The "personal" condition in filters now includes a test for the
7616 Auto-Submitted: header. If it contains the string "auto-" the message it
7617 not considered personal.
7619 64. Added rcpt_include_affixes as a generic transport option.
7621 65. Added queue_only_override (default true).
7623 66. Added the syslog_duplication option.
7625 67. If what should have been the first header line of a message consisted of
7626 a space followed by a colon, Exim was mis-interpreting it as a header line.
7627 It isn't of course - it is syntactically invalid and should therefore be
7628 treated as the start of the message body. The misbehaviour could have
7629 caused a number of strange effects, including loss of data in subsequent
7630 header lines, and spool format errors.
7632 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7633 client host had authenticated. This control can now be exercised by an ACL
7634 for more flexibility.
7636 69. By default, callouts do not happen when testing with -bh. There is now a
7637 variant, -bhc, which does actually run the callout code, including
7638 consulting and updating the callout cache.
7640 70. Added support for saslauthd authentication, courtesy of Alexander
7643 71. If statvfs() failed on the spool or log directories while checking their
7644 size for availability, Exim confusingly gave the error "space shortage".
7645 Furthermore, in debugging mode it crashed with a floating point exception.
7646 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7647 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7648 serious problem, Exim now writes to the main and panic logs when this
7649 happens, with details of the failure. It then refuses to accept the
7650 incoming message, giving the message "spool directory problem" or "log
7651 directory problem" with a 421 code for SMTP messages.
7653 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7654 0, 1, and 2 exist, because some OS complain for execs without them (see
7655 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7656 descriptors. However, the code omitted to check that the open succeeded,
7657 causing mysterious errors if for some reason the permissions on /dev/null
7658 got screwed. Now Exim writes a message to the main and panic logs, and
7659 bombs out if it can't open /dev/null.
7661 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7662 interact so that it is all more flexible. It is supposed to remain
7663 backwards compatible. Also added extra_local_interfaces.
7665 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7666 to bomb out with an assertion failure - to the client this appears as a
7667 connection drop. This problem occurs in the part of the code that was taken
7668 from the Samba project. Fortunately, the assertion is in a very simple
7669 function, so I have fixed this by reproducing the function inline in the
7670 one place where it is called, and arranging for authentication to fail
7671 instead of killing the process with assert().
7673 75. The SPA client code was not working when the server requested OEM rather
7674 than Unicode encoding.
7676 76. Added code to make require_files with a specific uid setting more usable in
7677 the case where statting the file as root fails - usually a non-root-mounted
7678 NFS file system. When this happens and the failure is EACCES, Exim now
7679 forks a subprocess and does the per-uid checking as the relevant uid.
7681 77. Added process_log_path.
7683 78. If log_file_path was not explicitly set, a setting of check_log_space or
7684 check_log_inodes was ignored.
7686 79. If a space check for the spool or log partitions fails, the incident is now
7687 logged. Of course, in the latter case the data may get lost...
7689 80. Added the %p formatting code to string_format() so that it can be used to
7690 print addresses in debug_print(). Adjusted all the address printing in the
7691 debugging in store.c to use %p rather than %d.
7693 81. There was a concern that a line of code in smtp_in.c could overflow a
7694 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7695 initially expressed, the concern was not well-founded, because trailing
7696 spaces are removed early. However, if the trailing spaces were followed by
7697 a NULL, they did not get removed, so the overflow was possible. Two fixes
7700 (a) I re-wrote the offending code in a cleaner fashion.
7701 (b) If an incoming SMTP command contains a NULL character, it is rejected
7704 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7705 runs initgroups(), so that if the Exim user is in any additional groups,
7706 they will be used during message reception.
7712 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob