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.
133 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
134 side, to decode a TLS 1.3 packet capture.
136 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
137 Previously the default library behaviour applied, sending two, each in
140 JH/03 Debug output for ACL now gives the config file name and line number for
143 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
145 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
147 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
148 buffer overrun for (non-chunking) other transports.
150 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
151 TLS1.3, means that a server rejecting a client certificate is not visible
152 to the client until the first read of encrypted data (typically the
153 response to EHLO). Add detection for that case and treat it as a failed
154 TLS connection attempt, so that the normal retry-in-clear can work (if
155 suitably configured).
157 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
158 and/or domain. Found and fixed by Jason Betts.
160 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
161 configuration). If a CNAME target was not a wellformed name pattern, a
164 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
165 the OS reports them interleaved with other addresses.
167 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
168 used both for input and for a verify callout, both encrypted, SMTP
169 responses being sent by the server could be lost. This resulted in
170 dropped connections and sometimes bounces generated by a peer sending
173 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
174 string. Previously it was possible to cause undefined behaviour in a
175 library routine (usually a crash). Found by "zerons".
177 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
180 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
181 API was removed, so update to use the newer ones.
183 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
184 any timeout set, is taking a long time. Previously we would hang on to a
185 rotated logfile "forever" if the input was arriving with long gaps
186 (a previous attempt to fix addressed lack, for a long time, of initial
189 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
190 shared (NFS) environment. The length of the tempfile name is now
191 4 + 16 ("hdr.$message_exim_id") which might break on file
192 systems which restrict the file name length to lower values.
195 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
196 shared (NFS) environment.
198 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
199 did for all versions <4.90). Notably -M, -m, --invert, -I may be
202 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
203 on some platforms for bit 31.
205 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
206 to changes apparently associated with TLS1.3 handling some of the APIs
207 previously used were either nonfunctional or inappropriate. Strings
208 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
209 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
210 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
211 This affects log line X= elements, the $tls_{in,out}_cipher variables,
212 and the use of specific cipher names in the encrypted= ACL condition.
214 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
216 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
217 verification result was not updated unless hosts_require_ocsp applied.
219 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
220 queue_list_requires_admin set to false, non-admin users were denied the
223 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
224 directory-of-certs mode. Previously they were advertised despite the
227 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
228 A single TCP connection by a client will now hold a TLS connection open
229 for multiple message deliveries, by default. Previoud the default was to
232 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
233 default. If built with the facility, DANE will be used. The facility
234 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
236 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
237 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
238 must be defined and you must still, unless you define DISABLE_TLS, manage
239 the the include-dir and library-file requirements that go with that
240 choice. Non-TLS builds are still supported.
242 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
245 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
246 default. If the platform supports and has the facility enabled, it will
247 be requested on all coneections.
249 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
250 controlled by the build-time option SUPPORT_PIPE_CONNECT.
252 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
254 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
255 success-DSN messages. Previously the From: header was always the default
256 one for these; the option was ignored.
258 JH/28 Fix the timeout on smtp response to apply to the whole response.
259 Previously it was reset for every read, so a teergrubing peer sending
260 single bytes within the time limit could extend the connection for a
261 long time. Credit to Qualsys Security Advisory Team for the discovery.
263 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
264 delivery address, which leaked information of the results of local
265 forwarding. Change to the original envelope recipient address, per
268 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
269 requested. Previously not bounce was generated and a log entry of
270 error ignored was made.
272 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
274 JH/32 Introduce a general tainting mechanism for values read from the input
275 channel, and values derived from them. Refuse to expand any tainted
276 values, to catch one form of exploit.
278 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
279 was unused and the unexpanded text used for the test. Found and
280 fixed by Ruben Jenster.
282 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
283 an attempt to use a TLS library read routine dereffed a nul pointer,
286 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
287 them in our smtp response.
289 JH/36 Have the general router option retry_use_local_part default to true when
290 any of the restrictive preconditions are set (to anything). Previously it
291 was only for check_local user. The change removes one item of manual
292 configuration which is required for proper retries when a remote router
293 handles a subset of addresses for a domain.
295 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
296 link count into consideration.
298 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
299 caused the extension of big_buffer, the following lines were ignored.
301 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
302 accordance with RFC 2308. Previously there was no expiry, so a longlived
303 receive process (eg. due to ACL delays) versus a short SOA value could
306 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
308 JH/39 Promote DMARC support to mainline.
310 JH/40 Bug 2452: Add a References: header to DSNs.
312 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
313 parameters. The relevant library call is documented as "Deprecated: This
314 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
315 3.6.0, DH parameters are negotiated following RFC7919."
317 HS/06 Change the default of dnssec_request_domains to "*"
319 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
320 carried on and emitted a BDAT command, even when PIPELINING was not
323 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
324 buffer was used for the filename, resulting in a trap when tainted
325 arguments (eg. $domain) were used.
327 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
328 recommended to avoid a possible server-load attack. The feature can be
329 re-enabled via the openssl_options main cofiguration option.
331 JH/45 local_scan API: documented the current smtp_printf() call. This changed
332 for version 4.90 - adding a "more data" boolean to the arguments.
333 Bumped the ABI version number also, this having been missed previously;
334 release versions 4.90 to 4.92.3 inclusive were effectively broken in
335 respect of usage of smtp_printf() by either local_scan code or libraries
336 accessed via the ${dlfunc } expansion item. Both will need coding
337 adjustment for any calls to smtp_printf() to match the new function
338 signature; a FALSE value for the new argument is always safe.
340 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
341 the file-offset (which the Linux syscall does, and exim expects); this
342 resulted in an indefinite loop.
344 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
345 to do ARC verification. The Authentication-Results: header line added
346 by the configuration then had no ARC item.
352 JH/01 Remove code calling the customisable local_scan function, unless a new
353 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
355 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
356 non-signal-safe functions being used.
358 JH/03 Bug 2269: When presented with a received message having a stupidly large
359 number of DKIM-Signature headers, disable DKIM verification to avoid
360 a resource-consumption attack. The limit is set at twenty.
362 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
363 report of oldest_pass in ${authres } in consequence, and separate out
364 some descriptions of reasons for verification fail.
366 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
367 files in the spool were present and unlocked. A queue-runner could spot
368 them, resulting in a duplicate delivery. Fix that by doing the unlock
369 after the unlink. Investigation by Tim Stewart. Take the opportunity to
370 add more error-checking on spoolfile handling while that code is being
373 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
374 No known attacks, no CVE, this is defensive hardening.
376 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
377 a queue-runner could start a delivery while other operations were ongoing.
378 Cutthrough delivery was a common victim, resulting in duplicate delivery.
379 Found and investigated by Tim Stewart. Fix by using the open message data
380 file handle rather than opening another, and not locally closing it (which
381 releases a lock) for that case, while creating the temporary .eml format
382 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
384 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
385 $sender_verify_failure/$recipient_verify_failure to "random".
387 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
390 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
391 Previously this would segfault.
393 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
396 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
397 like zero, since the resolver should be doing this for us, But we need one
398 as a CNAME but no MX presence gets the CNAME returned; we need to check
399 that doesn't point to an MX to declare it "no MX returned" rather than
400 "error, loop". A new main option is added so the older capability of
401 following some limited number of chain links is maintained.
403 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
405 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
406 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
407 now-deprecated earlier definitions used only specified the range up to TLS
408 1.2 (in the older-version library docs).
410 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
412 JH/15 Rework TLS client-side context management. Stop using a global, and
413 explicitly pass a context around. This enables future use of TLS for
414 connections to service-daemons (eg. malware scanning) while a client smtp
415 connection is using TLS; with cutthrough connections this is quite likely.
417 JH/16 Fix ARC verification to do AS checks in reverse order.
419 JH/17 Support a "tls" option on the ${readsocket } expansion item.
421 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
422 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
423 Previously the "utf8" would be re-prepended for every additional message.
425 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
426 Previously thery were accepted, resulting in issues when attempting to
427 forward messages to a non-supporting MTA.
429 PP/02 Let -n work with printing macros too, not just options.
431 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
432 one parent address was copied, and bogus data was used at delivery-logging
433 time. Either a crash (after delivery) or bogus log data could result.
434 Discovery and analysis by Tim Stewart.
436 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
437 Previously if the string ended mid-character, we did not insert the
438 promised '?' replacement.
440 PP/04 Documentation: current string operators work on bytes, not codepoints.
442 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
443 should pack well giving a smaller memory footprint so better caching and
444 therefore performance. Group the declarations where this can't be done so
445 that the byte-sized flag variables are not interspersed among pointer
446 variables, giving a better chance of good packing by the compiler.
448 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
449 non-null, to avoid issues with sites running BATV. Previously reports were
450 sent with an empty envelope sender so looked like bounces.
452 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
453 The ignore_error flag wasn't being returned from the filter subprocess so
454 was not set for later routers. Investigation and fix by Matthias Kurz.
456 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
457 and a msg:complete for the whole, when a message is manually removed using
458 -Mrm. Developement by Matthias Kurz, hacked on by JH.
460 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
461 a "Gnu special" function, asprintf() in the DB utility binary builds; I
462 hope that is portable enough.
464 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
465 requiring a known-CA anchor certificate; make it now rely entirely on the
466 TLSA as an anchor. Checking the name on the leaf cert against the name
467 on the A-record for the host is still done for TA (but not for EE mode).
469 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
470 included in delivery lines for non-proxied connections, when compiled with
471 SUPPORT_SOCKS and running with proxy logging enabled.
473 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
474 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
475 move the existing event to fire before the normal logging of message
476 failure so that custom logging is bracketed by normal logging.
478 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
479 msg:fail:internal event. Developement by Matthias Kurz.
481 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
482 far too small for todays use of crypto signatures stored there. Go all
483 the way to the max DNS message size of 64kB, even though this might be
484 overmuch for IOT constrained device use.
486 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
487 copy a string over itself. The library routine is documented as not
488 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
490 JH/32 For main options check_spool_space and check_inode_space, where the
491 platform supports 64b integers, support more than the previous 2^31 kB
492 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
493 the previous G, M, k.
495 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
496 $authenticated_fail_id variable on authentication failure. Previously
499 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
500 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
501 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
502 GNUTLS_SEC_PARAM_MEDIUM.
504 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
505 side. Previously we would continue as if no SNI had been received.
507 JH/36 Harden the handling of string-lists. When a list consisted of a sole
508 "<" character, which should be a list-separator specification, we walked
509 off past the nul-terimation.
511 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
512 causes) even when the retry time is not yet met. Previously they were
513 not, meaning that when (say) an account was over-quota and temp-rejecting,
514 and multiple senders' messages were queued, only one sender would get
515 notified on each configured delay_warning cycle.
517 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
519 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
520 cipher-suites, an error can be left on the stack even for a succeeding
521 accept; this results in impossible error messages when a later operation
524 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
525 return error codes indicating retry. Under TLS1.3 this becomes required.
527 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
528 it only wrote the new authenticators, resulting in a lack of tracking of
529 peer changes of ESMTP extensions until the next cache flush.
531 JH/41 Fix the loop reading a message header line to check for integer overflow,
532 and more-often against header_maxsize. Previously a crafted message could
533 induce a crash of the recive process; now the message is cleanly rejected.
535 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
536 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
542 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
543 When redis_servers is set to a list of > 1 element, and the Redis servers
544 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
545 case of MOVED into a DEFER case instead, thus moving the query onto the
546 next server in the list. For a cluster of N elements, all N servers must
547 be defined in redis_servers.
549 GF/02 Catch and remove uninitialized value warning in exiqsumm
550 Check for existence of @ARGV before looking at $ARGV[0]
552 JH/01 Replace the store_release() internal interface with store_newblock(),
553 which internalises the check required to safely use the old one, plus
554 the allocate and data copy operations duplicated in both (!) of the
555 extant use locations.
557 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
558 modifier. This matches the restriction on the commandline.
560 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
561 Previously only the last row was returned.
563 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
564 we assumed that tags in the header were well-formed, and parsed the
565 element content after inspecting only the first char of the tag.
566 Assumptions at that stage could crash the receive process on malformed
569 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
570 While running the DKIM ACL we operate on the Permanent memory pool so that
571 variables created with "set" persist to the DATA ACL. Also (at any time)
572 DNS lookups that fail create cache records using the Permanent pool. But
573 expansions release any allocations made on the current pool - so a dnsdb
574 lookup expansion done in the DKIM ACL releases the memory used for the
575 DNS negative-cache, and bad things result. Solution is to switch to the
576 Main pool for expansions.
577 While we're in that code, add checks on the DNS cache during store_reset,
578 active in the testsuite.
579 Problem spotted, and debugging aided, by Wolfgang Breyha.
581 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
582 When none of the hosts presented to a transport match an already-open
583 connection, close it and proceed with the list. Previously we would
584 queue the message. Spotted by Lena with Yahoo, probably involving
587 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
588 Previously a spurious "250 OK id=" response was appended to the proper
591 JH/08 The "support for" informational output now, which built with Content
592 Scanning support, has a line for the malware scanner interfaces compiled
593 in. Interface can be individually included or not at build time.
595 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
596 by the template makefile "src/EDITME". The "STREAM" support for an older
597 ClamAV interface method is removed.
599 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
600 rows affected is given instead).
602 JH/11 The runtime Berkeley DB library version is now additionally output by
603 "exim -d -bV". Previously only the compile-time version was shown.
605 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
606 SMTP connection. Previously, when one had more receipients than the
607 first, an abortive onward connection was made. Move to full support for
608 multiple onward connections in sequence, handling cutthrough connection
609 for all multi-message initiating connections.
611 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
612 routers. Previously, a multi-recipient message would fail to match the
613 onward-connection opened for the first recipient, and cause its closure.
615 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
616 a timeout on read on a GnuTLS initiating connection, resulting in the
617 initiating connection being dropped. This mattered most when the callout
618 was marked defer_ok. Fix to keep the two timeout-detection methods
621 JH/15 Relax results from ACL control request to enable cutthrough, in
622 unsupported situations, from error to silently (except under debug)
623 ignoring. This covers use with PRDR, frozen messages, queue-only and
626 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
628 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
629 metadata, resulting in a crash in free().
631 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
632 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
633 Broken also in d185889f4, with init system revamp.
635 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
636 Previously we abruptly closed the connection after reading a malware-
637 found indication; now we go on to read the "scan ok" response line,
640 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
641 ACL. Previously, a crash would result.
643 JH/19 Speed up macro lookups during configuration file read, by skipping non-
644 macro text after a replacement (previously it was only once per line) and
645 by skipping builtin macros when searching for an uppercase lead character.
647 JH/20 DANE support moved from Experimental to mainline. The Makefile control
648 for the build is renamed.
650 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
651 was allocated for every new TLS startup, meaning one per message. Fix
652 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
654 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
655 reported the original. Fix to report (as far as possible) the ACL
656 result replacing the original.
658 JH/23 Fix memory leak during multi-message connections using STARTTLS under
659 OpenSSL. Certificate information is loaded for every new TLS startup,
660 and the resources needed to be freed.
662 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
664 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
667 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
668 DATA response info to the (existing) per-recipient response info for
669 the "C=" log element. It can have useful tracking info from the
670 destination system. Patch from Simon Arlott.
672 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
673 length value. Previously this would segfault.
675 HS/02 Support Avast multiline protoocol, this allows passing flags to
676 newer versions of the scanner.
678 JH/28 Ensure that variables possibly set during message acceptance are marked
679 dead before release of memory in the daemon loop. This stops complaints
680 about them when the debug_store option is enabled. Discovered specifically
681 for sender_rate_period, but applies to a whole set of variables.
682 Do the same for the queue-runner and queue-list loops, for variables set
683 from spool message files. Do the same for the SMTP per-message loop, for
684 certain variables indirectly set in ACL operations.
686 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
687 as a multi-recipient message from a mailinglist manager). The coding had
688 an arbitrary cutoff number of characters while checking for more input;
689 enforced by writing a NUL into the buffer. This corrupted long / fast
690 input. The problem was exposed more widely when more pipelineing of SMTP
691 responses was introduced, and one Exim system was feeding another.
692 The symptom is log complaints of SMTP syntax error (NUL chars) on the
693 receiving system, and refused recipients seen by the sending system
694 (propating to people being dropped from mailing lists).
695 Discovered and pinpointed by David Carter.
697 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
698 replaced by the ${authresults } expansion.
700 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
702 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
703 allows proper process termination in container environments.
705 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
706 Previously the "final dot" had a newline after it; ensure it is CR,LF.
708 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
709 and "err_perm", deprecated since 4.83 when the RFC-defined words
710 "temperror" and "permerror" were introduced.
712 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
713 transport-filters or DKIM-signing. The restriction was lost in the
714 consolidation of verify-callout and delivery SMTP handling.
715 Extend the restriction to also cover ARC-signing.
717 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
718 in defer=pass mode supply a 450 to the initiator. Previously the message
721 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
722 tls_require_ciphers is used as before.
724 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
725 "pass_unscanned". Only tmpfails from the scanner are written to
726 the paniclog, as they may require admin intervention (permission
727 denied, license issues). Other scanner errors (like decompression
728 bombs) do not cause a paniclog entry.
730 JH/36 Fix reinitialisation of DKIM logging variable between messages.
731 Previously it was possible to log spurious information in receive log
734 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
735 triggered odd behaviour from Outlook Express clients.
737 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
740 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
741 since the IETF WG has not yet settled on that versus the original
742 "bare" representation.
744 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
745 Previously the millisecond value corrupted the output.
746 Fix also for syslog_pid=no and log_selector +pid, for which the pid
747 corrupted the output.
753 JH/01 Rework error string handling in TLS interface so that the caller in
754 more cases is responsible for logging. This permits library-sourced
755 string to be attached to addresses during delivery, and collapses
756 pairs of long lines into single ones.
758 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
759 during configuration. Wildcards are allowed and expanded.
761 JH/02 Rework error string handling in DKIM to pass more info back to callers.
762 This permits better logging.
764 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
765 do not close it down and have the child transport start it up again on
766 the passed-on TCP connection. Instead, proxy the child (and any
767 subsequent ones) for TLS via a unix-domain socket channel. Logging is
768 affected: the continued delivery log lines do not have any DNSSEC, TLS
769 Certificate or OCSP information. TLS cipher information is still logged.
771 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
772 identical IP addresses on different listening ports. Will also affect
775 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
776 add noisy ifdef guards to special-case this sillyness.
777 Patch from Bernd Kuhls.
779 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
780 than 255 are no longer allowed.
782 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
783 Disable the session-cache too, which might reduce our load. Since we
784 currrectly use a new context for every connection, both as server and
785 client, there is no benefit for these.
786 GnuTLS appears to not support tickets server-side by default (we don't
787 call gnutls_session_ticket_enable_server()) but client side is enabled
788 by default on recent versions (3.1.3 +) unless the PFS priority string
791 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
792 <https://reproducible-builds.org/specs/source-date-epoch/>.
794 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
795 the check for any unsuccessful recipients did not notice the limit, and
796 erroneously found still-pending ones.
798 JH/08 Pipeline CHUNKING command and data together, on kernels that support
799 MSG_MORE. Only in-clear (not on TLS connections).
801 JH/09 Avoid using a temporary file during transport using dkim. Unless a
802 transport-filter is involved we can buffer the headers in memory for
803 creating the signature, and read the spool data file once for the
804 signature and again for transmission.
806 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
807 4.77 as the kernel support then wasn't solid, having issues in 64bit
808 mode. Now, it's been long enough. Add support for FreeBSD also.
810 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
811 case where the routing stage had gathered several addresses to send to
812 a host before calling the transport for the first, we previously failed
813 to close down TLS in the old transport process before passing the TCP
814 connection to the new process. The new one sent a STARTTLS command
815 which naturally failed, giving a failed delivery and bloating the retry
816 database. Investigation and fix prototype from Wolfgang Breyha.
818 JH/12 Fix check on SMTP command input synchronisation. Previously there were
819 false-negatives in the check that the sender had not preempted a response
820 or prompt from Exim (running as a server), due to that code's lack of
821 awareness of the SMTP input buffering.
823 PP/04 Add commandline_checks_require_admin option.
824 Exim drops privileges sanely, various checks such as -be aren't a
825 security problem, as long as you trust local users with access to their
826 own account. When invoked by services which pass untrusted data to
827 Exim, this might be an issue. Set this option in main configuration
828 AND make fixes to the calling application, such as using `--` to stop
831 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
832 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
833 responses to those, into a single TLS record each way (this usually means
834 a single packet). As a side issue, smtp_enforce_sync now works on TLS
837 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
838 affects you only if you're dancing at the edge of the param size limits.
839 If you are, and this message makes sense to you, then: raise the
840 configured limit or use OpenSSL 1.1. Nothing we can do for older
843 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
844 cmdline element to get the documented default one. Previously it was
847 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
848 are unsuable for non-message-specific reasons - eg. connection timeout,
849 banner-time rejection.
851 JH/16 Fix logging of delivery remote port, when specified by router, under
854 PP/06 Repair manualroute's ability to take options in any order, even if one
855 is the name of a transport.
858 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
860 JH/17 Change the list-building routines interface to use the expanding-string
861 triplet model, for better allocation and copying behaviour.
863 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
864 Previously it was constructed the first time a possibly-matching string
865 was met in the configuration file input during startup; now it is done
868 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
869 compatible one, to avoid the (poorly documented) possibility of a config
870 file in the working directory redirecting the DB files, possibly correpting
871 some existing file. CVE-2017-10140 assigned for BDB.
873 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
874 cache-hot. Previously, although the result was properly cached, the
875 initial verify call returned a defer.
877 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
878 the main verify for receipient in uncached-mode.
880 JH/22 Retire historical build files to an "unsupported" subdir. These are
881 defined as "ones for which we have no current evidence of testing".
883 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
884 if present. Previously it was ignored.
886 JH/24 Start using specified-initialisers in C structure init coding. This is
887 a C99 feature (it's 2017, so now considered safe).
889 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
890 if was a fixed-sized field and bitmask ops via macros; it is now more
893 PP/07 GitHub PR 56: Apply MariaDB build fix.
894 Patch provided by Jaroslav Škarvada.
896 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
897 during Coverity cleanups [4.87 JH/47]
898 Diagnosis and fix provided by Michael Fischer v. Mollard.
900 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
901 the right size to place the terminating semicolon on its own folded
902 line, the header hash was calculated to an incorrect value thanks to
903 the (relaxed) space the fold became.
905 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
906 and confused the parent.
908 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
909 which could crash as a result. This could lead to undeliverable messages.
911 JH/28 Logging: "next input sent too soon" now shows where input was truncated
914 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
915 matters on fast-turnover and PID-randomising systems, which were getting
916 out-of-order delivery.
918 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
919 a possibly-overlapping copy. The symptom was that "Remote host closed
920 connection in response to HELO" was logged instead of the actual 4xx
923 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
924 Previously only that bufferd was discarded, resulting in SYMTP command
927 JH/32 DKIM: when a message has multiple signatures matching an identity given
928 in dkim_verify_signers, run the dkim acl once for each. Previously only
929 one run was done. Bug 2189.
931 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
932 "panic the current process" to "deliberately defer". The panic log is
933 still written with the problem list name; the mail and reject logs now
934 get a temp-reject line for the message that was being handled, saying
935 something like "domains check lookup or other defer". The SMTP 451
936 message is still "Temporary local problem".
938 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
939 A crafted sequence of BDAT commands could result in in-use memory beeing
940 freed. CVE-2017-16943.
942 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
943 from SMTP input. Previously it was always done; now only done for DATA
944 and not BDAT commands. CVE-2017-16944.
946 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
947 to the message (such as an overlong header line). Previously this was
948 not done and we did not exit BDAT mode. Followon from the previous item
949 though a different problem.
955 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
956 than -2003 did; needs libidn2 in addition to libidn.
958 JH/02 The path option on a pipe transport is now expanded before use.
960 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
961 Patch provided by "Björn", documentation fix added too.
963 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
964 missing a wire-to-host endian conversion.
966 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
967 close after a BDAT command line could be taken as a following command,
968 giving a synch failure. Fix by only checking for synch immediately
969 before acknowledging the chunk.
971 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
972 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
973 macro. Patches provided by Josh Soref.
975 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
976 Previously we did not; the RFC seems ambiguous and VRFY is not listed
977 by IANA as a service extension. However, John Klensin suggests that we
980 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
981 the dkim code may be unix-mode line endings rather than smtp wire-format
982 CRLF, so prepend a CR to any bare LF.
984 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
985 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
987 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
988 the first were themselves being wrongly included in the feed into dkim
989 processing; with most chunk sizes in use this resulted in an incorrect
990 body hash calculated value.
992 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
993 DKIM signature block, for verification. Although advised against by
994 standards it is specifically not ruled illegal.
996 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
998 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
999 missing a body hash (the bh= tag).
1001 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1002 It seems that HAProxy sends the Proxy Protocol information in clear and
1003 only then does a TLS startup, so do the same.
1005 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1006 TCP connections (such as for Spamd) unless the daemon successfully set
1007 Fast Open mode on its listening sockets. This fixes breakage seen on
1008 too-old kernels or those not configured for Fast Open, at the cost of
1009 requiring both directions being enabled for TFO, and TFO never being used
1010 by non-daemon-related Exim processes.
1012 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1013 endings, at least on the first header line. Try to canonify any that get
1014 past that check, despite the cost.
1016 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1017 now limited to an arbitrary five deep, while parsing addresses with the
1018 strip_excess_angle_brackets option enabled.
1020 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1021 instead leave the unprompted TLS handshake in socket buffer for the
1022 TLS library to consume.
1024 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1026 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1028 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1029 region in one operation, for speed, and this leaves assigned pointers
1030 dangling. Add checks run only under the testsuite which checks all
1031 variables at a store-reset and panics on a dangling pointer; add code
1032 explicitly nulling out all the variables discovered. Fixes one known
1033 bug: a transport crash, where a dangling pointer for $sending_ip_address
1034 originally assigned in a verify callout, is re-used.
1036 PP/06 Drop '.' from @INC in various Perl scripts.
1038 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1040 PP/08 Reduce a number of compilation warnings under clang; building with
1041 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1042 should be warning-free.
1044 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1046 HS/01 Fix portability problems introduced by PP/08 for platforms where
1047 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1049 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1050 chunk. This allows us to accept broken chunked messages. We need a more
1051 general solution here.
1053 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1054 already-broken messages in the queue.
1056 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1058 JH/19 Fix reference counting bug in routing-generated-address tracking.
1064 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1065 supports it and a size is available (ie. the sending peer gave us one).
1067 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1068 years of being deprecated). The replacements are the ACLs
1069 acl_smtp_mime and acl_not_smtp_mime.
1071 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1072 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1073 or even an in-clear connection were permitted. Now, if the host lookup
1074 was dnssec and dane was requested then the host is only used if the
1075 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1076 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1077 if one fails this test.
1078 This means that a poorly-configured remote DNS will make it incommunicado;
1079 but it protects against a DNS-interception attack on it.
1081 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1082 non-noisy when a race steals the message being considered.
1084 JH/05 If main configuration option tls_certificate is unset, generate a
1085 self-signed certificate for inbound TLS connections.
1087 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1088 in rewrites and routers.
1090 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1091 and logged a warning sing 4.83; now they are a configuration file error.
1093 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1094 (lacking @domain). Apply the same qualification processing as RCPT.
1096 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1098 JH/10 Support ${sha256:} applied to a string (as well as the previous
1101 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1102 a cutthrough deliver is pending, as we always want to make a connection.
1103 This also avoids re-routing the message when later placing the cutthrough
1104 connection after a verify cache hit.
1105 Do not update it with the verify result either.
1107 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1108 when routing results in more than one destination address.
1110 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1111 signing (which inhibits the cutthrough capability). Previously only
1112 the presence of an option was tested; now an expansion evaluating as
1113 empty is permissible (obviously it should depend only on data available
1114 when the cutthrough connection is made).
1116 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1117 the relevant preceding SMTP command did not note the pipelining mode.
1119 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1120 Previously they were not counted.
1122 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1123 as one having no matching records. Previously we deferred the message
1124 that needed the lookup.
1126 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1127 distinguished as "(=".
1129 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1130 for missing MX records. Previously it only worked for missing A records.
1132 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1134 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1135 after the data-go-ahead and data-ack. Patch from Jason Betts.
1137 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1138 even for a "none" policy. Patch from Tony Meyer.
1140 JH/22 Fix continued use of a connection for further deliveries. If a port was
1141 specified by a router, it must also match for the delivery to be
1144 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1145 When one of the recipients of a message was unsuitable for the connection
1146 (has no matching addresses), we lost track of needing to mark it
1147 deferred. As a result mail would be lost.
1149 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1151 JH/25 Decoding ACL controls is now done using a binary search; the source code
1152 takes up less space and should be simpler to maintain. Merge the ACL
1153 condition decode tables also, with similar effect.
1155 JH/26 Fix problem with one_time used on a redirect router which returned the
1156 parent address unchanged. A retry would see the parent address marked as
1157 delivered, so not attempt the (identical) child. As a result mail would
1160 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1161 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1162 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1165 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1166 Main config options check_{log,spool}_{inodes,space} are now
1167 100 inodes, 10MB unless set otherwise in the configuration.
1169 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1170 Previously it only applied to the main-section connection policy
1173 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1175 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1176 by me. Added RFC7919 DH primes as an alternative.
1178 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1179 are not in the system include path.
1181 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1182 GnuTLS, when a session startup failed (eg because the client disconnected)
1183 Exim did stdio operations after fclose. This was exposed by a recent
1184 change which nulled out the file handle after the fclose.
1186 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1187 signed directly by the cert-signing cert, rather than an intermediate
1188 OCSP-signing cert. This is the model used by LetsEncrypt.
1190 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1192 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1193 an incoming connection.
1195 HS/02 Bug 1802: Do not half-close the connection after sending a request
1198 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1199 fallback to "prime256v1".
1201 JH/34 SECURITY: Use proper copy of DATA command in error message.
1202 Could leak key material. Remotely exploitable. CVE-2016-9963.
1208 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1209 and 3.4.4 - once the server is enabled to respond to an OCSP request
1210 it does even when not requested, resulting in a stapling non-aware
1211 client dropping the TLS connection.
1213 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1214 support variable-length bit vectors. No functional change.
1216 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1217 The I= interface field on outgoing lines is now after the H= remote
1218 host field, same as incoming lines. There is a separate
1219 outgoing_interface log selector which allows you to disable the
1222 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1223 If not running log_selector +smtp_connection the mainlog would be held
1224 open indefinitely after a "too many connections" event, including to a
1225 deleted file after a log rotate. Leave the per net connection logging
1226 leaving it open for efficiency as that will be quickly detected by the
1227 check on the next write.
1229 HS/01 Bug 1671: Fix post transport crash.
1230 Processing the wait-<transport> messages could crash the delivery
1231 process if the message IDs didn't exist for some reason. When
1232 using 'split_spool_directory=yes' the construction of the spool
1233 file name failed already, exposing the same netto behaviour.
1235 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1236 mime_regex ACL conditions.
1238 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1239 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1240 to HELO, local diagnostic string.
1242 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1243 log line to debug. Even when configured with a tls authenticator many
1244 client connections are expected to not authenticate in this way, so
1245 an authenticate fail is not an error.
1247 HS/02 Add the Exim version string to the process info. This way exiwhat
1248 gives some more detail about the running daemon.
1250 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1251 matter for fast-change records such as DNSBLs.
1253 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1254 retry record, even if constant. There may be multiple transports with
1255 different interface settings and the retry behaviour needs to be kept
1258 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1260 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1262 JH/10 Bug 840: fix log_defer_output option of pipe transport
1264 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1265 hosts_randomize. This matters a lot when combined with mua_wrapper.
1267 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1268 ${quote_pgsql:<string>} operator.
1270 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1272 JH/14 Change header file name for freeradius-client. Relevant if compiling
1273 with Radius support; from the Gentoo tree and checked under Fedora.
1275 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1277 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1278 option transport_filter, meaning no filtering.
1280 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1282 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1283 defaults to "*" (all hosts). The variable is now available when not built
1284 with TLS, default unset, mainly to enable keeping the testsuite sane.
1285 If a server certificate is not supplied (via tls_certificate) an error is
1286 logged, and clients will find TLS connections fail on startup. Presumably
1287 they will retry in-clear.
1288 Packagers of Exim are strongly encouraged to create a server certificate
1289 at installation time.
1291 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1292 with the $config_file variable.
1294 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1295 in transport context, after the attempt, and per-recipient. The latter type
1296 is per host attempted. The event data is the error message, and the errno
1297 information encodes the lookup type (A vs. MX) used for the (first) host,
1298 and the trailing two digits of the smtp 4xx response.
1300 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1301 to write to mainlog (or rejectlog, paniclog) in the window between file
1302 creation and permissions/ownership being changed. Particularly affects
1303 installations where exicyclog is run as root, rather than exim user;
1304 result is that the running daemon panics and dies.
1306 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1308 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1309 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1310 "pri" and "weight". Note that the previous implicit priority given by the
1311 list order is no longer honoured.
1313 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1314 for DKIM processing.
1316 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1317 by defining SUPPORT_SOCKS.
1319 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1320 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1321 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1322 variables are renamed to proxy_{local,external}_{address,port}.
1324 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1325 for a build by defining SUPPORT_I18N
1327 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1328 of the query string, and make ${quote_redis:} do that quoting.
1330 JH/29 Move Events support from Experimental to mainline, enabled by default
1331 and removable for a build by defining DISABLE_EVENT.
1333 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1335 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1336 cached by the daemon.
1338 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1339 by defining LOOKUP_REDIS. The libhiredis library is required.
1341 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1342 keys are given for lookup.
1344 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1345 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1346 only supported when built with TLS support. The PolarSSL SHA routines
1347 are still used when the TLS library is too old for convenient support.
1349 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1350 openssl_options), for security. OpenSSL forces this from version 1.1.0
1351 server-side so match that on older versions.
1353 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1354 allocation for $value could be released as the expansion processing
1355 concluded, but leaving the global pointer active for it.
1357 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1358 and to use the domains and local_parts ACL conditions.
1360 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1361 incorrectly not doubled on cutthrough transmission, hence seen as a
1362 body-termination at the receiving system - resulting in truncated mails.
1363 Commonly the sender saw a TCP-level error, and retransmitted the message
1364 via the normal store-and-forward channel. This could result in duplicates
1365 received - but deduplicating mailstores were liable to retain only the
1366 initial truncated version.
1368 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1370 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1372 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1373 we're in there, support oversigning also; bug 1309.
1375 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1377 HS/04 Add support for keep_environment and add_environment options.
1379 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1380 either intentional arithmetic overflow during PRNG, or testing config-
1383 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1384 delivery resulted in actual delivery. Cancel cutthrough before DATA
1387 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1388 rejects a recipient: pass the reject to the originator.
1390 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1391 Many were false-positives and ignorable, but it's worth fixing the
1394 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1395 for the new environment-manipulation done at startup. Move the routines
1396 from being local to tls.c to being global via the os.c file.
1398 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1399 an extract embedded as result-arg for a map, the first arg for extract
1400 is unavailable so we cannot tell if this is a numbered or keyed
1401 extraction. Accept either.
1407 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1410 JH/02 The smtp transport option "multi_domain" is now expanded.
1412 JH/03 The smtp transport now requests PRDR by default, if the server offers
1415 JH/04 Certificate name checking on server certificates, when exim is a client,
1416 is now done by default. The transport option tls_verify_cert_hostnames
1417 can be used to disable this per-host. The build option
1418 EXPERIMENTAL_CERTNAMES is withdrawn.
1420 JH/05 The value of the tls_verify_certificates smtp transport and main options
1421 default to the word "system" to access the system default CA bundle.
1422 For GnuTLS, only version 3.0.20 or later.
1424 JH/06 Verification of the server certificate for a TLS connection is now tried
1425 (but not required) by default. The verification status is now logged by
1426 default, for both outbound TLS and client-certificate supplying inbound
1429 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1432 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1433 Status Notification (bounce) messages are now MIME format per RFC 3464.
1434 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1435 under the control of the dsn_advertise_hosts option, and routers may
1436 have a dsn_lasthop option.
1438 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1439 default, modifiable by a malware= option. The list separator for
1440 the options can now be changed in the usual way. Bug 68.
1442 JH/10 The smtp_receive_timeout main option is now expanded before use.
1444 JH/11 The incoming_interface log option now also enables logging of the
1445 local interface on delivery outgoing connections.
1447 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1448 if the interface and destination host and port all match.
1450 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1453 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1454 Patch from Andrew Lewis.
1456 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1457 now supports optional time-restrictions, weighting, and priority
1458 modifiers per server. Patch originally by <rommer@active.by>.
1460 JH/16 The spamd_address main option now supports a mixed list of local
1461 and remote servers. Remote servers can be IPv6 addresses, and
1462 specify a port-range.
1464 JH/17 Bug 68: The spamd_address main option now supports an optional
1465 timeout value per server.
1467 JH/18 Bug 1581: Router and transport options headers_add/remove can
1468 now have the list separator specified.
1470 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1473 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1476 JH/21 Support for the A6 type of dns record is withdrawn.
1478 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1479 rather than the verbs used.
1481 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1482 from 255 to 1024 chars.
1484 JH/24 Verification callouts now attempt to use TLS by default.
1486 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1487 are generic router options now. The defaults didn't change.
1489 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1490 Original patch from Alexander Shikoff, worked over by JH.
1492 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1493 files if ZCAT_COMMAND is not executable.
1495 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1497 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1499 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1500 Normally benign, it bites when the pair was led to by a CNAME;
1501 modern usage is to not canonicalize the domain to a CNAME target
1502 (and we were inconsistent anyway for A-only vs AAAA+A).
1504 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1506 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1507 when evaluating $sender_host_dnssec.
1509 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1510 $sender_helo_dnssec variable.
1512 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1514 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1516 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1518 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1519 documented as working, but never had. Support all but $spam_report.
1521 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1522 added for tls authenticator.
1524 HS/03 Add perl_taintmode main config option
1530 TL/01 When running the test suite, the README says that variables such as
1531 no_msglog_check are global and can be placed anywhere in a specific
1532 test's script, however it was observed that placement needed to be near
1533 the beginning for it to behave that way. Changed the runtest perl
1534 script to read through the entire script once to detect and set these
1535 variables, reset to the beginning of the script, and then run through
1536 the script parsing/test process like normal.
1538 TL/02 The BSD's have an arc4random API. One of the functions to induce
1539 adding randomness was arc4random_stir(), but it has been removed in
1540 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1541 function when detected.
1543 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1544 cause callback expansion.
1546 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1547 syntax errors in an expansion can be treated as a string instead of
1548 logging or causing an error, due to the internal use of bool_lax
1549 instead of bool when processing it.
1551 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1552 server certificates when making smtp deliveries.
1554 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1556 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1558 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1560 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1561 Merged patch from Sebastian Wiedenroth.
1563 JH/05 Fix results-pipe from transport process. Several recipients, combined
1564 with certificate use, exposed issues where response data items split
1565 over buffer boundaries were not parsed properly. This eventually
1566 resulted in duplicates being sent. This issue only became common enough
1567 to notice due to the introduction of connection certificate information,
1568 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1570 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1571 size buffer was used, resulting in syntax errors when an expansion
1574 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1575 version 3.3.6 or later.
1577 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1578 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1579 both become "event_action", the variables become $event_name, $event_data
1580 and $event_defer_errno. There is a new variable $verify_mode, usable in
1581 routers, transports and related events. The tls:cert event is now also
1582 raised for inbound connections, if the main configuration event_action
1585 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1586 early OCSP support, but no stapling (appears to be less than 1.0.0).
1588 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1589 server certificate names available under the smtp transport option
1590 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1593 JH/10 Time-related extraction expansions from certificates now use the main
1594 option "timezone" setting for output formatting, and are consistent
1595 between OpenSSL and GnuTLS compilations. Bug 1541.
1597 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1598 encoded parameter in the incoming message. Bug 1558.
1600 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1601 include certificate info, eximon was claiming there were spoolfile
1604 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1606 JH/14 Log delivery-related information more consistently, using the sequence
1607 "H=<name> [<ip>]" wherever possible.
1609 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1610 are problematic for Debian distribution, omit them from the release
1613 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1615 JH/16 Fix string representation of time values on 64bit time_t architectures.
1618 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1619 output list separator was used.
1624 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1625 checkers that were complaining about end of non-void function with no
1628 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1629 This was a regression introduced in 4.83 by another bugfix.
1631 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1633 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1634 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1640 TF/01 Correctly close the server side of TLS when forking for delivery.
1642 When a message was received over SMTP with TLS, Exim failed to clear up
1643 the incoming connection properly after forking off the child process to
1644 deliver the message. In some situations the subsequent outgoing
1645 delivery connection happened to have the same fd number as the incoming
1646 connection previously had. Exim would try to use TLS and fail, logging
1647 a "Bad file descriptor" error.
1649 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1650 utilities have not been installed.
1652 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1653 temporary space as the ACL may create new global variables.
1655 TL/01 LDAP support uses per connection or global context settings, depending
1656 upon the detected version of the libraries at build time.
1658 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1659 to extract and use the src ip:port in logging and expansions as if it
1660 were a direct connection from the outside internet. PPv2 support was
1661 updated based on HAProxy spec change in May 2014.
1663 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1665 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1666 Properly escape header and check for NULL return.
1668 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1671 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1673 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1674 characters in header names, implemented as a verify condition.
1675 Contributed by Michael Fischer v. Mollard.
1677 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1678 results permerror and temperror. Previous values are deprecated but
1679 still accepted. In a future release, err_perm and err_temp will be
1680 completely removed, which will be a backward incompatibility if the
1681 ACL tests for either of these two old results. Patch contributed by
1682 user bes-internal on the mailing list.
1684 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1686 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1687 selectors, in both main and reject logs.
1689 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1692 JH/07 Add malware type "sock" for talking to simple daemon.
1694 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1696 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1697 routers/transports under cutthrough routing.
1699 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1700 numbers. Touch up "bool" conditional to keep the same definition.
1702 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1704 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1706 JH/12 Expand items in router/transport headers_add or headers_remove lists
1707 individually rather than the list as a whole. Bug 1452.
1709 Required for reasonable handling of multiple headers_ options when
1710 they may be empty; requires that headers_remove items with embedded
1711 colons must have them doubled (or the list-separator changed).
1713 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1714 view the policy declared in the DMARC record. Currently, $dmarc_status
1715 is a combined value of both the record presence and the result of the
1718 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1720 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1721 dnslookup router and the smtp transport (applying to the forward
1724 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1725 of ldap servers used for a specific lookup. Patch provided by Heiko
1728 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1729 New variable $lookup_dnssec_authenticated for observability.
1731 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1732 Patch submitted by Lars Timman.
1734 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1736 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1737 Requires trusted mode and valid format message id, aborts otherwise.
1738 Patch contributed by Heiko Schlichting.
1740 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1741 certextract with support for various fields. Bug 1358.
1743 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1744 is requested by default, modifiable by smtp transport option
1747 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1748 operate on certificate variables to give certificate fingerprints
1749 Also new ${sha256:cert_variable}.
1751 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1753 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1754 Christian Aistleitner.
1756 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1758 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1759 file. Patch from Wolfgang Breyha.
1761 JH/25 Expand the coverage of the delivery $host and $host_address to
1762 client authenticators run in verify callout. Bug 1476.
1764 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1765 align with daemon_smtp_ports. Bug 72.
1767 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1768 support and error reporting did not work properly.
1770 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1771 and is readable. Patch from Andrew Colin Kissa.
1773 TL/14 Enhance documentation of ${run expansion and how it parses the
1774 commandline after expansion, particularly in the case when an
1775 unquoted variable expansion results in an empty value.
1777 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1779 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1780 by renaming away from it. Addresses GH issue 15, reported by
1783 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1785 TL/15 SECURITY: prevent double expansion in math comparison functions
1786 (can expand unsanitized data). Not remotely exploitable.
1793 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1795 PP/02 Make -n do something, by making it not do something.
1796 When combined with -bP, the name of an option is not output.
1798 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1801 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1802 $sender_host_name and config options to manage this, and basic check
1805 PP/05 DSCP support for outbound connections and control modifier for inbound.
1807 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1808 (Only plugin which currently uses this is kerberos4, which nobody should
1809 be using, but we should make it available and other future plugins might
1810 conceivably use it, even though it would break NAT; stuff *should* be
1811 using channel bindings instead).
1813 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1814 name; added for Sendmail compatibility; requires admin caller.
1815 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1816 just ignore it); requires trusted caller.
1817 Also parse but ignore: -Ac -Am -X<logfile>
1820 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1822 TL/02 Add +smtp_confirmation as a default logging option.
1824 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1825 Patch by Magnus Holmgren from 2007-02-20.
1827 TL/04 Bugzilla 1281 - Spec typo.
1828 Bugzilla 1283 - Spec typo.
1829 Bugzilla 1290 - Spec grammar fixes.
1831 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1833 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1835 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1836 mailing list by by Dmitry Isaikin.
1838 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1840 JH/02 Support "G" suffix to numbers in ${if comparisons.
1842 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1844 NM/01 Bugzilla 1197 - Spec typo
1845 Bugzilla 1196 - Spec examples corrections
1847 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1849 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1850 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1853 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1854 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1856 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1857 "acl {{name}{arg}...}", and optional args on acl condition
1860 JH/05 Permit multiple router/transport headers_add/remove lines.
1862 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1864 JH/07 Avoid using a waiting database for a single-message-only transport.
1865 Performance patch from Paul Fisher. Bugzilla 1262.
1867 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1870 JH/09 Add $headers_added variable, with content from use of ACL modifier
1871 add_header (but not yet added to the message). Bugzilla 199.
1873 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1874 Pulled from Bugzilla 817 by Wolfgang Breyha.
1876 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1878 (nb: this is the same fix as in Exim 4.80.1)
1880 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1883 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1886 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1888 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1889 advertise SMTP AUTH mechanism to us, instead of a generic
1890 protocol violation error. Also, make Exim more robust to bad
1891 data from the Dovecot auth socket.
1893 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1895 When a queue runner is handling a message, Exim first routes the
1896 recipient addresses, during which it prunes them based on the retry
1897 hints database. After that it attempts to deliver the message to
1898 any remaining recipients. It then updates the hints database using
1901 So if a recipient address works intermittently, it can get repeatedly
1902 deferred at routing time. The retry hints record remains fresh so the
1903 address never reaches the final cutoff time.
1905 This is a fairly common occurrence when a user is bumping up against
1906 their storage quota. Exim had some logic in its local delivery code
1907 to deal with this. However it did not apply to per-recipient defers
1908 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1910 This change adds a proper retry rule check during routing so that the
1911 final cutoff time is checked against the message's age. We only do
1912 this check if there is an address retry record and there is not a
1913 domain retry record; this implies that previous attempts to handle
1914 the address had the retry_use_local_parts option turned on. We use
1915 this as an approximation for the destination being like a local
1916 delivery, as in LMTP.
1918 I suspect this new check makes the old local delivery cutoff check
1919 redundant, but I have not verified this so I left the code in place.
1921 TF/02 Correct gecos expansion when From: is a prefix of the username.
1923 Test 0254 submits a message to Exim with the header
1927 When I ran the test suite under the user fanf2, Exim expanded
1928 the header to contain my full name, whereas it should have added
1929 a Resent-Sender: header. It erroneously treats any prefix of the
1930 username as equal to the username.
1932 This change corrects that bug.
1934 GF/01 DCC debug and logging tidyup
1935 Error conditions log to paniclog rather than rejectlog.
1936 Debug lines prefixed by "DCC: " to remove any ambiguity.
1938 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1940 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1941 Bug spotted by Jeremy Harris; was flawed since initial commit.
1942 Would have resulted in OCSP responses post-SNI triggering an Exim
1943 NULL dereference and crash.
1945 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1947 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1948 Bug detection, analysis and fix by Samuel Thibault.
1949 Bugzilla 1331, Debian bug #698092.
1951 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1953 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1954 Server implementation by Todd Lyons, client by JH.
1955 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1956 config variable "prdr_enable" controls whether the server
1957 advertises the facility. If the client requests PRDR a new
1958 acl_data_smtp_prdr ACL is called once for each recipient, after
1959 the body content is received and before the acl_smtp_data ACL.
1960 The client is controlled by both of: a hosts_try_prdr option
1961 on the smtp transport, and the server advertisement.
1962 Default client logging of deliveries and rejections involving
1963 PRDR are flagged with the string "PRDR".
1965 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1966 fclose(). Diagnosis by Todd Lyons.
1968 PP/17 Update configure.default to handle IPv6 localhost better.
1969 Patch by Alain Williams (plus minor tweaks).
1972 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1973 This is now consistent with GnuTLS, and is now documented: the
1974 previous undocumented portable approach to treating the option as
1975 unset was to force an expansion failure. That still works, and
1976 an empty string is now equivalent.
1978 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1979 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1980 not performing validation itself.
1982 PP/20 Added force_command boolean option to pipe transport.
1983 Patch from Nick Koston, of cPanel Inc.
1985 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1988 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1990 PP/21 Fix eximon continuous updating with timestamped log-files.
1991 Broken in a format-string cleanup in 4.80, missed when I repaired the
1992 other false fix of the same issue.
1993 Report and fix from Heiko Schlichting.
1996 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1997 Report from Prashanth Katuri.
1999 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2000 It's SecureTransport, so affects any MacOS clients which use the
2001 system-integrated TLS libraries, including email clients.
2003 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2004 using a MIME ACL for non-SMTP local injection.
2005 Report and assistance in diagnosis by Warren Baker.
2007 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2009 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2011 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2012 last id that failed so it may be referenced in subsequent ACL's.
2014 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2017 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2018 ldap library initialization, allowing self-signed CA's to be
2019 used. Also properly sets require_cert option later in code by
2020 using NULL (global ldap config) instead of ldap handle (per
2021 session). Bug diagnosis and testing by alxgomz.
2023 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2024 the src/util/ subdirectory.
2026 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2027 renamed to Transport Post Delivery Action by Jeremy Harris, as
2030 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2031 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2032 redis_servers = needs to be configured which will be used by the redis
2033 lookup. Patch from Warren Baker, of The Packet Hub.
2035 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2037 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2038 hostname or reverse DNS when processing a host list. Used suggestions
2039 from multiple comments on this bug.
2041 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2043 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2044 Missed a few lines, added it to make the runtest require no keyboard
2047 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2048 contains upper case chars. Make router use caseful_local_part.
2050 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2051 support when GnuTLS has been built with p11-kit.
2057 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2059 This, or similar/improved, will also be change PP/11 of 4.82.
2065 PP/01 Handle short writes when writing local log-files.
2066 In practice, only affects FreeBSD (8 onwards).
2067 Bugzilla 1053, with thanks to Dmitry Isaikin.
2069 NM/01 Bugzilla 949 - Documentation tweak
2071 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2074 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2076 PP/02 Implemented gsasl authenticator.
2078 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2080 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2081 `pkg-config foo` for cflags/libs.
2083 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2084 with rest of GSASL and with heimdal_gssapi.
2086 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2087 `pkg-config foo` for cflags/libs for the TLS implementation.
2089 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2090 properties get this fed in as external SSF. A number of robustness
2091 and debugging improvements to the cyrus_sasl authenticator.
2093 PP/08 cyrus_sasl server now expands the server_realm option.
2095 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2096 Patch by Jeremy Harris.
2098 PP/10 Added dbmjz lookup type.
2100 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2102 PP/12 MAIL args handles TAB as well as SP, for better interop with
2103 non-compliant senders.
2104 Analysis and variant patch by Todd Lyons.
2106 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2107 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2108 Patch from Dirk Mueller <dmueller@suse.com>
2110 PP/13 tls_peerdn now print-escaped for spool files.
2111 Observed some $tls_peerdn in wild which contained \n, which resulted
2112 in spool file corruption.
2114 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2115 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2116 or write after TLS renegotiation, which otherwise led to messages
2119 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2120 as a tracking header (ie: a signed header comes before the signature).
2121 Patch from Wolfgang Breyha.
2123 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2124 comma-sep list; embedded commas doubled.
2126 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2128 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2130 Report and patch from Dmitry Banschikov.
2132 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2133 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2134 failures. We appear to now support TLS1.1+ with Exim.
2136 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2137 lets Exim select keys and certificates based upon TLS SNI from client.
2138 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2139 before an outbound SMTP session. New log_selector, +tls_sni.
2141 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2142 NULL dereference. Report and patch from Alun Jones.
2144 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2145 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2146 Not seeing resolver debug output on NetBSD, but suspect this is a
2147 resolver implementation change.
2149 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2150 Left warnings. Added "eximon gdb" invocation mode.
2152 PP/21 Defaulting "accept_8bitmime" to true, not false.
2154 PP/22 Added -bw for inetd wait mode support.
2156 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2157 locate the relevant includes and libraries. Made this the default.
2159 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2160 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2162 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2163 This may cause build issues on older platforms.
2165 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2166 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2167 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2168 Added SNI support via GnuTLS too.
2169 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2171 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2173 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2174 Applied second patch from Janne, implementing suggestion to default
2175 multiple-strings-in-record handling to match SPF spec.
2177 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2179 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2180 read-only, out of scope).
2181 Patch from Wolfgang Breyha, report from Stuart Northfield.
2183 PP/29 Fix three issues highlighted by clang analyser static analysis.
2184 Only crash-plausible issue would require the Cambridge-specific
2185 iplookup router and a misconfiguration.
2186 Report from Marcin Mirosław.
2188 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2190 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2191 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2192 As part of this, removing so much warning spew let me fix some minor
2193 real issues in debug logging.
2195 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2196 assignment on my part. Fixed.
2198 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2199 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2200 Janne Snabb (who went above and beyond: thank you).
2202 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2203 string otherwise requires a connection and a bunch more work and it's
2204 relatively easy to get wrong. Should also expose TLS library linkage
2207 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2208 64-bit ${eval} (JH/03).
2210 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2211 GNU libc to support some of the 64-bit stuff, should not lead to
2212 conflicts. Defined before os.h is pulled in, so if a given platform
2213 needs to override this, it can.
2215 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2216 protection layer was required, which is not implemented.
2217 Bugzilla 1254, patch from Wolfgang Breyha.
2219 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2220 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2221 tls_dhparam take prime identifiers. Also unbreak combination of
2222 OpenSSL+DH_params+TLSSNI.
2224 PP/39 Disable SSLv2 by default in OpenSSL support.
2230 PP/01 Solaris build fix for Oracle's LDAP libraries.
2231 Bugzilla 1109, patch from Stephen Usher.
2233 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2235 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2238 TF/02 Fix a couple more cases where we did not log the error message
2239 when unlink() failed. See also change 4.74-TF/03.
2241 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2242 lock up or crash if it happened to be inside a call to libc when it
2243 got a SIGUSR1 from exiwhat.
2245 The SIGUSR1 handler appends the current process status to the process
2246 log which is later printed by exiwhat. It used to use the general
2247 purpose logging code to do this, but several functions it calls are
2248 not safe for signals.
2250 The new output code in the SIGUSR1 handler is specific to the process
2251 log, and simple enough that it's easy to inspect for signal safety.
2252 Removing some special cases also simplifies the general logging code.
2253 Removing the spurious timestamps from the process log simplifies
2256 TF/04 Improved ratelimit ACL condition.
2258 The /noupdate option has been deprecated in favour of /readonly which
2259 has clearer semantics. The /leaky, /strict, and /readonly update modes
2260 are mutually exclusive. The update mode is no longer included in the
2261 database key; it just determines when the database is updated. (This
2262 means that when you upgrade Exim will forget old rate measurements.)
2264 Exim now checks that the per_* options are used with an update mode that
2265 makes sense for the current ACL. For example, when Exim is processing a
2266 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2267 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2268 must specify per_mail/readonly. If you omit the update mode it defaults to
2269 /leaky where that makes sense (as before) or /readonly where required.
2271 The /noupdate option is now undocumented but still supported for
2272 backwards compatibility. It is equivalent to /readonly except that in
2273 ACLs where /readonly is required you may specify /leaky/noupdate or
2274 /strict/noupdate which are treated the same as /readonly.
2276 A useful new feature is the /count= option. This is a generalization
2277 of the per_byte option, so that you can measure the throughput of other
2278 aggregate values. For example, the per_byte option is now equivalent
2279 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2281 The per_rcpt option has been generalized using the /count= mechanism
2282 (though it's more complicated than the per_byte equivalence). When it is
2283 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2284 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2285 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2286 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2287 Note that using per_rcpt with a non-readonly update mode in more than
2288 one ACL will cause the recipients to be double-counted. (The per_mail
2289 and per_byte options don't have this problem.)
2291 The handling of very low rates has changed slightly. If the computed rate
2292 is less than the event's count (usually one) then this event is the first
2293 after a long gap. In this case the rate is set to the same as this event's
2294 count, so that the first message of a spam run is counted properly.
2296 The major new feature is a mechanism for counting the rate of unique
2297 events. The new per_addr option counts the number of different
2298 recipients that someone has sent messages to in the last time period. It
2299 behaves like per_rcpt if all the recipient addresses are different, but
2300 duplicate recipient addresses do not increase the measured rate. Like
2301 the /count= option this is a general mechanism, so the per_addr option
2302 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2303 example, measure the rate that a client uses different sender addresses
2304 with the options per_mail/unique=$sender_address. There are further
2305 details in the main documentation.
2307 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2309 TF/06 Removed a few PCRE remnants.
2311 TF/07 Automatically extract Exim's version number from tags in the git
2312 repository when doing development or release builds.
2314 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2315 Bugzilla 879. Patch from Paul Fisher.
2317 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2318 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2321 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2323 PP/05 Variable $av_failed, true if the AV scanner deferred.
2324 Bugzilla 1078. Patch from John Horne.
2326 PP/06 Stop make process more reliably on build failure.
2327 Bugzilla 1087. Patch from Heiko Schlittermann.
2329 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2330 Bugzilla 1089. Patch from Heiko Schlittermann.
2332 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2333 Bugzilla 1131. Patch from Holger Weiß.
2335 PP/09 Handle IPv6 addresses with SPF.
2336 Bugzilla 860. Patch from Wolfgang Breyha.
2338 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2340 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2343 PP/11 match_* no longer expand right-hand-side by default.
2344 New compile-time build option, EXPAND_LISTMATCH_RHS.
2345 New expansion conditions, "inlist", "inlisti".
2347 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2349 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2351 PP/14 fix log_write() format string regression from TF/03.
2352 Bugzilla 1152. Patch from Dmitry Isaikin.
2358 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2360 PP/02 Harmonised TLS library version reporting; only show if debugging.
2361 Layout now matches that introduced for other libraries in 4.74 PP/03.
2363 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2365 PP/04 New "dns_use_edns0" global option.
2367 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2370 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2371 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2373 TK/01 Updated PolarSSL code to 0.14.2.
2374 Bugzilla 1097. Patch from Andreas Metzler.
2376 PP/07 Catch divide-by-zero in ${eval:...}.
2377 Fixes bugzilla 1102.
2379 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2382 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2383 format-string attack -- SECURITY: remote arbitrary code execution.
2385 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2386 time unintentionally subject to list matching rules, letting the header
2387 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2388 arbitrary string expansion). This allowed for information disclosure.
2390 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2391 INT_MIN/-1 -- value coerced to INT_MAX.
2397 NM/01 Workaround for PCRE version dependency in version reporting
2400 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2401 This fixes portability to compilers other than gcc, notably
2402 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2404 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2405 makefiles for portability to HP-UX and POSIX correctness.
2407 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2408 Also via indented variable definition in the Makefile.
2409 (Debugging by Oliver Heesakkers).
2411 PP/02 Restore caching of spamd results with expanded spamd_address.
2412 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2414 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2415 Improves build reliability. Fix from: Frank Elsner
2417 NM/02 Fix wide character breakage in the rfc2047 coding
2418 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2420 NM/03 Allow underscore in dnslist lookups
2421 Fixes bug 1026. Patch from Graeme Fowler
2423 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2424 Code patches from Adam Ciarcinski of NetBSD.
2426 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2429 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2430 is logged, to avoid truncation. Patch from John Horne.
2432 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2433 Patch from Jakob Hirsch.
2435 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2436 SQL string expansion failure details.
2437 Patch from Andrey Oktyabrski.
2439 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2440 Patch from Simon Arlott.
2442 PP/09 New lookups functionality failed to compile on old gcc which rejects
2443 extern declarations in function scope.
2444 Patch from Oliver Fleischmann
2446 PP/10 Use sig_atomic_t for flags set from signal handlers.
2447 Check getgroups() return and improve debugging.
2448 Fixed developed for diagnosis in bug 927 (which turned out to be
2451 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2452 Patch from Mark Zealey.
2454 PP/12 Bugzilla 1056: Improved spamd server selection.
2455 Patch from Mark Zealey.
2457 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2458 Based on patch from Heiko Schlittermann.
2460 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2461 Patch from Uwe Doering, sign-off by Michael Haardt.
2463 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2464 variable declaration deep within a block. Bug and patch from
2467 PP/15 lookups-Makefile IRIX compatibility coercion.
2469 PP/16 Make DISABLE_DKIM build knob functional.
2471 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2472 Patch by Simon Arlott
2474 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2475 variable argument macros. Our copy now differs from upstream.
2481 TF/01 Failure to get a lock on a hints database can have serious
2482 consequences so log it to the panic log.
2484 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2485 controlled using the smtp_confirmation log selector.
2487 TF/03 Include the error message when we fail to unlink a spool file.
2489 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2490 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2491 for maintaining out-of-tree patches for some time.
2493 PP/01 Bugzilla 139: Documentation and portability issues.
2494 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2495 Handle per-OS dynamic-module compilation flags.
2497 PP/02 Let /dev/null have normal permissions.
2498 The 4.73 fixes were a little too stringent and complained about the
2499 permissions on /dev/null. Exempt it from some checks.
2500 Reported by Andreas M. Kirchwitz.
2502 PP/03 Report version information for many libraries, including
2503 Exim version information for dynamically loaded libraries. Created
2504 version.h, now support a version extension string for distributors
2505 who patch heavily. Dynamic module ABI change.
2507 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2508 privilege escalation vulnerability whereby the Exim run-time user
2509 can cause root to append content of the attacker's choosing to
2512 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2515 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2516 If dropping privileges for untrusted macros, we disabled normal logging
2517 on the basis that it would fail; for the Exim run-time user, this is not
2518 the case, and it resulted in successful deliveries going unlogged.
2519 Fixed. Reported by Andreas Metzler.
2525 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2526 only prepend for the Resent-* case. Fixes regression introduced in
2527 Exim 4.70 by NM/22 for Bugzilla 607.
2529 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2530 increasing numbers of administrators be bitten by this.
2532 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2534 PP/03 Bugzilla 994: added openssl_options main configuration option.
2536 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2538 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2540 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2542 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2543 without a peer certificate, leading to a segfault because of an
2544 assumption that peers always have certificates. Be a little more
2545 paranoid. Problem reported by Martin Tscholak.
2547 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2548 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2549 NB: ClamAV planning to remove STREAM in "middle of 2010".
2550 CL also introduces -bmalware, various -d+acl logging additions and
2551 more caution in buffer sizes.
2553 PP/09 Implemented reverse_ip expansion operator.
2555 PP/10 Bugzilla 937: provide a "debug" ACL control.
2557 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2559 PP/12 Bugzilla 973: Implement --version.
2561 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2563 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2565 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2567 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2568 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2569 ignore trailing whitespace.
2571 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2573 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2576 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2577 Notification from Dr Andrew Aitchison.
2579 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2580 ExtendedDetectionInfo response format.
2581 Notification from John Horne.
2583 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2586 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2587 XSL and documented dependency on system catalogs, with examples of how
2590 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2593 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2594 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2595 configuration file which is writeable by the Exim user or group.
2597 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2598 of configuration files to cover files specified with the -C option if
2599 they are going to be used with root privileges, not just the default
2602 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2603 option (effectively making it always true).
2605 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2606 files to be used while preserving root privileges.
2608 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2609 that rogue child processes cannot use them.
2611 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2612 run-time user, instead of root.
2614 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2615 Exim run-time user without dropping privileges.
2617 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2618 result string, instead of calling string_vformat() twice with the same
2621 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2622 for other users. Others should always drop root privileges if they use
2623 -C on the command line, even for a whitelisted configure file.
2625 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2627 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2633 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2634 $data_path, and $header_path variables; fixed documentation bugs and
2637 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2638 exipick to access non-standard spools, including the "frozen" queue
2641 NM/01 Bugzilla 965: Support mysql stored procedures.
2642 Patch from Alain Williams
2644 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2646 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2647 Patch from Andreas Metzler
2649 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2650 Patch from Kirill Miazine
2652 NM/05 Bugzilla 671: Added umask to procmail example.
2654 JJ/03 installed exipick 20100323.0, fixing doc bug
2656 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2657 directory. Notification and patch from Dan Rosenberg.
2659 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2661 TK/02 Improve log output when DKIM signing operation fails.
2663 MH/01 Treat the transport option dkim_domain as a colon separated
2664 list, not as a single string, and sign the message with each element,
2665 omitting multiple occurences of the same signer.
2667 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2668 Bugzilla 985, 986. Patch by Simon Arlott
2670 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2671 Patch by Simon Arlott
2673 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2674 MBX locking. Notification from Dan Rosenberg.
2680 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2682 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2684 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2686 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2688 NM/04 Bugzilla 915: Flag broken perl installation during build.
2694 TK/01 Added patch by Johannes Berg that expands the main option
2695 "spamd_address" if it starts with a dollar sign.
2697 TK/02 Write list of recipients to X-Envelope-Sender header when building
2698 the mbox-format spool file for content scanning (suggested by Jakob
2701 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2702 (http://www.dcc-servers.net/) support via dccifd. Activated by
2703 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2705 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2706 by Mark Daniel Reidel <mr@df.eu>.
2708 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2709 When building exim an external PCRE library is now needed -
2710 PCRE is a system library on the majority of modern systems.
2711 See entry on PCRE_LIBS in EDITME file.
2713 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2714 conversation. Added nologin parameter to request.
2715 Patch contributed by Kirill Miazine.
2717 TF/01 Do not log submission mode rewrites if they do not change the address.
2719 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2721 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2722 log files in place. Contributed by Roberto Lima.
2724 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2726 TF/03 Bugzilla 615: When checking the local_parts router precondition
2727 after a local_part_suffix or local_part_prefix option, Exim now
2728 does not use the address's named list lookup cache, since this
2729 contains cached lookups for the whole local part.
2731 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2732 Robert Millan. Documentation is in experimental-spec.txt.
2734 TF/04 Bugzilla 668: Fix parallel build (make -j).
2736 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2738 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2739 Patch provided by Jan Srzednicki.
2741 TF/05 Leading white space used to be stripped from $spam_report which
2742 wrecked the formatting. Now it is preserved.
2744 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2745 that they are available at delivery time.
2747 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2749 TF/08 TLS error reporting now respects the incoming_interface and
2750 incoming_port log selectors.
2752 TF/09 Produce a more useful error message if an SMTP transport's hosts
2753 setting expands to an empty string.
2755 NM/06 Bugzilla 744: EXPN did not work under TLS.
2756 Patch provided by Phil Pennock.
2758 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2759 Patch provided by Richard Godbee.
2761 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2762 acl_smtp_notquit, added index entry.
2764 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2765 Patch provided by Eugene Bujak.
2767 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2768 accept(). Patch provided by Maxim Dounin.
2770 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2771 Patch provided by Phil Pennock.
2773 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2775 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2776 Patch provided by Brad "anomie" Jorsch.
2778 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2779 Patch provided by Dean Brooks.
2781 TK/05 Add native DKIM support (does not depend on external libraries).
2783 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2784 Patch provided by Graeme Fowler.
2786 NM/16 Bugzilla 851: Documentation example syntax fix.
2788 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2790 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2793 NM/19 Bugzilla 745: TLS version reporting.
2794 Patch provided by Phil Pennock.
2796 NM/20 Bugzilla 167: bool: condition support.
2797 Patch provided by Phil Pennock.
2799 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2800 clients. Patch provided by Phil Pennock.
2802 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2803 Patch provided by Brad "anomie" Jorsch.
2805 NM/23 Bugzilla 687: Fix misparses in eximstats.
2806 Patch provided by Heiko Schlittermann.
2808 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2809 Patch provided by Heiko Schlittermann.
2811 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2812 plus update to original patch.
2814 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2816 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2817 Patch provided by David Brownlee.
2819 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2821 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2823 NM/30 Bugzilla 888: TLS documentation bugfixes.
2825 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2827 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2828 Unlike the original bugzilla I have changed all shell scripts in src tree.
2830 NM/33 Bugzilla 898: Transport filter timeout fix.
2831 Patch by Todd Rinaldo.
2833 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2834 Patch by Serge Demonchaux.
2836 NM/35 Bugzilla 39: Base64 decode bug fixes.
2837 Patch by Jakob Hirsch.
2839 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2841 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2843 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2845 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2851 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2852 ALT-N's libdkim that I have put here:
2853 http://duncanthrax.net/exim-experimental/
2855 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2856 were called 'true' and it seems that C99 defines that as a reserved
2857 keyword to be used with 'bool' variable types. That means you could
2858 not include C99-style headers which use bools without triggering
2859 build errors in sieve.c.
2861 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2862 as mailq or other aliases. Changed the --help handling significantly
2863 to do whats expected. exim_usage() emits usage/help information.
2865 SC/01 Added the -bylocaldomain option to eximstats.
2867 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2869 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2871 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2877 PH/01 Another patch from the Sieve maintainer.
2879 PH/02 When an IPv6 address is converted to a string for single-key lookup
2880 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2881 dots are used instead of colons so that keys in lsearch files need not
2882 contain colons. This was done some time before quoting was made available
2883 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2884 (notated using the quote facility) so as to distinguish them from IPv4
2885 keys. This meant that lookups for IP addresses in host lists did not work
2886 for iplsearch lookups.
2888 This has been fixed by arranging for IPv6 addresses to be expressed with
2889 colons if the lookup type is iplsearch. This is not incompatible, because
2890 previously such lookups could never work.
2892 The situation is now rather anomalous, since one *can* have colons in
2893 ordinary lsearch keys. However, making the change in all cases is
2894 incompatible and would probably break a number of configurations.
2896 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2899 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2900 conversion specification without a maximum field width, thereby enabling
2901 a rogue spamd server to cause a buffer overflow. While nobody in their
2902 right mind would setup Exim to query an untrusted spamd server, an
2903 attacker that gains access to a server running spamd could potentially
2904 exploit this vulnerability to run arbitrary code as the Exim user.
2906 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2907 $primary_hostname instead of what libspf2 thinks the hosts name is.
2909 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2910 a directory entry by the name of the lookup key. Previously, if a
2911 symlink pointed to a non-existing file or a file in a directory that
2912 Exim lacked permissions to read, a lookup for a key matching that
2913 symlink would fail. Now it is enough that a matching directory entry
2914 exists, symlink or not. (Bugzilla 503.)
2916 PH/03 The body_linecount and body_zerocount variables are now exported in the
2919 PH/04 Added the $dnslist_matched variable.
2921 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2922 This means they are set thereafter only if the connection becomes
2925 PH/06 Added the client_condition to authenticators so that some can be skipped
2926 by clients under certain conditions.
2928 PH/07 The error message for a badly-placed control=no_multiline_responses left
2929 "_responses" off the end of the name.
2931 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2933 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2934 (without spaces) instead of just copying the configuration text.
2936 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2938 PH/11 Added $max_received_linelength.
2940 PH/12 Added +ignore_defer and +include_defer to host lists.
2942 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2943 way in which PCRE > 7.0 is built.
2945 PH/14 Implemented queue_only_load_latch.
2947 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2948 MAIL command. The effect was to mangle the value on 64-bit systems.
2950 PH/16 Another patch from the Sieve maintainer.
2952 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2954 PH/18 If a system quota error occurred while trying to create the file for
2955 a maildir delivery, the message "Mailbox is full" was not appended to the
2956 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2957 only to a quota excession during the actual writing of the file.
2959 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2960 characters?) which causes problems in log lines. The DN values are now
2961 passed through string_printing() before being added to log lines.
2963 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2964 and InterBase are left for another time.)
2966 PH/21 Added message_body_newlines option.
2968 PH/22 Guard against possible overflow in moan_check_errorcopy().
2970 PH/23 POSIX allows open() to be a macro; guard against that.
2972 PH/24 If the recipient of an error message contained an @ in the local part
2973 (suitably quoted, of course), incorrect values were put in $domain and
2974 $local_part during the evaluation of errors_copy.
2980 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2981 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2984 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2985 issue a MAIL command.
2987 PH/02 In an ACL statement such as
2989 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2991 if a client was not listed at all, or was listed with a value other than
2992 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2993 the condition was not true (as it should be), so access was not denied.
2994 The bug was that the ! inversion was incorrectly passed on to the second
2995 item. This has been fixed.
2997 PH/03 Added additional dnslists conditions == and =& which are different from
2998 = and & when the dns lookup returns more than one IP address.
3000 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3001 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3003 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3004 FSYNC, which compiles an option called disable_fsync that allows for
3005 bypassing fsync(). The documentation is heavily laced with warnings.
3007 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3009 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3010 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3011 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3012 including adding "make clean"; (3) Added -fPIC when compiling the test
3013 dynamically loaded module, to get rid of a warning.
3015 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3016 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3017 = 0s. The bug is otherwise harmless.
3019 PH/07 There was a bug in the dovecot authenticator such that the value of
3020 $auth1 could be overwritten, and so not correctly preserved, after a
3021 successful authentication. This usually meant that the value preserved by
3022 the server_setid option was incorrect.
3024 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3026 PH/09 Installed PCRE release 7.0.
3028 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3029 run for batched SMTP input. It is now run at the start of every message
3030 in the batch. While fixing this I discovered that the process information
3031 (output by running exiwhat) was not always getting set for -bs and -bS
3032 input. This is fixed, and it now also says "batched" for BSMTP.
3034 PH/11 Added control=no_pipelining.
3036 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3037 patch, slightly modified), and move the expansion of helo_data till after
3038 the connection is made in the smtp transport (so it can use these
3041 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3043 PH/14 Added log_selector = +pid.
3045 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3047 PH/16 Add ${if forany and ${if forall.
3049 PH/17 Added dsn_from option to vary the From: line in DSNs.
3051 PH/18 Flush SMTP output before performing a callout, unless control =
3052 no_callout_flush is set.
3054 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3055 was true (the default) a successful delivery failed to delete the retry
3056 item, thus causing premature timeout of the address. The bug is now
3059 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3061 PH/21 Long custom messages for fakedefer and fakereject are now split up
3062 into multiline responses in the same way that messages for "deny" and
3063 other ACL rejections are.
3065 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3066 with slight modification.
3068 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3069 draft, changing the syntax and factoring some duplicate code".
3071 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3072 for deliveries of the second and subsequent messages over the same SMTP
3075 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3076 ${reduce, with only minor "tidies".
3078 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3080 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3081 expansion side effects.
3083 PH/27 When a message times out after an over-quota error from an Exim-imposed
3084 quota, the bounce message says "mailbox is full". This message was not
3085 being given when it was a system quota that was exceeded. It now should
3088 MH/03 Made $recipients available in local_scan(). local_scan() already has
3089 better access to the recipient list through recipients_list[], but
3090 $recipients can be useful in postmaster-provided expansion strings.
3092 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3093 in the case of a MAIL command with additional options following the
3094 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3095 were accidentally chopped off.
3097 PH/29 SMTP synchronization checks are implemented when a command is read -
3098 there is a check that no more input is waiting when there shouldn't be
3099 any. However, for some commands, a delay in an ACL can mean that it is
3100 some time before the response is written. In this time, more input might
3101 arrive, invalidly. So now there are extra checks after an ACL has run for
3102 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3103 pipelining has not been advertised.
3105 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3107 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3108 returned in the SMTP response when smtp_return_error_details was false.
3109 This has been fixed.
3111 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3112 instead of the C I/O that was originally supplied, because problems were
3113 reported on Solaris.
3115 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3116 Exim which did not show up earlier: it was assuming that a call to
3117 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3118 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3119 was a macro that became an assignment, so it seemed to work. This has
3120 changed to a proper function call with a void return, hence the compile
3121 error. Exim's code has been fixed.
3123 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3126 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3128 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3130 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3131 because it turns out to be too big or there is a timeout) there is no
3132 "Completed" line in the log. When some messages of this type were
3133 selected by exigrep, they were listed as "not completed". Others were
3134 picked up by some special patterns. I have improved the selection
3135 criteria to be more general.
3137 PH/38 The host_find_failed option in the manualroute router can now be set
3138 to "ignore", to completely ignore a host whose IP address cannot be
3139 found. If all hosts are ignored, the behaviour is controlled by the new
3140 host_all_ignored option.
3142 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3143 homing or because of multiple MX records with /mx) generated more than
3144 one IP address, and the following item turned out to be the local host,
3145 all the secondary addresses of the first item were incorrectly removed
3146 from the list, along with the local host and any following hosts (which
3147 is what is supposed to happen).
3149 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3150 whoever called Exim into the -H file. In the case of the daemon it was
3151 behaving confusingly. When first started, it used values for whoever
3152 started the daemon, but after a SIGHUP it used the Exim user (because it
3153 calls itself on a restart). I have changed the code so that it now always
3156 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3157 message are rejected with the same error (e.g. no authentication or bad
3158 sender address), and a DATA command is nevertheless sent (as can happen
3159 with PIPELINING or a stupid MUA), the error message that was given to the
3160 RCPT commands is included in the rejection of the DATA command. This is
3161 intended to be helpful for MUAs that show only the final error to their
3164 PH/42 Another patch from the Sieve maintainer.
3166 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3167 Eximstats - Fixed some broken HTML links and added missing column headers
3169 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3170 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3172 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3174 PH/43 Yet another patch from the Sieve maintainer.
3176 PH/44 I found a way to check for a TCP/IP connection going away before sending
3177 the response to the final '.' that terminates a message, but only in the
3178 case where the client has not sent further data following the '.'
3179 (unfortunately, this is allowed). However, in many cases there won't be
3180 any further data because there won't be any more messages to send. A call
3181 to select() can be used: if it shows that the input is "ready", there is
3182 either input waiting, or the socket has been closed. An attempt to read
3183 the next input character can distinguish the two cases. Previously, Exim
3184 would have sent an OK response which the client would never have see.
3185 This could lead to message repetition. This fix should cure that, at
3186 least in a lot of common cases.
3188 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3189 advertised in response to EHLO.
3195 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3196 fixed by 4.65/MH/01 (is this a record?) are fixed:
3198 (i) An empty string was always treated as zero by the numeric comparison
3199 operators. This behaviour has been restored.
3201 (ii) It is documented that the numeric comparison operators always treat
3202 their arguments as decimal numbers. This was broken in that numbers
3203 starting with 0 were being interpreted as octal.
3205 While fixing these problems I realized that there was another issue that
3206 hadn't been noticed. Values of message_size_limit (both the global option
3207 and the transport option) were treated as octal if they started with 0.
3208 The documentation was vague. These values are now always treated as
3209 decimal, and I will make that clear in the documentation.
3215 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3216 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3219 MH/01 Don't check that the operands of numeric comparison operators are
3220 integers when their expansion is in "skipping" mode (fixes bug
3221 introduced by 4.64-PH/07).
3223 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3224 child addresses, Exim now panics and dies. Previously, because the count
3225 is held in a short int, deliveries were likely to be lost. As such a
3226 large number of recipients for a single message is ridiculous
3227 (performance will be very, very poor), I have chosen to impose a limit
3228 rather than extend the field.
3234 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3235 leftover -K file (the existence of which was triggered by #402).
3236 While we were at it, introduced process PID as part of the -K
3237 filename. This should rule out race conditions when creating
3240 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3241 processing considerably. Previous code took too long for large mails,
3242 triggering a timeout which in turn triggers #401.
3244 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3245 in the DK code in transports.c. sendfile() is not really portable,
3246 hence the _LINUX specificness.
3248 TF/01 In the add_headers option to the mail command in an Exim filter,
3249 there was a bug that Exim would claim a syntax error in any
3250 header after the first one which had an odd number of characters
3253 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3254 callout verification, Exim cached a "reject" for the entire domain. This
3255 is correct for most verifications, but it is not correct for a recipient
3256 verification with use_sender or use_postmaster set, because in that case
3257 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3258 case of MAIL FROM:<> rejection from other early rejections (e.g.
3259 rejection of HELO). When verifying a recipient using a non-null MAIL
3260 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3261 Whatever the result of the callout, the value of the domain cache is
3262 left unchanged (for any other kind of callout, getting as far as trying
3263 RCPT means that the domain itself is ok).
3265 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3268 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3269 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3270 session, the daemon ignores ECONNECT errors and logs others; it now
3271 ignores EPIPE as well.
3273 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3274 (quoted-printable decoding).
3276 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3277 later the small subsequent patch to fix an introduced bug.
3279 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3281 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3283 PH/08 An error is now given if message_size_limit is specified negative.
3285 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3286 to be given (somewhat) arbitrary names.
3288 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3291 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3292 miscellaneous code fixes
3294 PH/10 Added the log_reject_target ACL modifier to specify where to log
3297 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3298 hostname. This is wrong, because it relates to the incoming message (and
3299 probably the interface on which it is arriving) and not to the outgoing
3300 callout (which could be using a different interface). This has been
3301 changed to use the value of the helo_data option from the smtp transport
3302 instead - this is what is used when a message is actually being sent. If
3303 there is no remote transport (possible with a router that sets up host
3304 addresses), $smtp_active_hostname is used.
3306 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3307 tweaks were necessary in order to get it to work (see also 21 below):
3308 (a) The code assumed that strncpy() returns a negative number on buffer
3309 overflow, which isn't the case. Replaced with Exim's string_format()
3311 (b) There were several signed/unsigned issues. I just did the minimum
3312 hacking in of casts. There is scope for a larger refactoring.
3313 (c) The code used strcasecmp() which is not a standard C function.
3314 Replaced with Exim's strcmpic() function.
3315 (d) The code set only $1; it now sets $auth1 as well.
3316 (e) A simple test gave the error "authentication client didn't specify
3317 service in request". It would seem that Dovecot has changed its
3318 interface. Fortunately there's a specification; I followed it and
3319 changed what the client sends and it appears to be working now.
3321 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3324 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3325 address A is aliased to B and C, where B exists and C does not. Without
3326 -v the output is "A verified" because verification stops after a
3327 successful redirection if more than one address is generated. However,
3328 with -v the child addresses are also verified. Exim was outputting "A
3329 failed to verify" and then showing the successful verification for C,
3330 with its parentage. It now outputs "B failed to verify", showing B's
3331 parentage before showing the successful verification of C.
3333 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3334 look up a TXT record in a specific list after matching in a combined
3337 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3338 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3339 they consult the DNS. I had assumed they would set it the way they
3340 wanted; and indeed my experiments on Linux seem to show that in some
3341 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3342 To be on the safe side, however, I have now made the interface to
3343 host_find_byname() similar to host_find_bydns(), with an argument
3344 containing the DNS resolver options. The host_find_byname() function now
3345 sets these options at its start, just as host_find_bydns() does. The smtp
3346 transport options dns_qualify_single and dns_search_parents are passed to
3347 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3348 of host_find_byname() use the default settings of RES_DEFNAMES
3349 (qualify_single) but not RES_DNSRCH (search_parents).
3351 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3352 spool_read_header() do less string testing, by means of a preliminary
3353 switch on the second character of optional "-foo" lines. (This is
3354 overdue, caused by the large number of possibilities that now exist.
3355 Originally there were few.) While I was there, I also converted the
3356 str(n)cmp tests so they don't re-test the leading "-" and the first
3357 character, in the hope this might squeeze out yet more improvement.
3359 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3360 flag allowing group syntax was set by the header_syntax check but not
3361 turned off, possible causing trouble later; (2) The flag was not being
3362 set at all for the header_verify test, causing "group"-style headers to
3363 be rejected. I have now set it in this case, and also caused header_
3364 verify to ignore an empty address taken from a group. While doing this, I
3365 came across some other cases where the code for allowing group syntax
3366 while scanning a header line wasn't quite right (mostly, not resetting
3367 the flag correctly in the right place). These bugs could have caused
3368 trouble for malformed header lines. I hope it is now all correct.
3370 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3371 with the "reply" argument non-NULL. The code, however (which originally
3372 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3373 but it didn't always do it. This confused somebody who was copying the
3374 code for some other use. I have removed all the tests.
3376 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3377 feature that was used to support insecure browsers during the U.S. crypto
3378 embargo. It requires special client support, and Exim is probably the
3379 only MTA that supported it -- and would never use it because real RSA is
3380 always available. This code has been removed, because it had the bad
3381 effect of slowing Exim down by computing (never used) parameters for the
3382 RSA_EXPORT functionality.
3384 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3385 authenticator to fail if there's a tab character in the incoming data
3386 (there should never be unless someone is messing about, as it's supposed
3387 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3388 if the connection is using TLS or if the remote IP is the same as the
3389 local IP, and the "valid-client-cert option" if a client certificate has
3392 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3393 authenticators. This can be used for authorization after authentication
3394 succeeds. (In the case of plaintext, it servers for both authentication
3397 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3398 if any retry times were supplied.
3400 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3401 connection, where there is no client IP address to check. In this
3402 situation, the verify now always succeeds.
3404 PH/25 Applied John Jetmore's -Mset patch.
3406 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3408 PH/27 In a string expansion for a processed (not raw) header when multiple
3409 headers of the same name were present, leading whitespace was being
3410 removed from all of them, but trailing whitespace was being removed only
3411 from the last one. Now trailing whitespace is removed from each header
3412 before concatenation. Completely empty headers in a concatenation (as
3413 before) are ignored.
3415 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3416 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3418 PH/29 [Removed. This was a change that I later backed out, and forgot to
3419 correct the ChangeLog entry (that I had efficiently created) before
3420 committing the later change.]
3422 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3423 address errors (4xx response to RCPT) over the same connection as other
3424 messages routed to the same hosts. Such deliveries are always "forced",
3425 so retry times are not inspected. This resulted in far too many retries
3426 for the affected addresses. The effect occurred only when there were more
3427 hosts than the hosts_max_try setting in the smtp transport when it had
3428 the 4xx errors. Those hosts that it had tried were not added to the list
3429 of hosts for which the message was waiting, so if all were tried, there
3430 was no problem. Two fixes have been applied:
3432 (i) If there are any address or message errors in an SMTP delivery, none
3433 of the hosts (tried or untried) are now added to the list of hosts
3434 for which the message is waiting, so the message should not be a
3435 candidate for sending over the same connection that was used for a
3436 successful delivery of some other message. This seems entirely
3437 reasonable: after all the message is NOT "waiting for some host".
3438 This is so "obvious" that I'm not sure why it wasn't done
3439 previously. Hope I haven't missed anything, but it can't do any
3440 harm, as the worst effect is to miss an optimization.
3442 (ii) If, despite (i), such a delivery is accidentally attempted, the
3443 routing retry time is respected, so at least it doesn't keep
3444 hammering the server.
3446 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3447 in ${readsocket because some servers need this prod.
3449 PH/32 Added some extra debug output when updating a wait-xxx database.
3451 PH/33 The hint "could be header name not terminated by colon", which has been
3452 given for certain expansion errors for a long time, was not being given
3453 for the ${if def:h_colon_omitted{... case.
3455 PH/34 The spec says: "With one important exception, whenever a domain list is
3456 being scanned, $domain contains the subject domain." There was at least
3457 one case where this was not true.
3459 PH/35 The error "getsockname() failed: connection reset by peer" was being
3460 written to the panic log as well as the main log, but it isn't really
3461 panic-worthy as it just means the connection died rather early on. I have
3462 removed the panic log writing for the ECONNRESET error when getsockname()
3465 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3466 runs only) independently of the message's sender address. This meant
3467 that, if the 4xx error was in fact related to the sender, a different
3468 message to the same recipient with a different sender could confuse
3469 things. In particular, this can happen when sending to a greylisting
3470 server, but other circumstances could also provoke similar problems.
3471 I have changed the default so that the retry time for these errors is now
3472 based a combination of the sender and recipient addresses. This change
3473 can be overridden by setting address_retry_include_sender=false in the
3476 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3477 remote server are returned as part of bounce messages. This was not
3478 happening for LMTP over a pipe (the lmtp transport), but now it is the
3479 same for both kinds of LMTP.
3481 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3482 in header lines that were in fact CNAMEs. This is no longer the case.
3484 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3485 and queue runs started by the daemon processed all messages. This has
3486 been fixed so that -R and -S can now usefully be given with -q<time>.
3488 PH/40 Import PCRE release 6.7 (fixes some bugs).
3490 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3492 PH/42 Give an error if -q is specified more than once.
3494 PH/43 Renamed the variables $interface_address and $interface_port as
3495 $received_ip_address and $received_port, to make it clear that these
3496 values apply to message reception, and not to the outgoing interface when
3497 a message is delivered. (The old names remain recognized, of course.)
3499 PH/44 There was no timeout on the connect() call when using a Unix domain
3500 socket in the ${readsocket expansion. There now is.
3502 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3503 be meaningful with "accept".
3505 SC/01 Eximstats V1.43
3506 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3508 SC/02 Eximstats V1.44
3509 Use a glob alias rather than an array ref in the generated
3510 parser. This improves both readability and performance.
3512 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3513 Collect SpamAssassin and rejection statistics.
3514 Don't display local sender or destination tables unless
3515 there is data to show.
3516 Added average volumes into the top table text output.
3518 SC/04 Eximstats V1.46
3519 Collect data on the number of addresses (recipients)
3520 as well as the number of messages.
3522 SC/05 Eximstats V1.47
3523 Added 'Message too big' to the list of mail rejection
3524 reasons (thanks to Marco Gaiarin).
3526 SC/06 Eximstats V1.48
3527 Mainlog lines which have GMT offsets and are too short to
3528 have a flag are now skipped.
3530 SC/07 Eximstats V1.49 (Alain Williams)
3531 Added the -emptyok flag.
3533 SC/08 Eximstats V1.50
3534 Fixes for obtaining the IP address from reject messages.
3536 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3537 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3538 whitespace changes from 4.64-PH/27
3540 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3543 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3544 are found, allow negative numbers in numeric criteria)
3546 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3548 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3549 to match changes made in 4.64-PH/43
3551 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3553 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3554 "domain" is actually an IP address, and force a failure. This locks out
3555 those revolvers/nameservers that support "A-for-A" lookups, in
3556 contravention of the specifications.
3558 PH/48 When a host name was looked up from an IP address, and the subsequent
3559 forward lookup of the name timed out, the host name was left in
3560 $sender_host_name, contrary to the specification.
3562 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3563 restricted to single-key lookups, Exim was not diagnosing an error if
3564 * or *@ was used with a query-style lookup.
3566 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3568 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3569 long ago, but noone interested enough thought of it. Let's just say that
3570 the "1.1" means that there are some new functions that weren't there at
3571 some point in the past.
3573 PH/51 Error processing for expansion failure of helo_data from an smtp
3574 transport during callout processing was broken.
3576 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3577 tested/used via the -bh/-bhc/-bs options.
3579 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3580 bug, fixed in subsequent PCRE releases).
3582 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3583 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3585 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3591 SC/01 Use a glob alias rather than an array ref in eximstats generated
3592 parser. This improves both readability and performance.
3594 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3595 Don't display local sender or destination tables in eximstats unless
3596 there is data to show.
3597 Added average volumes into the eximstats top table text output.
3599 SC/03 Collect data on the number of addresses (recipients) as well
3600 as the number of messages in eximstats.
3602 TF/01 Correct an error in the documentation for the redirect router. Exim
3603 does (usually) call initgroups() when daemonizing.
3605 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3606 with consistent privilege compared to when running as a daemon.
3608 TF/03 Note in the spec that $authenticated_id is not set for local
3609 submissions from trusted users.
3611 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3612 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3614 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3615 by adding some example configuration directives to the default
3616 configuration file. A little bit of work is required to uncomment the
3617 directives and define how usernames and passwords are checked, but
3618 there is now a framework to start from.
3620 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3621 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3622 without this. I don't know how relevant this is to other LDAP libraries.
3624 PH/02 Add the verb name to the "unknown ACL verb" error.
3626 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3628 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3630 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3631 directory not expanded when it should be if an expanded home directory
3632 was set for the address (which is overridden by the transport).
3634 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3637 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3638 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3639 because it is too late at that time, and has no effect.
3641 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3642 security issue with \' (bugzilla #107). I could not use the
3643 PQescapeStringConn() function, because it needs a PGconn value as one of
3646 PH/08 When testing addresses using -bt, indicate those final addresses that
3647 are duplicates that would not cause an additional delivery. At least one
3648 person was confused, thinking that -bt output corresponded to deliveries.
3649 (Suppressing duplicates isn't a good idea as you lose the information
3650 about possibly different redirections that led to the duplicates.)
3652 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3653 systems where poll() doesn't work, in particular OS X.
3655 PH/10 Added more information to debugging output for retry time not reached.
3657 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3658 operations in malware.c.
3660 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3663 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3664 syslog_duplication set false, log lines that would normally be written
3665 both the the main log and to the reject log were not written to syslog at
3668 PH/14 In the default configuration, change the use of "message" in ACL warn
3669 statements to "add_header".
3671 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3672 not followed by a command (e.g. "seen endif").
3674 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3675 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3678 PH/17 Added extra conditions to the default value of delay_warning_condition
3682 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3683 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3684 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3687 The Auto-Submitted: and various List- headers are standardised, whereas I
3688 don't think Precedence: ever was.
3690 PH/18 Refactored debugging code in route_finduser() to show more information,
3691 in particular, the error code if getpwnam() issues one.
3693 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3694 This is apparently needed in addition to the PH/07 change above to avoid
3695 any possible encoding problems.
3697 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3698 but not after initializing Perl.
3700 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3701 output them only if debugging. By default they are written stderr,
3702 apparently, which is not desirable.
3704 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3707 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3710 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3712 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3713 authenticated or an ident call has been made. Suppress the default
3714 values for $authenticated_id and $authenticated_sender (but permit -oMai
3715 and -oMas) when testing with -bh.
3717 PH/24 Re-jigged the order of the tests in the default configuration so that the
3718 tests for valid domains and recipients precede the DNS black list and CSA
3719 tests, on the grounds that those ones are more expensive.
3721 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3722 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3723 This bug exists in every version of Exim that I still have, right back to
3726 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3727 However, an attempt to turn on case-sensitivity in a regex key by
3728 including (?-i) didn't work because the subject string was already
3729 lowercased, and the effects were non-intuitive. It turns out that a
3730 one-line patch can be used to allow (?-i) to work as expected.
3736 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3737 other effects) broke the use of negated acl sub-conditions.
3739 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3742 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3743 "Deny" causes Exim to reject the incoming connection with a 554 error.
3744 Unfortunately, if there is a major crisis, such as a disk failure,
3745 tcp-wrappers gives "deny", whereas what one would like would be some
3746 kind of temporary error. A kludge has been added to help with this.
3747 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3748 554 error is used if errno is still zero or contains ENOENT (which occurs
3749 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3752 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3754 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3755 errors. Otherwise a message that provokes a temporary error (when other
3756 messages do not) can cause a whole host to time out.
3758 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3759 addresses were routed directly to files or pipes from a redirect router.
3760 File deliveries just didn't batch; pipe deliveries might have suffered
3763 PH/06 A failure to get a lock for a hints database would erroneously always say
3764 "Failed to get write lock", even when it was really a read lock.
3766 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3767 of 0600. This has been changed to use the value of the lockfile_mode
3768 option (which defaults to 0600).
3770 PH/08 Applied small patch from the Sieve maintainer.
3772 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3773 folder from quota calculations, a direct delivery into this folder messed
3774 up the contents of the maildirsize file. This was because the regex was
3775 used only to exclude .Trash (or whatever) when the size of the mailbox
3776 was calculated. There was no check that a delivery was happening into an
3777 excluded directory. This bug has been fixed by ignoring all quota
3778 processing for deliveries into excluded directories.
3780 PH/10 Added the maildirfolder_create_regex option to appendfile.
3786 PH/01 The code for finding all the local interface addresses on a FreeBSD
3787 system running IPv6 was broken. This may well have applied to all BSD
3788 systems, as well as to others that have similar system calls. The broken
3789 code found IPv4 interfaces correctly, but gave incorrect values for the
3790 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3791 that it would not match correctly against @[] and not recognize the IPv6
3794 PH/02 The ipliteral router was not recognizing addresses of the form user@
3795 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3797 PH/03 Added disable_ipv6.
3799 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3800 decoded form, because it is most often used to construct To: headers
3801 lines in autoreplies, and the decoded form may well be syntactically
3802 invalid. However, $reply_address has leading white space removed, and all
3803 newlines turned into spaces so that the autoreply transport does not
3806 PH/05 If group was specified without a user on a router, and no group or user
3807 was specified on a transport, the group from the router was ignored.
3809 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3810 for visible compile-time settings that can be used to change these
3811 numbers, for those that want even more. Backwards compatibility with old
3812 spool files has been maintained. However, going back to a previous Exim
3813 release will lost any variables that are in spool files.
3815 PH/07 Two small changes when running in the test harness: increase delay when
3816 passing a TCP/IP connection to a new process, in case the original
3817 process has to generate a bounce, and remove special handling of
3818 127.0.0.2 (sic), which is no longer necessary.
3820 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3821 be the same on different OS.
3823 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3826 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3827 whether --show-vars was specified or not
3829 JJ/02 exipick: Added support for new ACL variable spool format introduced
3832 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3833 syntactically invalid From: or Reply-to: line, and a filter used this to
3834 generate an autoreply, and therefore failed to obtain an address for the
3835 autoreply, Exim could try to deliver to a non-existent relative file
3836 name, causing unrelated and misleading errors. What now happens is that
3837 it logs this as a hard delivery error, but does not attempt to create a
3840 PH/11 The exinext utility has a -C option for testing purposes, but although
3841 the given file was scanned by exinext itself; it wasn't being passed on
3842 when Exim was called.
3844 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3845 an end-of-file indication when reading a command response.
3847 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3848 compiled. In many other places in Exim, IPv6 addresses are always
3849 recognized, so I have changed this. It also means that IPv4 domain
3850 literals of the form [IPV4:n.n.n.n] are now always recognized.
3852 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3853 used if the router is not running as root, for example, when verifying at
3854 ACL time, or when using -bh. The debugging output from this situation was
3855 non-existent - all you got was a failure to exec. I have made two
3858 (a) Failures to set uid/gid, the current directory, or a process leader
3859 in a subprocess such as that created by queryprogram now generate
3860 suitable debugging output when -d is set.
3862 (b) The queryprogram router detects when it is not running as root,
3863 outputs suitable debugging information if -d is set, and then runs
3864 the subprocess without attempting to change uid/gid.
3866 PH/15 Minor change to Makefile for building test_host (undocumented testing
3869 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3870 additional section of a DNS packet that returns MX or SRV records.
3871 Instead, it always explicitly searches for A/AAAA records. This avoids
3872 major problems that occur when a DNS server includes only records of one
3873 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3874 fixed another bug: if SRV records were looked up and the corresponding
3875 address records were *not* found in the additional section, the port
3876 values from the SRV records were lost.
3878 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3879 using the correct key (the original address) when searching the retry
3880 rules in order to find which one to use for generating the retry hint.
3882 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3883 adding the default one. Similarly, if it contains a Reply-To: header, the
3884 errors_reply_to option, if set, is not used.
3886 PH/19 When calculating a retry time, Exim used to measure the "time since
3887 failure" by looking at the "first failed" field in the retry record. Now
3888 it does not use this if it is later than than the arrival time of the
3889 message. Instead it uses the arrival time. This makes for better
3890 behaviour in cases where some deliveries succeed, thus re-setting the
3891 "first failed" field. An example is a quota failure for a huge message
3892 when small messages continue to be delivered. Without this change, the
3893 "time since failure" will always be short, possible causing more frequent
3894 delivery attempts for the huge message than are intended.
3895 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3897 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3898 $1, $2, $3) because the numerical variables can be reset during some
3899 expansion items (e.g. "match"), thereby losing the authentication data.
3901 PH/21 Make -bV show the size of off_t variables so that the test suite can
3902 decide whether to run tests for quotas > 2G.
3904 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3905 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3906 filecount value is greater than 2G or if a quota value is greater than 2G
3907 on a system where the size of off_t is not greater than 4, a panic error
3910 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3911 never match. The debug and -bh output now contains an explicit error
3912 message indicating a malformed IPv4 address or mask.
3914 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3915 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3916 PH/23 above applies.
3918 PH/25 Do not write to syslog when running in the test harness. The only
3919 occasion when this arises is a failure to open the main or panic logs
3920 (for which there is an explicit test).
3922 PH/26 Added the /no_tell option to "control=freeze".
3924 PH/27 If a host name lookup failed very early in a connection, for example, if
3925 the IP address matched host_lookup and the reverse lookup yielded a name
3926 that did not have a forward lookup, an error message of the form "no IP
3927 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3928 could be logged. Now it outputs the IP address instead of "NULL".
3930 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3931 allows the sender and the authenticated sender to be set when
3932 submitting a message from within Exim. Since child_open_exim() is
3933 documented for local_scan(), the new function should be too.
3935 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3936 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3937 results in an empty string is now treated as unset.
3939 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3941 PH/31 Added sender_verify_fail logging option.
3943 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3944 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3945 tidied the source and removed it altogether.
3947 PH/33 When a queue run was abandoned because the load average was too high, a
3948 log line was always written; now it is written only if the queue_run log
3949 selector is set. In addition, the log line for abandonment now contains
3950 information about the queue run such as the pid. This is always present
3951 in "start" and "stop" lines but was omitted from the "abandon" line.
3953 PH/34 Omit spaces between a header name and the colon in the error message that
3954 is given when verify = headers_syntax fails (if there are lots of them,
3955 the message gets confusing).
3957 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3958 names, as there are now some PTR records that contain slashes. This check
3959 is only to protect against broken name servers that fall over on strange
3960 characters, so the fact that it applies to all lookups doesn't matter.
3962 PH/36 Now that the new test suite is complete, we can remove some of the
3963 special code in Exim that was needed for the old test suite. For example,
3964 sorting DNS records because real resolvers return them in an arbitrary
3965 order. The new test suite's fake resolver always returns records in the
3968 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3969 bounces) except when queue_only is set, to avoid logging races between
3970 the different processes.
3972 PH/38 Panic-die if .include specifies a non-absolute path.
3974 PH/39 A tweak to the "H" retry rule from its user.
3976 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3977 a label. They prevented compilation on older perls.
3979 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3980 a warning to be raised on newish perls.
3982 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3983 on queue. Changes to match documented behaviour of showing count of
3984 messages matching specified criteria.
3986 PH/40 Changed the default ident timeout from 30s to 5s.
3988 PH/41 Added support for the use of login_cap features, on those BSD systems
3989 that have them, for controlling the resources used by pipe deliveries.
3991 PH/42 The content-scanning code uses fopen() to create files in which to put
3992 message data. Previously it was not paying any attention to the mode of
3993 the files. Exim runs with umask(0) because the rest of the code creates
3994 files with open(), and sets the required mode explicitly. Thus, these
3995 files were ending up world-writeable. This was not a big issue, because,
3996 being within the spool directory, they were not world-accessible. I have
3997 created a function called modefopen, which takes an additional mode
3998 argument. It sets umask(777), creates the file, chmods it to the required
3999 mode, then resets the umask. All the relevant calls to fopen() in the
4000 content scanning code have been changed to use this function.
4002 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4003 to 24 hours. This avoids potential overflow problems when processing G
4004 and H retry rules. I suspect nobody ever tinkers with this value.
4006 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4008 PH/45 When the plaintext authenticator is running as a client, the server's
4009 challenges are checked to ensure they are valid base64 strings. By
4010 default, the authentication attempt is cancelled if an invalid string is
4011 received. Setting client_ignore_invalid_base64 true ignores these errors.
4012 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4013 they are received. Thus, the responses can be made to depend on the
4014 challenges. If an invalid string is ignored, an empty string is placed in
4017 PH/46 Messages that are created by the autoreply transport now contains a
4018 References: header, in accordance with RFCs 2822 and 3834.
4020 PH/47 Added authenticated_sender_force to the smtp transport.
4022 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4024 PH/49 Installed latest patch from the Sieve maintainer.
4026 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4027 was also set, the appendfile transport was unnecessarily scanning a
4028 directory of message files (e.g. for maildir delivery) to find the count
4029 of files (along with the size), even though it did not need this
4030 information. It now does the scan only if it needs to find either the
4031 size of the count of files.
4033 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4035 PH/52 Two bugs concerned with error handling when the smtp transport is
4038 (i) Exim was not creating retry information for temporary errors given
4039 for individual recipients after the DATA command when the smtp transport
4040 was used in LMTP mode. This meant that they could be retried too
4041 frequently, and not timed out correctly.
4043 (ii) Exim was setting the flag that allows error details to be returned
4044 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4045 recipients that were returned after the DATA command.
4047 PH/53 This is related to PH/52, but is more general: for any failing address,
4048 when detailed error information was permitted to be returned to the
4049 sender, but the error was temporary, then after the final timeout, only
4050 "retry timeout exceeded" was returned. Now it returns the full error as
4051 well as "retry timeout exceeded".
4053 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4054 do this, and (what is worse) MTAs that accept it.
4056 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4057 will now be deprecated.
4059 PH/56 New os.c-cygwin from the Cygwin maintainer.
4061 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4062 formats (previously only available in exim formats via -bpr, -bpru,
4063 and -bpra. Now also available in native and exiqgrep formats)
4065 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4066 with very large, slow to parse queues
4068 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4070 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4072 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4073 responses to authentication challenges, though it was showing the
4074 challenges; (ii) I've removed the CR characters from the debug output for
4077 PH/58 Allow for the insertion of a newline as well as a space when a string
4078 is turned into more than one encoded-word during RFC 2047 encoding. The
4079 Sieve code now uses this.
4081 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4082 data_4xx, lost_connection, tls_required.
4084 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4085 message was being sent as an SMTP response.
4087 PH/61 Add -l and -k options to exicyclog.
4089 PH/62 When verifying, if an address was redirected to one new address, so that
4090 verification continued, and the new address failed or deferred after
4091 having set something in $address_data, the value of $address_data was not
4092 passed back to the ACL. This was different to the case when no
4093 redirection occurred. The value is now passed back in both cases.
4095 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4096 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4097 use login_cap.h, so on its own it isn't the distinguishing feature. The
4098 new name refers directly to the setclassresources() function.
4100 PH/65 Added configuration files for NetBSD3.
4102 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4104 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4105 is preferred over IPv4.
4107 PH/68 The bounce_return_message and bounce_return_body options were not being
4108 honoured for bounces generated during the reception of non-SMTP messages.
4109 In particular, this applied to messages rejected by the ACL. This bug has
4110 been fixed. However, if bounce_return_message is true and bounce_return_
4111 body is false, the headers that are returned for a non-SMTP message
4112 include only those that have been read before the error was detected.
4113 (In the case of an ACL rejection, they have all been read.)
4115 PH/69 The HTML version of the specification is now built in a directory called
4116 spec_html instead of spec.html, because the latter looks like a path with
4117 a MIME-type, and this confuses some software.
4119 PH/70 Catch two compiler warnings in sieve.c.
4121 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4122 function verify_get_ident() calls ip_connect() to connect a socket, but
4123 if the "connect()" function timed out, ip_connect() used to close the
4124 socket. However, verify_get_ident() also closes the socket later, and in
4125 between Exim writes to the log, which may get opened at this point. When
4126 the socket was closed in ip_connect(), the log could get the same file
4127 descriptor number as the socket. This naturally causes chaos. The fix is
4128 not to close the socket in ip_connect(); the socket should be closed by
4129 the function that creates it. There was only one place in the code where
4130 this was missing, in the iplookup router, which I don't think anybody now
4131 uses, but I've fixed it anyway.
4133 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4134 well as to direct DNS lookups. Otherwise the handling of names in host
4135 lists is inconsistent and therefore confusing.
4141 PH/01 Two changes to the default runtime configuration:
4143 (1) Move the checks for relay_from_hosts and authenticated clients from
4144 after to before the (commented out) DNS black list checks.
4146 (2) Add control=submission to the relay_from_hosts and authenticated
4147 clients checks, on the grounds that messages accepted by these
4148 statements are most likely to be submissions.
4150 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4152 (1) Generate an error if the third argument for the ${prvs expansion is
4155 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4158 (3) Reset the variables that are obtained from the first argument of
4159 ${prvscheck and used in the second argument before leaving the code,
4160 because their memory is reclaimed, so using them afterwards may do
4163 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4164 one (it's much easier than Tom thought :-).
4166 (5) Because of (4), we can now allow for the use of $prvscheck_result
4167 inside the third argument.
4169 PH/03 For some reason, the default setting of PATH when running a command from
4170 a pipe transport was just "/usr/bin". I have changed it to
4173 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4174 anything to be listed in the output from -bV.
4176 PH/05 When a filter generated an autoreply, the entire To: header line was
4177 quoted in the delivery log line, like this:
4179 => >A.N.Other <ano@some.domain> <original@ddress> ...
4181 This has been changed so that it extracts the operative address. There
4182 may be more than one such address. If so, they are comma-separated, like
4185 => >ano@some.domain,ona@other.domain <original@ddress> ...
4187 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4188 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4189 not being looked up in the rDNS to get a host name, Exim was showing the
4190 IP address twice in Received: lines, even though the IP addresses were
4191 identical. For example:
4193 Received: from [1.2.3.4] (helo=[1.2.3.4])
4195 However, if the real host name was known, it was omitting the HELO data
4196 if it matched the actual IP address. This has been tidied up so that it
4197 doesn't show the same IP address twice.
4199 PH/07 When both +timestamp and +memory debugging was on, the value given by
4200 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4201 called by the debug printing, thereby overwriting the timestamp buffer.
4202 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4204 PH/08 When the original message was included in an autoreply transport, it
4205 always said "this is a copy of the message, including all the headers",
4206 even if body_only or headers_only was set. It now gives an appropriate
4209 PH/09 Applied a patch from the Sieve maintainer which:
4211 o fixes some comments
4212 o adds the (disabled) notify extension core
4213 o adds some debug output for the result of if/elsif tests
4214 o points to the current vacation draft in the documentation
4215 and documents the missing references header update
4219 o fixes a bug in processing the envelope test (when testing
4220 multiple envelope elements, the last element determined the
4223 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4224 Electronic Mail") by including:
4226 Auto-submitted: auto-generated
4228 in the messages that it generates (bounce messages and others, such as
4229 warnings). In the case of bounce messages for non-SMTP messages, there was
4230 also a typo: it was using "Auto_submitted" (underscore instead of
4231 hyphen). Since every message generated by Exim is necessarily in response
4232 to another message, thes have all been changed to:
4234 Auto-Submitted: auto-replied
4236 in accordance with these statements in the RFC:
4238 The auto-replied keyword:
4240 - SHOULD be used on messages sent in direct response to another
4241 message by an automatic process,
4243 - MUST NOT be used on manually-generated messages,
4245 - MAY be used on Delivery Status Notifications (DSNs) and Message
4246 Disposition Notifications (MDNs),
4248 - MUST NOT be used on messages generated by automatic or periodic
4249 processes, except for messages which are automatic responses to
4252 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4253 to the default Received: header definition.
4255 PH/12 Added log selector acl_warn_skipped (default on).
4257 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4258 variables because (a) they are in the wrong storage pool and (b) even if
4259 they were copied, it wouldn't work properly because of the caching.
4261 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4262 checking when decoding. Apparently there are clients that generate
4263 overlong encoded strings. Why am I not surprised?
4265 PH/15 If the first argument of "${if match_address" was not empty, but did not
4266 contain an "@" character, Exim crashed. Now it writes a panic log message
4267 and treats the condition as false.
4269 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4271 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4272 extension "envelope-auth". The code is finished and in agreement with
4273 other implementations, but there is no documentation so far and in fact,
4274 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4275 not changing the active code.
4277 Print executed "if" and "elsif" statements when debugging is used. This
4278 helps a great deal to understand what a filter does.
4280 Document more things not specified clearly in RFC3028. I had all this
4281 sorted out, when out of a sudden new issues came to my mind. Oops."
4283 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4286 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4287 canonical form (as documented). However, after a host name lookup from
4288 the IP address, check_host() was doing a simple string comparison with
4289 addresses acquired from the DNS when checking that the found name did
4290 have the original IP as one of its addresses. Since any found IPv6
4291 addresses are likely to be in abbreviated form, the comparison could
4292 fail. Luckily, there already exists a function for doing the comparison
4293 by converting both addresses to binary, so now that is used instead of
4294 the text comparison.
4296 PH/20 There was another similar case to PH/19, when a complete host name was
4297 given in a host list; looking up its IP address could give an abbreviated
4298 form, whereas the current host's name might or might not be abbreviated.
4299 The same fix has been applied.
4305 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4306 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4309 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4310 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4312 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4314 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4315 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4316 Also, for "Auto-Submitted", treat anything other than "no" as
4317 non-personal, in accordance with RFC 3834. (Previously it treated
4318 anything starting "auto-" as non-personal.)
4320 TF/01 The control=submission/name=... option had a problem with syntax
4321 errors if the name included a slash character. The /name= option
4322 now slurps the rest of the string, so it can include any characters
4323 but it must come last in the list of options (after /sender_retain
4326 PH/05 Some modifications to the interface to the fake nameserver for the new
4334 TK/01 Added the "success_on_redirect" address verification option. See
4335 NewStuff for rationale and an example.
4337 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4339 PH/02 Patch to exigrep to allow it to work on syslog lines.
4341 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4342 fread() to skip over the body file's header line, because in Cygwin the
4343 header line is locked and is inaccessible.
4345 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4346 co-exist for some time) to make it clear that it is the Exim ID that is
4347 referenced, not the Message-ID: header line.
4349 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4350 string_format() function, because snprintf() does not exist on all
4353 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4354 ${stat: expansion item.
4356 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4357 protocol synchronization error", to keep the pedants happy.
4359 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4360 well as for IRIX systems, when gcc is being used. See the host.c source
4363 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4365 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4368 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4369 transport and to the smtp transport in LMTP mode.
4371 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4373 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4374 run a filter in a subprocess. This could lead to confusion in subsequent
4375 lookups in the parent process. There should also be a search_tidyup() at
4376 the end of the subprocess.
4378 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4379 only if the host matched helo_try_verify_hosts, which caused the
4380 verification to occur when the EHLO/HELO command was issued. The ACL just
4381 tested the remembered result. Now, if a previous verification attempt has
4382 not happened, "verify = helo" does it there and then.
4384 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4386 TK/03 Fix log output including CR from clamd.
4388 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4389 exist provoked a memory error which could cause a segfault.
4391 PH/15 Installed PCRE 6.2
4393 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4395 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4396 of the problem. Specifically, suggested +O2 rather than +O1 for the
4399 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4401 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4402 record, the port was not correctly logged when the outgoing_port log
4403 selector was set (it logged the transort's default port).
4405 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4406 fallback_hosts, and "hosts" in the smtp transport.
4408 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4409 host errors such as "Connection refused".
4411 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4412 authentication with radiusclient 0.4.9:
4414 - Error returned from rc_read_config was caught wrongly
4415 - Username/password not passed on to radius server due to wrong length.
4417 The presumption is that some radiusclient API changes for 4.51/PH/17
4418 were not taken care of correctly. The code is still untested by me (my
4419 Linux distribution still has 0.3.2 of radiusclient), but it was
4420 contributed by a Radius user.
4422 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4423 expanding the "port" option of the smtp transport.
4425 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4426 while reading a MIME header. Thanks to Tom Hughes for a patch.
4428 PH/24 Include config.h inside local_scan.h so that configuration settings are
4431 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4432 that in an ACL for RCPT (for example), you can examine exactly what was
4435 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4436 commands, but it was not correctly comparing the address with the actual
4437 client host address. Thus, it would show the EHLO address in Received:
4438 header lines when this was not necessary.
4440 PH/27 Added the % operator to ${eval:}.
4442 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4443 it should be ignoring failures (because with -C, for example, it has lost
4444 privilege). It wasn't ignoring creation failures other than "already
4447 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4450 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4451 because an input line was too long, either on its own, or by virtue of
4452 too many continuations, the temporary file was not being removed, and the
4453 return code was incorrect.
4455 PH/31 Missing "BOOL" in function definition in filtertest.c.
4457 PH/32 Applied Sieve patches from the maintainer.
4459 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4461 PH/33 Added "verify = not_blind".
4463 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4464 Local/Makefile (with some defaults set). These are used in built scripts
4465 such as exicyclog, but they have never been used in the exim_install
4466 script (though there are many overriding facilities there). I have
4467 arranged that the exim_install script now takes note of these two
4470 PH/35 Installed configuration files for Dragonfly.
4472 PH/36 When a locally submitted message by a trusted user did not contain a
4473 From: header, and the sender address was obtained from -f or from an SMTP
4474 MAIL command, and the trusted user did not use -F to supply a sender
4475 name, $originator_name was incorrectly used when constructing a From:
4476 header. Furthermore, $originator_name was used for submission mode
4477 messages from external hosts without From: headers in a similar way,
4478 which is clearly wrong.
4480 PH/37 Added control=suppress_local_fixups.
4482 PH/38 When log_selector = +received_sender was set, and the addition of the
4483 sender made the log line's construction buffer exactly full, or one byte
4484 less than full, an overflow happened when the terminating "\n" was
4487 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4488 when the result of a list match is failure because a DNS lookup failed.
4490 PH/40 RM_COMMAND is now used in the building process.
4492 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4493 the "build-* directories that it finds.
4495 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4496 address in a domain literal was a prefix of an interface address.
4498 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4499 when verifying a sender address, unless rewrite_headers is false.
4501 PH/44 Wrote a long comment about why errors_to addresses are verified as
4502 recipients, not senders.
4504 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4505 the ratelimit ACL was added.
4507 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4509 PH/46 Added extra information about PostgreSQL errors to the error string.
4511 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4512 avoiding the need to install special zones in a real server. This is
4513 backwards compatible; if it can't find the fake resolver, it drops back.
4514 Thus, both old and new test suites can be run.
4516 TF/02 Added util/ratelimit.pl
4518 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4519 clock is set back in time.
4521 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4522 Candler <B.Candler@pobox.com>.
4524 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4525 OK for addresses that are the result of redirections.
4527 PH/48 A number of further additions for the benefit of the new test suite,
4528 including a fake gethostbyname() that interfaces to the fake DNS resolver
4531 TF/06 The fix for widen_domains has also been applied to qualify_single and
4532 search_parents which are the other dnslookup options that can cause
4535 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4538 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4540 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4541 service" if it encountered an MX record with an empty target hostname.
4542 The message is now "an MX or SRV record indicated no SMTP service".
4544 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4545 if the DNS of the sending site is misconfigured. This is quite a
4546 common situation. This change restores the behaviour of treating a
4547 helo verification defer as a failure.
4549 PH/51 If self=fail was set on a router, the bounce message did not include the
4550 actual error message.
4556 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4558 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4559 command itself ended in error, the underlying message about the transport
4560 filter timeout was being overwritten with the pipe command error. Now the
4561 underlying error message should be appended to the second error message.
4563 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4565 PH/02 Reduce the amount of output that "make" produces by default. Full output
4566 can still be requested.
4568 PH/03 The warning log line about a condition test deferring for a "warn" verb
4569 was being output only once per connection, rather than after each
4570 occurrence (because it was using the same function as for successful
4571 "warn" verbs). This seems wrong, so I have changed it.
4573 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4574 it should not have, which might have caused a crash in the right
4575 circumstances, but probably never did.
4577 PH/04 Installed a modified version of Tony Finch's patch to make submission
4578 mode fix the return path as well as the Sender: header line, and to
4579 add a /name= option so that you can make the user's friendly name appear
4582 TF/03 Added the control = fakedefer ACL modifier.
4584 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4585 Mark Lowes for thorough testing.
4587 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4589 TK/03 Merged latest SRS patch from Miles Wilton.
4591 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4592 with the definition in sysexits.h (which is #included earlier).
4593 Fortunately, Exim does not actually use EX_OK. The code used to try to
4594 preserve the sysexits.h value, by assuming that macro definitions were
4595 scanned for macro replacements. I have been disabused of this notion,
4596 so now the code just undefines EX_OK before #including unistd.h.
4598 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4599 in the smtp transport. When a block could not be written in a single
4600 write() function, the timeout was being re-applied to each part-write.
4601 This seems wrong - if the receiver was accepting one byte at a time it
4602 would take for ever. The timeout is now adjusted when this happens. It
4603 doesn't have to be particularly precise.
4605 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4606 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4608 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4609 without a domain if the check to <postmaster@domain> fails.
4611 SC/01 Eximstats: added -xls and the ability to specify output files
4612 (patch written by Frank Heydlauf).
4614 SC/02 Eximstats: use FileHandles for outputting results.
4616 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4618 SC/04 Eximstats: fixed display of large numbers with -nvr option
4620 SC/05 Eximstats: fixed merging of reports with empty tables.
4622 SC/06 Eximstats: added the -include_original_destination flag
4624 SC/07 Eximstats: removed tabs and trailing whitespace.
4626 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4628 TK/06 MBOX spool code: Add real "From " MBOX separator line
4629 so the .eml file is really in mbox format (even though
4630 most programs do not really care). Patch from Alex Miller.
4632 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4633 The latter is generated from $received_to and is only set if the
4634 message has one envelope recipient. SA can use these headers,
4635 obviously out-of-the-box. Patch from Alex Miller.
4637 PH/08 The ${def test on a variable was returning false if the variable's
4638 value was "0", contrary to what the specification has always said!
4639 The result should be true unless the variable is empty.
4641 PH/09 The syntax error of a character other than { following "${if
4642 def:variable_name" (after optional whitespace) was not being diagnosed.
4643 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4644 accidental colon was present, for example, could give incorrect results.
4646 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4647 result is used (not including appendfile, where other changes are about
4650 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4651 This involved changing a lot of size variables from int to off_t. It
4652 should work with maildirs and everything.
4654 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4655 spamd dying while we are connected to it.
4657 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4660 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4661 function for BDB 4.3.
4663 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4665 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4666 the sources for PCRE have been re-arranged and more files are now
4669 PH/15 The code I had for printing potentially long long variables in PH/11
4670 above was not the best (it lost precision). The length of off_t variables
4671 is now inspected at build time, and an appropriate printing format (%ld
4672 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4673 to be "long long int" or "long int". This is needed for the internal
4674 formatting function string_vformat().
4676 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4677 the configuration file to be ":syslog", then the script "guesses" where
4678 the logs files are, rather than using the compiled in default. In our
4679 case the guess is not the same as the compiled default, so the script
4680 suddenly stopped working when I started to use syslog. The patch checks
4681 to see if log_file_path is "". If so, it attempts to read it from exim
4682 with no configuration file to get the compiled in version, before it
4683 falls back to the previous guessing code."
4685 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4686 implementing BATV in an Exim configuration. See NewStuff for the gory
4689 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4690 Makefile that are specific to HP-UX.
4692 PH/18 If the "use_postmaster" option was set for a recipient callout together
4693 with the "random" option, the postmaster address was used as the MAIL
4694 FROM address for the random test, but not for the subsequent recipient
4695 test. It is now used for both.
4697 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4698 patch removes a few documentation additions to RFC 3028, because the
4699 latest draft now contains them. It adds the new en;ascii-case comparator
4700 and a new error check for 8bit text in MIME parts. Comparator and
4701 require names are now matched exactly. I enabled the subaddress
4702 extension, but it is not well tested yet (read: it works for me)."
4704 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4705 rework some of the code of TK/09 above to avoid the hardwired use of
4706 "%lld" and "long long". Replaced the call to snprintf() with a call to
4709 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4710 records point to non-existent hosts", "retry timeout exceeded", and
4711 "retry time not reached for any host after a long failure period".
4713 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4714 experimental DomainKeys support:
4716 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4717 (2) On an error such as an illegally used "control", the wrong name for
4718 the control was given.
4720 These problems did NOT occur unless DomainKeys support was compiled.
4722 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4724 PH/24 Added ${if match_ip condition.
4726 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4727 databases so that it will be absolutely obvious if a crash occurs in the
4728 DB library. This is a regular occurrence (often caused by mis-matched
4731 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4732 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4733 picked up on a user's system that detects such things. There doesn't seem
4734 to be a gcc warning option for this - only an attribute that has to be
4735 put on the function's prototype. It seems that in Fedora Core 4 they have
4736 set this on a number of new functions. No doubt there will be more in due
4739 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4740 specify a transport. However, if an address that was verified by such a
4741 router was the subject of a callout, Exim crashed because it tried to
4742 read the rcpt_include_affixes from the non-existent transport. Now it
4743 just assumes that the setting of that option is false. This bug was
4744 introduced by 4.51/PH/31.
4746 PH/28 Changed -d+all to exclude +memory, because that information is very
4747 rarely of interest, but it makes the output a lot bigger. People tend to
4748 do -d+all out of habit.
4750 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4751 code in os-type was giving problems when libc.so lives in lib64, like on
4754 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4755 aren't the modern standard, and it seems that some systems' include files
4756 don't always have them. Exim was already checking for some of the newer
4757 ones like T_AAAA, and defining it itself. I've added checks for all the
4758 record types that Exim uses.
4760 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4761 not automatically generating a new one, as it is supposed to. This
4762 prevented TLS from working. If the file did exist, but contained invalid
4763 data, a new version was generated, as expected. It was only the case of a
4764 non-existent file that was broken.
4766 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4767 with a change in libdomainkeys > 0.64.
4769 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4770 from DNS. If the selector record carries the flag, it now has
4771 precedence over the domain-wide flag.
4773 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4775 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4776 the use of a port name that isn't defined in /etc/services) occurred, the
4777 message was deferred as in a normal delivery, and thus remained on the
4778 spool, instead of being failed because of the mua_wrapper setting. This
4779 is now fixed, and I tidied up some of the mua_wrapper messages at the
4782 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4783 the messages in a hash of arrays rather than using individual hashes.
4784 This is a bit cleaner and results in dramatic memory savings, albeit
4785 at a slight CPU cost.
4787 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4788 as requested by Marc Sherman.
4790 SC/10 Eximstats: added histograms for user specified patterns as requested
4793 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4795 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4796 fopen() in the content-scanning modules that did not already have it.
4802 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4803 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4805 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4807 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4809 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4810 to test Sieve filters that use "vacation".
4812 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4813 that changes the way the GnuTLS parameters are stored in the cache file.
4814 The new format can be generated externally. For backward compatibility,
4815 if the data in the cache doesn't make sense, Exim assumes it has read an
4816 old-format file, and it generates new data and writes a new file. This
4817 means that you can't go back to an older release without removing the
4820 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4821 work if there are any delivery delays because "one_time" forces the
4822 parent to be marked "delivered", so its unseen clone is never tried
4823 again. For this reason, Exim now forbids the simultaneous setting of
4826 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4827 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4828 entry if you want to know the details. The fix, however, neglected to
4829 consider the case when local delivery batching is involved. The test for
4830 "previously delivered" was not happening when checking to see if an
4831 address could be batched with a previous (undelivered) one; under
4832 certain circumstances this could lead to multiple deliveries to the same
4835 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4836 in its include files, and this causes problems building Exim.
4838 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4839 header_syntax) but Exim was just ignoring anything given after a slash.
4840 In particular, this caused confusion with an attempt to use "verify =
4841 reverse_host_lookup/defer_ok". An error is now given when options are
4842 supplied for verify items that do not have them. (Maybe reverse_host_
4843 lookup should have a defer_ok option, but that's a different point.)
4845 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4846 defined by RFC 821) to 2048, because there were problems with some AUTH
4847 commands, and RFC 1869 says the size should be increased for extended
4848 SMTP commands that take arguments.
4850 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4853 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4854 "unknown" error; now it says that the functionality isn't in the binary.
4856 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4857 an address' error message when a string expansion fails (syntax or
4858 whatever). Otherwise the password may appear in the log. Following change
4859 PH/42 below, there is no longer a chance of it appearing in a bounce
4862 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4864 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4865 of its addresses was ever tried. (Bugzilla bug #2.)
4867 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4868 the result incorrectly in the debug output. (It correctly added a newline
4869 to what was transported.)
4871 TF/01 Added $received_time.
4873 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4874 commented out examples of how to interface to a virus scanner and to
4875 SpamAssassin. Also added commented examples of av_scanner and
4876 spamd_address settings.
4878 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4879 and controls are allowed in which ACLs. There were a couple of minor
4880 errors. Some of the entries in the conditions table (which is a table of
4881 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4882 negation of where the condition IS allowed.
4884 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4886 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4887 header file does not have a version number, so I've had to invent a new
4888 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4889 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4890 radiusclient), but it was contributed by a Radius user.
4892 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4893 files or directories, for OpenSSL.
4895 PH/19 When an Exim process that is running as root has to create an Exim log
4896 file, it does so in a subprocess that runs as exim:exim so as to get the
4897 ownership right at creation (otherwise, other Exim processes might see
4898 the file with the wrong ownership). There was no test for failure of this
4899 fork() call, which would lead to the process getting stuck as it waited
4900 for a non-existent subprocess. Forks do occasionally fail when resources
4901 run out. I reviewed all the other calls to fork(); they all seem to check
4904 PH/20 When checking for unexpected SMTP input at connect time (before writing
4905 the banner), Exim was not dealing correctly with a non-positive return
4906 from the read() function. If the client had disconnected by this time,
4907 the result was a log entry for a synchronization error with an empty
4908 string after "input=" when read() returned zero. If read() returned -1
4909 (an event I could not check), uninitialized data bytes were printed.
4910 There were reports of junk text (parts of files, etc) appearing after
4913 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4915 PH/22 Added support for macro redefinition, and (re)definition in between
4916 driver and ACL definitions.
4918 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4919 forgetting to use the resulting value; it was using the unexpanded value.
4921 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4922 hadn't been configured. The fix is from Juergen Kreileder, who
4923 understands it better than I do:
4925 "Here's what I see happening with three configured cyrus_sasl
4926 authenticators configured (plain, login, cram-md5):
4928 On startup auth_cyrus_sasl_init() gets called for each of these.
4929 This means three calls to sasl_listmech() without a specified mech_list.
4930 => SASL tests which mechs of all available mechs actually work
4931 => three warnings about OTP not working
4932 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4934 With the patch, sasl_listmech() also gets called three times. But now
4935 SASL's mech_list option is set to the server_mech specified in the the
4936 authenticator. Or in other words, the answer from sasl_listmech()
4937 gets limited to just the mech you're testing for (which is different
4939 => the return list contains just 'plain' or 'login', 'cram-md5' or
4940 nothing depending on the value of ob->server_mech.
4942 I've just tested the patch: Authentication still works fine,
4943 unavailable mechs specified in the exim configuration are still
4944 caught, and the auth.log warnings about OTP are gone."
4946 PH/25 When debugging is enabled, the contents of the command line are added
4947 to the debugging output, even when log_selector=+arguments is not
4950 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4951 answer is "GNU", and only if the return is "GNU/something" is the answer
4954 PH/27 $acl_verify_message is now set immediately after the failure of a
4955 verification in an ACL, and so is available in subsequent modifiers. In
4956 particular, the message can be preserved by coding like this:
4958 warn !verify = sender
4959 set acl_m0 = $acl_verify_message
4961 Previously, $acl_verify_message was set only while expanding "message"
4962 and "log_message" when a very denied access.
4964 PH/28 Modified OS/os.c-Linux with
4966 -#ifndef OS_LOAD_AVERAGE
4967 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4969 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4970 nomenclature these days.)
4972 PH/29 Installed patch from the Sieve maintainer that adds the options
4973 sieve_useraddress and sieve_subaddress to the redirect router.
4975 PH/30 In these circumstances:
4976 . Two addresses routed to the same list of hosts;
4977 . First host does not offer TLS;
4978 . First host accepts first address;
4979 . First host gives temporary error to second address;
4980 . Second host offers TLS and a TLS session is established;
4981 . Second host accepts second address.
4982 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4983 and peerdn, if requested) that were in fact used only for the second
4986 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4987 attention to any local part prefix or suffix that was matched by the
4988 router that accepted the address. It now behaves in the same way as it
4989 does for delivery: the affixes are removed from the local part unless
4990 rcpt_include_affixes is set on the transport.
4992 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4993 timeout during the DATA phase of an incoming message.
4995 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4996 applied a patch sanctioned by the Sieve maintainer.
4998 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4999 the uid or gid is negative. A case of a negative gid caused this to be
5000 noticed. The fix allows for either to be negative.
5002 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5003 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5006 PH/36 The change PH/12 above was broken. Fixed it.
5008 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5009 the grounds that routing the same address twice would always produce the
5010 same answer. This might have been true once, but it is certainly no
5011 longer true now. Routing a child address may depend on the previous
5012 routing that produced that child. Some complicated redirection strategies
5013 went wrong when messages had multiple recipients, and made Exim's
5014 behaviour dependent on the order in which the addresses were given.
5016 I have moved the duplicate checking until after the routing is complete.
5017 Exim scans the addresses that are assigned to local and remote
5018 transports, and removes any duplicates. This means that more work will be
5019 done, as duplicates will always all be routed, but duplicates are
5020 presumably rare, so I don't expect this is of any significance.
5022 For deliveries to pipes, files, and autoreplies, the duplicate checking
5023 still happens during the routing process, since they are not going to be
5026 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5027 It corrects a timeout issue with spamd. This is Ian's comment: "The
5028 background is that sometimes spamd either never reads data from a
5029 connection it has accepted, or it never writes response data. The exiscan
5030 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5031 blindly assumes that writes won't block so it may never time out."
5033 PH/39 Allow G after quota size as well as K and M.
5035 PH/40 The value set for $authenticated_id in an authenticator may not contain
5036 binary zeroes or newlines because the value is written to log lines and
5037 to spool files. There was no check on this. Now the value is run through
5038 the string_printing() function so that such characters are converted to
5039 printable escape sequences.
5041 PH/41 $message_linecount is a new variable that contains the total number of
5042 lines in the message. Compare $body_linecount, which is the count for the
5045 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5046 bounce and delay warning messages, except in certain special cases, which
5049 (a) An SMTP error message from a remote host;
5050 (b) A message specified in a :fail: redirection;
5051 (c) A message specified in a "fail" command in a system filter;
5052 (d) A message specified in a FAIL return from the queryprogram router;
5053 (e) A message specified by the cannot_route_message router option.
5055 In these cases only, Exim does include the error details in bounce and
5056 warning messages. There are also a few cases where bland messages such
5057 as "unrouteable address" or "local delivery error" are given.
5059 PH/43 $value is now also set for the "else" part of a ${run expansion.
5061 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5062 being worked on, but at least Exim now implements the latest version to
5065 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5066 process to complete was treated as a delivery failure, a timeout while
5067 writing the message to the pipe was logged, but erroneously treated as a
5068 successful delivery. Such timeouts include transport filter timeouts. For
5069 consistency with the overall process timeout, these timeouts are now
5070 treated as errors, giving rise to delivery failures by default. However,
5071 there is now a new Boolean option for the pipe transport called
5072 timeout_defer, which, if set TRUE, converts the failures into defers for
5073 both kinds of timeout. A transport filter timeout is now identified in
5076 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5077 systems where "make" and "gmake" are different, calling "gmake" at top
5078 level broke things. I've arranged for the value of $(MAKE) to be passed
5079 from the Makefile to this script so that it can call the same version of
5083 A note about Exim versions 4.44 and 4.50
5084 ----------------------------------------
5086 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5087 changes of various kinds. As a consequence, a big documentation update was
5088 needed. This delayed the release for rather longer than seemed good, especially
5089 in the light of a couple of (minor) security issues. Therefore, the changes
5090 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5091 release. So 4.44 and 4.50 are in effect two different branches that both start
5094 I have left the 4.50 change log unchanged; it contains all the changes since
5095 4.43. The change log for 4.44 is below; many of its items are identical to
5096 those for 4.50. This seems to be the most sensible way to preserve the
5097 historical information.
5103 1. Minor wording change to the doc/README.SIEVE file.
5105 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5106 computation of the current number of files was incorrect.
5108 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5109 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5112 4. Give more explanation in the error message when the command for a transport
5113 filter fails to execute.
5115 5. There are several places where Exim runs a non-Exim command in a
5116 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5117 was being done only for the command run by the queryprogram router. It is
5118 now done for all such subprocesses. The other cases are: ${run, transport
5119 filters, and the commands run by the lmtp and pipe transports.
5121 6. Added CONFIGURE_GROUP build-time option.
5123 7. Some older OS have a limit of 256 on the maximum number of file
5124 descriptors. Exim was using setrlimit() to set 1000 as a large value
5125 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5126 systems. I've change it so that if it can't get 1000, it tries for 256.
5128 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5129 was an oversight, and furthermore, ever since the addition of extra
5130 controls (e.g. 4.43/32), the checks on when to allow different forms of
5131 "control" were broken. There should now be diagnostics for all cases when a
5132 control that does not make sense is encountered.
5134 9. Added the /retain_sender option to "control=submission".
5136 10. $recipients is now available in the predata ACL (oversight).
5138 11. Tidy the search cache before the fork to do a delivery from a message
5139 received from the command line. Otherwise the child will trigger a lookup
5140 failure and thereby defer the delivery if it tries to use (for example) a
5141 cached ldap connection that the parent has called unbind on.
5143 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5144 of $address_data from the recipient verification was clobbered by the
5145 sender verification.
5147 13. The value of address_data from a sender verification is now available in
5148 $sender_address_data in subsequent conditions in the ACL statement.
5150 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5152 15. Added a new option "connect=<time>" to callout options, to set a different
5155 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5156 was its contents. (It was OK if the option was not defined at all.)
5158 17. A "Completed" log line is now written for messages that are removed from
5159 the spool by the -Mrm option.
5161 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5162 information about exactly what failed.
5164 19. Added -dd to debug only the daemon process.
5166 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5167 handles timeouts, both on the server side and network timeouts. Renamed the
5168 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5170 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5171 It is now set to "smtps".
5173 22. $host_address is now set to the target address during the checking of
5174 ignore_target_hosts.
5176 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5177 being passed; this would have caused $sender_host_name to have been used if
5178 matching the list had actually called for a host name (not very likely,
5179 since this list is usually IP addresses). A host name is now passed as
5182 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5183 code that specifies a non-restarting handler (typically sigaction() in
5184 modern systems) in an attempt to fix a rare and obscure crash bug.
5186 25. Narrowed the window for a race in the daemon that could cause it to ignore
5187 SIGCHLD signals. This is not a major problem, because they are used only to
5188 wake it up if nothing else does.
5190 26. A malformed maildirsize file could cause Exim to calculate negative values
5191 for the mailbox size or file count. Odd effects could occur as a result.
5192 The maildirsize information is now recalculated if the size or filecount
5195 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5196 support for a long time. Removed HAVE_SYS_VFS_H.
5198 28. Installed the latest version of exipick from John Jetmore.
5200 29. In an address list, if the pattern was not a regular expression, an empty
5201 subject address (from a bounce message) matched only if the pattern was an
5202 empty string. Non-empty patterns were not even tested. This was the wrong
5203 because it is perfectly reasonable to use an empty address as part of a
5204 database query. An empty address is now tested by patterns that are
5205 lookups. However, all the other forms of pattern expect the subject to
5206 contain a local part and a domain, and therefore, for them, an empty
5207 address still always fails if the pattern is not itself empty.
5209 30. Exim went into a mad DNS loop when attempting to do a callout where the
5210 host was specified on an smtp transport, and looking it up yielded more
5211 than one IP address.
5213 31. Re-factored the code for checking spool and log partition space into a
5214 function that finds that data and another that does the check. The former
5215 is then used to implement four new variables: $spool_space, $log_space,
5216 $spool_inodes, and $log_inodes.
5218 32. The RFC2047 encoding function was originally intended for short strings
5219 such as real names; it was not keeping to the 75-character limit for
5220 encoded words that the RFC imposes. It now respects the limit, and
5221 generates multiple encoded words if necessary. To be on the safe side, I
5222 have increased the buffer size for the ${rfc2047: expansion operator from
5225 33. It is now permitted to omit both strings after an "if" condition; if the
5226 condition is true, the result is "true". As before, when the second string
5227 is omitted, a false condition yields an empty string. This makes it less
5228 cumbersome to write custom ACL and router conditions.
5230 34. Failure to deliver a bounce message always caused it to be frozen, even if
5231 there was an errors_to setting on the router. The errors_to setting is now
5234 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5235 canonical form (fully expanded) before being placed in
5236 $sender_host_address.
5238 36. The table in the code that translates DNS record types into text (T_A to
5239 "A" for instance) was missing entries for NS and CNAME. It is just possible
5240 that this could have caused confusion if both these types were looked up
5241 for the same domain, because the text type is used as part of Exim's
5242 per-process caching. But the chance of anyone hitting this buglet seems
5245 37. The dnsdb lookup has been extended in a number of ways.
5247 (1) There is a new type, "zns", which walks up the domain tree until it
5248 finds some nameserver records. It should be used with care.
5250 (2) There is a new type, "mxh", which is like "mx" except that it returns
5251 just the host names, not the priorities.
5253 (3) It is now possible to give a list of domains (or IP addresses) to be
5254 looked up. The behaviour when one of the lookups defers can be
5255 controlled by a keyword.
5257 (4) It is now possible to specify the separator character for use when
5258 multiple records are returned.
5260 38. The dnslists ACL condition has been extended: it is now possible to supply
5261 a list of IP addresses and/or domains to be looked up in a particular DNS
5264 39. Added log_selector=+queue_time_overall.
5266 40. When running the queue in the test harness, wait just a tad after forking a
5267 delivery process, to get repeatability of debugging output.
5269 41. Include certificate and key file names in error message when GnuTLS fails
5270 to set them up, because the GnuTLS error message doesn't include the name
5271 of the failing file when there is a problem reading it.
5273 42. Allow both -bf and -bF in the same test run.
5275 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5277 44. The "Exiscan patch" is now merged into the mainline Exim source.
5279 45. Sometimes the final signoff response after QUIT could fail to get
5280 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5281 < 0 before doing a fflush(). This bug looks as though it goes back to the
5282 introduction of TLS in release 3.20, but "sometimes" must have been rare
5283 because the tests only now provoked it.
5285 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5286 (this can affect the format of dates).
5288 47. exim_tidydb, when checking for the continued existence of a message for
5289 which it has found a message-specific retry record, was not finding
5290 messages that were in split spool directories. Consequently, it was
5291 deleting retry records that should have stayed in existence.
5293 48. Steve fixed some bugs in eximstats.
5295 49. The SPA authentication driver was not abandoning authentication and moving
5296 on to the next authenticator when an expansion was forced to fail,
5297 contradicting the general specification for all authenticators. Instead it
5298 was generating a temporary error. It now behaves as specified.
5300 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5301 (the order specifies the preference for clients). The order is now AES256,
5302 AES128, 3DES, ARCFOUR128.
5304 51. Small patch to Sieve code - explicitly set From: when generating an
5307 52. Exim crashed if a remote delivery caused a very long error message to be
5308 recorded - for instance if somebody sent an entire SpamAssassin report back
5309 as a large number of 550 error lines. This bug was coincidentally fixed by
5310 increasing the size of one of Exim's internal buffers (big_buffer) that
5311 happened as part of the Exiscan merge. However, to be on the safe side, I
5312 have made the code more robust (and fixed the comments that describe what
5315 53. Now that there can be additional text after "Completed" in log lines (if
5316 the queue_time_overall log selector is set), a one-byte patch to exigrep
5317 was needed to allow it to recognize "Completed" as not the last thing in
5320 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5321 patch that reportedly fixes this has been added. I am not expert enough to
5322 create a test for it. This is what the patch creator wrote:
5324 "I found a little strange behaviour of ldap code when working with
5325 Windows 2003 AD Domain, where users was placed in more than one
5326 Organization Units. When I tried to give exim partial DN, the exit code
5327 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5328 But simultaneously result of request was absolutely normal ldap result,
5329 so I produce this patch..."
5331 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5332 REFERENCE, so I have modified the code to exclude the patch when that macro
5335 55. Some experimental protocols are using DNS PTR records for new purposes. The
5336 keys for these records are domain names, not reversed IP addresses. The
5337 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5338 leaves it alone. Component reversal etc. now happens only for IP addresses.
5341 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5343 57. Double the size of the debug message buffer (to 2048) so that more of very
5344 long debug lines gets shown.
5346 58. The exicyclog utility now does better if the number of log files to keep
5347 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5349 59. Two changes related to the smtp_active_hostname option:
5351 (1) $smtp_active_hostname is now available as a variable.
5352 (2) The default for smtp_banner uses $smtp_active_hostname instead
5353 of $primary_hostname.
5355 60. The host_aton() function is supposed to be passed a string that is known
5356 to be a valid IP address. However, in the case of IPv6 addresses, it was
5357 not checking this. This is a hostage to fortune. Exim now panics and dies
5358 if the condition is not met. A case was found where this could be provoked
5359 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5360 components; fortuitously, this particular loophole had already been fixed
5361 by change 4.50/55 above.
5363 If there are any other similar loopholes, the new check in host_aton()
5364 itself should stop them being exploited. The report I received stated that
5365 data on the command line could provoke the exploit when Exim was running as
5366 exim, but did not say which command line option was involved. All I could
5367 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5368 running as the user.
5371 61. There was a buffer overflow vulnerability in the SPA authentication code
5372 (which came originally from the Samba project). I have added a test to the
5373 spa_base64_to_bits() function which I hope fixes it.
5376 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5377 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5379 63. The daemon start-up calls getloadavg() while still root for those OS that
5380 need the first call to be done as root, but it missed one case: when
5381 deliver_queue_load_max is set with deliver_drop_privilege. This is
5382 necessary for the benefit of the queue runner, because there is no re-exec
5383 when deliver_drop_privilege is set.
5385 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5386 This has been fixed.
5388 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5389 was in use, was not putting the data itself into the right store pool;
5390 consequently, it could be overwritten for a subsequent message in the same
5391 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5394 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5396 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5397 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5398 the function were treating the return as a boolean value, which happened to
5399 work because 0=false and not-0=true, but is not correct code.
5401 68. The host_aton() function was not handling scoped IPv6 addresses (those
5402 with, for example, "%eth0" on the end) correctly.
5404 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5405 negated items (that is, ~something) in unsigned ints. Some compilers
5406 apparently mutter when there is no cast.
5408 70. If an address verification called from an ACL failed, and did not produce a
5409 user-specific message (i.e. there was only a "system" message), nothing was
5410 put in $acl_verify_message. In this situation, it now puts the system
5413 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5414 session; change 4.31/43 added the unwanted input to the log line - except
5415 that it did not do this in the start of session case. It now does.
5417 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5418 This is wrong and can cause the other end to generate a synchronization
5419 error if it is another Exim or anything else that does the synchronization
5420 check. A QUIT command is no longer sent after a timeout.
5422 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5423 during host lookups.
5425 74. The defer_ok option of callout verification was not working if it was used
5426 when verifying addresses in header lines, that is, for this case:
5428 verify = header_sender/callout=defer_ok
5430 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5431 those file descriptors could be used for SMTP connections. If anything
5432 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5433 could be sent to the SMTP client, causing chaos. The daemon now opens
5434 stdin, stdout, and stderr to /dev/null when it puts itself into the
5437 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5438 log by default. The user can override this with suitable Perl magic.
5440 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5441 the log message when discard triggers, was not working for the DATA ACL or
5442 for the non-SMTP ACL.
5444 78. Error message wording change in sieve.c.
5446 79. If smtp_accept_max_per_host was set, the number of connections could be
5447 restricted to fewer than expected, because the daemon was trying to set up
5448 a new connection before checking whether the processes handling previous
5449 connections had finished. The check for completed processes is now done
5450 earlier. On busy systems, this bug wouldn't be noticed because something
5451 else would have woken the daemon, and it would have reaped the completed
5454 80. If a message was submitted locally by a user whose login name contained one
5455 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5456 It caused a spool format error. I have fixed the spool reading code. A
5457 related problem was that the "from" clause in the Received: line became
5458 illegal because of the space(s). It is now covered by ${quote_local_part.
5460 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5463 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5465 83. Patch from Sieve maintainer to fix unterminated string problem in
5466 "vacation" handling.
5468 84. Some minor changes to the Linux configuration files to help with other
5469 OS variants using glibc.
5471 85. One more patch for Sieve to update vacation handling to latest spec.
5474 ----------------------------------------------------
5475 See the note above about the 4.44 and 4.50 releases.
5476 ----------------------------------------------------
5482 1. Change 4.43/35 introduced a bug that caused file counts to be
5483 incorrectly computed when quota_filecount was set in an appendfile
5486 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5487 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5490 3. Give more explanation in the error message when the command for a transport
5491 filter fails to execute.
5493 4. There are several places where Exim runs a non-Exim command in a
5494 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5495 was being done only for the command run by the queryprogram router. It is
5496 now done for all such subprocesses. The other cases are: ${run, transport
5497 filters, and the commands run by the lmtp and pipe transports.
5499 5. Some older OS have a limit of 256 on the maximum number of file
5500 descriptors. Exim was using setrlimit() to set 1000 as a large value
5501 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5502 systems. I've change it so that if it can't get 1000, it tries for 256.
5504 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5505 was an oversight, and furthermore, ever since the addition of extra
5506 controls (e.g. 4.43/32), the checks on when to allow different forms of
5507 "control" were broken. There should now be diagnostics for all cases when a
5508 control that does not make sense is encountered.
5510 7. $recipients is now available in the predata ACL (oversight).
5512 8. Tidy the search cache before the fork to do a delivery from a message
5513 received from the command line. Otherwise the child will trigger a lookup
5514 failure and thereby defer the delivery if it tries to use (for example) a
5515 cached ldap connection that the parent has called unbind on.
5517 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5518 of $address_data from the recipient verification was clobbered by the
5519 sender verification.
5521 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5522 was its contents. (It was OK if the option was not defined at all.)
5524 11. A "Completed" log line is now written for messages that are removed from
5525 the spool by the -Mrm option.
5527 12. $host_address is now set to the target address during the checking of
5528 ignore_target_hosts.
5530 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5531 being passed; this would have caused $sender_host_name to have been used if
5532 matching the list had actually called for a host name (not very likely,
5533 since this list is usually IP addresses). A host name is now passed as
5536 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5537 code that specifies a non-restarting handler (typically sigaction() in
5538 modern systems) in an attempt to fix a rare and obscure crash bug.
5540 15. Narrowed the window for a race in the daemon that could cause it to ignore
5541 SIGCHLD signals. This is not a major problem, because they are used only to
5542 wake it up if nothing else does.
5544 16. A malformed maildirsize file could cause Exim to calculate negative values
5545 for the mailbox size or file count. Odd effects could occur as a result.
5546 The maildirsize information is now recalculated if the size or filecount
5549 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5550 support for a long time. Removed HAVE_SYS_VFS_H.
5552 18. Updated exipick to current release from John Jetmore.
5554 19. Allow an empty sender to be matched against a lookup in an address list.
5555 Previously the only cases considered were a regular expression, or an
5558 20. Exim went into a mad DNS lookup loop when doing a callout where the
5559 host was specified on the transport, if the DNS lookup yielded more than
5562 21. The RFC2047 encoding function was originally intended for short strings
5563 such as real names; it was not keeping to the 75-character limit for
5564 encoded words that the RFC imposes. It now respects the limit, and
5565 generates multiple encoded words if necessary. To be on the safe side, I
5566 have increased the buffer size for the ${rfc2047: expansion operator from
5569 22. Failure to deliver a bounce message always caused it to be frozen, even if
5570 there was an errors_to setting on the router. The errors_to setting is now
5573 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5574 canonical form (fully expanded) before being placed in
5575 $sender_host_address.
5577 24. Updated eximstats to version 1.33
5579 25. Include certificate and key file names in error message when GnuTLS fails
5580 to set them up, because the GnuTLS error message doesn't include the name
5581 of the failing file when there is a problem reading it.
5583 26. Expand error message when OpenSSL has problems setting up cert/key files.
5586 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5587 (this can affect the format of dates).
5589 28. exim_tidydb, when checking for the continued existence of a message for
5590 which it has found a message-specific retry record, was not finding
5591 messages that were in split spool directories. Consequently, it was
5592 deleting retry records that should have stayed in existence.
5594 29. eximstats updated to version 1.35
5595 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5596 1.35 - bugfix such that pie charts by volume are generated correctly
5598 30. The SPA authentication driver was not abandoning authentication and moving
5599 on to the next authenticator when an expansion was forced to fail,
5600 contradicting the general specification for all authenticators. Instead it
5601 was generating a temporary error. It now behaves as specified.
5603 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5604 (the order specifies the preference for clients). The order is now AES256,
5605 AES128, 3DES, ARCFOUR128.
5607 31. Small patch to Sieve code - explicitly set From: when generating an
5610 32. Exim crashed if a remote delivery caused a very long error message to be
5611 recorded - for instance if somebody sent an entire SpamAssassin report back
5612 as a large number of 550 error lines. This bug was coincidentally fixed by
5613 increasing the size of one of Exim's internal buffers (big_buffer) that
5614 happened as part of the Exiscan merge. However, to be on the safe side, I
5615 have made the code more robust (and fixed the comments that describe what
5618 33. Some experimental protocols are using DNS PTR records for new purposes. The
5619 keys for these records are domain names, not reversed IP addresses. The
5620 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5621 leaves it alone. Component reversal etc. now happens only for IP addresses.
5624 34. The host_aton() function is supposed to be passed a string that is known
5625 to be a valid IP address. However, in the case of IPv6 addresses, it was
5626 not checking this. This is a hostage to fortune. Exim now panics and dies
5627 if the condition is not met. A case was found where this could be provoked
5628 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5629 components; fortuitously, this particular loophole had already been fixed
5630 by change 4.50/55 or 4.44/33 above.
5632 If there are any other similar loopholes, the new check in host_aton()
5633 itself should stop them being exploited. The report I received stated that
5634 data on the command line could provoke the exploit when Exim was running as
5635 exim, but did not say which command line option was involved. All I could
5636 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5637 running as the user.
5640 35. There was a buffer overflow vulnerability in the SPA authentication code
5641 (which came originally from the Samba project). I have added a test to the
5642 spa_base64_to_bits() function which I hope fixes it.
5645 36. The daemon start-up calls getloadavg() while still root for those OS that
5646 need the first call to be done as root, but it missed one case: when
5647 deliver_queue_load_max is set with deliver_drop_privilege. This is
5648 necessary for the benefit of the queue runner, because there is no re-exec
5649 when deliver_drop_privilege is set.
5651 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5652 was in use, was not putting the data itself into the right store pool;
5653 consequently, it could be overwritten for a subsequent message in the same
5654 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5657 38. Sometimes the final signoff response after QUIT could fail to get
5658 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5659 < 0 before doing a fflush(). This bug looks as though it goes back to the
5660 introduction of TLS in release 3.20, but "sometimes" must have been rare
5661 because the tests only now provoked it.
5667 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5668 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5669 Now it returns an integer. A number of calls were still expecting a T/F
5670 return. Fortuitously, in all cases, the tests worked in OK situations,
5671 which is the norm. However, things would have gone wrong on any write
5672 failures on the smtp file descriptor. This function is used when sending
5673 messages over SMTP and also when doing verify callouts.
5675 2. When Exim is called to do synchronous delivery of a locally submitted
5676 message (the -odf or -odi options), it no longer closes stderr before doing
5679 3. Implemented the mua_wrapper option.
5681 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5683 5. Implemented the functions header_remove(), header_testname(),
5684 header_add_at_position(), and receive_remove_recipient(), and exported them
5687 6. If an ACL "warn" statement specified the addition of headers, Exim already
5688 inserted X-ACL-Warn: at the start if there was no header name. However, it
5689 was not making this test for the second and subsequent header lines if
5690 there were newlines in the string. This meant that an invalid header could
5691 be inserted if Exim was badly configured.
5693 7. Allow an ACL "warn" statement to add header lines at the start or after all
5694 the Received: headers, as well as at the end.
5696 8. Added the rcpt_4xx retry error code.
5698 9. Added postmaster_mailfrom=xxx to callout verification option.
5700 10. Added mailfrom=xxxx to the callout verification option, for verify=
5703 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5704 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5706 12. Inserted some casts to stop certain compilers warning when using pointer
5707 differences as field lengths or precisions in printf-type calls (mostly
5708 affecting debugging statements).
5710 13. Added optional readline() support for -be (dynamically loaded).
5712 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5713 same clock tick as a message's arrival, so that its received time was the
5714 same as the "first fail" time on the retry record, and that message
5715 remained on the queue past the ultimate address timeout, every queue runner
5716 would try a delivery (because it was past the ultimate address timeout) but
5717 after another failure, the ultimate address timeout, which should have then
5718 bounced the address, did not kick in. This was a "< instead of <=" error;
5719 in most cases the first failure would have been in the next clock tick
5720 after the received time, and all would be well.
5722 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5723 being recognized when the domain list was tested by the match_domain
5724 condition in an expansion string.
5726 16. Added the ${str2b64: operator.
5728 17. Exim was always calling setrlimit() to set a large limit for the number of
5729 processes, without checking whether the existing limit was already
5730 adequate. (It did check for the limit on file descriptors.) Furthermore,
5731 errors from getrlimit() and setrlimit() were being ignored. Now they are
5732 logged to the main and panic logs, but Exim does carry on, to try to do its
5733 job under whatever limits there are.
5735 18. Imported PCRE 5.0.
5737 19. Trivial typo in log message " temporarily refused connection" (the leading
5740 20. If the log selector return_path_on_delivery was set and an address was
5741 redirected to /dev/null, the delivery process crashed because it assumed
5742 that a return path would always be set for a "successful" delivery. In this
5743 case, the whole delivery is bypassed as an optimization, and therefore no
5746 21. Internal re-arrangement: the function for sending a challenge and reading
5747 a response while authentication was assuming a zero-terminated challenge
5748 string. It's now changed to take a pointer and a length, to allow for
5749 binary data in such strings.
5751 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5753 23. Exim was not respecting finduser_retries when seeking the login of the
5754 uid under which it was called; it was always trying 10 times. (The default
5755 setting of finduser_retries is zero.) Also, it was sleeping after the final
5756 failure, which is pointless.
5758 24. Implemented tls_on_connect_ports.
5760 25. Implemented acl_smtp_predata.
5762 26. If the domain in control=submission is set empty, Exim assumes that the
5763 authenticated id is a complete email address when it generates From: or
5764 Sender: header lines.
5766 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5767 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5768 chown and chgrp in /bin and hostname in /usr/bin.
5770 28. Exim was keeping the "process log" file open after each use, just as it
5771 does for the main log. This opens the possibility of it remaining open for
5772 long periods when the USR1 signal hits a daemon. Occasional processlog
5773 errors were reported, that could have been caused by this. Anyway, it seems
5774 much more sensible not to leave this file open at all, so that is what now
5777 29. The long-running daemon process does not normally write to the log once it
5778 has entered its main loop, and it closes the log before doing so. This is
5779 so that log files can straightforwardly be renamed and moved. However,
5780 there are a couple of unusual error situations where the daemon does write
5781 log entries, and I had neglected to close the log afterwards.
5783 30. The text of an SMTP error response that was received during a remote
5784 delivery was being truncated at 512 bytes. This is too short for some of
5785 the long messages that one sometimes sees. I've increased the limit to
5788 31. It is now possible to make retry rules that apply only when a message has a
5789 specific sender, in particular, an empty sender.
5791 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5792 it possible to be selective about when SMTP synchronization is enforced.
5794 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5796 32. Implemented hosts_connection_nolog.
5798 33. Added an ACL for QUIT.
5800 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5803 35. Added mailbox_size and mailbox_filecount to appendfile.
5805 36. Added control = no_multiline_responses to ACLs.
5807 37. There was a bug in the logic of the code that waits for the clock to tick
5808 in the case where the clock went backwards by a substantial amount such
5809 that the microsecond fraction of "now" was more than the microsecond
5810 fraction of "then" (but the whole seconds number was less).
5812 38. Added support for the libradius Radius client library this is found on
5813 FreeBSD (previously only the radiusclient library was supported).
5819 1. When certain lookups returned multiple values in the form name=value, the
5820 quoting of the values was not always being done properly. Specifically:
5821 (a) If the value started with a double quote, but contained no whitespace,
5823 (b) If the value contained whitespace other than a space character (i.e.
5824 tabs or newlines or carriage returns) it was not quoted.
5825 This fix has been applied to the mysql and pgsql lookups by writing a
5826 separate quoting function and calling it from the lookup code. The fix
5827 should probably also be applied to nisplus, ibase and oracle lookups, but
5828 since I cannot test any of those, I have not disturbed their existing code.
5830 2. A hit in the callout cache for a specific address caused a log line with no
5831 reason for rejecting RCPT. Now it says "Previous (cached) callout
5832 verification failure".
5834 3. There was an off-by-one bug in the queryprogram router. An over-long
5835 return line was truncated at 256 instead of 255 characters, thereby
5836 overflowing its buffer with the terminating zero. As well as fixing this, I
5837 have increased the buffer size to 1024 (and made a note to document this).
5839 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5840 when Exim is waiting for an SMTP response from a remote server, Exim
5841 restarts its select() call on the socket, thereby resetting its timeout.
5842 This is not a problem when such interrupts are rare. Somebody set up a cron
5843 job to run exiwhat every 2 minutes, which is less than the normal select()
5844 timeout (5 or 10 minutes). This meant that the select() timeout never
5845 kicked in because it was always reset. I have fixed this by comparing the
5846 time when an interrupt arrives with the time at the start of the first call
5847 to select(). If more time than the timeout has elapsed, the interrupt is
5848 treated as a timeout.
5850 5. Some internal re-factoring in preparation for the addition of Sieve
5851 extensions (by MH). In particular, the "personal" test is moved to a
5852 separate function, and given an option for scanning Cc: and Bcc: (which is
5853 not set for Exim filters).
5855 6. When Exim created an email address using the login of the caller as the
5856 local part (e.g. when creating a From: or Sender: header line), it was not
5857 quoting the local part when it contained special characters such as @.
5859 7. Installed new OpenBSD configuration files.
5861 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5862 try to make them clearer.
5864 9. Callout options, other than the timeout value, were being ignored when
5865 verifying sender addresses in header lines. For example, when using
5867 verify = header_sender/callout=no_cache
5869 the cache was (incorrectly) being used.
5871 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5872 only the Cygwin environment.
5874 11. When return_path_on_delivery was set as a log selector, if different remote
5875 addresses in the same message used different return paths and parallel
5876 remote delivery occurred, the wrong values would sometimes be logged.
5877 (Whenever a remote delivery process finished, the return path value from
5878 the most recently started remote delivery process was logged.)
5880 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5881 lines when AUTH and/or TLS are in use. This is the "received protocol"
5882 field. Exim used to use "asmtp" for authenticated SMTP, without any
5883 indication (in the protocol name) for TLS use. Now it follows the RFC and
5884 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5885 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5886 names appear in log lines as well as in Received: header lines.
5888 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5889 extensions, and comparison tests, and to fix some bugs.
5891 14. Changes to the "personal" filter test:
5893 (1) The test was buggy in that it was just doing the equivalent of
5894 "contains" tests on header lines. For example, if a user's address was
5895 anne@some.where, the "personal" test would incorrectly be true for
5897 To: susanne@some.where
5899 This test is now done by extracting each address from the header in turn,
5900 and checking the entire address. Other tests that are part of "personal"
5901 are now done using regular expressions (for example, to check local parts
5902 of addresses in From: header lines).
5904 (2) The list of non-personal local parts in From: addresses has been
5905 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5906 taken from the Sieve specification recommendations.
5908 (3) If the message contains any header line starting with "List-" it is
5909 treated as non-personal.
5911 (4) The test for "circular" in the Subject: header line has been removed
5912 because it now seems ill-conceived.
5914 15. Minor typos in src/EDITME comments corrected.
5916 16. Installed latest exipick from John Jetmore.
5918 17. If headers_add on a router specified a text string that was too long for
5919 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5920 of string_sprintf() is now avoided.
5922 18. $message_body_size was not set (it was always zero) when running the DATA
5923 ACL and the local_scan() function.
5925 19. For the "mail" command in an Exim filter, no default was being set for
5926 the once_repeat time, causing a random time value to be used if "once" was
5927 specified. (If the value happened to be <= 0, no repeat happened.) The
5928 default is now 0s, meaning "never repeat". The "vacation" command was OK
5929 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5930 (I found it when inspecting the code).
5932 20. There is now an overall timeout for performing a callout verification. It
5933 defaults to 4 times the callout timeout, which applies to individual SMTP
5934 commands during the callout. The overall timeout applies when there is more
5935 than one host that can be tried. The timeout is checked before trying the
5936 next host. This prevents very long delays if there are a large number of
5937 hosts and all are timing out (e.g. when the network connections are timing
5938 out). The value of the overall timeout can be changed by specifying an
5939 additional sub-option for "callout", called "maxwait". For example:
5941 verify = sender/callout=5s,maxwait=20s
5943 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5944 to the end before writing, but this should make it even safer).
5946 22. Exim was forgetting that it had advertised PIPELINING for the second and
5947 subsequent messages on an SMTP connection. It was also not resetting its
5948 memory on STARTTLS and an internal HELO.
5950 23. When Exim logs an SMTP synchronization error within a session, it now
5951 records whether PIPELINING has been advertised or not.
5953 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5954 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5955 rather than long int.
5957 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5959 26. Added the never_mail option to autoreply.
5965 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5966 crash if the getsockname() call failed; this can happen if a connection is
5967 closed very soon after it is established. The problem was simply in the
5968 order in which certain operations were done, causing Exim to try to write
5969 to the SMTP stream before it had set up the file descriptor. The bug has
5970 been fixed by making things happen in the correct order.
5976 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5977 before the connection was closed, thus losing the rejection response.
5979 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5980 some early Solaris releases, but causes trouble in current releases where
5981 socklen_t is defined.
5983 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5986 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5989 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5990 that releasing the top of it at the end releases what was used for sub-
5991 expansions (unless the block got too big). However, discard this block if
5992 the first thing is a variable or header, so that we can use its block when
5993 it is dynamic (useful for very large $message_headers, for example).
5995 6. Lookups now cache *every* query, not just the most recent. A new, separate
5996 store pool is used for this. It can be recovered when all lookup caches are
5997 flushed. Lookups now release memory at the end of their result strings.
5998 This has involved some general refactoring of the lookup sources.
6000 7. Some code has been added to the store_xxx() functions to reduce the amount
6001 of flapping under certain conditions.
6003 8. log_incoming_interface used to affect only the <= reception log lines. Now
6004 it causes the local interface and port to be added to several more SMTP log
6005 lines, for example "SMTP connection from", and rejection lines.
6007 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6009 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6011 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6012 could be overwritten at the end of the current message (or the start of a
6013 new message if it was set in a HELO ACL). The value is now preserved for
6014 the duration of the SMTP connection.
6016 12. If a transport had a headers_rewrite setting, and a matching header line
6017 contained an unqualified address, that address was qualified, even if it
6018 did not match any rewriting rules. The underlying bug was that the values
6019 of the flags that permit the existence of unqualified sender and recipient
6020 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6021 non-local messages, and by -bnq for local messages) were not being
6022 preserved with the message after it was received.
6024 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6025 "next input=" as part of the text comprising the host identity instead of
6026 the correct text. The code was using the same buffer for two different
6027 strings. However, depending on which order the printing function evaluated
6028 its arguments, the bug did not always show up. Under Linux, for example, my
6029 test suite worked just fine.
6031 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6032 doesn't work with some older versions of Perl. It has been changed to "my",
6033 which in any case is probably the better facility to use.
6035 15. A really picky compiler found some instances of statements for creating
6036 error messages that either had too many or two few arguments for the format
6039 16. The size of the buffer for calls to the DNS resolver has been increased
6040 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6041 for addresses that have a lot of PTR records. This alleviates a problem; it
6042 does not fully solve it.
6044 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6045 buffer now truncates the list and logs the incident, which is the same
6046 action as happens when Exim is looking up a host name and its aliases.
6047 Previously in this situation something unpredictable would happen;
6048 sometimes it was "internal error: store_reset failed".
6050 18. If a server dropped the connection unexpectedly when an Exim client was
6051 using GnuTLS and trying to read a response, the client delivery process
6052 crashed while trying to generate an error log message.
6054 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6055 string, for example:
6057 warn message = H1: something\nH2: something
6059 the text was added as a single header line from Exim's point of view
6060 though it ended up OK in the delivered message. However, searching for the
6061 second and subsequent header lines using $h_h2: did not work. This has been
6062 fixed. Similarly, if a system filter added multiple headers in this way,
6063 the routers could not see them.
6065 20. Expanded the error message when iplsearch is called with an invalid key to
6066 suggest using net-iplsearch in a host list.
6068 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6069 ACLs are no longer actually imposed (and a message to that effect is
6072 22. If a "gecos" field in a passwd entry contained escaped characters, in
6073 particular, if it contained a \" sequence, Exim got it wrong when building
6074 a From: or a Sender: header from that name. A second bug also caused
6075 incorrect handling when an unquoted " was present following a character
6076 that needed quoting.
6078 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6079 was not being matched caselessly.
6081 24. Arranged for all hyphens in the exim.8 source to be escaped with
6084 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6085 properly. Recipient callout cache records were still being keyed to include
6086 the sender, even when use_sender was set false. This led to far more
6087 callouts that were necessary. The sender is no longer included in the key
6088 when use_sender is false.
6090 26. Added "control = submission" modifier to ACLs.
6092 27. Added the ${base62d: operator to decode base 62 numbers.
6094 28. dnsdb lookups can now access SRV records.
6096 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6097 the configuration file.
6099 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6100 (-v) mode. This makes the output for a verbose queue run more intelligible.
6102 31. Added a use_postmaster feature to recipient callouts.
6104 32. Added the $body_zerocount variable, containing the number of binary zero
6105 bytes in the message body.
6107 33. The time of last modification of the "new" subdirectory is now used as the
6108 "mailbox time last read" when there is a quota error for a maildir
6111 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6113 35. Added +ignore_unknown as a special item in host lists.
6115 36. Code for decoding IPv6 addresses in host lists is now included, even if
6116 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6117 address was recognized as an IP address, but was then not correctly decoded
6118 into binary, causing unexpected and incorrect effects when compared with
6125 1. Very minor rewording of debugging text in manualroute to say "list of
6126 hosts" instead of "hostlist".
6128 2. If verify=header_syntax was set, and a header line with an unqualified
6129 address (no domain) and a large number of spaces between the end of the
6130 name and the colon was received, the reception process suffered a buffer
6131 overflow, and (when I tested it) crashed. This was caused by some obsolete
6132 code that should have been removed. The fix is to remove it!
6134 3. When running in the test harness, delay a bit after writing a bounce
6135 message to get a bit more predictability in the log output.
6137 4. Added a call to search_tidyup() just before forking a reception process. In
6138 theory, someone could use a lookup in the expansion of smtp_accept_max_
6139 per_host which, without the tidyup, could leave open a database connection.
6141 5. Added the variables $recipient_data and $sender_data which get set from a
6142 lookup success in an ACL "recipients" or "senders" condition, or a router
6143 "senders" option, similar to $domain_data and $local_part_data.
6145 6. Moved the writing of debug_print from before to after the "senders" test
6148 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6149 problems for message scanning, either using a data ACL, or using
6150 local_scan() because the Received: header was not generated till after they
6151 were called (in order to set the time as the time of reception completion).
6152 I have revised the way this works. The header is now generated after the
6153 body is received, but before the ACL or local_scan() are called. After they
6154 are run, the timestamp in the header is updated.
6160 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6161 before starting a queue runner without re-exec. This happened only when
6162 deliver_drop_privilege was set or when the Exim user was set to root. The
6163 effect of the bug was that timeouts during subsequent deliveries caused
6164 crashes instead of being properly handled. The handler is now left at its
6165 default (and expected) setting.
6167 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6168 message, again when deliver_drop_privilege is set or Exim is run as root.
6169 The bug described in (1) was not present in this case, but the tidying up
6170 of the other signals was missing. I have made the two cases consistent.
6172 3. The ignore_target_hosts setting on a manualroute router was being ignored
6173 for hosts that were looked up using the /MX notation.
6175 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6178 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6179 operated on the sender address. After changing the $sender_address to <>
6180 for the sender address verify, Exim was re-instated it as the original
6181 (before rewriting) address, but remembering that it had rewritten it, so it
6182 wasn't rewriting it again. This bug also had the effect of breaking the
6183 sender address verification caching when the sender address was rewritten.
6185 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6186 This has been changed so that if the ip literal address matches
6187 ignore_target_hosts, the router declines.
6189 7. Added expansion conditions match_domain, match_address, and match_local_
6190 part (NOT match_host).
6192 8. The placeholder for the Received: header didn't have a length field set.
6194 9. Added code to Exim itself and to exim_lock to test for a specific race
6195 condition that could lead to file corruption when using MBX delivery. The
6196 issue is with the lockfile that is created in /tmp. If this file is removed
6197 after a process has opened it but before that process has acquired a lock,
6198 there is the potential for a second process to recreate the file and also
6199 acquire a lock. This could lead to two Exim processes writing to the file
6200 at the same time. The added code performs the same test as UW imapd; it
6201 checks after acquiring the lock that its file descriptor still refers to
6202 the same named file.
6204 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6205 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6208 11. Added the smtp_active_hostname option. If used, this will typically be made
6209 to depend on the incoming interface address. Because $interface_address is
6210 not set up until the daemon has forked a reception process, error responses
6211 that can happen earlier (such as "too many connections") no longer contain
6214 12. If an expansion in a condition on a "warn" statement fails because a lookup
6215 defers, the "warn" statement is abandoned, and the next ACL statement is
6216 processed. Previously this caused the whole ACL to be aborted.
6218 13. Added the iplsearch lookup type.
6220 14. Added ident_timeout as a log selector.
6222 15. Added tls_certificate_verified as a log selector.
6224 16. Added a global option tls_require_ciphers (compare the smtp transport
6225 option of the same name). This controls incoming TLS connections.
6227 17. I finally figured out how to make tls_require_ciphers do a similar thing
6228 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6229 before starting the TLS session.
6231 18. Tabs are now shown as \t in -bP output.
6233 19. If the log selector return_path_on_delivery was set, Exim crashed when
6234 bouncing a message because it had too many Received: header lines.
6236 20. If two routers both had headers_remove settings, and the first one included
6237 a superfluous trailing colon, the final name in the first list and the
6238 first name in the second list were incorrectly joined into one item (with a
6239 colon in the middle).
6245 1. Added -C and -D options to the exinext utility, mainly to make it easier
6246 to include in the automated testing, but these could be helpful when
6247 multiple configurations are in use.
6249 2. The exinext utility was not formatting the output nicely when there was
6250 an alternate port involved in the retry record key, nor when there was a
6251 message id as well (for retries that were specific to a specific message
6252 and a specific host). It was also confused by IPv6 addresses, because of
6253 the additional colons they contain. I have fixed the IPv4 problem, and
6254 patched it up to do a reasonable job for IPv6.
6256 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6257 delivery, the log line now contains "pipelined" if PIPELINING was used.
6259 4. An SMTP transport process used to panic and die if the bind() call to set
6260 an explicit outgoing interface failed. This has been changed; it is now
6261 treated in the same way as a connect() failure.
6263 5. A reference to $sender_host_name in the part of a conditional expansion
6264 that was being skipped was still causing a DNS lookup. This no longer
6267 6. The def: expansion condition was not recognizing references to header lines
6268 that used bh_ and bheader_.
6270 7. Added the _cache feature to named lists.
6272 8. The code for checking quota_filecount in the appendfile transport was
6273 allowing one more file than it should have been.
6275 9. For compatibility with Sendmail, the command line option
6283 and sets the incoming protocol and host name (for trusted callers). The
6284 host name and its colon can be omitted when only the protocol is to be set.
6285 Note the Exim already has two private options, -pd and -ps, that refer to
6286 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6287 "s", but I don't think that's a major issue.
6289 10. A number of refactoring changes to the code, none of which should affect
6292 (a) The number of logging options was getting close to filling up the
6293 32-bit word that was used as a bit map. I have split them into two classes:
6294 those that are passed in the argument to log_write(), and those that are
6295 only ever tested independently outside of that function. These are now in
6296 separate 32-bit words, so there is plenty of room for expansion again.
6297 There is no change in the user interface or the logging behaviour.
6299 (b) When building, for example, log lines, the code previously used a
6300 macro that called string_cat() twice, in order to add two strings. This is
6301 not really sufficiently general. Furthermore, there was one instance where
6302 it was actually wrong because one of the argument was used twice, and in
6303 one call a function was used. (As it happened, calling the function twice
6304 did not affect the overall behaviour.) The macro has been replaced by a
6305 function that can join an arbitrary number of extra strings onto a growing
6308 (c) The code for expansion conditions now uses a table and a binary chop
6309 instead of a serial search (which was left over from when there were very
6310 few conditions). Also, it now recognizes conditions like "pam" even when
6311 the relevant support is not compiled in: a suitably worded error message is
6312 given if an attempt is made to use such a condition.
6314 11. Added ${time_interval:xxxxx}.
6316 12. A bug was causing one of the ddress fields not to be passed back correctly
6317 from remote delivery subprocesses. The field in question was not being
6318 subsequently used, so this caused to problems in practice.
6320 13. Added new log selectors queue_time and deliver_time.
6322 14. Might have fixed a bug in maildirsizefile handling that threw up
6323 "unexpected character" debug warnings, and recalculated the data
6324 unnecessarily. In any case, I expanded the warning message to give more
6327 15. Added the message "Restricted characters in address" to the statements in
6328 the default ACL that block characters like @ and % in local parts.
6330 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6331 Three changes have been made:
6333 (a) There was a serious bug; a negative response to MAIL caused the whole
6334 recipient domain to be cached as invalid, thereby blocking all messages
6335 to all local parts at the same domain, from all senders. This bug has
6336 been fixed. The domain is no longer cached after a negative response to
6337 MAIL if the sender used is not empty.
6339 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6342 (c) A new callout option, "use_sender" has been added for people who want
6343 the modified behaviour.
6349 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6352 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6353 indeed breaks things for older releases.
6355 3. Added additional logging to the case where there is a problem reading data
6356 from a filter that is running in a subprocess using a pipe, in order to
6357 try to track down a specific problem.
6359 4. Testing facility fudge: when running in the test harness and attempting
6360 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6361 getting "No route to host". Convert this to a timeout.
6363 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6366 6. Some OS don't have socklen_t but use size_t instead. This affects the
6367 fifth argument of getsockopt() amongst other things. This is now
6368 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6369 can be set for individual OS. I have set it for SunOS5, OSF1, and
6370 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6371 some earlier ones do not.
6373 7. Change 4.30/15 was not doing the test caselessly.
6375 8. The standard form for an IPv6 address literal was being rejected by address
6376 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6377 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6378 this, as well as the form without the "IPv6" on the front (but only when
6379 address literals are enabled, of course).
6381 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6383 10. Exim crashed if a message with an empty sender address specified by -f
6384 encountered a router with an errors_to setting. This could be provoked only
6385 by a command such as
6389 where an empty string was supplied; "<>" did not hit this bug.
6391 11. Installed PCRE release 4.5.
6393 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6394 remained set. It is now erased.
6396 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6397 times from message ids (which are base 36 rather than the normal 62).
6399 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6400 were being counted as actual protocol errors, and logged if the log
6401 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6402 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6403 and DATA following a set of rejected RCPTs do not count as protocol errors.
6404 In other words, Exim assumes they were pipelined, though this may not
6405 actually be the case. Of course, in all cases the client gets an
6406 appropriate error code.
6408 15. If a lookup fails in an ACL condition, a message about the failure may
6409 be available; it is used if testing the ACL cannot continue, because most
6410 such messages specify what the cause of the deferral is. However, some
6411 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6412 that caused an old message to be retained and used if a later statement
6413 caused a defer, replacing the real cause of the deferral.
6415 16. If an IP address had so many PTR records that the DNS lookup buffer
6416 was not large enough to hold them, Exim could crash while trying to process
6417 the truncated data. It now detects and logs this case.
6419 17. Further to 4.21/58, another change has been made: if (and only if) the
6420 first line of a message (the first header line) ends with CRLF, a bare LF
6421 in a subsequent header line has a space inserted after it, so as not to
6422 terminate the header.
6424 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6425 unnecessarily, used atoi() instead of strtol(), and didn't check the
6426 termination when getting file sizes from file names by regex.
6428 19. Completely re-implemented the support for maildirsize files, in the light
6429 of a number of problems with the previous contributed implementation
6430 (4.30/29). In particular:
6432 . If the quota is zero, the maildirsize file is maintained, but no quota is
6435 . If the maildir directory does not exist, it is created before any attempt
6436 to write a maildirsize file.
6438 . The quota value in the file is just a cache; if the quota is changed in
6439 the transport, the new value overrides.
6441 . A regular expression is available for excluding directories from the
6444 20. The autoreply transport checks the characters in options that define the
6445 message's headers; it allows continued headers, but it was checking with
6446 isspace() after an embedded newline instead of explicitly looking for a
6449 21. If all the "regular" hosts to which an address was routed had passed their
6450 expiry times, and had not reached their retry times, the address was
6451 bounced, even if fallback hosts were defined. Now Exim should go on to try
6454 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6455 equivalent code in the SMTP transport. Some hosts send humungous responses
6456 to HELO/EHLO, more than 1024 it seems.
6458 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6459 gives compiler warnings in some environments. I've now done it "properly",
6462 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6463 (because of problems with the built-in one) was declared to return uschar *
6464 instead of char *, causing compiler failure.
6466 25. Fixed a file descriptor leak when processing alias/forward files.
6468 26. Fixed a minor format string issue in dbfn.c.
6470 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6472 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6473 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6476 29. When a sender address is verified, it is cached, to save repeating the test
6477 when there is more than one recipient in a message. However, when the
6478 verification involves a callout, it is possible for different callout
6479 options to be set for different recipients. It is too complicated to keep
6480 track of this in the cache, so now Exim always runs a verification when a
6481 callout is required, relying on the callout cache for the optimization.
6482 The overhead is duplication of the address routing, but this should not be
6485 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6486 to be verified with callout=postmaster, and the main callout worked but the
6487 postmaster check failed, the verification correctly failed. However, if a
6488 subsequent RCPT command asked for sender verification *without* the
6489 postmaster check, incorrect caching caused this verification also to fail,
6492 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6493 it was not caching the DNS options (qualify_single, search_parents) that
6494 were used when the lookup failed. A subsequent lookup with different
6495 options therefore always gave the same answer, though there were cases
6496 where it should not have. (Example: a "domains = !$mx_any" option on a
6497 dnslookup router: the "domains" option is always processed without any
6498 widening, but the router might have qualify_single set.) Now Exim uses the
6499 cached value only when the same options are set.
6501 32. Added John Jetmore's "exipick" utility to the distribution.
6503 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6504 than a timeout (e.g. a certificate is required, and is not provided), an
6505 Exim server now closes the connection immediately. Previously it waited for
6506 the client to close - but if the client is SSL, it seems that they each
6507 wait for each other, leading to a delay before one of them times out.
6509 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6510 maintained 0.8.x compatibility because I don't think many are using it, and
6511 it is clearly obsolete.
6513 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6516 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6517 client certificate was expired. A simple patch fixes this, though I don't
6518 understand the full logic of why the verify callback is called multiple
6521 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6522 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6523 which causes problems with some clients (such as the Certicom SSL Plus
6524 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6525 disables the coutermeasure allowing Eudora to connect."
6527 38. Exim was not checking that a write() to a log file succeeded. This could
6528 lead to Bad Things if a log got too big, in particular if it hit a file
6529 size limit. Exim now panics and dies if it cannot write to a log file, just
6530 as it does if it cannot open a log file.
6532 39. Modified OS/Makefile-Linux so that it now contains
6534 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6536 The two -D definitions ensure that Exim is compiled with large file
6537 support, which makes it possible to handle log files that are bigger than
6540 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6541 instance) a domain was checked against a named list that involved a lookup,
6542 causing $domain_data to be set, then another domain was checked against the
6543 same list, then the first domain was re-checked, the value of $domain_data
6544 after the final check could be wrong. In particular, if the second check
6545 failed, it could be set empty. This bug probably also applied to
6548 41. The strip_trailing_dot option was not being applied to the address given
6549 with the -f command-line option.
6551 42. The code for reading a message's header from the spool was incrementing
6552 $received_count, but never initializing it. This meant that the value was
6553 incorrect (doubled) while delivering a message in the same process in which
6554 it was received. In the most common configuration of Exim, this never
6555 happens - a fresh exec is done - but it can happen when
6556 deliver_drop_privilege is set.
6558 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6559 it now includes up to 150 characters of the unexpected data in the log
6562 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6563 and building data strings. The size of both of these buffers was 10 000
6564 bytes - far larger than anybody would *ever* want, thought I. Needless to
6565 say, somebody hit the limit. I have increased the maximum line length to
6566 20 000 and the maximum data length of concatenated lines to 100 000. I have
6567 also fixed two bugs, because there was no checking on these buffers. Tsk,
6568 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6569 buffer is too small.
6571 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6572 lsearch lookups. Now it does.
6574 46. When parsing a route_list item in a manualroute router, a fixed-length
6575 buffer was used for the list of hosts. I made this 1024 bytes long,
6576 thinking that nobody would ever have a list of hosts that long. Wrong.
6577 Somebody had a whole pile of complicated expansion conditions, and the
6578 string was silently truncated, leading to an expansion error. It turns out
6579 that it is easier to change to an unlimited length (owing to other changes
6580 that have happened since this code was originally written) than to build
6581 structure for giving a limitation error. The length of the item that
6582 expands into the list of hosts is now unlimited.
6584 47. The lsearch lookup could not handle data where the length of text line was
6585 more than 4095 characters. Such lines were truncated, leading to shortened
6586 data being returned. It should now handle lines of any length.
6588 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6589 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6592 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6593 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6594 no longer happens in comment lines. A list of replacements is now placed
6595 at the head of all of the source files, except those whose only change is
6596 to replace PERL_COMMAND in the very first #! line.
6598 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6599 messages on the queue, with a bottom-up merge sort, using code contributed
6600 by Michael Haardt. This should make operations like -bp somewhat faster on
6601 large queues. It won't affect queue runners, except when queue_run_in_order
6604 51. Installed eximstats 1.31 in the distribution.
6606 52. Added support for SRV lookups to the dnslookup router.
6608 53. If an ACL referred to $message_body or $message_body_end, the value was not
6609 reset for any messages that followed in the same SMTP session.
6611 54. The store-handling optimization for building very long strings was not
6612 differentiating between the different store pools. I don't think this
6613 actually made any difference in practice, but I've tidied it.
6615 55. While running the routers to verify a sender address, $sender_address
6616 was still set to the sender address. This is wrong, because when routing to
6617 send a bounce to the sender, it would be empty. Therefore, I have changed
6618 it so that, while verifying a sender address, $sender_address is set to <>.
6619 (There is no change to what happens when verifying a recipient address.)
6621 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6622 target A or AAAA records (if not already returned) without resetting the
6623 qualify_single or search_parents options of the DNS resolver. These are
6624 inappropriate in this case because the targets of MX and SRV records must
6625 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6626 target that, when qualified, matched something in the local domain. These
6627 two options are now turned off when doing these lookups.
6629 57. It seems that at least some releases of Reiserfs (which does not have the
6630 concept of a fixed number of inodes) returns zero and not -1 for the
6631 number of available inodes. This interacted badly with check_spool_inodes,
6632 which assumed that -1 was the "no such thing" setting. What I have done is
6633 to check that the total number of inodes is greater than zero before doing
6634 the test of how many are available.
6636 58. When a "warn" ACL statement has a log_message modifier, the message is
6637 remembered, and not repeated. This is to avoid a lot of repetition when a
6638 message has many recipients that cause the same warning to be written.
6639 However, Exim was preserving the list of already written lines for an
6640 entire SMTP session, which doesn't seem right. The memory is now reset if a
6641 new message is started.
6643 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6644 debugging output unless log_rewrite was also set.
6646 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6647 of (int)(handle) when we know that handle contains (void *)(-1).
6649 61. The Exim daemon panic-logs an error return when it closes the incoming
6650 connection. However "connection reset by peer" seems to be common, and
6651 isn't really an error worthy of noting specially, so that particular error
6654 62. When Exim is trying to find all the local interfaces, it used to panic and
6655 die if the ioctl to get the interface flags failed. However, it seems that
6656 on at least one OS (Solaris 9) it is possible to have an interface that is
6657 included in the list of interfaces, but for which you get a failure error
6658 for this call. This happens when the interface is not "plumbed" into a
6659 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6660 failure of the "get flags" call assumes that the interface is down.
6662 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6663 makes life easier for people who are doing arithmetic on fields extracted
6664 from dates, where you often get leading zeros that should not be
6665 interpreted as octal.
6667 64. Added qualify_domain to the redirect router, to override the global
6670 65. If a pathologically long header line contained very many addresses (the
6671 report of this problem mentioned 10 000) and each of them was rewritten,
6672 Exim could use up a very large amount of memory. (It kept on making new
6673 copies of the header line as it rewrote, and never released the old ones.)
6674 At the expense of a bit more processing, the header rewriting function has
6675 been changed so that it no longer eats memory in this way.
6677 66. The generation of the Received: header has been moved from the time that a
6678 message starts to be received, to the time that it finishes. The timestamp
6679 in the Received: header should now be very close to that of the <= log
6680 line. There are two side-effects of this change:
6682 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6683 logged header lines no longer include the local Received: line, because
6684 it has not yet been created. The same applies to a copy of the message
6685 that is returned to a non-SMTP sender when a message is rejected.
6687 (b) When a filter file is tested using -bf, no additional Received: header
6688 is added to the test message. After some thought, I decided that this
6691 This change does not affect the value of $received_for. It is still set
6692 after address rewriting, but before local_scan() is called.
6694 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6696 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6697 gave an unhelpful panic error message, and a defer error. I have managed to
6698 change this behaviour so that it now rejects any supplied certificate,
6699 which seems right, as the list of acceptable certificates is empty.
6701 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6702 gave an unhelpful defer error. I have not managed to make this reject any
6703 supplied certificates, but the error message it gives is "no certificate
6704 supplied", which is not helpful.
6706 70. exigrep's output now also includes lines that are not associated with any
6707 message, but which match the given pattern. Implemented by a patch from
6708 Martin Sluka, which also tidied up the Perl a bit.
6710 71. Recipient callout verification, like sender verification, was using <> in
6711 the MAIL FROM command. This isn't really the right thing, since the actual
6712 sender may affect whether the remote host accepts the recipient or not. I
6713 have changed it to use the actual sender in the callout; this means that
6714 the cache record is now keyed on a recipient/sender pair, not just the
6715 recipient address. There doesn't seem to be a real danger of callout loops,
6716 since a callout by the remote host to check the sender would use <>.
6717 [SEE ABOVE: changed after hitting problems.]
6719 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6720 temporary errors. However, in the case of such a code being given after
6721 the end of a data transmission (i.e. after ".") Exim was failing to write
6722 a retry record for the message. (Yes, there was some broken host that was
6723 actually sending 8xx at this point.)
6725 73. An unknown lookup type in a host list could cause Exim to panic-die when
6726 the list was checked. (An example that provoked this was putting <; in the
6727 middle of a list instead of at the start.) If this happened during a DATA
6728 ACL check, a -D file could be left lying around. This kind of configuration
6729 error no longer causes Exim to die; instead it causes a defer error. The
6730 incident is still logged to the main and panic logs.
6732 74. Buglet left over from Exim 3 conversion. The message "too many messages
6733 in one connection" was written to the rejectlog but not the mainlog, except
6734 when address rewriting (yes!) was being logged.
6736 75. Added write_rejectlog option.
6738 76. When a system filter was run not as root (that is, when system_filter_user
6739 was set), the values of the $n variables were not being returned to the
6740 main process; thus, they were not subsequently available in the $sn
6743 77. Added +return_path_on_delivery log selector.
6745 78. A connection timeout was being treated differently from recipients deferred
6746 when testing hosts_max_try with a message that was older than the host's
6747 retry timeout. (The host should not be counted, thus allowing all hosts to
6748 be tried at least once before bouncing.) This may have been the cause of an
6749 occasionally reported bug whereby a message would remain on the queue
6750 longer than the retry timeout, but would be bounced if a delivery was
6751 forced. I say "may" because I never totally pinned down the problem;
6752 setting up timeout/retry tests is difficult. See also the next item.
6754 79. The ultimate address timeout was not being applied to errors that involved
6755 a combination of host plus message (for example, a timeout on a MAIL
6756 command). When an address resolved to a number of possible hosts, and they
6757 were not all tried for each delivery (e.g. because of hosts_max_try), a
6758 message could remain on the queue longer than the retry timeout.
6760 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6763 81. Fixed an obscure SMTP outgoing bug which required at least the following
6764 conditions: (a) there was another message waiting for the same server;
6765 (b) the server returned 5xx to all RCPT commands in the first message so
6766 that the message was not completed; (c) the server dropped the connection
6767 or gave a negative response to the RSET that Exim sends to abort the
6768 transaction. The observed case was a dropped connection after DATA that had
6769 been sent in pipelining mode. That is, the server had advertised PIPELINING
6770 but was not implementing it correctly. The effect of the bug was incorrect
6771 behaviour, such as trying another host, and this could lead to a crash.
6777 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6778 and daemon.c were passed as pointers to ints; they should have been
6779 pointers to socklen_t variables (which are typically unsigned ints).
6781 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6784 3. Fixed a really odd bug that affected only the testing scheme; patching a
6785 certain fixed string in the binary changed the value of another string that
6786 happened to be identical to the end of the original first string.
6788 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6789 name", it returns that address as the IP address. On some operating
6790 systems (e.g. Solaris), it also passes back the IP address string as the
6791 "host name". However, on others (e.g. Linux), it passes back an empty
6792 string. Exim wasn't checking for this, and was changing the host name to an
6793 empty string, assuming it had been canonicalized.
6795 5. Although rare, it is permitted to have more than one PTR record for a given
6796 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6797 all the names associated with an address, because they do in Solaris.
6798 However, it seems that they do not in Linux for data that comes from the
6799 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6800 I found this out when I moved to a new Linux workstation and tried to run
6801 the Exim test suite.
6803 To get round this problem I have changed the code so that it now does its
6804 own call to the DNS to look up PTR records when searching for a host name.
6805 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6806 addresses that are only in /etc/hosts are still found.
6808 This behaviour is, however, controlled by an option called host_lookup_
6809 order, which defaults to "bydns:byaddr". If people want to use the other
6810 order, or indeed, just use one or the other means of lookup, they can
6811 specify it in this variable.
6813 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6814 some operating systems, this comes back from gethostbyaddr() as an empty
6815 string, and this is what Exim used to test for. However, it seems that in
6816 other systems, "." is yielded. Exim now tests for this case too.
6818 7. The values of check_spool_space and check_log_space are now held internally
6819 as a number of kilobytes instead of an absolute number of bytes. If a
6820 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6821 kilobyte. This means that much larger values can be stored.
6823 8. Exim monitor: an attempt to get the action menu when not actually pointing
6824 at a message produces an empty menu entitled "No message selected". This
6825 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6826 no entries in it ("Shell widget menu has zero width and/or height"). So I
6827 have added a single, blank menu entry in this case.
6829 9. Added ${quote_local_part.
6831 10. MIME decoding is now applied to the contents of Subject: header lines when
6834 11. Now that a reference to $sender_host_address automatically causes a reverse
6835 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6836 host lookup before query-style lookups in lists that might use this
6837 variable. This has therefore been abolished, and the "net-" prefix is no
6838 longer necessary for query-style lookups.
6840 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6841 have been a typo for LFLAGS, so it has been changed.
6843 13. The install script calls Exim with "-C /dev/null" in order to find the
6844 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6845 to be output. However, since Exim outputs its version number before the
6846 error, it didn't break the script. It just looked ugly. I fixed this by
6847 always allowing "-C /dev/null" if the caller is root.
6849 14. Ignore overlarge ACL variable number when reading spool file - insurance
6850 against a later release with more variables having written the file.
6852 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6853 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6854 this, as well as the form without the "IPv6" on the front.
6856 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6857 OS/Makefile-Darwin file.
6859 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6860 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6862 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6863 variable. However, some OS use ints for inodes. I've added cast to long int
6864 to get rid of the compiler warning.
6866 19. I had forgotten to lock out "/../" in configuration file names when
6867 ALT_CONFIG_PREFIX was set.
6869 20. Routers used for verification do not need to specify transports. However,
6870 if such a router generated a host list, and callout was configured, Exim
6871 crashed, because it could not find a port number from the (non-existent)
6872 transport. It now assumes port 25 in this circumstance.
6874 21. Added the -t option to exigrep.
6876 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6877 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6878 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6879 them is set, LOOKUP_LSEARCH is forced.
6881 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6883 24. Added sender and host information to the "rejected by local_scan()" log
6884 line; previously there was no indication of these.
6886 25. Added .include_if_exists.
6888 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6889 Linux. It turns out that not all file systems support this. Apparently some
6890 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6891 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6892 not supported on the file descriptor, is now ignored when Exim is trying to
6893 sync a directory. This applies only to Linux.
6895 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6897 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6898 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6901 29. Added support for maildirsize files from supplied patch (modified a bit).
6903 30. The use of :fail: followed by an empty string could lead Exim to respond to
6904 sender verification failures with (e.g.):
6906 550 Verification failed for <xxx>
6907 550 Sender verify failed
6909 where the first response line was missing the '-' that indicates it is not
6910 the final line of the response.
6912 31. The loop for finding the name of the user that called Exim had a hardwired
6913 limit of 10; it now uses the value of finduser_retries, which is used for
6914 all other user lookups.
6916 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6919 33. Exim was neglecting to zero errno before one call of strtol() when
6920 expanding a string and expecting an integer value. On some systems this
6921 resulted in spurious "integer overflow" errors. Also, it was casting the
6922 result into an int without checking.
6924 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6925 did not work. The code looks as if it has *never* worked, though it appears
6926 to have been documented since at least release 1.62. I have made it work.
6928 35. The "timeout_DNS" error in retry rules, also documented since at least
6929 1.62, also never worked. As it isn't clear exactly what this means, and
6930 clearly it isn't a major issue, I have abolished the feature by treating it
6931 as "timeout", and writing a warning to the main and panic logs.
6933 36. The display of retry rules for -brt wasn't always showing the error code
6936 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6937 timeout_connect_A, timeout_connect_MX.
6939 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6940 to the empty sender.
6942 39. The daemon was not analysing the content of -oX till after it had closed
6943 stderr and disconnected from the controlling terminal. This meant that any
6944 syntax errors were only noted on the panic log, and the return code from
6945 the command was 0. By re-arranging the code a little, I've made the
6946 decoding happen first, so such errors now appear on stderr, and the return
6947 code is 1. However, the actual setting up of the sockets still happens in
6948 the disconnected process, so errors there are still only recorded on the
6951 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6952 connections (as happens on some IP stacks) was logged at start up time as
6953 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6954 it from "IPv6 and IPv4", which means that two separate sockets are being
6957 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6958 says whether AF_INET or AF_INET6 was passed as an argument.
6960 42. Exiwhat output was messed up when time zones were included in log
6963 43. Exiwhat now gives more information about the daemon's listening ports,
6964 and whether -tls-on-connect was used.
6966 44. The "port" option of the smtp transport is now expanded.
6968 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6969 silently ignored. Now an error message is written to the main and panic
6972 46. There's a new ACL modifier called "logwrite" which writes to a log file
6973 as soon as it is encountered.
6975 47. Added $local_user_uid and $local_user_gid at routing time.
6977 48. Exim crashed when trying to verify a sender address that was being
6980 49. Exim was recognizing only a space character after ".include". It now also
6981 recognizes a tab character.
6983 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6984 extracting the relevant information from the specification. The man page no
6985 longer contains scrambled data for the -d option, and I've added a section
6986 at the front about calling Exim under different names.
6988 51. Added "extra_headers" argument to the "mail" command in filter files.
6990 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6993 53. Installed eximstats 1.29.
6995 54. Added transport_filter_timeout as a generic transport option.
6997 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6998 Cc: header lines. This was required by RFC 822, but it not required by RFC
7001 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7002 incoming messages that did not have them. Now it does so only if the
7003 message originates locally, that is, if there is no associated remote host
7004 address. When Resent- header lines are present, this applies to the Resent-
7005 lines rather than the non-Resent- lines.
7007 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7008 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7010 58. Messages for configuration errors now include the name of the main
7011 configuration file - useful now that there may be more than one file in a
7012 list (.included file names were always shown).
7014 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7015 for those rare installations that do not start the daemon as root or run it
7016 setuid root. I've cut out the call to initgroups() if the daemon is not
7019 60. The Exim user and group can now be bound into the binary as text strings
7020 that are looked up at the start of Exim's processing.
7022 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7024 62. Added $mailstore_basename variable.
7026 63. Installed patch to sieve.c from Michael Haardt.
7028 64. When Exim failed to open the panic log after failing to open the main log,
7029 the original message it was trying to log was written to stderr and debug
7030 output, but if they were not available (the usual case in production), it
7031 was lost. Now it is written to syslog before the two lines that record the
7032 failures to open the logs.
7034 65. Users' Exim filters run in subprocesses under the user's uid. It is
7035 possible for a "deliver" command or an alias in a "personal" command to
7036 provoke an address rewrite. If logging of address rewriting is configured,
7037 this fails because the process is not running as root or exim. There may be
7038 a better way of dealing with this, but for the moment (because 4.30 needs
7039 to be released), I have disabled address rewrite logging when running a
7040 filter in a non-root, non-exim process.
7046 1. The buildconfig auxiliary program wasn't quoting the value set for
7047 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7048 not defined. This bug was masked in 4.22 by the effect that was fixed in
7051 2. Some messages that were rejected after a message id was allocated were
7052 shown as "incomplete" by exigrep. It no longer does this for messages that
7053 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7055 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7056 have allow_domain_literals set, the ID did not get logged in the <= line.
7057 Domain literals are now always recognized in Message-ID: header lines.
7059 4. The first argument for a ${extract expansion item is the key name or field
7060 number. Leading and trailing spaces in this item were not being ignored,
7061 causing some misleading effects.
7063 5. When deliver_drop_privilege was set, single queue runner processes started
7064 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7065 same command in the process it spins off) were not dropping privilege.
7067 6. When the daemon running as "exim" started a queue runner, it always
7068 re-executed Exim in the spun-off process. This is a waste of effort when
7069 deliver_drop_privilege is set. The new process now just calls the
7070 queue-runner function directly.
7076 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7079 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7080 ignored. Though the use of -oP was forcing the writing of a pid file, it
7081 was always written to the default place.
7083 3. If the message "no IP address found for host xxxx" is generated during
7084 incoming verification, it is now followed by identification of the incoming
7085 connection (so you can more easily find what provoked it).
7087 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7089 5. Added some features to "harden" Exim a bit more against certain attacks:
7091 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7092 be put in Local/Makefile. This is like the never_users runtime option,
7093 but it cannot be overridden. The default setting is "root".
7095 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7096 prefix string with which any file named in a -C command line option
7099 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7100 is retained for -C and -D only if the caller of Exim is root. Without
7101 it, the exim user may also use -C and -D and retain privilege.
7103 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7104 command line option is disabled.
7106 6. Macro names set by the -D option must start with an upper case letter, just
7107 like macro names defined in the configuration file.
7109 7. Added "dereference=" facility to LDAP.
7111 8. Two instances of the typo "uknown" in the source files are fixed.
7113 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7114 the Configure-Makefile script screwed up while processing it.
7116 10. Incorporated PCRE 4.4.
7118 11. The SMTP synchronization check was not operating right at the start of an
7119 SMTP session. For example, it could not catch a HELO sent before the client
7120 waited for the greeting. There is now a check for outstanding input at the
7121 point when the greeting is written. Because of the duplex, asynchronous
7122 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7123 way, but not yet received, when the check is performed.
7125 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7126 on TCP/IP sockets, because this apparently causes some broken clients to
7129 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7130 unchanged) from the Cygwin maintainer.
7132 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7133 was supported instead of testing for mbx. Effectively a typo.
7135 15. The spa authenticator server code was not checking that the input it
7136 received was valid base64.
7138 16. The debug output line for the "set" modifier in ACLs was not showing the
7139 name of the variable that was being set.
7141 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7143 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7144 lookup on its own. Something else was needed to trigger the lookup. For
7145 example, a match in host_lookup or the need for a host name in a host list.
7146 Now, if $sender_host_name is referenced and the host name has not yet been
7147 looked up, a lookup is performed. If the lookup fails, the variable remains
7148 empty, and $host_lookup_failed is set to "1".
7150 19. Added "eqi" as a case-independent comparison operator.
7152 20. The saslauthd authentication condition could segfault if neither service
7153 nor realm was specified.
7155 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7156 error message that was logged was misleading, and incoming SMTP
7157 connections were dropped. The message is now more accurate, and temporary
7158 errors are given to SMTP connections.
7160 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7161 (except RSET) with a 421 error, until QUIT is received. However, it was
7162 failing to send a response to QUIT.
7164 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7165 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7166 ACL is now run after the helo_try_verify_hosts code.
7168 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7169 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7170 case-independent, so other case variants are also recognized). Apparently
7171 some systems use these upper case variants.
7173 25. If more than two messages were waiting for the same host, and a transport
7174 filter was specified for the transport, Exim sent two messages over the
7175 same TCP/IP connection, and then failed with "socket operation on non-
7176 socket" when it tried to send the third.
7178 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7180 27. The extern definition of crypt16() in expand.c was not being excluded when
7181 the OS had its own crypt16() function.
7183 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7184 as a preferred synonym for return_size_limit, both as an option and as an
7187 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7189 30. Changed the default configuration ACL to relax the local part checking rule
7190 for addresses that are not in any local domains. For these addresses,
7191 slashes and pipe symbols are allowed within local parts, but the sequence
7192 /../ is explicitly forbidden.
7194 31. SPA server authentication was not clearing the challenge buffer before
7197 32. log_message in a "warn" ACL statement was writing to the reject log as
7198 well as to the main log, which contradicts the documentation and doesn't
7199 seem right (because no rejection is happening). So I have stopped it.
7201 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7202 However, I am unable to do any testing of this.
7204 34. Fixed an infelicity in the appendfile transport. When checking directories
7205 for a mailbox, to see if any needed to be created, it was accidentally
7206 using path names with one or more superfluous leading slashes; tracing
7207 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7209 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7210 contents are added to the log line that is written for every discarded
7211 recipient. (Previously a log_message setting was ignored.)
7213 36. The ${quote: operator now quotes the string if it is empty.
7215 37. The install script runs exim in order to find its version number. If for
7216 some reason other than non-existence or emptiness, which it checks, it
7217 could not run './exim', it was installing it with an empty version number,
7218 i.e. as "exim-". This error state is now caught, and the installation is
7221 38. An argument was missing from the function that creates an error message
7222 when Exim fails to connect to the socket for saslauthd authentication.
7223 This could cause Exim to crash, or give a corrupted message.
7225 39. Added isip, isip4, and isip6 to ${if conditions.
7227 40. The ACL variables $acl_xx are now saved with the message, and can be
7228 accessed later in routers, transports, and filters.
7230 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7231 strings in the file are not string-expanded.
7233 42. If a MAIL command specified a SIZE value that was too large to fit into an
7234 int variable, the check against message_size_limit failed. Such values are
7235 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7236 day this will have to be increased, but I don't think I want to be around
7237 when emails are that large.
7244 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7245 iconv() is not standard in FreeBSD.
7247 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7248 IPv6 enabled. The observed symptom was a segmentation fault on return from
7249 the function os_common_find_running_interfaces() in src/os.c.
7251 3. In the check_special_case() function in daemon.c I had used "errno" as an
7252 argument name, which causes warnings on some systems. This was basically a
7253 typo, since it was named "eno" in the comments!
7255 4. The code that waits for the clock to tick (at a resolution of some fraction
7256 of a second) so as to ensure message-id uniqueness was always waiting for
7257 at least one whole tick, when it could have waited for less. [This is
7258 almost certainly not relevant at current processor speeds, where it is
7259 unlikely to ever wait at all. But we try to future-proof.]
7261 5. The function that sleeps for a time interval that includes fractions of a
7262 second contained a race. It did not block SIGALRM between setting the
7263 timer, and suspending (a couple of lines later). If the interval was short
7264 and the sigsuspend() was delayed until after it had expired, the suspension
7265 never ended. On busy systems this could lead to processes getting stuck for
7268 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7269 process, before it forks any delivery processes. The open lookup caching
7270 mechanism meant that the open file or database connection was passed into
7271 the delivery process. The problem was that delivery processes always tidy
7272 up cached lookup data. This could cause a problem for the next delivery
7273 process started by the queue runner, because the external queue runner
7274 process does not know about the closure. So the next delivery process
7275 still has data in the lookup cache. In the case of a file lookup, there was
7276 no problem because closing a file descriptor in a subprocess doesn't affect
7277 the parent. However, if the lookup was caching a connection to a database,
7278 the connection was closed, and the second delivery process was likely to
7279 see errors such as "PGSQL: query failed: server closed the connection
7280 unexpectedly". The problem has been fixed by closing all cached lookups
7281 in a queue runner before running a delivery process.
7283 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7284 seem to have the "const" qualifier which it has on other OS. I've
7287 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7288 *of the same type* (client or server) with the same public name that an
7289 error should be diagnosed.
7291 9. When Exim looked up a host name for an IP address, but failed to find the
7292 original IP address when looking up the host name (a safety check), it
7293 output the message "<ip address> does not match any IP for NULL", which was
7294 confusing, to say the least. The bug was that the host name should have
7295 appeared instead of "NULL".
7297 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7298 user that is built into the binary, and the -C or -D options is used, root
7299 privilege is dropped before the configuration file is read. In addition,
7300 logging is switched to stderr instead of the normal log files. If the
7301 configuration then re-defines the Exim user, the unprivileged environment
7302 is probably not what is expected, so Exim logs a panic warning message (but
7305 However, if deliver_drop_privilege is set, the unprivileged state may well
7306 be exactly what is intended, so the warning has been cut out in that case,
7307 and Exim is allowed to try to write to its normal log files.
7313 1. smtp_return_error_details was not giving details for temporary sender
7314 or receiver verification errors.
7316 2. Diagnose a configuration error if two authenticators have the same public
7319 3. Exim used not to create the message log file for a message until the first
7320 delivery attempt. This could be confusing when incoming messages were held
7321 for policy or load reasons. The message log file is now created at the time
7322 the message is received, and an initial "Received" line is written to it.
7324 4. The automatically generated man page for command line options had a minor
7325 bug that caused no ill effects; however, a more serious problem was that
7326 the procedure for building the man page automatically didn't always
7327 operate. Consequently, release 4.20 contains an out-of-date version. This
7328 shouldn't happen again.
7330 5. When building Exim with embedded Perl support, the script that builds the
7331 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7332 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7334 6. The freeze_tell option was not being used for messages that were frozen on
7335 arrival, either by an ACL or by local_scan().
7337 7. Added the smtp_incomplete_transaction log selector.
7339 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7340 was accepting AUTH without a new EHLO.
7342 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7343 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7346 10. Logging of TCP/IP connections (when configured) now happens in the main
7347 daemon process instead of the child process, so that the TCP/IP connection
7348 count is more accurate (but it can never be perfect).
7350 11. The use of "drop" in a nested ACL was not being handled correctly in the
7351 outer ACL. Now, if condition failure induced by the nested "drop" causes
7352 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7353 or "require"), the connection is dropped.
7355 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7356 that yield "discard" can now be used with an "accept" or a "discard" verb,
7357 but an error is generated for any others (because I can't see a useful way
7358 to define what should happen).
7360 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7361 are now processed in the same way as lines in the Exim configuration file.
7362 In particular, continuation lines are supported.
7364 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7366 15. Added -ti meaning -t -i.
7368 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7369 domains, and warn by logging if others are found.
7371 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7372 returned from GIFCONF when Exim is trying to find the list of interfaces on
7373 a host. The code in os.c has been modified to copy each ifreq to an aligned
7374 structure in all cases.
7376 Also, in some cases, the returned ifreq's were being copied to a 'struct
7377 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7378 means the last couple of bytes of an IPv6 address could be chopped if the
7379 ifreq contained only a normal sockaddr (14 bytes storage).
7381 18. Named domain lists were not supported in the hosts_treat_as_local option.
7382 An entry such as +xxxx was not recognized, and was treated as a literal
7385 19. Ensure that header lines added by a DATA ACL are included in the reject log
7386 if the ACL subsequently rejects the message.
7388 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7389 MD5 (which is deprecated).
7391 21. When testing a filter file using -bf, Exim was writing a message when it
7392 took the sender from a "From " line in the message, but it was not doing so
7393 when it took $return_path from a Return-Path: header line. It now does.
7395 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7396 with a valid header line field name (a series of printing characters
7397 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7399 23. Changed "disc" in the source to "disk" to conform to the documentation and
7400 the book and for uniformity.
7402 24. Ignore Sendmail's -Ooption=value command line item.
7404 25. When execve() failed while trying to run a command in a pipe transport,
7405 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7406 could be confused with a return value of 69 from the command itself. This
7407 has been changed to 127, the value the shell returns if it is asked to run
7408 a non-existent command. The wording for the related log line suggests a
7409 non-existent command as the problem.
7411 26. If received_header_text expands to an empty string, do not add a Received:
7412 header line to the message. (Well, it adds a token one on the spool, but
7413 marks it "old" so that it doesn't get used or transmitted.)
7415 27. Installed eximstats 1.28 (addition of -nt option).
7417 28. There was no check for failure on the call to getsockname() in the daemon
7418 code. This can fail if there is a shortage of resources on the system, with
7419 ENOMEM, for example. A temporary error is now given on failure.
7421 29. Contrary to the C standard, it seems that in some environments, the
7422 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7423 program. Exim now does this explicitly; it affects the formatting of
7424 timestamps using strftime().
7426 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7427 complaints. I've now initialized all the variables, to avoid this.
7429 32. Header lines added by a system filter were not being "seen" during
7430 transport-time rewrites.
7432 33. The info_callback() function passed to OpenSSL is set up with type void
7433 (*)(SSL *, int, int), as described somewhere. However, when calling the
7434 function (actually a macro) that sets it up, the type void(*)() is
7435 expected. I've put in a cast to prevent warnings from picky compilers.
7437 34. If a DNS black list lookup found a CNAME record, but there were no A
7438 records associated with the domain it pointed at, Exim crashed.
7440 35. If a DNS black list lookup returned more than one A record, Exim ignored
7441 all but the first. It now scans all returned addresses if a particular IP
7442 value is being sought. In this situation, the contents of the
7443 $dnslist_value variable are a list of all the addresses, separated by a
7446 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7447 to accept a host name and all its aliases if the forward lookup for any of
7448 them yielded the IP address of the incoming connection. Now it accepts only
7449 those names whose forward lookup yields the correct IP address. Any other
7450 names are discarded. This closes a loophole whereby a rogue DNS
7451 administrator could create reverse DNS records to break through a
7452 wildcarded host restriction in an ACL.
7454 37. If a user filter or a system filter that ran in a subprocess used any of
7455 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7456 the wrong values were passed to the pipe command ($thisaddress had the
7457 value of $0, $0 had the value of $1, etc). This bug was introduced by
7458 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7460 38. Improved the line breaking for long SMTP error messages from ACLs.
7461 Previously, if there was no break point between 40 and 75 characters, Exim
7462 left the rest of the message alone. Two changes have been made: (a) I've
7463 reduced the minimum length to 35 characters; (b) if it can't find a break
7464 point between 35 and 75 characters, it looks ahead and uses the first one
7465 that it finds. This may give the occasional overlong line, but at least the
7466 remaining text gets split now.
7468 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7469 file descriptors might be low, and that setting 1000 would always raise it.
7470 It turns out that in some environments, the limit is already over 1000 and
7471 that lowering it causes trouble. So now Exim takes care not to decrease it.
7473 40. When delivering a message, the value of $return_path is set to $sender_
7474 address at the start of routing (routers may change the value). By an
7475 oversight, this default was not being set up when an address was tested by
7476 -bt or -bv, which affected the outcome if any router or filter referred to
7479 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7480 log only when "message" or "log_message" are set. However, if one of the
7481 conditions was an address verification, or a call to a nested ACL, the
7482 messages generated by the underlying test were being passed through. This
7483 no longer happens. The underlying message is available in $acl_verify_
7484 message for both "message" and "log_message" expansions, so it can be
7485 passed through if needed.
7487 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7488 new expansion $bh_ to give the encoded byte string without charset
7489 translation. Translation happens only if iconv() is available; HAVE_ICONV
7490 indicates this at build time. HEADERS_CHARSET gives the charset to
7491 translate to; headers_charset can change it in the configuration, and
7492 "headers charset" can change it in an individual filter file.
7494 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7495 that creates RFC 2047 encoded "words" labels them as that charset instead
7496 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7497 expansion operator; (ii) when Exim creates a From: line for a local
7498 message; (iii) when a header line is rewritten to include a "phrase" part.
7500 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7501 buggy, causing it to skip the first lines of messages whose message ID
7502 ended in 'D'. This would not have bitten before Exim release 4.14, because
7503 message IDs were unlikely to end in 'D' before then. The effect was to have
7504 incorrect size information for certain domains.
7506 45. #include "config.h" was missing at the start of the crypt16.c module. This
7507 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7510 46. If there was a timeout during a "random" callout check, Exim treated it as
7511 a failure of the random address, and carried on sending RSET and the real
7512 address. If the delay was just some slowness somewhere, the response to the
7513 original RCPT would be taken as a response to RSET and so on, causing
7514 mayhem of various kinds.
7516 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7517 when I implemented it. It didn't allow for the fact that some option values
7518 may legitimately be negative (e.g. size_addition), and it didn't even do
7519 the right test for positive values.
7521 48. Domain names in DNS records are case-independent. Exim always looks them up
7522 in lower case. Some resolvers return domain names in exactly the case they
7523 appear in the zone file, that is, they may contain uppercase letters. Not
7524 all resolvers do this - some return always lower case. Exim was treating a
7525 change of case by a resolver as a change of domain, similar to a widening
7526 of a domain abbreviation. This triggered its re-routing code and so it was
7527 trying to route what was effectively the same domain again. This normally
7528 caused routing to fail (because the router wouldn't handle the domain
7529 twice). Now Exim checks for this case specially, and just changes the
7530 casing of the domain that it ultimately uses when it transmits the message
7533 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7536 50. If a filter generated a file delivery with a non-absolute name (possible if
7537 no home directory exists for the router), the forbid_file option was not
7540 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7541 the existing equality matching.
7543 52. Exim was using ints instead of ino_t variables in some places where it was
7544 dealing with inode numbers.
7546 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7547 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7548 called TMPDIR, and if it finds it is different, it changes its value.
7550 54. The smtp_printf() function is now made available to local_scan() so
7551 additional output lines can be written before returning. There is also an
7552 smtp_fflush() function to enable the detection of a dropped connection.
7553 The variables smtp_input and smtp_batched_input are exported to
7556 55. Changed the default runtime configuration: the message "Unknown user"
7557 has been removed from the ACL, and instead placed on the localuser router,
7558 using the cannot_route_message feature. This means that any verification
7559 failures that generate their own messages won't get overridden. Similarly,
7560 the "Unrouteable address" message that was in the ACL for unverifiable
7561 relay addresses has also been removed.
7563 56. Added hosts_avoid_esmtp to the smtp transport.
7565 57. The exicyclog script was not checking for the esoteric option
7566 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7567 will work only if exicyclog is run under the appropriate euid.
7569 58. Following a discussion on the list, the rules by which Exim recognises line
7570 endings on incoming messages have been changed. The -dropcr and drop_cr
7571 options are now no-ops, retained only for backwards compatibility. The
7572 following line terminators are recognized: LF CRLF CR. However, special
7573 processing applies to CR:
7575 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7576 nor a local message in the state where . is a terminator.
7578 (ii) If a bare CR is encountered in a header line, an extra space is added
7579 after the line terminator so as not to end the header. The reasoning
7580 behind this is that bare CRs in header lines are most likely either
7581 to be mistakes, or people trying to play silly games.
7583 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7584 was being incorrectly given as 18 bytes larger than it should have been.
7585 This is a VOB (very old bug).
7587 60. This may never have affected anything current, but just in case it has:
7588 When the local host is found other than at the start of a list of hosts,
7589 the local host, those with the same MX, and any that follow, are discarded.
7590 When the list in question was part of a longer list of hosts, the following
7591 hosts (not currently being processed) were also being discarded. This no
7592 longer happens. I'm not sure if this situation could ever has previously
7595 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7598 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7599 header. This is something that is recommended in a new Internet Draft, and
7600 is something that is documented as being done by Sendmail. There are two
7601 possible values. For messages generated by the autoreply transport, Exim
7604 Auto-Submitted: auto-replied
7606 whereas for all other generated messages (e.g. bounces) it adds
7608 Auto-Submitted: auto-generated
7610 63. The "personal" condition in filters now includes a test for the
7611 Auto-Submitted: header. If it contains the string "auto-" the message it
7612 not considered personal.
7614 64. Added rcpt_include_affixes as a generic transport option.
7616 65. Added queue_only_override (default true).
7618 66. Added the syslog_duplication option.
7620 67. If what should have been the first header line of a message consisted of
7621 a space followed by a colon, Exim was mis-interpreting it as a header line.
7622 It isn't of course - it is syntactically invalid and should therefore be
7623 treated as the start of the message body. The misbehaviour could have
7624 caused a number of strange effects, including loss of data in subsequent
7625 header lines, and spool format errors.
7627 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7628 client host had authenticated. This control can now be exercised by an ACL
7629 for more flexibility.
7631 69. By default, callouts do not happen when testing with -bh. There is now a
7632 variant, -bhc, which does actually run the callout code, including
7633 consulting and updating the callout cache.
7635 70. Added support for saslauthd authentication, courtesy of Alexander
7638 71. If statvfs() failed on the spool or log directories while checking their
7639 size for availability, Exim confusingly gave the error "space shortage".
7640 Furthermore, in debugging mode it crashed with a floating point exception.
7641 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7642 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7643 serious problem, Exim now writes to the main and panic logs when this
7644 happens, with details of the failure. It then refuses to accept the
7645 incoming message, giving the message "spool directory problem" or "log
7646 directory problem" with a 421 code for SMTP messages.
7648 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7649 0, 1, and 2 exist, because some OS complain for execs without them (see
7650 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7651 descriptors. However, the code omitted to check that the open succeeded,
7652 causing mysterious errors if for some reason the permissions on /dev/null
7653 got screwed. Now Exim writes a message to the main and panic logs, and
7654 bombs out if it can't open /dev/null.
7656 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7657 interact so that it is all more flexible. It is supposed to remain
7658 backwards compatible. Also added extra_local_interfaces.
7660 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7661 to bomb out with an assertion failure - to the client this appears as a
7662 connection drop. This problem occurs in the part of the code that was taken
7663 from the Samba project. Fortunately, the assertion is in a very simple
7664 function, so I have fixed this by reproducing the function inline in the
7665 one place where it is called, and arranging for authentication to fail
7666 instead of killing the process with assert().
7668 75. The SPA client code was not working when the server requested OEM rather
7669 than Unicode encoding.
7671 76. Added code to make require_files with a specific uid setting more usable in
7672 the case where statting the file as root fails - usually a non-root-mounted
7673 NFS file system. When this happens and the failure is EACCES, Exim now
7674 forks a subprocess and does the per-uid checking as the relevant uid.
7676 77. Added process_log_path.
7678 78. If log_file_path was not explicitly set, a setting of check_log_space or
7679 check_log_inodes was ignored.
7681 79. If a space check for the spool or log partitions fails, the incident is now
7682 logged. Of course, in the latter case the data may get lost...
7684 80. Added the %p formatting code to string_format() so that it can be used to
7685 print addresses in debug_print(). Adjusted all the address printing in the
7686 debugging in store.c to use %p rather than %d.
7688 81. There was a concern that a line of code in smtp_in.c could overflow a
7689 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7690 initially expressed, the concern was not well-founded, because trailing
7691 spaces are removed early. However, if the trailing spaces were followed by
7692 a NULL, they did not get removed, so the overflow was possible. Two fixes
7695 (a) I re-wrote the offending code in a cleaner fashion.
7696 (b) If an incoming SMTP command contains a NULL character, it is rejected
7699 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7700 runs initgroups(), so that if the Exim user is in any additional groups,
7701 they will be used during message reception.
7707 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob