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.
84 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
85 side, to decode a TLS 1.3 packet capture.
87 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
88 Previously the default library behaviour applied, sending two, each in
91 JH/03 Debug output for ACL now gives the config file name and line number for
94 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
96 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
98 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
99 buffer overrun for (non-chunking) other transports.
101 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
102 TLS1.3, means that a server rejecting a client certificate is not visible
103 to the client until the first read of encrypted data (typically the
104 response to EHLO). Add detection for that case and treat it as a failed
105 TLS connection attempt, so that the normal retry-in-clear can work (if
106 suitably configured).
108 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
109 and/or domain. Found and fixed by Jason Betts.
111 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
112 configuration). If a CNAME target was not a wellformed name pattern, a
115 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
116 the OS reports them interleaved with other addresses.
118 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
119 used both for input and for a verify callout, both encrypted, SMTP
120 responses being sent by the server could be lost. This resulted in
121 dropped connections and sometimes bounces generated by a peer sending
124 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
125 string. Previously it was possible to cause undefined behaviour in a
126 library routine (usually a crash). Found by "zerons".
128 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
131 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
132 API was removed, so update to use the newer ones.
134 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
135 any timeout set, is taking a long time. Previously we would hang on to a
136 rotated logfile "forever" if the input was arriving with long gaps
137 (a previous attempt to fix addressed lack, for a long time, of initial
140 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
141 shared (NFS) environment. The length of the tempfile name is now
142 4 + 16 ("hdr.$message_exim_id") which might break on file
143 systems which restrict the file name length to lower values.
146 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
147 shared (NFS) environment.
149 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
150 did for all versions <4.90). Notably -M, -m, --invert, -I may be
153 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
154 on some platforms for bit 31.
156 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
157 to changes apparently associated with TLS1.3 handling some of the APIs
158 previously used were either nonfunctional or inappropriate. Strings
159 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
160 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
161 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
162 This affects log line X= elements, the $tls_{in,out}_cipher variables,
163 and the use of specific cipher names in the encrypted= ACL condition.
165 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
167 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
168 verification result was not updated unless hosts_require_ocsp applied.
170 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
171 queue_list_requires_admin set to false, non-admin users were denied the
174 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
175 directory-of-certs mode. Previously they were advertised despite the
178 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
179 A single TCP connection by a client will now hold a TLS connection open
180 for multiple message deliveries, by default. Previoud the default was to
183 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
184 default. If built with the facility, DANE will be used. The facility
185 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
187 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
188 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
189 must be defined and you must still, unless you define DISABLE_TLS, manage
190 the the include-dir and library-file requirements that go with that
191 choice. Non-TLS builds are still supported.
193 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
196 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
197 default. If the platform supports and has the facility enabled, it will
198 be requested on all coneections.
200 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
201 controlled by the build-time option SUPPORT_PIPE_CONNECT.
203 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
205 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
206 success-DSN messages. Previously the From: header was always the default
207 one for these; the option was ignored.
209 JH/28 Fix the timeout on smtp response to apply to the whole response.
210 Previously it was reset for every read, so a teergrubing peer sending
211 single bytes within the time limit could extend the connection for a
212 long time. Credit to Qualsys Security Advisory Team for the discovery.
214 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
215 delivery address, which leaked information of the results of local
216 forwarding. Change to the original envelope recipient address, per
219 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
220 requested. Previously not bounce was generated and a log entry of
221 error ignored was made.
223 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
225 JH/32 Introduce a general tainting mechanism for values read from the input
226 channel, and values derived from them. Refuse to expand any tainted
227 values, to catch one form of exploit.
229 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
230 was unused and the unexpanded text used for the test. Found and
231 fixed by Ruben Jenster.
233 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
234 an attempt to use a TLS library read routine dereffed a nul pointer,
237 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
238 them in our smtp response.
240 JH/36 Have the general router option retry_use_local_part default to true when
241 any of the restrictive preconditions are set (to anything). Previously it
242 was only for check_local user. The change removes one item of manual
243 configuration which is required for proper retries when a remote router
244 handles a subset of addresses for a domain.
246 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
247 link count into consideration.
249 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
250 caused the extension of big_buffer, the following lines were ignored.
252 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
253 accordance with RFC 2308. Previously there was no expiry, so a longlived
254 receive process (eg. due to ACL delays) versus a short SOA value could
257 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
259 JH/39 Promote DMARC support to mainline.
261 JH/40 Bug 2452: Add a References: header to DSNs.
263 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
264 parameters. The relevant library call is documented as "Deprecated: This
265 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
266 3.6.0, DH parameters are negotiated following RFC7919."
268 HS/06 Change the default of dnssec_request_domains to "*"
270 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
271 carried on and emitted a BDAT command, even when PIPELINING was not
274 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
275 buffer was used for the filename, resulting in a trap when tainted
276 arguments (eg. $domain) were used.
278 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
279 recommended to avoid a possible server-load attack. The feature can be
280 re-enabled via the openssl_options main cofiguration option.
282 JH/45 local_scan API: documented the current smtp_printf() call. This changed
283 for version 4.90 - adding a "more data" boolean to the arguments.
284 Bumped the ABI version number also, this having been missed previously;
285 release versions 4.90 to 4.92.3 inclusive were effectively broken in
286 respect of usage of smtp_printf() by either local_scan code or libraries
287 accessed via the ${dlfunc } expansion item. Both will need coding
288 adjustment for any calls to smtp_printf() to match the new function
289 signature; a FALSE value for the new argument is always safe.
291 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
292 the file-offset (which the Linux syscall does, and exim expects); this
293 resulted in an indefinite loop.
295 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
296 to do ARC verification. The Authentication-Results: header line added
297 by the configuration then had no ARC item.
303 JH/01 Remove code calling the customisable local_scan function, unless a new
304 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
306 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
307 non-signal-safe functions being used.
309 JH/03 Bug 2269: When presented with a received message having a stupidly large
310 number of DKIM-Signature headers, disable DKIM verification to avoid
311 a resource-consumption attack. The limit is set at twenty.
313 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
314 report of oldest_pass in ${authres } in consequence, and separate out
315 some descriptions of reasons for verification fail.
317 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
318 files in the spool were present and unlocked. A queue-runner could spot
319 them, resulting in a duplicate delivery. Fix that by doing the unlock
320 after the unlink. Investigation by Tim Stewart. Take the opportunity to
321 add more error-checking on spoolfile handling while that code is being
324 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
325 No known attacks, no CVE, this is defensive hardening.
327 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
328 a queue-runner could start a delivery while other operations were ongoing.
329 Cutthrough delivery was a common victim, resulting in duplicate delivery.
330 Found and investigated by Tim Stewart. Fix by using the open message data
331 file handle rather than opening another, and not locally closing it (which
332 releases a lock) for that case, while creating the temporary .eml format
333 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
335 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
336 $sender_verify_failure/$recipient_verify_failure to "random".
338 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
341 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
342 Previously this would segfault.
344 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
347 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
348 like zero, since the resolver should be doing this for us, But we need one
349 as a CNAME but no MX presence gets the CNAME returned; we need to check
350 that doesn't point to an MX to declare it "no MX returned" rather than
351 "error, loop". A new main option is added so the older capability of
352 following some limited number of chain links is maintained.
354 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
356 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
357 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
358 now-deprecated earlier definitions used only specified the range up to TLS
359 1.2 (in the older-version library docs).
361 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
363 JH/15 Rework TLS client-side context management. Stop using a global, and
364 explicitly pass a context around. This enables future use of TLS for
365 connections to service-daemons (eg. malware scanning) while a client smtp
366 connection is using TLS; with cutthrough connections this is quite likely.
368 JH/16 Fix ARC verification to do AS checks in reverse order.
370 JH/17 Support a "tls" option on the ${readsocket } expansion item.
372 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
373 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
374 Previously the "utf8" would be re-prepended for every additional message.
376 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
377 Previously thery were accepted, resulting in issues when attempting to
378 forward messages to a non-supporting MTA.
380 PP/02 Let -n work with printing macros too, not just options.
382 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
383 one parent address was copied, and bogus data was used at delivery-logging
384 time. Either a crash (after delivery) or bogus log data could result.
385 Discovery and analysis by Tim Stewart.
387 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
388 Previously if the string ended mid-character, we did not insert the
389 promised '?' replacement.
391 PP/04 Documentation: current string operators work on bytes, not codepoints.
393 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
394 should pack well giving a smaller memory footprint so better caching and
395 therefore performance. Group the declarations where this can't be done so
396 that the byte-sized flag variables are not interspersed among pointer
397 variables, giving a better chance of good packing by the compiler.
399 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
400 non-null, to avoid issues with sites running BATV. Previously reports were
401 sent with an empty envelope sender so looked like bounces.
403 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
404 The ignore_error flag wasn't being returned from the filter subprocess so
405 was not set for later routers. Investigation and fix by Matthias Kurz.
407 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
408 and a msg:complete for the whole, when a message is manually removed using
409 -Mrm. Developement by Matthias Kurz, hacked on by JH.
411 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
412 a "Gnu special" function, asprintf() in the DB utility binary builds; I
413 hope that is portable enough.
415 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
416 requiring a known-CA anchor certificate; make it now rely entirely on the
417 TLSA as an anchor. Checking the name on the leaf cert against the name
418 on the A-record for the host is still done for TA (but not for EE mode).
420 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
421 included in delivery lines for non-proxied connections, when compiled with
422 SUPPORT_SOCKS and running with proxy logging enabled.
424 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
425 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
426 move the existing event to fire before the normal logging of message
427 failure so that custom logging is bracketed by normal logging.
429 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
430 msg:fail:internal event. Developement by Matthias Kurz.
432 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
433 far too small for todays use of crypto signatures stored there. Go all
434 the way to the max DNS message size of 64kB, even though this might be
435 overmuch for IOT constrained device use.
437 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
438 copy a string over itself. The library routine is documented as not
439 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
441 JH/32 For main options check_spool_space and check_inode_space, where the
442 platform supports 64b integers, support more than the previous 2^31 kB
443 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
444 the previous G, M, k.
446 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
447 $authenticated_fail_id variable on authentication failure. Previously
450 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
451 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
452 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
453 GNUTLS_SEC_PARAM_MEDIUM.
455 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
456 side. Previously we would continue as if no SNI had been received.
458 JH/36 Harden the handling of string-lists. When a list consisted of a sole
459 "<" character, which should be a list-separator specification, we walked
460 off past the nul-terimation.
462 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
463 causes) even when the retry time is not yet met. Previously they were
464 not, meaning that when (say) an account was over-quota and temp-rejecting,
465 and multiple senders' messages were queued, only one sender would get
466 notified on each configured delay_warning cycle.
468 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
470 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
471 cipher-suites, an error can be left on the stack even for a succeeding
472 accept; this results in impossible error messages when a later operation
475 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
476 return error codes indicating retry. Under TLS1.3 this becomes required.
478 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
479 it only wrote the new authenticators, resulting in a lack of tracking of
480 peer changes of ESMTP extensions until the next cache flush.
482 JH/41 Fix the loop reading a message header line to check for integer overflow,
483 and more-often against header_maxsize. Previously a crafted message could
484 induce a crash of the recive process; now the message is cleanly rejected.
486 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
487 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
493 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
494 When redis_servers is set to a list of > 1 element, and the Redis servers
495 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
496 case of MOVED into a DEFER case instead, thus moving the query onto the
497 next server in the list. For a cluster of N elements, all N servers must
498 be defined in redis_servers.
500 GF/02 Catch and remove uninitialized value warning in exiqsumm
501 Check for existence of @ARGV before looking at $ARGV[0]
503 JH/01 Replace the store_release() internal interface with store_newblock(),
504 which internalises the check required to safely use the old one, plus
505 the allocate and data copy operations duplicated in both (!) of the
506 extant use locations.
508 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
509 modifier. This matches the restriction on the commandline.
511 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
512 Previously only the last row was returned.
514 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
515 we assumed that tags in the header were well-formed, and parsed the
516 element content after inspecting only the first char of the tag.
517 Assumptions at that stage could crash the receive process on malformed
520 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
521 While running the DKIM ACL we operate on the Permanent memory pool so that
522 variables created with "set" persist to the DATA ACL. Also (at any time)
523 DNS lookups that fail create cache records using the Permanent pool. But
524 expansions release any allocations made on the current pool - so a dnsdb
525 lookup expansion done in the DKIM ACL releases the memory used for the
526 DNS negative-cache, and bad things result. Solution is to switch to the
527 Main pool for expansions.
528 While we're in that code, add checks on the DNS cache during store_reset,
529 active in the testsuite.
530 Problem spotted, and debugging aided, by Wolfgang Breyha.
532 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
533 When none of the hosts presented to a transport match an already-open
534 connection, close it and proceed with the list. Previously we would
535 queue the message. Spotted by Lena with Yahoo, probably involving
538 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
539 Previously a spurious "250 OK id=" response was appended to the proper
542 JH/08 The "support for" informational output now, which built with Content
543 Scanning support, has a line for the malware scanner interfaces compiled
544 in. Interface can be individually included or not at build time.
546 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
547 by the template makefile "src/EDITME". The "STREAM" support for an older
548 ClamAV interface method is removed.
550 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
551 rows affected is given instead).
553 JH/11 The runtime Berkeley DB library version is now additionally output by
554 "exim -d -bV". Previously only the compile-time version was shown.
556 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
557 SMTP connection. Previously, when one had more receipients than the
558 first, an abortive onward connection was made. Move to full support for
559 multiple onward connections in sequence, handling cutthrough connection
560 for all multi-message initiating connections.
562 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
563 routers. Previously, a multi-recipient message would fail to match the
564 onward-connection opened for the first recipient, and cause its closure.
566 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
567 a timeout on read on a GnuTLS initiating connection, resulting in the
568 initiating connection being dropped. This mattered most when the callout
569 was marked defer_ok. Fix to keep the two timeout-detection methods
572 JH/15 Relax results from ACL control request to enable cutthrough, in
573 unsupported situations, from error to silently (except under debug)
574 ignoring. This covers use with PRDR, frozen messages, queue-only and
577 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
579 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
580 metadata, resulting in a crash in free().
582 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
583 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
584 Broken also in d185889f4, with init system revamp.
586 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
587 Previously we abruptly closed the connection after reading a malware-
588 found indication; now we go on to read the "scan ok" response line,
591 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
592 ACL. Previously, a crash would result.
594 JH/19 Speed up macro lookups during configuration file read, by skipping non-
595 macro text after a replacement (previously it was only once per line) and
596 by skipping builtin macros when searching for an uppercase lead character.
598 JH/20 DANE support moved from Experimental to mainline. The Makefile control
599 for the build is renamed.
601 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
602 was allocated for every new TLS startup, meaning one per message. Fix
603 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
605 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
606 reported the original. Fix to report (as far as possible) the ACL
607 result replacing the original.
609 JH/23 Fix memory leak during multi-message connections using STARTTLS under
610 OpenSSL. Certificate information is loaded for every new TLS startup,
611 and the resources needed to be freed.
613 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
615 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
618 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
619 DATA response info to the (existing) per-recipient response info for
620 the "C=" log element. It can have useful tracking info from the
621 destination system. Patch from Simon Arlott.
623 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
624 length value. Previously this would segfault.
626 HS/02 Support Avast multiline protoocol, this allows passing flags to
627 newer versions of the scanner.
629 JH/28 Ensure that variables possibly set during message acceptance are marked
630 dead before release of memory in the daemon loop. This stops complaints
631 about them when the debug_store option is enabled. Discovered specifically
632 for sender_rate_period, but applies to a whole set of variables.
633 Do the same for the queue-runner and queue-list loops, for variables set
634 from spool message files. Do the same for the SMTP per-message loop, for
635 certain variables indirectly set in ACL operations.
637 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
638 as a multi-recipient message from a mailinglist manager). The coding had
639 an arbitrary cutoff number of characters while checking for more input;
640 enforced by writing a NUL into the buffer. This corrupted long / fast
641 input. The problem was exposed more widely when more pipelineing of SMTP
642 responses was introduced, and one Exim system was feeding another.
643 The symptom is log complaints of SMTP syntax error (NUL chars) on the
644 receiving system, and refused recipients seen by the sending system
645 (propating to people being dropped from mailing lists).
646 Discovered and pinpointed by David Carter.
648 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
649 replaced by the ${authresults } expansion.
651 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
653 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
654 allows proper process termination in container environments.
656 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
657 Previously the "final dot" had a newline after it; ensure it is CR,LF.
659 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
660 and "err_perm", deprecated since 4.83 when the RFC-defined words
661 "temperror" and "permerror" were introduced.
663 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
664 transport-filters or DKIM-signing. The restriction was lost in the
665 consolidation of verify-callout and delivery SMTP handling.
666 Extend the restriction to also cover ARC-signing.
668 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
669 in defer=pass mode supply a 450 to the initiator. Previously the message
672 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
673 tls_require_ciphers is used as before.
675 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
676 "pass_unscanned". Only tmpfails from the scanner are written to
677 the paniclog, as they may require admin intervention (permission
678 denied, license issues). Other scanner errors (like decompression
679 bombs) do not cause a paniclog entry.
681 JH/36 Fix reinitialisation of DKIM logging variable between messages.
682 Previously it was possible to log spurious information in receive log
685 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
686 triggered odd behaviour from Outlook Express clients.
688 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
691 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
692 since the IETF WG has not yet settled on that versus the original
693 "bare" representation.
695 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
696 Previously the millisecond value corrupted the output.
697 Fix also for syslog_pid=no and log_selector +pid, for which the pid
698 corrupted the output.
704 JH/01 Rework error string handling in TLS interface so that the caller in
705 more cases is responsible for logging. This permits library-sourced
706 string to be attached to addresses during delivery, and collapses
707 pairs of long lines into single ones.
709 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
710 during configuration. Wildcards are allowed and expanded.
712 JH/02 Rework error string handling in DKIM to pass more info back to callers.
713 This permits better logging.
715 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
716 do not close it down and have the child transport start it up again on
717 the passed-on TCP connection. Instead, proxy the child (and any
718 subsequent ones) for TLS via a unix-domain socket channel. Logging is
719 affected: the continued delivery log lines do not have any DNSSEC, TLS
720 Certificate or OCSP information. TLS cipher information is still logged.
722 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
723 identical IP addresses on different listening ports. Will also affect
726 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
727 add noisy ifdef guards to special-case this sillyness.
728 Patch from Bernd Kuhls.
730 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
731 than 255 are no longer allowed.
733 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
734 Disable the session-cache too, which might reduce our load. Since we
735 currrectly use a new context for every connection, both as server and
736 client, there is no benefit for these.
737 GnuTLS appears to not support tickets server-side by default (we don't
738 call gnutls_session_ticket_enable_server()) but client side is enabled
739 by default on recent versions (3.1.3 +) unless the PFS priority string
742 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
743 <https://reproducible-builds.org/specs/source-date-epoch/>.
745 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
746 the check for any unsuccessful recipients did not notice the limit, and
747 erroneously found still-pending ones.
749 JH/08 Pipeline CHUNKING command and data together, on kernels that support
750 MSG_MORE. Only in-clear (not on TLS connections).
752 JH/09 Avoid using a temporary file during transport using dkim. Unless a
753 transport-filter is involved we can buffer the headers in memory for
754 creating the signature, and read the spool data file once for the
755 signature and again for transmission.
757 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
758 4.77 as the kernel support then wasn't solid, having issues in 64bit
759 mode. Now, it's been long enough. Add support for FreeBSD also.
761 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
762 case where the routing stage had gathered several addresses to send to
763 a host before calling the transport for the first, we previously failed
764 to close down TLS in the old transport process before passing the TCP
765 connection to the new process. The new one sent a STARTTLS command
766 which naturally failed, giving a failed delivery and bloating the retry
767 database. Investigation and fix prototype from Wolfgang Breyha.
769 JH/12 Fix check on SMTP command input synchronisation. Previously there were
770 false-negatives in the check that the sender had not preempted a response
771 or prompt from Exim (running as a server), due to that code's lack of
772 awareness of the SMTP input buffering.
774 PP/04 Add commandline_checks_require_admin option.
775 Exim drops privileges sanely, various checks such as -be aren't a
776 security problem, as long as you trust local users with access to their
777 own account. When invoked by services which pass untrusted data to
778 Exim, this might be an issue. Set this option in main configuration
779 AND make fixes to the calling application, such as using `--` to stop
782 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
783 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
784 responses to those, into a single TLS record each way (this usually means
785 a single packet). As a side issue, smtp_enforce_sync now works on TLS
788 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
789 affects you only if you're dancing at the edge of the param size limits.
790 If you are, and this message makes sense to you, then: raise the
791 configured limit or use OpenSSL 1.1. Nothing we can do for older
794 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
795 cmdline element to get the documented default one. Previously it was
798 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
799 are unsuable for non-message-specific reasons - eg. connection timeout,
800 banner-time rejection.
802 JH/16 Fix logging of delivery remote port, when specified by router, under
805 PP/06 Repair manualroute's ability to take options in any order, even if one
806 is the name of a transport.
809 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
811 JH/17 Change the list-building routines interface to use the expanding-string
812 triplet model, for better allocation and copying behaviour.
814 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
815 Previously it was constructed the first time a possibly-matching string
816 was met in the configuration file input during startup; now it is done
819 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
820 compatible one, to avoid the (poorly documented) possibility of a config
821 file in the working directory redirecting the DB files, possibly correpting
822 some existing file. CVE-2017-10140 assigned for BDB.
824 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
825 cache-hot. Previously, although the result was properly cached, the
826 initial verify call returned a defer.
828 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
829 the main verify for receipient in uncached-mode.
831 JH/22 Retire historical build files to an "unsupported" subdir. These are
832 defined as "ones for which we have no current evidence of testing".
834 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
835 if present. Previously it was ignored.
837 JH/24 Start using specified-initialisers in C structure init coding. This is
838 a C99 feature (it's 2017, so now considered safe).
840 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
841 if was a fixed-sized field and bitmask ops via macros; it is now more
844 PP/07 GitHub PR 56: Apply MariaDB build fix.
845 Patch provided by Jaroslav Škarvada.
847 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
848 during Coverity cleanups [4.87 JH/47]
849 Diagnosis and fix provided by Michael Fischer v. Mollard.
851 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
852 the right size to place the terminating semicolon on its own folded
853 line, the header hash was calculated to an incorrect value thanks to
854 the (relaxed) space the fold became.
856 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
857 and confused the parent.
859 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
860 which could crash as a result. This could lead to undeliverable messages.
862 JH/28 Logging: "next input sent too soon" now shows where input was truncated
865 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
866 matters on fast-turnover and PID-randomising systems, which were getting
867 out-of-order delivery.
869 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
870 a possibly-overlapping copy. The symptom was that "Remote host closed
871 connection in response to HELO" was logged instead of the actual 4xx
874 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
875 Previously only that bufferd was discarded, resulting in SYMTP command
878 JH/32 DKIM: when a message has multiple signatures matching an identity given
879 in dkim_verify_signers, run the dkim acl once for each. Previously only
880 one run was done. Bug 2189.
882 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
883 "panic the current process" to "deliberately defer". The panic log is
884 still written with the problem list name; the mail and reject logs now
885 get a temp-reject line for the message that was being handled, saying
886 something like "domains check lookup or other defer". The SMTP 451
887 message is still "Temporary local problem".
889 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
890 A crafted sequence of BDAT commands could result in in-use memory beeing
891 freed. CVE-2017-16943.
893 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
894 from SMTP input. Previously it was always done; now only done for DATA
895 and not BDAT commands. CVE-2017-16944.
897 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
898 to the message (such as an overlong header line). Previously this was
899 not done and we did not exit BDAT mode. Followon from the previous item
900 though a different problem.
906 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
907 than -2003 did; needs libidn2 in addition to libidn.
909 JH/02 The path option on a pipe transport is now expanded before use.
911 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
912 Patch provided by "Björn", documentation fix added too.
914 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
915 missing a wire-to-host endian conversion.
917 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
918 close after a BDAT command line could be taken as a following command,
919 giving a synch failure. Fix by only checking for synch immediately
920 before acknowledging the chunk.
922 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
923 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
924 macro. Patches provided by Josh Soref.
926 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
927 Previously we did not; the RFC seems ambiguous and VRFY is not listed
928 by IANA as a service extension. However, John Klensin suggests that we
931 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
932 the dkim code may be unix-mode line endings rather than smtp wire-format
933 CRLF, so prepend a CR to any bare LF.
935 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
936 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
938 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
939 the first were themselves being wrongly included in the feed into dkim
940 processing; with most chunk sizes in use this resulted in an incorrect
941 body hash calculated value.
943 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
944 DKIM signature block, for verification. Although advised against by
945 standards it is specifically not ruled illegal.
947 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
949 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
950 missing a body hash (the bh= tag).
952 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
953 It seems that HAProxy sends the Proxy Protocol information in clear and
954 only then does a TLS startup, so do the same.
956 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
957 TCP connections (such as for Spamd) unless the daemon successfully set
958 Fast Open mode on its listening sockets. This fixes breakage seen on
959 too-old kernels or those not configured for Fast Open, at the cost of
960 requiring both directions being enabled for TFO, and TFO never being used
961 by non-daemon-related Exim processes.
963 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
964 endings, at least on the first header line. Try to canonify any that get
965 past that check, despite the cost.
967 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
968 now limited to an arbitrary five deep, while parsing addresses with the
969 strip_excess_angle_brackets option enabled.
971 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
972 instead leave the unprompted TLS handshake in socket buffer for the
973 TLS library to consume.
975 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
977 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
979 JH/16 Drop variables when they go out of scope. Memory management drops a whole
980 region in one operation, for speed, and this leaves assigned pointers
981 dangling. Add checks run only under the testsuite which checks all
982 variables at a store-reset and panics on a dangling pointer; add code
983 explicitly nulling out all the variables discovered. Fixes one known
984 bug: a transport crash, where a dangling pointer for $sending_ip_address
985 originally assigned in a verify callout, is re-used.
987 PP/06 Drop '.' from @INC in various Perl scripts.
989 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
991 PP/08 Reduce a number of compilation warnings under clang; building with
992 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
993 should be warning-free.
995 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
997 HS/01 Fix portability problems introduced by PP/08 for platforms where
998 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1000 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1001 chunk. This allows us to accept broken chunked messages. We need a more
1002 general solution here.
1004 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1005 already-broken messages in the queue.
1007 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1009 JH/19 Fix reference counting bug in routing-generated-address tracking.
1015 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1016 supports it and a size is available (ie. the sending peer gave us one).
1018 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1019 years of being deprecated). The replacements are the ACLs
1020 acl_smtp_mime and acl_not_smtp_mime.
1022 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1023 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1024 or even an in-clear connection were permitted. Now, if the host lookup
1025 was dnssec and dane was requested then the host is only used if the
1026 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1027 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1028 if one fails this test.
1029 This means that a poorly-configured remote DNS will make it incommunicado;
1030 but it protects against a DNS-interception attack on it.
1032 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1033 non-noisy when a race steals the message being considered.
1035 JH/05 If main configuration option tls_certificate is unset, generate a
1036 self-signed certificate for inbound TLS connections.
1038 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1039 in rewrites and routers.
1041 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1042 and logged a warning sing 4.83; now they are a configuration file error.
1044 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1045 (lacking @domain). Apply the same qualification processing as RCPT.
1047 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1049 JH/10 Support ${sha256:} applied to a string (as well as the previous
1052 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1053 a cutthrough deliver is pending, as we always want to make a connection.
1054 This also avoids re-routing the message when later placing the cutthrough
1055 connection after a verify cache hit.
1056 Do not update it with the verify result either.
1058 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1059 when routing results in more than one destination address.
1061 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1062 signing (which inhibits the cutthrough capability). Previously only
1063 the presence of an option was tested; now an expansion evaluating as
1064 empty is permissible (obviously it should depend only on data available
1065 when the cutthrough connection is made).
1067 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1068 the relevant preceding SMTP command did not note the pipelining mode.
1070 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1071 Previously they were not counted.
1073 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1074 as one having no matching records. Previously we deferred the message
1075 that needed the lookup.
1077 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1078 distinguished as "(=".
1080 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1081 for missing MX records. Previously it only worked for missing A records.
1083 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1085 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1086 after the data-go-ahead and data-ack. Patch from Jason Betts.
1088 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1089 even for a "none" policy. Patch from Tony Meyer.
1091 JH/22 Fix continued use of a connection for further deliveries. If a port was
1092 specified by a router, it must also match for the delivery to be
1095 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1096 When one of the recipients of a message was unsuitable for the connection
1097 (has no matching addresses), we lost track of needing to mark it
1098 deferred. As a result mail would be lost.
1100 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1102 JH/25 Decoding ACL controls is now done using a binary search; the source code
1103 takes up less space and should be simpler to maintain. Merge the ACL
1104 condition decode tables also, with similar effect.
1106 JH/26 Fix problem with one_time used on a redirect router which returned the
1107 parent address unchanged. A retry would see the parent address marked as
1108 delivered, so not attempt the (identical) child. As a result mail would
1111 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1112 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1113 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1116 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1117 Main config options check_{log,spool}_{inodes,space} are now
1118 100 inodes, 10MB unless set otherwise in the configuration.
1120 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1121 Previously it only applied to the main-section connection policy
1124 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1126 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1127 by me. Added RFC7919 DH primes as an alternative.
1129 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1130 are not in the system include path.
1132 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1133 GnuTLS, when a session startup failed (eg because the client disconnected)
1134 Exim did stdio operations after fclose. This was exposed by a recent
1135 change which nulled out the file handle after the fclose.
1137 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1138 signed directly by the cert-signing cert, rather than an intermediate
1139 OCSP-signing cert. This is the model used by LetsEncrypt.
1141 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1143 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1144 an incoming connection.
1146 HS/02 Bug 1802: Do not half-close the connection after sending a request
1149 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1150 fallback to "prime256v1".
1152 JH/34 SECURITY: Use proper copy of DATA command in error message.
1153 Could leak key material. Remotely exploitable. CVE-2016-9963.
1159 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1160 and 3.4.4 - once the server is enabled to respond to an OCSP request
1161 it does even when not requested, resulting in a stapling non-aware
1162 client dropping the TLS connection.
1164 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1165 support variable-length bit vectors. No functional change.
1167 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1168 The I= interface field on outgoing lines is now after the H= remote
1169 host field, same as incoming lines. There is a separate
1170 outgoing_interface log selector which allows you to disable the
1173 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1174 If not running log_selector +smtp_connection the mainlog would be held
1175 open indefinitely after a "too many connections" event, including to a
1176 deleted file after a log rotate. Leave the per net connection logging
1177 leaving it open for efficiency as that will be quickly detected by the
1178 check on the next write.
1180 HS/01 Bug 1671: Fix post transport crash.
1181 Processing the wait-<transport> messages could crash the delivery
1182 process if the message IDs didn't exist for some reason. When
1183 using 'split_spool_directory=yes' the construction of the spool
1184 file name failed already, exposing the same netto behaviour.
1186 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1187 mime_regex ACL conditions.
1189 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1190 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1191 to HELO, local diagnostic string.
1193 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1194 log line to debug. Even when configured with a tls authenticator many
1195 client connections are expected to not authenticate in this way, so
1196 an authenticate fail is not an error.
1198 HS/02 Add the Exim version string to the process info. This way exiwhat
1199 gives some more detail about the running daemon.
1201 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1202 matter for fast-change records such as DNSBLs.
1204 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1205 retry record, even if constant. There may be multiple transports with
1206 different interface settings and the retry behaviour needs to be kept
1209 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1211 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1213 JH/10 Bug 840: fix log_defer_output option of pipe transport
1215 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1216 hosts_randomize. This matters a lot when combined with mua_wrapper.
1218 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1219 ${quote_pgsql:<string>} operator.
1221 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1223 JH/14 Change header file name for freeradius-client. Relevant if compiling
1224 with Radius support; from the Gentoo tree and checked under Fedora.
1226 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1228 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1229 option transport_filter, meaning no filtering.
1231 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1233 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1234 defaults to "*" (all hosts). The variable is now available when not built
1235 with TLS, default unset, mainly to enable keeping the testsuite sane.
1236 If a server certificate is not supplied (via tls_certificate) an error is
1237 logged, and clients will find TLS connections fail on startup. Presumably
1238 they will retry in-clear.
1239 Packagers of Exim are strongly encouraged to create a server certificate
1240 at installation time.
1242 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1243 with the $config_file variable.
1245 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1246 in transport context, after the attempt, and per-recipient. The latter type
1247 is per host attempted. The event data is the error message, and the errno
1248 information encodes the lookup type (A vs. MX) used for the (first) host,
1249 and the trailing two digits of the smtp 4xx response.
1251 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1252 to write to mainlog (or rejectlog, paniclog) in the window between file
1253 creation and permissions/ownership being changed. Particularly affects
1254 installations where exicyclog is run as root, rather than exim user;
1255 result is that the running daemon panics and dies.
1257 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1259 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1260 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1261 "pri" and "weight". Note that the previous implicit priority given by the
1262 list order is no longer honoured.
1264 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1265 for DKIM processing.
1267 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1268 by defining SUPPORT_SOCKS.
1270 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1271 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1272 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1273 variables are renamed to proxy_{local,external}_{address,port}.
1275 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1276 for a build by defining SUPPORT_I18N
1278 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1279 of the query string, and make ${quote_redis:} do that quoting.
1281 JH/29 Move Events support from Experimental to mainline, enabled by default
1282 and removable for a build by defining DISABLE_EVENT.
1284 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1286 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1287 cached by the daemon.
1289 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1290 by defining LOOKUP_REDIS. The libhiredis library is required.
1292 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1293 keys are given for lookup.
1295 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1296 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1297 only supported when built with TLS support. The PolarSSL SHA routines
1298 are still used when the TLS library is too old for convenient support.
1300 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1301 openssl_options), for security. OpenSSL forces this from version 1.1.0
1302 server-side so match that on older versions.
1304 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1305 allocation for $value could be released as the expansion processing
1306 concluded, but leaving the global pointer active for it.
1308 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1309 and to use the domains and local_parts ACL conditions.
1311 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1312 incorrectly not doubled on cutthrough transmission, hence seen as a
1313 body-termination at the receiving system - resulting in truncated mails.
1314 Commonly the sender saw a TCP-level error, and retransmitted the message
1315 via the normal store-and-forward channel. This could result in duplicates
1316 received - but deduplicating mailstores were liable to retain only the
1317 initial truncated version.
1319 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1321 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1323 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1324 we're in there, support oversigning also; bug 1309.
1326 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1328 HS/04 Add support for keep_environment and add_environment options.
1330 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1331 either intentional arithmetic overflow during PRNG, or testing config-
1334 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1335 delivery resulted in actual delivery. Cancel cutthrough before DATA
1338 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1339 rejects a recipient: pass the reject to the originator.
1341 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1342 Many were false-positives and ignorable, but it's worth fixing the
1345 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1346 for the new environment-manipulation done at startup. Move the routines
1347 from being local to tls.c to being global via the os.c file.
1349 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1350 an extract embedded as result-arg for a map, the first arg for extract
1351 is unavailable so we cannot tell if this is a numbered or keyed
1352 extraction. Accept either.
1358 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1361 JH/02 The smtp transport option "multi_domain" is now expanded.
1363 JH/03 The smtp transport now requests PRDR by default, if the server offers
1366 JH/04 Certificate name checking on server certificates, when exim is a client,
1367 is now done by default. The transport option tls_verify_cert_hostnames
1368 can be used to disable this per-host. The build option
1369 EXPERIMENTAL_CERTNAMES is withdrawn.
1371 JH/05 The value of the tls_verify_certificates smtp transport and main options
1372 default to the word "system" to access the system default CA bundle.
1373 For GnuTLS, only version 3.0.20 or later.
1375 JH/06 Verification of the server certificate for a TLS connection is now tried
1376 (but not required) by default. The verification status is now logged by
1377 default, for both outbound TLS and client-certificate supplying inbound
1380 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1383 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1384 Status Notification (bounce) messages are now MIME format per RFC 3464.
1385 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1386 under the control of the dsn_advertise_hosts option, and routers may
1387 have a dsn_lasthop option.
1389 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1390 default, modifiable by a malware= option. The list separator for
1391 the options can now be changed in the usual way. Bug 68.
1393 JH/10 The smtp_receive_timeout main option is now expanded before use.
1395 JH/11 The incoming_interface log option now also enables logging of the
1396 local interface on delivery outgoing connections.
1398 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1399 if the interface and destination host and port all match.
1401 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1404 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1405 Patch from Andrew Lewis.
1407 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1408 now supports optional time-restrictions, weighting, and priority
1409 modifiers per server. Patch originally by <rommer@active.by>.
1411 JH/16 The spamd_address main option now supports a mixed list of local
1412 and remote servers. Remote servers can be IPv6 addresses, and
1413 specify a port-range.
1415 JH/17 Bug 68: The spamd_address main option now supports an optional
1416 timeout value per server.
1418 JH/18 Bug 1581: Router and transport options headers_add/remove can
1419 now have the list separator specified.
1421 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1424 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1427 JH/21 Support for the A6 type of dns record is withdrawn.
1429 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1430 rather than the verbs used.
1432 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1433 from 255 to 1024 chars.
1435 JH/24 Verification callouts now attempt to use TLS by default.
1437 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1438 are generic router options now. The defaults didn't change.
1440 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1441 Original patch from Alexander Shikoff, worked over by JH.
1443 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1444 files if ZCAT_COMMAND is not executable.
1446 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1448 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1450 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1451 Normally benign, it bites when the pair was led to by a CNAME;
1452 modern usage is to not canonicalize the domain to a CNAME target
1453 (and we were inconsistent anyway for A-only vs AAAA+A).
1455 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1457 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1458 when evaluating $sender_host_dnssec.
1460 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1461 $sender_helo_dnssec variable.
1463 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1465 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1467 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1469 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1470 documented as working, but never had. Support all but $spam_report.
1472 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1473 added for tls authenticator.
1475 HS/03 Add perl_taintmode main config option
1481 TL/01 When running the test suite, the README says that variables such as
1482 no_msglog_check are global and can be placed anywhere in a specific
1483 test's script, however it was observed that placement needed to be near
1484 the beginning for it to behave that way. Changed the runtest perl
1485 script to read through the entire script once to detect and set these
1486 variables, reset to the beginning of the script, and then run through
1487 the script parsing/test process like normal.
1489 TL/02 The BSD's have an arc4random API. One of the functions to induce
1490 adding randomness was arc4random_stir(), but it has been removed in
1491 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1492 function when detected.
1494 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1495 cause callback expansion.
1497 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1498 syntax errors in an expansion can be treated as a string instead of
1499 logging or causing an error, due to the internal use of bool_lax
1500 instead of bool when processing it.
1502 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1503 server certificates when making smtp deliveries.
1505 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1507 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1509 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1511 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1512 Merged patch from Sebastian Wiedenroth.
1514 JH/05 Fix results-pipe from transport process. Several recipients, combined
1515 with certificate use, exposed issues where response data items split
1516 over buffer boundaries were not parsed properly. This eventually
1517 resulted in duplicates being sent. This issue only became common enough
1518 to notice due to the introduction of connection certificate information,
1519 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1521 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1522 size buffer was used, resulting in syntax errors when an expansion
1525 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1526 version 3.3.6 or later.
1528 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1529 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1530 both become "event_action", the variables become $event_name, $event_data
1531 and $event_defer_errno. There is a new variable $verify_mode, usable in
1532 routers, transports and related events. The tls:cert event is now also
1533 raised for inbound connections, if the main configuration event_action
1536 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1537 early OCSP support, but no stapling (appears to be less than 1.0.0).
1539 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1540 server certificate names available under the smtp transport option
1541 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1544 JH/10 Time-related extraction expansions from certificates now use the main
1545 option "timezone" setting for output formatting, and are consistent
1546 between OpenSSL and GnuTLS compilations. Bug 1541.
1548 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1549 encoded parameter in the incoming message. Bug 1558.
1551 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1552 include certificate info, eximon was claiming there were spoolfile
1555 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1557 JH/14 Log delivery-related information more consistently, using the sequence
1558 "H=<name> [<ip>]" wherever possible.
1560 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1561 are problematic for Debian distribution, omit them from the release
1564 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1566 JH/16 Fix string representation of time values on 64bit time_t architectures.
1569 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1570 output list separator was used.
1575 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1576 checkers that were complaining about end of non-void function with no
1579 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1580 This was a regression introduced in 4.83 by another bugfix.
1582 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1584 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1585 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1591 TF/01 Correctly close the server side of TLS when forking for delivery.
1593 When a message was received over SMTP with TLS, Exim failed to clear up
1594 the incoming connection properly after forking off the child process to
1595 deliver the message. In some situations the subsequent outgoing
1596 delivery connection happened to have the same fd number as the incoming
1597 connection previously had. Exim would try to use TLS and fail, logging
1598 a "Bad file descriptor" error.
1600 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1601 utilities have not been installed.
1603 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1604 temporary space as the ACL may create new global variables.
1606 TL/01 LDAP support uses per connection or global context settings, depending
1607 upon the detected version of the libraries at build time.
1609 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1610 to extract and use the src ip:port in logging and expansions as if it
1611 were a direct connection from the outside internet. PPv2 support was
1612 updated based on HAProxy spec change in May 2014.
1614 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1616 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1617 Properly escape header and check for NULL return.
1619 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1622 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1624 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1625 characters in header names, implemented as a verify condition.
1626 Contributed by Michael Fischer v. Mollard.
1628 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1629 results permerror and temperror. Previous values are deprecated but
1630 still accepted. In a future release, err_perm and err_temp will be
1631 completely removed, which will be a backward incompatibility if the
1632 ACL tests for either of these two old results. Patch contributed by
1633 user bes-internal on the mailing list.
1635 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1637 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1638 selectors, in both main and reject logs.
1640 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1643 JH/07 Add malware type "sock" for talking to simple daemon.
1645 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1647 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1648 routers/transports under cutthrough routing.
1650 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1651 numbers. Touch up "bool" conditional to keep the same definition.
1653 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1655 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1657 JH/12 Expand items in router/transport headers_add or headers_remove lists
1658 individually rather than the list as a whole. Bug 1452.
1660 Required for reasonable handling of multiple headers_ options when
1661 they may be empty; requires that headers_remove items with embedded
1662 colons must have them doubled (or the list-separator changed).
1664 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1665 view the policy declared in the DMARC record. Currently, $dmarc_status
1666 is a combined value of both the record presence and the result of the
1669 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1671 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1672 dnslookup router and the smtp transport (applying to the forward
1675 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1676 of ldap servers used for a specific lookup. Patch provided by Heiko
1679 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1680 New variable $lookup_dnssec_authenticated for observability.
1682 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1683 Patch submitted by Lars Timman.
1685 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1687 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1688 Requires trusted mode and valid format message id, aborts otherwise.
1689 Patch contributed by Heiko Schlichting.
1691 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1692 certextract with support for various fields. Bug 1358.
1694 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1695 is requested by default, modifiable by smtp transport option
1698 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1699 operate on certificate variables to give certificate fingerprints
1700 Also new ${sha256:cert_variable}.
1702 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1704 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1705 Christian Aistleitner.
1707 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1709 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1710 file. Patch from Wolfgang Breyha.
1712 JH/25 Expand the coverage of the delivery $host and $host_address to
1713 client authenticators run in verify callout. Bug 1476.
1715 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1716 align with daemon_smtp_ports. Bug 72.
1718 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1719 support and error reporting did not work properly.
1721 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1722 and is readable. Patch from Andrew Colin Kissa.
1724 TL/14 Enhance documentation of ${run expansion and how it parses the
1725 commandline after expansion, particularly in the case when an
1726 unquoted variable expansion results in an empty value.
1728 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1730 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1731 by renaming away from it. Addresses GH issue 15, reported by
1734 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1736 TL/15 SECURITY: prevent double expansion in math comparison functions
1737 (can expand unsanitized data). Not remotely exploitable.
1744 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1746 PP/02 Make -n do something, by making it not do something.
1747 When combined with -bP, the name of an option is not output.
1749 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1752 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1753 $sender_host_name and config options to manage this, and basic check
1756 PP/05 DSCP support for outbound connections and control modifier for inbound.
1758 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1759 (Only plugin which currently uses this is kerberos4, which nobody should
1760 be using, but we should make it available and other future plugins might
1761 conceivably use it, even though it would break NAT; stuff *should* be
1762 using channel bindings instead).
1764 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1765 name; added for Sendmail compatibility; requires admin caller.
1766 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1767 just ignore it); requires trusted caller.
1768 Also parse but ignore: -Ac -Am -X<logfile>
1771 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1773 TL/02 Add +smtp_confirmation as a default logging option.
1775 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1776 Patch by Magnus Holmgren from 2007-02-20.
1778 TL/04 Bugzilla 1281 - Spec typo.
1779 Bugzilla 1283 - Spec typo.
1780 Bugzilla 1290 - Spec grammar fixes.
1782 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1784 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1786 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1787 mailing list by by Dmitry Isaikin.
1789 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1791 JH/02 Support "G" suffix to numbers in ${if comparisons.
1793 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1795 NM/01 Bugzilla 1197 - Spec typo
1796 Bugzilla 1196 - Spec examples corrections
1798 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1800 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1801 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1804 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1805 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1807 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1808 "acl {{name}{arg}...}", and optional args on acl condition
1811 JH/05 Permit multiple router/transport headers_add/remove lines.
1813 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1815 JH/07 Avoid using a waiting database for a single-message-only transport.
1816 Performance patch from Paul Fisher. Bugzilla 1262.
1818 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1821 JH/09 Add $headers_added variable, with content from use of ACL modifier
1822 add_header (but not yet added to the message). Bugzilla 199.
1824 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1825 Pulled from Bugzilla 817 by Wolfgang Breyha.
1827 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1829 (nb: this is the same fix as in Exim 4.80.1)
1831 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1834 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1837 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1839 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1840 advertise SMTP AUTH mechanism to us, instead of a generic
1841 protocol violation error. Also, make Exim more robust to bad
1842 data from the Dovecot auth socket.
1844 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1846 When a queue runner is handling a message, Exim first routes the
1847 recipient addresses, during which it prunes them based on the retry
1848 hints database. After that it attempts to deliver the message to
1849 any remaining recipients. It then updates the hints database using
1852 So if a recipient address works intermittently, it can get repeatedly
1853 deferred at routing time. The retry hints record remains fresh so the
1854 address never reaches the final cutoff time.
1856 This is a fairly common occurrence when a user is bumping up against
1857 their storage quota. Exim had some logic in its local delivery code
1858 to deal with this. However it did not apply to per-recipient defers
1859 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1861 This change adds a proper retry rule check during routing so that the
1862 final cutoff time is checked against the message's age. We only do
1863 this check if there is an address retry record and there is not a
1864 domain retry record; this implies that previous attempts to handle
1865 the address had the retry_use_local_parts option turned on. We use
1866 this as an approximation for the destination being like a local
1867 delivery, as in LMTP.
1869 I suspect this new check makes the old local delivery cutoff check
1870 redundant, but I have not verified this so I left the code in place.
1872 TF/02 Correct gecos expansion when From: is a prefix of the username.
1874 Test 0254 submits a message to Exim with the header
1878 When I ran the test suite under the user fanf2, Exim expanded
1879 the header to contain my full name, whereas it should have added
1880 a Resent-Sender: header. It erroneously treats any prefix of the
1881 username as equal to the username.
1883 This change corrects that bug.
1885 GF/01 DCC debug and logging tidyup
1886 Error conditions log to paniclog rather than rejectlog.
1887 Debug lines prefixed by "DCC: " to remove any ambiguity.
1889 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1891 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1892 Bug spotted by Jeremy Harris; was flawed since initial commit.
1893 Would have resulted in OCSP responses post-SNI triggering an Exim
1894 NULL dereference and crash.
1896 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1898 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1899 Bug detection, analysis and fix by Samuel Thibault.
1900 Bugzilla 1331, Debian bug #698092.
1902 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1904 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1905 Server implementation by Todd Lyons, client by JH.
1906 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1907 config variable "prdr_enable" controls whether the server
1908 advertises the facility. If the client requests PRDR a new
1909 acl_data_smtp_prdr ACL is called once for each recipient, after
1910 the body content is received and before the acl_smtp_data ACL.
1911 The client is controlled by both of: a hosts_try_prdr option
1912 on the smtp transport, and the server advertisement.
1913 Default client logging of deliveries and rejections involving
1914 PRDR are flagged with the string "PRDR".
1916 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1917 fclose(). Diagnosis by Todd Lyons.
1919 PP/17 Update configure.default to handle IPv6 localhost better.
1920 Patch by Alain Williams (plus minor tweaks).
1923 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1924 This is now consistent with GnuTLS, and is now documented: the
1925 previous undocumented portable approach to treating the option as
1926 unset was to force an expansion failure. That still works, and
1927 an empty string is now equivalent.
1929 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1930 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1931 not performing validation itself.
1933 PP/20 Added force_command boolean option to pipe transport.
1934 Patch from Nick Koston, of cPanel Inc.
1936 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1939 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1941 PP/21 Fix eximon continuous updating with timestamped log-files.
1942 Broken in a format-string cleanup in 4.80, missed when I repaired the
1943 other false fix of the same issue.
1944 Report and fix from Heiko Schlichting.
1947 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1948 Report from Prashanth Katuri.
1950 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1951 It's SecureTransport, so affects any MacOS clients which use the
1952 system-integrated TLS libraries, including email clients.
1954 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1955 using a MIME ACL for non-SMTP local injection.
1956 Report and assistance in diagnosis by Warren Baker.
1958 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1960 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1962 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1963 last id that failed so it may be referenced in subsequent ACL's.
1965 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1968 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1969 ldap library initialization, allowing self-signed CA's to be
1970 used. Also properly sets require_cert option later in code by
1971 using NULL (global ldap config) instead of ldap handle (per
1972 session). Bug diagnosis and testing by alxgomz.
1974 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1975 the src/util/ subdirectory.
1977 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1978 renamed to Transport Post Delivery Action by Jeremy Harris, as
1981 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1982 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1983 redis_servers = needs to be configured which will be used by the redis
1984 lookup. Patch from Warren Baker, of The Packet Hub.
1986 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1988 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1989 hostname or reverse DNS when processing a host list. Used suggestions
1990 from multiple comments on this bug.
1992 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1994 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1995 Missed a few lines, added it to make the runtest require no keyboard
1998 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1999 contains upper case chars. Make router use caseful_local_part.
2001 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2002 support when GnuTLS has been built with p11-kit.
2008 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2010 This, or similar/improved, will also be change PP/11 of 4.82.
2016 PP/01 Handle short writes when writing local log-files.
2017 In practice, only affects FreeBSD (8 onwards).
2018 Bugzilla 1053, with thanks to Dmitry Isaikin.
2020 NM/01 Bugzilla 949 - Documentation tweak
2022 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2025 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2027 PP/02 Implemented gsasl authenticator.
2029 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2031 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2032 `pkg-config foo` for cflags/libs.
2034 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2035 with rest of GSASL and with heimdal_gssapi.
2037 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2038 `pkg-config foo` for cflags/libs for the TLS implementation.
2040 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2041 properties get this fed in as external SSF. A number of robustness
2042 and debugging improvements to the cyrus_sasl authenticator.
2044 PP/08 cyrus_sasl server now expands the server_realm option.
2046 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2047 Patch by Jeremy Harris.
2049 PP/10 Added dbmjz lookup type.
2051 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2053 PP/12 MAIL args handles TAB as well as SP, for better interop with
2054 non-compliant senders.
2055 Analysis and variant patch by Todd Lyons.
2057 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2058 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2059 Patch from Dirk Mueller <dmueller@suse.com>
2061 PP/13 tls_peerdn now print-escaped for spool files.
2062 Observed some $tls_peerdn in wild which contained \n, which resulted
2063 in spool file corruption.
2065 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2066 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2067 or write after TLS renegotiation, which otherwise led to messages
2070 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2071 as a tracking header (ie: a signed header comes before the signature).
2072 Patch from Wolfgang Breyha.
2074 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2075 comma-sep list; embedded commas doubled.
2077 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2079 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2081 Report and patch from Dmitry Banschikov.
2083 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2084 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2085 failures. We appear to now support TLS1.1+ with Exim.
2087 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2088 lets Exim select keys and certificates based upon TLS SNI from client.
2089 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2090 before an outbound SMTP session. New log_selector, +tls_sni.
2092 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2093 NULL dereference. Report and patch from Alun Jones.
2095 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2096 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2097 Not seeing resolver debug output on NetBSD, but suspect this is a
2098 resolver implementation change.
2100 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2101 Left warnings. Added "eximon gdb" invocation mode.
2103 PP/21 Defaulting "accept_8bitmime" to true, not false.
2105 PP/22 Added -bw for inetd wait mode support.
2107 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2108 locate the relevant includes and libraries. Made this the default.
2110 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2111 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2113 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2114 This may cause build issues on older platforms.
2116 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2117 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2118 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2119 Added SNI support via GnuTLS too.
2120 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2122 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2124 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2125 Applied second patch from Janne, implementing suggestion to default
2126 multiple-strings-in-record handling to match SPF spec.
2128 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2130 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2131 read-only, out of scope).
2132 Patch from Wolfgang Breyha, report from Stuart Northfield.
2134 PP/29 Fix three issues highlighted by clang analyser static analysis.
2135 Only crash-plausible issue would require the Cambridge-specific
2136 iplookup router and a misconfiguration.
2137 Report from Marcin Mirosław.
2139 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2141 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2142 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2143 As part of this, removing so much warning spew let me fix some minor
2144 real issues in debug logging.
2146 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2147 assignment on my part. Fixed.
2149 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2150 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2151 Janne Snabb (who went above and beyond: thank you).
2153 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2154 string otherwise requires a connection and a bunch more work and it's
2155 relatively easy to get wrong. Should also expose TLS library linkage
2158 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2159 64-bit ${eval} (JH/03).
2161 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2162 GNU libc to support some of the 64-bit stuff, should not lead to
2163 conflicts. Defined before os.h is pulled in, so if a given platform
2164 needs to override this, it can.
2166 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2167 protection layer was required, which is not implemented.
2168 Bugzilla 1254, patch from Wolfgang Breyha.
2170 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2171 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2172 tls_dhparam take prime identifiers. Also unbreak combination of
2173 OpenSSL+DH_params+TLSSNI.
2175 PP/39 Disable SSLv2 by default in OpenSSL support.
2181 PP/01 Solaris build fix for Oracle's LDAP libraries.
2182 Bugzilla 1109, patch from Stephen Usher.
2184 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2186 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2189 TF/02 Fix a couple more cases where we did not log the error message
2190 when unlink() failed. See also change 4.74-TF/03.
2192 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2193 lock up or crash if it happened to be inside a call to libc when it
2194 got a SIGUSR1 from exiwhat.
2196 The SIGUSR1 handler appends the current process status to the process
2197 log which is later printed by exiwhat. It used to use the general
2198 purpose logging code to do this, but several functions it calls are
2199 not safe for signals.
2201 The new output code in the SIGUSR1 handler is specific to the process
2202 log, and simple enough that it's easy to inspect for signal safety.
2203 Removing some special cases also simplifies the general logging code.
2204 Removing the spurious timestamps from the process log simplifies
2207 TF/04 Improved ratelimit ACL condition.
2209 The /noupdate option has been deprecated in favour of /readonly which
2210 has clearer semantics. The /leaky, /strict, and /readonly update modes
2211 are mutually exclusive. The update mode is no longer included in the
2212 database key; it just determines when the database is updated. (This
2213 means that when you upgrade Exim will forget old rate measurements.)
2215 Exim now checks that the per_* options are used with an update mode that
2216 makes sense for the current ACL. For example, when Exim is processing a
2217 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2218 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2219 must specify per_mail/readonly. If you omit the update mode it defaults to
2220 /leaky where that makes sense (as before) or /readonly where required.
2222 The /noupdate option is now undocumented but still supported for
2223 backwards compatibility. It is equivalent to /readonly except that in
2224 ACLs where /readonly is required you may specify /leaky/noupdate or
2225 /strict/noupdate which are treated the same as /readonly.
2227 A useful new feature is the /count= option. This is a generalization
2228 of the per_byte option, so that you can measure the throughput of other
2229 aggregate values. For example, the per_byte option is now equivalent
2230 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2232 The per_rcpt option has been generalized using the /count= mechanism
2233 (though it's more complicated than the per_byte equivalence). When it is
2234 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2235 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2236 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2237 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2238 Note that using per_rcpt with a non-readonly update mode in more than
2239 one ACL will cause the recipients to be double-counted. (The per_mail
2240 and per_byte options don't have this problem.)
2242 The handling of very low rates has changed slightly. If the computed rate
2243 is less than the event's count (usually one) then this event is the first
2244 after a long gap. In this case the rate is set to the same as this event's
2245 count, so that the first message of a spam run is counted properly.
2247 The major new feature is a mechanism for counting the rate of unique
2248 events. The new per_addr option counts the number of different
2249 recipients that someone has sent messages to in the last time period. It
2250 behaves like per_rcpt if all the recipient addresses are different, but
2251 duplicate recipient addresses do not increase the measured rate. Like
2252 the /count= option this is a general mechanism, so the per_addr option
2253 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2254 example, measure the rate that a client uses different sender addresses
2255 with the options per_mail/unique=$sender_address. There are further
2256 details in the main documentation.
2258 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2260 TF/06 Removed a few PCRE remnants.
2262 TF/07 Automatically extract Exim's version number from tags in the git
2263 repository when doing development or release builds.
2265 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2266 Bugzilla 879. Patch from Paul Fisher.
2268 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2269 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2272 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2274 PP/05 Variable $av_failed, true if the AV scanner deferred.
2275 Bugzilla 1078. Patch from John Horne.
2277 PP/06 Stop make process more reliably on build failure.
2278 Bugzilla 1087. Patch from Heiko Schlittermann.
2280 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2281 Bugzilla 1089. Patch from Heiko Schlittermann.
2283 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2284 Bugzilla 1131. Patch from Holger Weiß.
2286 PP/09 Handle IPv6 addresses with SPF.
2287 Bugzilla 860. Patch from Wolfgang Breyha.
2289 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2291 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2294 PP/11 match_* no longer expand right-hand-side by default.
2295 New compile-time build option, EXPAND_LISTMATCH_RHS.
2296 New expansion conditions, "inlist", "inlisti".
2298 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2300 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2302 PP/14 fix log_write() format string regression from TF/03.
2303 Bugzilla 1152. Patch from Dmitry Isaikin.
2309 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2311 PP/02 Harmonised TLS library version reporting; only show if debugging.
2312 Layout now matches that introduced for other libraries in 4.74 PP/03.
2314 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2316 PP/04 New "dns_use_edns0" global option.
2318 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2321 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2322 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2324 TK/01 Updated PolarSSL code to 0.14.2.
2325 Bugzilla 1097. Patch from Andreas Metzler.
2327 PP/07 Catch divide-by-zero in ${eval:...}.
2328 Fixes bugzilla 1102.
2330 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2333 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2334 format-string attack -- SECURITY: remote arbitrary code execution.
2336 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2337 time unintentionally subject to list matching rules, letting the header
2338 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2339 arbitrary string expansion). This allowed for information disclosure.
2341 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2342 INT_MIN/-1 -- value coerced to INT_MAX.
2348 NM/01 Workaround for PCRE version dependency in version reporting
2351 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2352 This fixes portability to compilers other than gcc, notably
2353 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2355 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2356 makefiles for portability to HP-UX and POSIX correctness.
2358 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2359 Also via indented variable definition in the Makefile.
2360 (Debugging by Oliver Heesakkers).
2362 PP/02 Restore caching of spamd results with expanded spamd_address.
2363 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2365 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2366 Improves build reliability. Fix from: Frank Elsner
2368 NM/02 Fix wide character breakage in the rfc2047 coding
2369 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2371 NM/03 Allow underscore in dnslist lookups
2372 Fixes bug 1026. Patch from Graeme Fowler
2374 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2375 Code patches from Adam Ciarcinski of NetBSD.
2377 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2380 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2381 is logged, to avoid truncation. Patch from John Horne.
2383 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2384 Patch from Jakob Hirsch.
2386 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2387 SQL string expansion failure details.
2388 Patch from Andrey Oktyabrski.
2390 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2391 Patch from Simon Arlott.
2393 PP/09 New lookups functionality failed to compile on old gcc which rejects
2394 extern declarations in function scope.
2395 Patch from Oliver Fleischmann
2397 PP/10 Use sig_atomic_t for flags set from signal handlers.
2398 Check getgroups() return and improve debugging.
2399 Fixed developed for diagnosis in bug 927 (which turned out to be
2402 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2403 Patch from Mark Zealey.
2405 PP/12 Bugzilla 1056: Improved spamd server selection.
2406 Patch from Mark Zealey.
2408 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2409 Based on patch from Heiko Schlittermann.
2411 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2412 Patch from Uwe Doering, sign-off by Michael Haardt.
2414 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2415 variable declaration deep within a block. Bug and patch from
2418 PP/15 lookups-Makefile IRIX compatibility coercion.
2420 PP/16 Make DISABLE_DKIM build knob functional.
2422 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2423 Patch by Simon Arlott
2425 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2426 variable argument macros. Our copy now differs from upstream.
2432 TF/01 Failure to get a lock on a hints database can have serious
2433 consequences so log it to the panic log.
2435 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2436 controlled using the smtp_confirmation log selector.
2438 TF/03 Include the error message when we fail to unlink a spool file.
2440 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2441 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2442 for maintaining out-of-tree patches for some time.
2444 PP/01 Bugzilla 139: Documentation and portability issues.
2445 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2446 Handle per-OS dynamic-module compilation flags.
2448 PP/02 Let /dev/null have normal permissions.
2449 The 4.73 fixes were a little too stringent and complained about the
2450 permissions on /dev/null. Exempt it from some checks.
2451 Reported by Andreas M. Kirchwitz.
2453 PP/03 Report version information for many libraries, including
2454 Exim version information for dynamically loaded libraries. Created
2455 version.h, now support a version extension string for distributors
2456 who patch heavily. Dynamic module ABI change.
2458 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2459 privilege escalation vulnerability whereby the Exim run-time user
2460 can cause root to append content of the attacker's choosing to
2463 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2466 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2467 If dropping privileges for untrusted macros, we disabled normal logging
2468 on the basis that it would fail; for the Exim run-time user, this is not
2469 the case, and it resulted in successful deliveries going unlogged.
2470 Fixed. Reported by Andreas Metzler.
2476 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2477 only prepend for the Resent-* case. Fixes regression introduced in
2478 Exim 4.70 by NM/22 for Bugzilla 607.
2480 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2481 increasing numbers of administrators be bitten by this.
2483 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2485 PP/03 Bugzilla 994: added openssl_options main configuration option.
2487 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2489 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2491 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2493 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2494 without a peer certificate, leading to a segfault because of an
2495 assumption that peers always have certificates. Be a little more
2496 paranoid. Problem reported by Martin Tscholak.
2498 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2499 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2500 NB: ClamAV planning to remove STREAM in "middle of 2010".
2501 CL also introduces -bmalware, various -d+acl logging additions and
2502 more caution in buffer sizes.
2504 PP/09 Implemented reverse_ip expansion operator.
2506 PP/10 Bugzilla 937: provide a "debug" ACL control.
2508 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2510 PP/12 Bugzilla 973: Implement --version.
2512 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2514 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2516 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2518 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2519 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2520 ignore trailing whitespace.
2522 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2524 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2527 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2528 Notification from Dr Andrew Aitchison.
2530 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2531 ExtendedDetectionInfo response format.
2532 Notification from John Horne.
2534 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2537 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2538 XSL and documented dependency on system catalogs, with examples of how
2541 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2544 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2545 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2546 configuration file which is writeable by the Exim user or group.
2548 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2549 of configuration files to cover files specified with the -C option if
2550 they are going to be used with root privileges, not just the default
2553 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2554 option (effectively making it always true).
2556 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2557 files to be used while preserving root privileges.
2559 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2560 that rogue child processes cannot use them.
2562 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2563 run-time user, instead of root.
2565 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2566 Exim run-time user without dropping privileges.
2568 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2569 result string, instead of calling string_vformat() twice with the same
2572 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2573 for other users. Others should always drop root privileges if they use
2574 -C on the command line, even for a whitelisted configure file.
2576 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2578 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2584 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2585 $data_path, and $header_path variables; fixed documentation bugs and
2588 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2589 exipick to access non-standard spools, including the "frozen" queue
2592 NM/01 Bugzilla 965: Support mysql stored procedures.
2593 Patch from Alain Williams
2595 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2597 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2598 Patch from Andreas Metzler
2600 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2601 Patch from Kirill Miazine
2603 NM/05 Bugzilla 671: Added umask to procmail example.
2605 JJ/03 installed exipick 20100323.0, fixing doc bug
2607 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2608 directory. Notification and patch from Dan Rosenberg.
2610 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2612 TK/02 Improve log output when DKIM signing operation fails.
2614 MH/01 Treat the transport option dkim_domain as a colon separated
2615 list, not as a single string, and sign the message with each element,
2616 omitting multiple occurences of the same signer.
2618 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2619 Bugzilla 985, 986. Patch by Simon Arlott
2621 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2622 Patch by Simon Arlott
2624 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2625 MBX locking. Notification from Dan Rosenberg.
2631 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2633 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2635 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2637 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2639 NM/04 Bugzilla 915: Flag broken perl installation during build.
2645 TK/01 Added patch by Johannes Berg that expands the main option
2646 "spamd_address" if it starts with a dollar sign.
2648 TK/02 Write list of recipients to X-Envelope-Sender header when building
2649 the mbox-format spool file for content scanning (suggested by Jakob
2652 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2653 (http://www.dcc-servers.net/) support via dccifd. Activated by
2654 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2656 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2657 by Mark Daniel Reidel <mr@df.eu>.
2659 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2660 When building exim an external PCRE library is now needed -
2661 PCRE is a system library on the majority of modern systems.
2662 See entry on PCRE_LIBS in EDITME file.
2664 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2665 conversation. Added nologin parameter to request.
2666 Patch contributed by Kirill Miazine.
2668 TF/01 Do not log submission mode rewrites if they do not change the address.
2670 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2672 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2673 log files in place. Contributed by Roberto Lima.
2675 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2677 TF/03 Bugzilla 615: When checking the local_parts router precondition
2678 after a local_part_suffix or local_part_prefix option, Exim now
2679 does not use the address's named list lookup cache, since this
2680 contains cached lookups for the whole local part.
2682 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2683 Robert Millan. Documentation is in experimental-spec.txt.
2685 TF/04 Bugzilla 668: Fix parallel build (make -j).
2687 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2689 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2690 Patch provided by Jan Srzednicki.
2692 TF/05 Leading white space used to be stripped from $spam_report which
2693 wrecked the formatting. Now it is preserved.
2695 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2696 that they are available at delivery time.
2698 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2700 TF/08 TLS error reporting now respects the incoming_interface and
2701 incoming_port log selectors.
2703 TF/09 Produce a more useful error message if an SMTP transport's hosts
2704 setting expands to an empty string.
2706 NM/06 Bugzilla 744: EXPN did not work under TLS.
2707 Patch provided by Phil Pennock.
2709 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2710 Patch provided by Richard Godbee.
2712 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2713 acl_smtp_notquit, added index entry.
2715 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2716 Patch provided by Eugene Bujak.
2718 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2719 accept(). Patch provided by Maxim Dounin.
2721 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2722 Patch provided by Phil Pennock.
2724 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2726 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2727 Patch provided by Brad "anomie" Jorsch.
2729 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2730 Patch provided by Dean Brooks.
2732 TK/05 Add native DKIM support (does not depend on external libraries).
2734 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2735 Patch provided by Graeme Fowler.
2737 NM/16 Bugzilla 851: Documentation example syntax fix.
2739 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2741 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2744 NM/19 Bugzilla 745: TLS version reporting.
2745 Patch provided by Phil Pennock.
2747 NM/20 Bugzilla 167: bool: condition support.
2748 Patch provided by Phil Pennock.
2750 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2751 clients. Patch provided by Phil Pennock.
2753 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2754 Patch provided by Brad "anomie" Jorsch.
2756 NM/23 Bugzilla 687: Fix misparses in eximstats.
2757 Patch provided by Heiko Schlittermann.
2759 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2760 Patch provided by Heiko Schlittermann.
2762 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2763 plus update to original patch.
2765 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2767 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2768 Patch provided by David Brownlee.
2770 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2772 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2774 NM/30 Bugzilla 888: TLS documentation bugfixes.
2776 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2778 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2779 Unlike the original bugzilla I have changed all shell scripts in src tree.
2781 NM/33 Bugzilla 898: Transport filter timeout fix.
2782 Patch by Todd Rinaldo.
2784 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2785 Patch by Serge Demonchaux.
2787 NM/35 Bugzilla 39: Base64 decode bug fixes.
2788 Patch by Jakob Hirsch.
2790 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2792 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2794 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2796 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2802 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2803 ALT-N's libdkim that I have put here:
2804 http://duncanthrax.net/exim-experimental/
2806 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2807 were called 'true' and it seems that C99 defines that as a reserved
2808 keyword to be used with 'bool' variable types. That means you could
2809 not include C99-style headers which use bools without triggering
2810 build errors in sieve.c.
2812 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2813 as mailq or other aliases. Changed the --help handling significantly
2814 to do whats expected. exim_usage() emits usage/help information.
2816 SC/01 Added the -bylocaldomain option to eximstats.
2818 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2820 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2822 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2828 PH/01 Another patch from the Sieve maintainer.
2830 PH/02 When an IPv6 address is converted to a string for single-key lookup
2831 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2832 dots are used instead of colons so that keys in lsearch files need not
2833 contain colons. This was done some time before quoting was made available
2834 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2835 (notated using the quote facility) so as to distinguish them from IPv4
2836 keys. This meant that lookups for IP addresses in host lists did not work
2837 for iplsearch lookups.
2839 This has been fixed by arranging for IPv6 addresses to be expressed with
2840 colons if the lookup type is iplsearch. This is not incompatible, because
2841 previously such lookups could never work.
2843 The situation is now rather anomalous, since one *can* have colons in
2844 ordinary lsearch keys. However, making the change in all cases is
2845 incompatible and would probably break a number of configurations.
2847 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2850 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2851 conversion specification without a maximum field width, thereby enabling
2852 a rogue spamd server to cause a buffer overflow. While nobody in their
2853 right mind would setup Exim to query an untrusted spamd server, an
2854 attacker that gains access to a server running spamd could potentially
2855 exploit this vulnerability to run arbitrary code as the Exim user.
2857 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2858 $primary_hostname instead of what libspf2 thinks the hosts name is.
2860 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2861 a directory entry by the name of the lookup key. Previously, if a
2862 symlink pointed to a non-existing file or a file in a directory that
2863 Exim lacked permissions to read, a lookup for a key matching that
2864 symlink would fail. Now it is enough that a matching directory entry
2865 exists, symlink or not. (Bugzilla 503.)
2867 PH/03 The body_linecount and body_zerocount variables are now exported in the
2870 PH/04 Added the $dnslist_matched variable.
2872 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2873 This means they are set thereafter only if the connection becomes
2876 PH/06 Added the client_condition to authenticators so that some can be skipped
2877 by clients under certain conditions.
2879 PH/07 The error message for a badly-placed control=no_multiline_responses left
2880 "_responses" off the end of the name.
2882 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2884 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2885 (without spaces) instead of just copying the configuration text.
2887 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2889 PH/11 Added $max_received_linelength.
2891 PH/12 Added +ignore_defer and +include_defer to host lists.
2893 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2894 way in which PCRE > 7.0 is built.
2896 PH/14 Implemented queue_only_load_latch.
2898 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2899 MAIL command. The effect was to mangle the value on 64-bit systems.
2901 PH/16 Another patch from the Sieve maintainer.
2903 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2905 PH/18 If a system quota error occurred while trying to create the file for
2906 a maildir delivery, the message "Mailbox is full" was not appended to the
2907 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2908 only to a quota excession during the actual writing of the file.
2910 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2911 characters?) which causes problems in log lines. The DN values are now
2912 passed through string_printing() before being added to log lines.
2914 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2915 and InterBase are left for another time.)
2917 PH/21 Added message_body_newlines option.
2919 PH/22 Guard against possible overflow in moan_check_errorcopy().
2921 PH/23 POSIX allows open() to be a macro; guard against that.
2923 PH/24 If the recipient of an error message contained an @ in the local part
2924 (suitably quoted, of course), incorrect values were put in $domain and
2925 $local_part during the evaluation of errors_copy.
2931 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2932 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2935 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2936 issue a MAIL command.
2938 PH/02 In an ACL statement such as
2940 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2942 if a client was not listed at all, or was listed with a value other than
2943 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2944 the condition was not true (as it should be), so access was not denied.
2945 The bug was that the ! inversion was incorrectly passed on to the second
2946 item. This has been fixed.
2948 PH/03 Added additional dnslists conditions == and =& which are different from
2949 = and & when the dns lookup returns more than one IP address.
2951 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2952 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2954 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2955 FSYNC, which compiles an option called disable_fsync that allows for
2956 bypassing fsync(). The documentation is heavily laced with warnings.
2958 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2960 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2961 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2962 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2963 including adding "make clean"; (3) Added -fPIC when compiling the test
2964 dynamically loaded module, to get rid of a warning.
2966 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2967 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2968 = 0s. The bug is otherwise harmless.
2970 PH/07 There was a bug in the dovecot authenticator such that the value of
2971 $auth1 could be overwritten, and so not correctly preserved, after a
2972 successful authentication. This usually meant that the value preserved by
2973 the server_setid option was incorrect.
2975 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2977 PH/09 Installed PCRE release 7.0.
2979 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2980 run for batched SMTP input. It is now run at the start of every message
2981 in the batch. While fixing this I discovered that the process information
2982 (output by running exiwhat) was not always getting set for -bs and -bS
2983 input. This is fixed, and it now also says "batched" for BSMTP.
2985 PH/11 Added control=no_pipelining.
2987 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2988 patch, slightly modified), and move the expansion of helo_data till after
2989 the connection is made in the smtp transport (so it can use these
2992 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2994 PH/14 Added log_selector = +pid.
2996 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2998 PH/16 Add ${if forany and ${if forall.
3000 PH/17 Added dsn_from option to vary the From: line in DSNs.
3002 PH/18 Flush SMTP output before performing a callout, unless control =
3003 no_callout_flush is set.
3005 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3006 was true (the default) a successful delivery failed to delete the retry
3007 item, thus causing premature timeout of the address. The bug is now
3010 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3012 PH/21 Long custom messages for fakedefer and fakereject are now split up
3013 into multiline responses in the same way that messages for "deny" and
3014 other ACL rejections are.
3016 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3017 with slight modification.
3019 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3020 draft, changing the syntax and factoring some duplicate code".
3022 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3023 for deliveries of the second and subsequent messages over the same SMTP
3026 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3027 ${reduce, with only minor "tidies".
3029 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3031 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3032 expansion side effects.
3034 PH/27 When a message times out after an over-quota error from an Exim-imposed
3035 quota, the bounce message says "mailbox is full". This message was not
3036 being given when it was a system quota that was exceeded. It now should
3039 MH/03 Made $recipients available in local_scan(). local_scan() already has
3040 better access to the recipient list through recipients_list[], but
3041 $recipients can be useful in postmaster-provided expansion strings.
3043 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3044 in the case of a MAIL command with additional options following the
3045 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3046 were accidentally chopped off.
3048 PH/29 SMTP synchronization checks are implemented when a command is read -
3049 there is a check that no more input is waiting when there shouldn't be
3050 any. However, for some commands, a delay in an ACL can mean that it is
3051 some time before the response is written. In this time, more input might
3052 arrive, invalidly. So now there are extra checks after an ACL has run for
3053 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3054 pipelining has not been advertised.
3056 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3058 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3059 returned in the SMTP response when smtp_return_error_details was false.
3060 This has been fixed.
3062 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3063 instead of the C I/O that was originally supplied, because problems were
3064 reported on Solaris.
3066 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3067 Exim which did not show up earlier: it was assuming that a call to
3068 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3069 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3070 was a macro that became an assignment, so it seemed to work. This has
3071 changed to a proper function call with a void return, hence the compile
3072 error. Exim's code has been fixed.
3074 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3077 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3079 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3081 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3082 because it turns out to be too big or there is a timeout) there is no
3083 "Completed" line in the log. When some messages of this type were
3084 selected by exigrep, they were listed as "not completed". Others were
3085 picked up by some special patterns. I have improved the selection
3086 criteria to be more general.
3088 PH/38 The host_find_failed option in the manualroute router can now be set
3089 to "ignore", to completely ignore a host whose IP address cannot be
3090 found. If all hosts are ignored, the behaviour is controlled by the new
3091 host_all_ignored option.
3093 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3094 homing or because of multiple MX records with /mx) generated more than
3095 one IP address, and the following item turned out to be the local host,
3096 all the secondary addresses of the first item were incorrectly removed
3097 from the list, along with the local host and any following hosts (which
3098 is what is supposed to happen).
3100 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3101 whoever called Exim into the -H file. In the case of the daemon it was
3102 behaving confusingly. When first started, it used values for whoever
3103 started the daemon, but after a SIGHUP it used the Exim user (because it
3104 calls itself on a restart). I have changed the code so that it now always
3107 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3108 message are rejected with the same error (e.g. no authentication or bad
3109 sender address), and a DATA command is nevertheless sent (as can happen
3110 with PIPELINING or a stupid MUA), the error message that was given to the
3111 RCPT commands is included in the rejection of the DATA command. This is
3112 intended to be helpful for MUAs that show only the final error to their
3115 PH/42 Another patch from the Sieve maintainer.
3117 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3118 Eximstats - Fixed some broken HTML links and added missing column headers
3120 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3121 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3123 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3125 PH/43 Yet another patch from the Sieve maintainer.
3127 PH/44 I found a way to check for a TCP/IP connection going away before sending
3128 the response to the final '.' that terminates a message, but only in the
3129 case where the client has not sent further data following the '.'
3130 (unfortunately, this is allowed). However, in many cases there won't be
3131 any further data because there won't be any more messages to send. A call
3132 to select() can be used: if it shows that the input is "ready", there is
3133 either input waiting, or the socket has been closed. An attempt to read
3134 the next input character can distinguish the two cases. Previously, Exim
3135 would have sent an OK response which the client would never have see.
3136 This could lead to message repetition. This fix should cure that, at
3137 least in a lot of common cases.
3139 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3140 advertised in response to EHLO.
3146 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3147 fixed by 4.65/MH/01 (is this a record?) are fixed:
3149 (i) An empty string was always treated as zero by the numeric comparison
3150 operators. This behaviour has been restored.
3152 (ii) It is documented that the numeric comparison operators always treat
3153 their arguments as decimal numbers. This was broken in that numbers
3154 starting with 0 were being interpreted as octal.
3156 While fixing these problems I realized that there was another issue that
3157 hadn't been noticed. Values of message_size_limit (both the global option
3158 and the transport option) were treated as octal if they started with 0.
3159 The documentation was vague. These values are now always treated as
3160 decimal, and I will make that clear in the documentation.
3166 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3167 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3170 MH/01 Don't check that the operands of numeric comparison operators are
3171 integers when their expansion is in "skipping" mode (fixes bug
3172 introduced by 4.64-PH/07).
3174 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3175 child addresses, Exim now panics and dies. Previously, because the count
3176 is held in a short int, deliveries were likely to be lost. As such a
3177 large number of recipients for a single message is ridiculous
3178 (performance will be very, very poor), I have chosen to impose a limit
3179 rather than extend the field.
3185 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3186 leftover -K file (the existence of which was triggered by #402).
3187 While we were at it, introduced process PID as part of the -K
3188 filename. This should rule out race conditions when creating
3191 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3192 processing considerably. Previous code took too long for large mails,
3193 triggering a timeout which in turn triggers #401.
3195 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3196 in the DK code in transports.c. sendfile() is not really portable,
3197 hence the _LINUX specificness.
3199 TF/01 In the add_headers option to the mail command in an Exim filter,
3200 there was a bug that Exim would claim a syntax error in any
3201 header after the first one which had an odd number of characters
3204 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3205 callout verification, Exim cached a "reject" for the entire domain. This
3206 is correct for most verifications, but it is not correct for a recipient
3207 verification with use_sender or use_postmaster set, because in that case
3208 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3209 case of MAIL FROM:<> rejection from other early rejections (e.g.
3210 rejection of HELO). When verifying a recipient using a non-null MAIL
3211 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3212 Whatever the result of the callout, the value of the domain cache is
3213 left unchanged (for any other kind of callout, getting as far as trying
3214 RCPT means that the domain itself is ok).
3216 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3219 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3220 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3221 session, the daemon ignores ECONNECT errors and logs others; it now
3222 ignores EPIPE as well.
3224 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3225 (quoted-printable decoding).
3227 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3228 later the small subsequent patch to fix an introduced bug.
3230 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3232 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3234 PH/08 An error is now given if message_size_limit is specified negative.
3236 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3237 to be given (somewhat) arbitrary names.
3239 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3242 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3243 miscellaneous code fixes
3245 PH/10 Added the log_reject_target ACL modifier to specify where to log
3248 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3249 hostname. This is wrong, because it relates to the incoming message (and
3250 probably the interface on which it is arriving) and not to the outgoing
3251 callout (which could be using a different interface). This has been
3252 changed to use the value of the helo_data option from the smtp transport
3253 instead - this is what is used when a message is actually being sent. If
3254 there is no remote transport (possible with a router that sets up host
3255 addresses), $smtp_active_hostname is used.
3257 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3258 tweaks were necessary in order to get it to work (see also 21 below):
3259 (a) The code assumed that strncpy() returns a negative number on buffer
3260 overflow, which isn't the case. Replaced with Exim's string_format()
3262 (b) There were several signed/unsigned issues. I just did the minimum
3263 hacking in of casts. There is scope for a larger refactoring.
3264 (c) The code used strcasecmp() which is not a standard C function.
3265 Replaced with Exim's strcmpic() function.
3266 (d) The code set only $1; it now sets $auth1 as well.
3267 (e) A simple test gave the error "authentication client didn't specify
3268 service in request". It would seem that Dovecot has changed its
3269 interface. Fortunately there's a specification; I followed it and
3270 changed what the client sends and it appears to be working now.
3272 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3275 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3276 address A is aliased to B and C, where B exists and C does not. Without
3277 -v the output is "A verified" because verification stops after a
3278 successful redirection if more than one address is generated. However,
3279 with -v the child addresses are also verified. Exim was outputting "A
3280 failed to verify" and then showing the successful verification for C,
3281 with its parentage. It now outputs "B failed to verify", showing B's
3282 parentage before showing the successful verification of C.
3284 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3285 look up a TXT record in a specific list after matching in a combined
3288 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3289 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3290 they consult the DNS. I had assumed they would set it the way they
3291 wanted; and indeed my experiments on Linux seem to show that in some
3292 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3293 To be on the safe side, however, I have now made the interface to
3294 host_find_byname() similar to host_find_bydns(), with an argument
3295 containing the DNS resolver options. The host_find_byname() function now
3296 sets these options at its start, just as host_find_bydns() does. The smtp
3297 transport options dns_qualify_single and dns_search_parents are passed to
3298 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3299 of host_find_byname() use the default settings of RES_DEFNAMES
3300 (qualify_single) but not RES_DNSRCH (search_parents).
3302 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3303 spool_read_header() do less string testing, by means of a preliminary
3304 switch on the second character of optional "-foo" lines. (This is
3305 overdue, caused by the large number of possibilities that now exist.
3306 Originally there were few.) While I was there, I also converted the
3307 str(n)cmp tests so they don't re-test the leading "-" and the first
3308 character, in the hope this might squeeze out yet more improvement.
3310 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3311 flag allowing group syntax was set by the header_syntax check but not
3312 turned off, possible causing trouble later; (2) The flag was not being
3313 set at all for the header_verify test, causing "group"-style headers to
3314 be rejected. I have now set it in this case, and also caused header_
3315 verify to ignore an empty address taken from a group. While doing this, I
3316 came across some other cases where the code for allowing group syntax
3317 while scanning a header line wasn't quite right (mostly, not resetting
3318 the flag correctly in the right place). These bugs could have caused
3319 trouble for malformed header lines. I hope it is now all correct.
3321 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3322 with the "reply" argument non-NULL. The code, however (which originally
3323 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3324 but it didn't always do it. This confused somebody who was copying the
3325 code for some other use. I have removed all the tests.
3327 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3328 feature that was used to support insecure browsers during the U.S. crypto
3329 embargo. It requires special client support, and Exim is probably the
3330 only MTA that supported it -- and would never use it because real RSA is
3331 always available. This code has been removed, because it had the bad
3332 effect of slowing Exim down by computing (never used) parameters for the
3333 RSA_EXPORT functionality.
3335 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3336 authenticator to fail if there's a tab character in the incoming data
3337 (there should never be unless someone is messing about, as it's supposed
3338 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3339 if the connection is using TLS or if the remote IP is the same as the
3340 local IP, and the "valid-client-cert option" if a client certificate has
3343 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3344 authenticators. This can be used for authorization after authentication
3345 succeeds. (In the case of plaintext, it servers for both authentication
3348 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3349 if any retry times were supplied.
3351 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3352 connection, where there is no client IP address to check. In this
3353 situation, the verify now always succeeds.
3355 PH/25 Applied John Jetmore's -Mset patch.
3357 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3359 PH/27 In a string expansion for a processed (not raw) header when multiple
3360 headers of the same name were present, leading whitespace was being
3361 removed from all of them, but trailing whitespace was being removed only
3362 from the last one. Now trailing whitespace is removed from each header
3363 before concatenation. Completely empty headers in a concatenation (as
3364 before) are ignored.
3366 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3367 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3369 PH/29 [Removed. This was a change that I later backed out, and forgot to
3370 correct the ChangeLog entry (that I had efficiently created) before
3371 committing the later change.]
3373 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3374 address errors (4xx response to RCPT) over the same connection as other
3375 messages routed to the same hosts. Such deliveries are always "forced",
3376 so retry times are not inspected. This resulted in far too many retries
3377 for the affected addresses. The effect occurred only when there were more
3378 hosts than the hosts_max_try setting in the smtp transport when it had
3379 the 4xx errors. Those hosts that it had tried were not added to the list
3380 of hosts for which the message was waiting, so if all were tried, there
3381 was no problem. Two fixes have been applied:
3383 (i) If there are any address or message errors in an SMTP delivery, none
3384 of the hosts (tried or untried) are now added to the list of hosts
3385 for which the message is waiting, so the message should not be a
3386 candidate for sending over the same connection that was used for a
3387 successful delivery of some other message. This seems entirely
3388 reasonable: after all the message is NOT "waiting for some host".
3389 This is so "obvious" that I'm not sure why it wasn't done
3390 previously. Hope I haven't missed anything, but it can't do any
3391 harm, as the worst effect is to miss an optimization.
3393 (ii) If, despite (i), such a delivery is accidentally attempted, the
3394 routing retry time is respected, so at least it doesn't keep
3395 hammering the server.
3397 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3398 in ${readsocket because some servers need this prod.
3400 PH/32 Added some extra debug output when updating a wait-xxx database.
3402 PH/33 The hint "could be header name not terminated by colon", which has been
3403 given for certain expansion errors for a long time, was not being given
3404 for the ${if def:h_colon_omitted{... case.
3406 PH/34 The spec says: "With one important exception, whenever a domain list is
3407 being scanned, $domain contains the subject domain." There was at least
3408 one case where this was not true.
3410 PH/35 The error "getsockname() failed: connection reset by peer" was being
3411 written to the panic log as well as the main log, but it isn't really
3412 panic-worthy as it just means the connection died rather early on. I have
3413 removed the panic log writing for the ECONNRESET error when getsockname()
3416 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3417 runs only) independently of the message's sender address. This meant
3418 that, if the 4xx error was in fact related to the sender, a different
3419 message to the same recipient with a different sender could confuse
3420 things. In particular, this can happen when sending to a greylisting
3421 server, but other circumstances could also provoke similar problems.
3422 I have changed the default so that the retry time for these errors is now
3423 based a combination of the sender and recipient addresses. This change
3424 can be overridden by setting address_retry_include_sender=false in the
3427 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3428 remote server are returned as part of bounce messages. This was not
3429 happening for LMTP over a pipe (the lmtp transport), but now it is the
3430 same for both kinds of LMTP.
3432 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3433 in header lines that were in fact CNAMEs. This is no longer the case.
3435 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3436 and queue runs started by the daemon processed all messages. This has
3437 been fixed so that -R and -S can now usefully be given with -q<time>.
3439 PH/40 Import PCRE release 6.7 (fixes some bugs).
3441 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3443 PH/42 Give an error if -q is specified more than once.
3445 PH/43 Renamed the variables $interface_address and $interface_port as
3446 $received_ip_address and $received_port, to make it clear that these
3447 values apply to message reception, and not to the outgoing interface when
3448 a message is delivered. (The old names remain recognized, of course.)
3450 PH/44 There was no timeout on the connect() call when using a Unix domain
3451 socket in the ${readsocket expansion. There now is.
3453 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3454 be meaningful with "accept".
3456 SC/01 Eximstats V1.43
3457 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3459 SC/02 Eximstats V1.44
3460 Use a glob alias rather than an array ref in the generated
3461 parser. This improves both readability and performance.
3463 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3464 Collect SpamAssassin and rejection statistics.
3465 Don't display local sender or destination tables unless
3466 there is data to show.
3467 Added average volumes into the top table text output.
3469 SC/04 Eximstats V1.46
3470 Collect data on the number of addresses (recipients)
3471 as well as the number of messages.
3473 SC/05 Eximstats V1.47
3474 Added 'Message too big' to the list of mail rejection
3475 reasons (thanks to Marco Gaiarin).
3477 SC/06 Eximstats V1.48
3478 Mainlog lines which have GMT offsets and are too short to
3479 have a flag are now skipped.
3481 SC/07 Eximstats V1.49 (Alain Williams)
3482 Added the -emptyok flag.
3484 SC/08 Eximstats V1.50
3485 Fixes for obtaining the IP address from reject messages.
3487 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3488 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3489 whitespace changes from 4.64-PH/27
3491 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3494 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3495 are found, allow negative numbers in numeric criteria)
3497 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3499 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3500 to match changes made in 4.64-PH/43
3502 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3504 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3505 "domain" is actually an IP address, and force a failure. This locks out
3506 those revolvers/nameservers that support "A-for-A" lookups, in
3507 contravention of the specifications.
3509 PH/48 When a host name was looked up from an IP address, and the subsequent
3510 forward lookup of the name timed out, the host name was left in
3511 $sender_host_name, contrary to the specification.
3513 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3514 restricted to single-key lookups, Exim was not diagnosing an error if
3515 * or *@ was used with a query-style lookup.
3517 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3519 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3520 long ago, but noone interested enough thought of it. Let's just say that
3521 the "1.1" means that there are some new functions that weren't there at
3522 some point in the past.
3524 PH/51 Error processing for expansion failure of helo_data from an smtp
3525 transport during callout processing was broken.
3527 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3528 tested/used via the -bh/-bhc/-bs options.
3530 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3531 bug, fixed in subsequent PCRE releases).
3533 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3534 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3536 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3542 SC/01 Use a glob alias rather than an array ref in eximstats generated
3543 parser. This improves both readability and performance.
3545 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3546 Don't display local sender or destination tables in eximstats unless
3547 there is data to show.
3548 Added average volumes into the eximstats top table text output.
3550 SC/03 Collect data on the number of addresses (recipients) as well
3551 as the number of messages in eximstats.
3553 TF/01 Correct an error in the documentation for the redirect router. Exim
3554 does (usually) call initgroups() when daemonizing.
3556 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3557 with consistent privilege compared to when running as a daemon.
3559 TF/03 Note in the spec that $authenticated_id is not set for local
3560 submissions from trusted users.
3562 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3563 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3565 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3566 by adding some example configuration directives to the default
3567 configuration file. A little bit of work is required to uncomment the
3568 directives and define how usernames and passwords are checked, but
3569 there is now a framework to start from.
3571 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3572 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3573 without this. I don't know how relevant this is to other LDAP libraries.
3575 PH/02 Add the verb name to the "unknown ACL verb" error.
3577 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3579 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3581 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3582 directory not expanded when it should be if an expanded home directory
3583 was set for the address (which is overridden by the transport).
3585 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3588 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3589 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3590 because it is too late at that time, and has no effect.
3592 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3593 security issue with \' (bugzilla #107). I could not use the
3594 PQescapeStringConn() function, because it needs a PGconn value as one of
3597 PH/08 When testing addresses using -bt, indicate those final addresses that
3598 are duplicates that would not cause an additional delivery. At least one
3599 person was confused, thinking that -bt output corresponded to deliveries.
3600 (Suppressing duplicates isn't a good idea as you lose the information
3601 about possibly different redirections that led to the duplicates.)
3603 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3604 systems where poll() doesn't work, in particular OS X.
3606 PH/10 Added more information to debugging output for retry time not reached.
3608 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3609 operations in malware.c.
3611 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3614 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3615 syslog_duplication set false, log lines that would normally be written
3616 both the the main log and to the reject log were not written to syslog at
3619 PH/14 In the default configuration, change the use of "message" in ACL warn
3620 statements to "add_header".
3622 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3623 not followed by a command (e.g. "seen endif").
3625 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3626 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3629 PH/17 Added extra conditions to the default value of delay_warning_condition
3633 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3634 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3635 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3638 The Auto-Submitted: and various List- headers are standardised, whereas I
3639 don't think Precedence: ever was.
3641 PH/18 Refactored debugging code in route_finduser() to show more information,
3642 in particular, the error code if getpwnam() issues one.
3644 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3645 This is apparently needed in addition to the PH/07 change above to avoid
3646 any possible encoding problems.
3648 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3649 but not after initializing Perl.
3651 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3652 output them only if debugging. By default they are written stderr,
3653 apparently, which is not desirable.
3655 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3658 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3661 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3663 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3664 authenticated or an ident call has been made. Suppress the default
3665 values for $authenticated_id and $authenticated_sender (but permit -oMai
3666 and -oMas) when testing with -bh.
3668 PH/24 Re-jigged the order of the tests in the default configuration so that the
3669 tests for valid domains and recipients precede the DNS black list and CSA
3670 tests, on the grounds that those ones are more expensive.
3672 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3673 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3674 This bug exists in every version of Exim that I still have, right back to
3677 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3678 However, an attempt to turn on case-sensitivity in a regex key by
3679 including (?-i) didn't work because the subject string was already
3680 lowercased, and the effects were non-intuitive. It turns out that a
3681 one-line patch can be used to allow (?-i) to work as expected.
3687 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3688 other effects) broke the use of negated acl sub-conditions.
3690 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3693 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3694 "Deny" causes Exim to reject the incoming connection with a 554 error.
3695 Unfortunately, if there is a major crisis, such as a disk failure,
3696 tcp-wrappers gives "deny", whereas what one would like would be some
3697 kind of temporary error. A kludge has been added to help with this.
3698 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3699 554 error is used if errno is still zero or contains ENOENT (which occurs
3700 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3703 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3705 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3706 errors. Otherwise a message that provokes a temporary error (when other
3707 messages do not) can cause a whole host to time out.
3709 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3710 addresses were routed directly to files or pipes from a redirect router.
3711 File deliveries just didn't batch; pipe deliveries might have suffered
3714 PH/06 A failure to get a lock for a hints database would erroneously always say
3715 "Failed to get write lock", even when it was really a read lock.
3717 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3718 of 0600. This has been changed to use the value of the lockfile_mode
3719 option (which defaults to 0600).
3721 PH/08 Applied small patch from the Sieve maintainer.
3723 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3724 folder from quota calculations, a direct delivery into this folder messed
3725 up the contents of the maildirsize file. This was because the regex was
3726 used only to exclude .Trash (or whatever) when the size of the mailbox
3727 was calculated. There was no check that a delivery was happening into an
3728 excluded directory. This bug has been fixed by ignoring all quota
3729 processing for deliveries into excluded directories.
3731 PH/10 Added the maildirfolder_create_regex option to appendfile.
3737 PH/01 The code for finding all the local interface addresses on a FreeBSD
3738 system running IPv6 was broken. This may well have applied to all BSD
3739 systems, as well as to others that have similar system calls. The broken
3740 code found IPv4 interfaces correctly, but gave incorrect values for the
3741 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3742 that it would not match correctly against @[] and not recognize the IPv6
3745 PH/02 The ipliteral router was not recognizing addresses of the form user@
3746 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3748 PH/03 Added disable_ipv6.
3750 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3751 decoded form, because it is most often used to construct To: headers
3752 lines in autoreplies, and the decoded form may well be syntactically
3753 invalid. However, $reply_address has leading white space removed, and all
3754 newlines turned into spaces so that the autoreply transport does not
3757 PH/05 If group was specified without a user on a router, and no group or user
3758 was specified on a transport, the group from the router was ignored.
3760 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3761 for visible compile-time settings that can be used to change these
3762 numbers, for those that want even more. Backwards compatibility with old
3763 spool files has been maintained. However, going back to a previous Exim
3764 release will lost any variables that are in spool files.
3766 PH/07 Two small changes when running in the test harness: increase delay when
3767 passing a TCP/IP connection to a new process, in case the original
3768 process has to generate a bounce, and remove special handling of
3769 127.0.0.2 (sic), which is no longer necessary.
3771 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3772 be the same on different OS.
3774 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3777 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3778 whether --show-vars was specified or not
3780 JJ/02 exipick: Added support for new ACL variable spool format introduced
3783 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3784 syntactically invalid From: or Reply-to: line, and a filter used this to
3785 generate an autoreply, and therefore failed to obtain an address for the
3786 autoreply, Exim could try to deliver to a non-existent relative file
3787 name, causing unrelated and misleading errors. What now happens is that
3788 it logs this as a hard delivery error, but does not attempt to create a
3791 PH/11 The exinext utility has a -C option for testing purposes, but although
3792 the given file was scanned by exinext itself; it wasn't being passed on
3793 when Exim was called.
3795 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3796 an end-of-file indication when reading a command response.
3798 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3799 compiled. In many other places in Exim, IPv6 addresses are always
3800 recognized, so I have changed this. It also means that IPv4 domain
3801 literals of the form [IPV4:n.n.n.n] are now always recognized.
3803 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3804 used if the router is not running as root, for example, when verifying at
3805 ACL time, or when using -bh. The debugging output from this situation was
3806 non-existent - all you got was a failure to exec. I have made two
3809 (a) Failures to set uid/gid, the current directory, or a process leader
3810 in a subprocess such as that created by queryprogram now generate
3811 suitable debugging output when -d is set.
3813 (b) The queryprogram router detects when it is not running as root,
3814 outputs suitable debugging information if -d is set, and then runs
3815 the subprocess without attempting to change uid/gid.
3817 PH/15 Minor change to Makefile for building test_host (undocumented testing
3820 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3821 additional section of a DNS packet that returns MX or SRV records.
3822 Instead, it always explicitly searches for A/AAAA records. This avoids
3823 major problems that occur when a DNS server includes only records of one
3824 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3825 fixed another bug: if SRV records were looked up and the corresponding
3826 address records were *not* found in the additional section, the port
3827 values from the SRV records were lost.
3829 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3830 using the correct key (the original address) when searching the retry
3831 rules in order to find which one to use for generating the retry hint.
3833 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3834 adding the default one. Similarly, if it contains a Reply-To: header, the
3835 errors_reply_to option, if set, is not used.
3837 PH/19 When calculating a retry time, Exim used to measure the "time since
3838 failure" by looking at the "first failed" field in the retry record. Now
3839 it does not use this if it is later than than the arrival time of the
3840 message. Instead it uses the arrival time. This makes for better
3841 behaviour in cases where some deliveries succeed, thus re-setting the
3842 "first failed" field. An example is a quota failure for a huge message
3843 when small messages continue to be delivered. Without this change, the
3844 "time since failure" will always be short, possible causing more frequent
3845 delivery attempts for the huge message than are intended.
3846 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3848 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3849 $1, $2, $3) because the numerical variables can be reset during some
3850 expansion items (e.g. "match"), thereby losing the authentication data.
3852 PH/21 Make -bV show the size of off_t variables so that the test suite can
3853 decide whether to run tests for quotas > 2G.
3855 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3856 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3857 filecount value is greater than 2G or if a quota value is greater than 2G
3858 on a system where the size of off_t is not greater than 4, a panic error
3861 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3862 never match. The debug and -bh output now contains an explicit error
3863 message indicating a malformed IPv4 address or mask.
3865 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3866 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3867 PH/23 above applies.
3869 PH/25 Do not write to syslog when running in the test harness. The only
3870 occasion when this arises is a failure to open the main or panic logs
3871 (for which there is an explicit test).
3873 PH/26 Added the /no_tell option to "control=freeze".
3875 PH/27 If a host name lookup failed very early in a connection, for example, if
3876 the IP address matched host_lookup and the reverse lookup yielded a name
3877 that did not have a forward lookup, an error message of the form "no IP
3878 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3879 could be logged. Now it outputs the IP address instead of "NULL".
3881 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3882 allows the sender and the authenticated sender to be set when
3883 submitting a message from within Exim. Since child_open_exim() is
3884 documented for local_scan(), the new function should be too.
3886 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3887 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3888 results in an empty string is now treated as unset.
3890 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3892 PH/31 Added sender_verify_fail logging option.
3894 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3895 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3896 tidied the source and removed it altogether.
3898 PH/33 When a queue run was abandoned because the load average was too high, a
3899 log line was always written; now it is written only if the queue_run log
3900 selector is set. In addition, the log line for abandonment now contains
3901 information about the queue run such as the pid. This is always present
3902 in "start" and "stop" lines but was omitted from the "abandon" line.
3904 PH/34 Omit spaces between a header name and the colon in the error message that
3905 is given when verify = headers_syntax fails (if there are lots of them,
3906 the message gets confusing).
3908 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3909 names, as there are now some PTR records that contain slashes. This check
3910 is only to protect against broken name servers that fall over on strange
3911 characters, so the fact that it applies to all lookups doesn't matter.
3913 PH/36 Now that the new test suite is complete, we can remove some of the
3914 special code in Exim that was needed for the old test suite. For example,
3915 sorting DNS records because real resolvers return them in an arbitrary
3916 order. The new test suite's fake resolver always returns records in the
3919 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3920 bounces) except when queue_only is set, to avoid logging races between
3921 the different processes.
3923 PH/38 Panic-die if .include specifies a non-absolute path.
3925 PH/39 A tweak to the "H" retry rule from its user.
3927 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3928 a label. They prevented compilation on older perls.
3930 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3931 a warning to be raised on newish perls.
3933 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3934 on queue. Changes to match documented behaviour of showing count of
3935 messages matching specified criteria.
3937 PH/40 Changed the default ident timeout from 30s to 5s.
3939 PH/41 Added support for the use of login_cap features, on those BSD systems
3940 that have them, for controlling the resources used by pipe deliveries.
3942 PH/42 The content-scanning code uses fopen() to create files in which to put
3943 message data. Previously it was not paying any attention to the mode of
3944 the files. Exim runs with umask(0) because the rest of the code creates
3945 files with open(), and sets the required mode explicitly. Thus, these
3946 files were ending up world-writeable. This was not a big issue, because,
3947 being within the spool directory, they were not world-accessible. I have
3948 created a function called modefopen, which takes an additional mode
3949 argument. It sets umask(777), creates the file, chmods it to the required
3950 mode, then resets the umask. All the relevant calls to fopen() in the
3951 content scanning code have been changed to use this function.
3953 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3954 to 24 hours. This avoids potential overflow problems when processing G
3955 and H retry rules. I suspect nobody ever tinkers with this value.
3957 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3959 PH/45 When the plaintext authenticator is running as a client, the server's
3960 challenges are checked to ensure they are valid base64 strings. By
3961 default, the authentication attempt is cancelled if an invalid string is
3962 received. Setting client_ignore_invalid_base64 true ignores these errors.
3963 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3964 they are received. Thus, the responses can be made to depend on the
3965 challenges. If an invalid string is ignored, an empty string is placed in
3968 PH/46 Messages that are created by the autoreply transport now contains a
3969 References: header, in accordance with RFCs 2822 and 3834.
3971 PH/47 Added authenticated_sender_force to the smtp transport.
3973 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3975 PH/49 Installed latest patch from the Sieve maintainer.
3977 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3978 was also set, the appendfile transport was unnecessarily scanning a
3979 directory of message files (e.g. for maildir delivery) to find the count
3980 of files (along with the size), even though it did not need this
3981 information. It now does the scan only if it needs to find either the
3982 size of the count of files.
3984 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3986 PH/52 Two bugs concerned with error handling when the smtp transport is
3989 (i) Exim was not creating retry information for temporary errors given
3990 for individual recipients after the DATA command when the smtp transport
3991 was used in LMTP mode. This meant that they could be retried too
3992 frequently, and not timed out correctly.
3994 (ii) Exim was setting the flag that allows error details to be returned
3995 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3996 recipients that were returned after the DATA command.
3998 PH/53 This is related to PH/52, but is more general: for any failing address,
3999 when detailed error information was permitted to be returned to the
4000 sender, but the error was temporary, then after the final timeout, only
4001 "retry timeout exceeded" was returned. Now it returns the full error as
4002 well as "retry timeout exceeded".
4004 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4005 do this, and (what is worse) MTAs that accept it.
4007 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4008 will now be deprecated.
4010 PH/56 New os.c-cygwin from the Cygwin maintainer.
4012 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4013 formats (previously only available in exim formats via -bpr, -bpru,
4014 and -bpra. Now also available in native and exiqgrep formats)
4016 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4017 with very large, slow to parse queues
4019 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4021 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4023 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4024 responses to authentication challenges, though it was showing the
4025 challenges; (ii) I've removed the CR characters from the debug output for
4028 PH/58 Allow for the insertion of a newline as well as a space when a string
4029 is turned into more than one encoded-word during RFC 2047 encoding. The
4030 Sieve code now uses this.
4032 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4033 data_4xx, lost_connection, tls_required.
4035 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4036 message was being sent as an SMTP response.
4038 PH/61 Add -l and -k options to exicyclog.
4040 PH/62 When verifying, if an address was redirected to one new address, so that
4041 verification continued, and the new address failed or deferred after
4042 having set something in $address_data, the value of $address_data was not
4043 passed back to the ACL. This was different to the case when no
4044 redirection occurred. The value is now passed back in both cases.
4046 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4047 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4048 use login_cap.h, so on its own it isn't the distinguishing feature. The
4049 new name refers directly to the setclassresources() function.
4051 PH/65 Added configuration files for NetBSD3.
4053 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4055 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4056 is preferred over IPv4.
4058 PH/68 The bounce_return_message and bounce_return_body options were not being
4059 honoured for bounces generated during the reception of non-SMTP messages.
4060 In particular, this applied to messages rejected by the ACL. This bug has
4061 been fixed. However, if bounce_return_message is true and bounce_return_
4062 body is false, the headers that are returned for a non-SMTP message
4063 include only those that have been read before the error was detected.
4064 (In the case of an ACL rejection, they have all been read.)
4066 PH/69 The HTML version of the specification is now built in a directory called
4067 spec_html instead of spec.html, because the latter looks like a path with
4068 a MIME-type, and this confuses some software.
4070 PH/70 Catch two compiler warnings in sieve.c.
4072 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4073 function verify_get_ident() calls ip_connect() to connect a socket, but
4074 if the "connect()" function timed out, ip_connect() used to close the
4075 socket. However, verify_get_ident() also closes the socket later, and in
4076 between Exim writes to the log, which may get opened at this point. When
4077 the socket was closed in ip_connect(), the log could get the same file
4078 descriptor number as the socket. This naturally causes chaos. The fix is
4079 not to close the socket in ip_connect(); the socket should be closed by
4080 the function that creates it. There was only one place in the code where
4081 this was missing, in the iplookup router, which I don't think anybody now
4082 uses, but I've fixed it anyway.
4084 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4085 well as to direct DNS lookups. Otherwise the handling of names in host
4086 lists is inconsistent and therefore confusing.
4092 PH/01 Two changes to the default runtime configuration:
4094 (1) Move the checks for relay_from_hosts and authenticated clients from
4095 after to before the (commented out) DNS black list checks.
4097 (2) Add control=submission to the relay_from_hosts and authenticated
4098 clients checks, on the grounds that messages accepted by these
4099 statements are most likely to be submissions.
4101 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4103 (1) Generate an error if the third argument for the ${prvs expansion is
4106 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4109 (3) Reset the variables that are obtained from the first argument of
4110 ${prvscheck and used in the second argument before leaving the code,
4111 because their memory is reclaimed, so using them afterwards may do
4114 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4115 one (it's much easier than Tom thought :-).
4117 (5) Because of (4), we can now allow for the use of $prvscheck_result
4118 inside the third argument.
4120 PH/03 For some reason, the default setting of PATH when running a command from
4121 a pipe transport was just "/usr/bin". I have changed it to
4124 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4125 anything to be listed in the output from -bV.
4127 PH/05 When a filter generated an autoreply, the entire To: header line was
4128 quoted in the delivery log line, like this:
4130 => >A.N.Other <ano@some.domain> <original@ddress> ...
4132 This has been changed so that it extracts the operative address. There
4133 may be more than one such address. If so, they are comma-separated, like
4136 => >ano@some.domain,ona@other.domain <original@ddress> ...
4138 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4139 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4140 not being looked up in the rDNS to get a host name, Exim was showing the
4141 IP address twice in Received: lines, even though the IP addresses were
4142 identical. For example:
4144 Received: from [1.2.3.4] (helo=[1.2.3.4])
4146 However, if the real host name was known, it was omitting the HELO data
4147 if it matched the actual IP address. This has been tidied up so that it
4148 doesn't show the same IP address twice.
4150 PH/07 When both +timestamp and +memory debugging was on, the value given by
4151 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4152 called by the debug printing, thereby overwriting the timestamp buffer.
4153 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4155 PH/08 When the original message was included in an autoreply transport, it
4156 always said "this is a copy of the message, including all the headers",
4157 even if body_only or headers_only was set. It now gives an appropriate
4160 PH/09 Applied a patch from the Sieve maintainer which:
4162 o fixes some comments
4163 o adds the (disabled) notify extension core
4164 o adds some debug output for the result of if/elsif tests
4165 o points to the current vacation draft in the documentation
4166 and documents the missing references header update
4170 o fixes a bug in processing the envelope test (when testing
4171 multiple envelope elements, the last element determined the
4174 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4175 Electronic Mail") by including:
4177 Auto-submitted: auto-generated
4179 in the messages that it generates (bounce messages and others, such as
4180 warnings). In the case of bounce messages for non-SMTP messages, there was
4181 also a typo: it was using "Auto_submitted" (underscore instead of
4182 hyphen). Since every message generated by Exim is necessarily in response
4183 to another message, thes have all been changed to:
4185 Auto-Submitted: auto-replied
4187 in accordance with these statements in the RFC:
4189 The auto-replied keyword:
4191 - SHOULD be used on messages sent in direct response to another
4192 message by an automatic process,
4194 - MUST NOT be used on manually-generated messages,
4196 - MAY be used on Delivery Status Notifications (DSNs) and Message
4197 Disposition Notifications (MDNs),
4199 - MUST NOT be used on messages generated by automatic or periodic
4200 processes, except for messages which are automatic responses to
4203 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4204 to the default Received: header definition.
4206 PH/12 Added log selector acl_warn_skipped (default on).
4208 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4209 variables because (a) they are in the wrong storage pool and (b) even if
4210 they were copied, it wouldn't work properly because of the caching.
4212 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4213 checking when decoding. Apparently there are clients that generate
4214 overlong encoded strings. Why am I not surprised?
4216 PH/15 If the first argument of "${if match_address" was not empty, but did not
4217 contain an "@" character, Exim crashed. Now it writes a panic log message
4218 and treats the condition as false.
4220 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4222 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4223 extension "envelope-auth". The code is finished and in agreement with
4224 other implementations, but there is no documentation so far and in fact,
4225 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4226 not changing the active code.
4228 Print executed "if" and "elsif" statements when debugging is used. This
4229 helps a great deal to understand what a filter does.
4231 Document more things not specified clearly in RFC3028. I had all this
4232 sorted out, when out of a sudden new issues came to my mind. Oops."
4234 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4237 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4238 canonical form (as documented). However, after a host name lookup from
4239 the IP address, check_host() was doing a simple string comparison with
4240 addresses acquired from the DNS when checking that the found name did
4241 have the original IP as one of its addresses. Since any found IPv6
4242 addresses are likely to be in abbreviated form, the comparison could
4243 fail. Luckily, there already exists a function for doing the comparison
4244 by converting both addresses to binary, so now that is used instead of
4245 the text comparison.
4247 PH/20 There was another similar case to PH/19, when a complete host name was
4248 given in a host list; looking up its IP address could give an abbreviated
4249 form, whereas the current host's name might or might not be abbreviated.
4250 The same fix has been applied.
4256 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4257 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4260 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4261 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4263 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4265 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4266 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4267 Also, for "Auto-Submitted", treat anything other than "no" as
4268 non-personal, in accordance with RFC 3834. (Previously it treated
4269 anything starting "auto-" as non-personal.)
4271 TF/01 The control=submission/name=... option had a problem with syntax
4272 errors if the name included a slash character. The /name= option
4273 now slurps the rest of the string, so it can include any characters
4274 but it must come last in the list of options (after /sender_retain
4277 PH/05 Some modifications to the interface to the fake nameserver for the new
4285 TK/01 Added the "success_on_redirect" address verification option. See
4286 NewStuff for rationale and an example.
4288 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4290 PH/02 Patch to exigrep to allow it to work on syslog lines.
4292 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4293 fread() to skip over the body file's header line, because in Cygwin the
4294 header line is locked and is inaccessible.
4296 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4297 co-exist for some time) to make it clear that it is the Exim ID that is
4298 referenced, not the Message-ID: header line.
4300 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4301 string_format() function, because snprintf() does not exist on all
4304 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4305 ${stat: expansion item.
4307 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4308 protocol synchronization error", to keep the pedants happy.
4310 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4311 well as for IRIX systems, when gcc is being used. See the host.c source
4314 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4316 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4319 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4320 transport and to the smtp transport in LMTP mode.
4322 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4324 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4325 run a filter in a subprocess. This could lead to confusion in subsequent
4326 lookups in the parent process. There should also be a search_tidyup() at
4327 the end of the subprocess.
4329 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4330 only if the host matched helo_try_verify_hosts, which caused the
4331 verification to occur when the EHLO/HELO command was issued. The ACL just
4332 tested the remembered result. Now, if a previous verification attempt has
4333 not happened, "verify = helo" does it there and then.
4335 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4337 TK/03 Fix log output including CR from clamd.
4339 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4340 exist provoked a memory error which could cause a segfault.
4342 PH/15 Installed PCRE 6.2
4344 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4346 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4347 of the problem. Specifically, suggested +O2 rather than +O1 for the
4350 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4352 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4353 record, the port was not correctly logged when the outgoing_port log
4354 selector was set (it logged the transort's default port).
4356 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4357 fallback_hosts, and "hosts" in the smtp transport.
4359 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4360 host errors such as "Connection refused".
4362 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4363 authentication with radiusclient 0.4.9:
4365 - Error returned from rc_read_config was caught wrongly
4366 - Username/password not passed on to radius server due to wrong length.
4368 The presumption is that some radiusclient API changes for 4.51/PH/17
4369 were not taken care of correctly. The code is still untested by me (my
4370 Linux distribution still has 0.3.2 of radiusclient), but it was
4371 contributed by a Radius user.
4373 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4374 expanding the "port" option of the smtp transport.
4376 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4377 while reading a MIME header. Thanks to Tom Hughes for a patch.
4379 PH/24 Include config.h inside local_scan.h so that configuration settings are
4382 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4383 that in an ACL for RCPT (for example), you can examine exactly what was
4386 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4387 commands, but it was not correctly comparing the address with the actual
4388 client host address. Thus, it would show the EHLO address in Received:
4389 header lines when this was not necessary.
4391 PH/27 Added the % operator to ${eval:}.
4393 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4394 it should be ignoring failures (because with -C, for example, it has lost
4395 privilege). It wasn't ignoring creation failures other than "already
4398 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4401 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4402 because an input line was too long, either on its own, or by virtue of
4403 too many continuations, the temporary file was not being removed, and the
4404 return code was incorrect.
4406 PH/31 Missing "BOOL" in function definition in filtertest.c.
4408 PH/32 Applied Sieve patches from the maintainer.
4410 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4412 PH/33 Added "verify = not_blind".
4414 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4415 Local/Makefile (with some defaults set). These are used in built scripts
4416 such as exicyclog, but they have never been used in the exim_install
4417 script (though there are many overriding facilities there). I have
4418 arranged that the exim_install script now takes note of these two
4421 PH/35 Installed configuration files for Dragonfly.
4423 PH/36 When a locally submitted message by a trusted user did not contain a
4424 From: header, and the sender address was obtained from -f or from an SMTP
4425 MAIL command, and the trusted user did not use -F to supply a sender
4426 name, $originator_name was incorrectly used when constructing a From:
4427 header. Furthermore, $originator_name was used for submission mode
4428 messages from external hosts without From: headers in a similar way,
4429 which is clearly wrong.
4431 PH/37 Added control=suppress_local_fixups.
4433 PH/38 When log_selector = +received_sender was set, and the addition of the
4434 sender made the log line's construction buffer exactly full, or one byte
4435 less than full, an overflow happened when the terminating "\n" was
4438 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4439 when the result of a list match is failure because a DNS lookup failed.
4441 PH/40 RM_COMMAND is now used in the building process.
4443 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4444 the "build-* directories that it finds.
4446 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4447 address in a domain literal was a prefix of an interface address.
4449 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4450 when verifying a sender address, unless rewrite_headers is false.
4452 PH/44 Wrote a long comment about why errors_to addresses are verified as
4453 recipients, not senders.
4455 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4456 the ratelimit ACL was added.
4458 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4460 PH/46 Added extra information about PostgreSQL errors to the error string.
4462 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4463 avoiding the need to install special zones in a real server. This is
4464 backwards compatible; if it can't find the fake resolver, it drops back.
4465 Thus, both old and new test suites can be run.
4467 TF/02 Added util/ratelimit.pl
4469 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4470 clock is set back in time.
4472 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4473 Candler <B.Candler@pobox.com>.
4475 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4476 OK for addresses that are the result of redirections.
4478 PH/48 A number of further additions for the benefit of the new test suite,
4479 including a fake gethostbyname() that interfaces to the fake DNS resolver
4482 TF/06 The fix for widen_domains has also been applied to qualify_single and
4483 search_parents which are the other dnslookup options that can cause
4486 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4489 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4491 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4492 service" if it encountered an MX record with an empty target hostname.
4493 The message is now "an MX or SRV record indicated no SMTP service".
4495 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4496 if the DNS of the sending site is misconfigured. This is quite a
4497 common situation. This change restores the behaviour of treating a
4498 helo verification defer as a failure.
4500 PH/51 If self=fail was set on a router, the bounce message did not include the
4501 actual error message.
4507 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4509 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4510 command itself ended in error, the underlying message about the transport
4511 filter timeout was being overwritten with the pipe command error. Now the
4512 underlying error message should be appended to the second error message.
4514 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4516 PH/02 Reduce the amount of output that "make" produces by default. Full output
4517 can still be requested.
4519 PH/03 The warning log line about a condition test deferring for a "warn" verb
4520 was being output only once per connection, rather than after each
4521 occurrence (because it was using the same function as for successful
4522 "warn" verbs). This seems wrong, so I have changed it.
4524 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4525 it should not have, which might have caused a crash in the right
4526 circumstances, but probably never did.
4528 PH/04 Installed a modified version of Tony Finch's patch to make submission
4529 mode fix the return path as well as the Sender: header line, and to
4530 add a /name= option so that you can make the user's friendly name appear
4533 TF/03 Added the control = fakedefer ACL modifier.
4535 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4536 Mark Lowes for thorough testing.
4538 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4540 TK/03 Merged latest SRS patch from Miles Wilton.
4542 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4543 with the definition in sysexits.h (which is #included earlier).
4544 Fortunately, Exim does not actually use EX_OK. The code used to try to
4545 preserve the sysexits.h value, by assuming that macro definitions were
4546 scanned for macro replacements. I have been disabused of this notion,
4547 so now the code just undefines EX_OK before #including unistd.h.
4549 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4550 in the smtp transport. When a block could not be written in a single
4551 write() function, the timeout was being re-applied to each part-write.
4552 This seems wrong - if the receiver was accepting one byte at a time it
4553 would take for ever. The timeout is now adjusted when this happens. It
4554 doesn't have to be particularly precise.
4556 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4557 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4559 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4560 without a domain if the check to <postmaster@domain> fails.
4562 SC/01 Eximstats: added -xls and the ability to specify output files
4563 (patch written by Frank Heydlauf).
4565 SC/02 Eximstats: use FileHandles for outputting results.
4567 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4569 SC/04 Eximstats: fixed display of large numbers with -nvr option
4571 SC/05 Eximstats: fixed merging of reports with empty tables.
4573 SC/06 Eximstats: added the -include_original_destination flag
4575 SC/07 Eximstats: removed tabs and trailing whitespace.
4577 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4579 TK/06 MBOX spool code: Add real "From " MBOX separator line
4580 so the .eml file is really in mbox format (even though
4581 most programs do not really care). Patch from Alex Miller.
4583 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4584 The latter is generated from $received_to and is only set if the
4585 message has one envelope recipient. SA can use these headers,
4586 obviously out-of-the-box. Patch from Alex Miller.
4588 PH/08 The ${def test on a variable was returning false if the variable's
4589 value was "0", contrary to what the specification has always said!
4590 The result should be true unless the variable is empty.
4592 PH/09 The syntax error of a character other than { following "${if
4593 def:variable_name" (after optional whitespace) was not being diagnosed.
4594 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4595 accidental colon was present, for example, could give incorrect results.
4597 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4598 result is used (not including appendfile, where other changes are about
4601 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4602 This involved changing a lot of size variables from int to off_t. It
4603 should work with maildirs and everything.
4605 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4606 spamd dying while we are connected to it.
4608 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4611 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4612 function for BDB 4.3.
4614 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4616 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4617 the sources for PCRE have been re-arranged and more files are now
4620 PH/15 The code I had for printing potentially long long variables in PH/11
4621 above was not the best (it lost precision). The length of off_t variables
4622 is now inspected at build time, and an appropriate printing format (%ld
4623 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4624 to be "long long int" or "long int". This is needed for the internal
4625 formatting function string_vformat().
4627 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4628 the configuration file to be ":syslog", then the script "guesses" where
4629 the logs files are, rather than using the compiled in default. In our
4630 case the guess is not the same as the compiled default, so the script
4631 suddenly stopped working when I started to use syslog. The patch checks
4632 to see if log_file_path is "". If so, it attempts to read it from exim
4633 with no configuration file to get the compiled in version, before it
4634 falls back to the previous guessing code."
4636 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4637 implementing BATV in an Exim configuration. See NewStuff for the gory
4640 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4641 Makefile that are specific to HP-UX.
4643 PH/18 If the "use_postmaster" option was set for a recipient callout together
4644 with the "random" option, the postmaster address was used as the MAIL
4645 FROM address for the random test, but not for the subsequent recipient
4646 test. It is now used for both.
4648 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4649 patch removes a few documentation additions to RFC 3028, because the
4650 latest draft now contains them. It adds the new en;ascii-case comparator
4651 and a new error check for 8bit text in MIME parts. Comparator and
4652 require names are now matched exactly. I enabled the subaddress
4653 extension, but it is not well tested yet (read: it works for me)."
4655 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4656 rework some of the code of TK/09 above to avoid the hardwired use of
4657 "%lld" and "long long". Replaced the call to snprintf() with a call to
4660 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4661 records point to non-existent hosts", "retry timeout exceeded", and
4662 "retry time not reached for any host after a long failure period".
4664 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4665 experimental DomainKeys support:
4667 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4668 (2) On an error such as an illegally used "control", the wrong name for
4669 the control was given.
4671 These problems did NOT occur unless DomainKeys support was compiled.
4673 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4675 PH/24 Added ${if match_ip condition.
4677 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4678 databases so that it will be absolutely obvious if a crash occurs in the
4679 DB library. This is a regular occurrence (often caused by mis-matched
4682 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4683 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4684 picked up on a user's system that detects such things. There doesn't seem
4685 to be a gcc warning option for this - only an attribute that has to be
4686 put on the function's prototype. It seems that in Fedora Core 4 they have
4687 set this on a number of new functions. No doubt there will be more in due
4690 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4691 specify a transport. However, if an address that was verified by such a
4692 router was the subject of a callout, Exim crashed because it tried to
4693 read the rcpt_include_affixes from the non-existent transport. Now it
4694 just assumes that the setting of that option is false. This bug was
4695 introduced by 4.51/PH/31.
4697 PH/28 Changed -d+all to exclude +memory, because that information is very
4698 rarely of interest, but it makes the output a lot bigger. People tend to
4699 do -d+all out of habit.
4701 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4702 code in os-type was giving problems when libc.so lives in lib64, like on
4705 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4706 aren't the modern standard, and it seems that some systems' include files
4707 don't always have them. Exim was already checking for some of the newer
4708 ones like T_AAAA, and defining it itself. I've added checks for all the
4709 record types that Exim uses.
4711 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4712 not automatically generating a new one, as it is supposed to. This
4713 prevented TLS from working. If the file did exist, but contained invalid
4714 data, a new version was generated, as expected. It was only the case of a
4715 non-existent file that was broken.
4717 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4718 with a change in libdomainkeys > 0.64.
4720 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4721 from DNS. If the selector record carries the flag, it now has
4722 precedence over the domain-wide flag.
4724 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4726 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4727 the use of a port name that isn't defined in /etc/services) occurred, the
4728 message was deferred as in a normal delivery, and thus remained on the
4729 spool, instead of being failed because of the mua_wrapper setting. This
4730 is now fixed, and I tidied up some of the mua_wrapper messages at the
4733 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4734 the messages in a hash of arrays rather than using individual hashes.
4735 This is a bit cleaner and results in dramatic memory savings, albeit
4736 at a slight CPU cost.
4738 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4739 as requested by Marc Sherman.
4741 SC/10 Eximstats: added histograms for user specified patterns as requested
4744 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4746 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4747 fopen() in the content-scanning modules that did not already have it.
4753 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4754 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4756 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4758 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4760 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4761 to test Sieve filters that use "vacation".
4763 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4764 that changes the way the GnuTLS parameters are stored in the cache file.
4765 The new format can be generated externally. For backward compatibility,
4766 if the data in the cache doesn't make sense, Exim assumes it has read an
4767 old-format file, and it generates new data and writes a new file. This
4768 means that you can't go back to an older release without removing the
4771 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4772 work if there are any delivery delays because "one_time" forces the
4773 parent to be marked "delivered", so its unseen clone is never tried
4774 again. For this reason, Exim now forbids the simultaneous setting of
4777 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4778 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4779 entry if you want to know the details. The fix, however, neglected to
4780 consider the case when local delivery batching is involved. The test for
4781 "previously delivered" was not happening when checking to see if an
4782 address could be batched with a previous (undelivered) one; under
4783 certain circumstances this could lead to multiple deliveries to the same
4786 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4787 in its include files, and this causes problems building Exim.
4789 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4790 header_syntax) but Exim was just ignoring anything given after a slash.
4791 In particular, this caused confusion with an attempt to use "verify =
4792 reverse_host_lookup/defer_ok". An error is now given when options are
4793 supplied for verify items that do not have them. (Maybe reverse_host_
4794 lookup should have a defer_ok option, but that's a different point.)
4796 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4797 defined by RFC 821) to 2048, because there were problems with some AUTH
4798 commands, and RFC 1869 says the size should be increased for extended
4799 SMTP commands that take arguments.
4801 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4804 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4805 "unknown" error; now it says that the functionality isn't in the binary.
4807 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4808 an address' error message when a string expansion fails (syntax or
4809 whatever). Otherwise the password may appear in the log. Following change
4810 PH/42 below, there is no longer a chance of it appearing in a bounce
4813 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4815 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4816 of its addresses was ever tried. (Bugzilla bug #2.)
4818 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4819 the result incorrectly in the debug output. (It correctly added a newline
4820 to what was transported.)
4822 TF/01 Added $received_time.
4824 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4825 commented out examples of how to interface to a virus scanner and to
4826 SpamAssassin. Also added commented examples of av_scanner and
4827 spamd_address settings.
4829 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4830 and controls are allowed in which ACLs. There were a couple of minor
4831 errors. Some of the entries in the conditions table (which is a table of
4832 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4833 negation of where the condition IS allowed.
4835 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4837 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4838 header file does not have a version number, so I've had to invent a new
4839 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4840 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4841 radiusclient), but it was contributed by a Radius user.
4843 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4844 files or directories, for OpenSSL.
4846 PH/19 When an Exim process that is running as root has to create an Exim log
4847 file, it does so in a subprocess that runs as exim:exim so as to get the
4848 ownership right at creation (otherwise, other Exim processes might see
4849 the file with the wrong ownership). There was no test for failure of this
4850 fork() call, which would lead to the process getting stuck as it waited
4851 for a non-existent subprocess. Forks do occasionally fail when resources
4852 run out. I reviewed all the other calls to fork(); they all seem to check
4855 PH/20 When checking for unexpected SMTP input at connect time (before writing
4856 the banner), Exim was not dealing correctly with a non-positive return
4857 from the read() function. If the client had disconnected by this time,
4858 the result was a log entry for a synchronization error with an empty
4859 string after "input=" when read() returned zero. If read() returned -1
4860 (an event I could not check), uninitialized data bytes were printed.
4861 There were reports of junk text (parts of files, etc) appearing after
4864 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4866 PH/22 Added support for macro redefinition, and (re)definition in between
4867 driver and ACL definitions.
4869 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4870 forgetting to use the resulting value; it was using the unexpanded value.
4872 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4873 hadn't been configured. The fix is from Juergen Kreileder, who
4874 understands it better than I do:
4876 "Here's what I see happening with three configured cyrus_sasl
4877 authenticators configured (plain, login, cram-md5):
4879 On startup auth_cyrus_sasl_init() gets called for each of these.
4880 This means three calls to sasl_listmech() without a specified mech_list.
4881 => SASL tests which mechs of all available mechs actually work
4882 => three warnings about OTP not working
4883 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4885 With the patch, sasl_listmech() also gets called three times. But now
4886 SASL's mech_list option is set to the server_mech specified in the the
4887 authenticator. Or in other words, the answer from sasl_listmech()
4888 gets limited to just the mech you're testing for (which is different
4890 => the return list contains just 'plain' or 'login', 'cram-md5' or
4891 nothing depending on the value of ob->server_mech.
4893 I've just tested the patch: Authentication still works fine,
4894 unavailable mechs specified in the exim configuration are still
4895 caught, and the auth.log warnings about OTP are gone."
4897 PH/25 When debugging is enabled, the contents of the command line are added
4898 to the debugging output, even when log_selector=+arguments is not
4901 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4902 answer is "GNU", and only if the return is "GNU/something" is the answer
4905 PH/27 $acl_verify_message is now set immediately after the failure of a
4906 verification in an ACL, and so is available in subsequent modifiers. In
4907 particular, the message can be preserved by coding like this:
4909 warn !verify = sender
4910 set acl_m0 = $acl_verify_message
4912 Previously, $acl_verify_message was set only while expanding "message"
4913 and "log_message" when a very denied access.
4915 PH/28 Modified OS/os.c-Linux with
4917 -#ifndef OS_LOAD_AVERAGE
4918 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4920 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4921 nomenclature these days.)
4923 PH/29 Installed patch from the Sieve maintainer that adds the options
4924 sieve_useraddress and sieve_subaddress to the redirect router.
4926 PH/30 In these circumstances:
4927 . Two addresses routed to the same list of hosts;
4928 . First host does not offer TLS;
4929 . First host accepts first address;
4930 . First host gives temporary error to second address;
4931 . Second host offers TLS and a TLS session is established;
4932 . Second host accepts second address.
4933 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4934 and peerdn, if requested) that were in fact used only for the second
4937 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4938 attention to any local part prefix or suffix that was matched by the
4939 router that accepted the address. It now behaves in the same way as it
4940 does for delivery: the affixes are removed from the local part unless
4941 rcpt_include_affixes is set on the transport.
4943 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4944 timeout during the DATA phase of an incoming message.
4946 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4947 applied a patch sanctioned by the Sieve maintainer.
4949 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4950 the uid or gid is negative. A case of a negative gid caused this to be
4951 noticed. The fix allows for either to be negative.
4953 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4954 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4957 PH/36 The change PH/12 above was broken. Fixed it.
4959 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4960 the grounds that routing the same address twice would always produce the
4961 same answer. This might have been true once, but it is certainly no
4962 longer true now. Routing a child address may depend on the previous
4963 routing that produced that child. Some complicated redirection strategies
4964 went wrong when messages had multiple recipients, and made Exim's
4965 behaviour dependent on the order in which the addresses were given.
4967 I have moved the duplicate checking until after the routing is complete.
4968 Exim scans the addresses that are assigned to local and remote
4969 transports, and removes any duplicates. This means that more work will be
4970 done, as duplicates will always all be routed, but duplicates are
4971 presumably rare, so I don't expect this is of any significance.
4973 For deliveries to pipes, files, and autoreplies, the duplicate checking
4974 still happens during the routing process, since they are not going to be
4977 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4978 It corrects a timeout issue with spamd. This is Ian's comment: "The
4979 background is that sometimes spamd either never reads data from a
4980 connection it has accepted, or it never writes response data. The exiscan
4981 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4982 blindly assumes that writes won't block so it may never time out."
4984 PH/39 Allow G after quota size as well as K and M.
4986 PH/40 The value set for $authenticated_id in an authenticator may not contain
4987 binary zeroes or newlines because the value is written to log lines and
4988 to spool files. There was no check on this. Now the value is run through
4989 the string_printing() function so that such characters are converted to
4990 printable escape sequences.
4992 PH/41 $message_linecount is a new variable that contains the total number of
4993 lines in the message. Compare $body_linecount, which is the count for the
4996 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4997 bounce and delay warning messages, except in certain special cases, which
5000 (a) An SMTP error message from a remote host;
5001 (b) A message specified in a :fail: redirection;
5002 (c) A message specified in a "fail" command in a system filter;
5003 (d) A message specified in a FAIL return from the queryprogram router;
5004 (e) A message specified by the cannot_route_message router option.
5006 In these cases only, Exim does include the error details in bounce and
5007 warning messages. There are also a few cases where bland messages such
5008 as "unrouteable address" or "local delivery error" are given.
5010 PH/43 $value is now also set for the "else" part of a ${run expansion.
5012 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5013 being worked on, but at least Exim now implements the latest version to
5016 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5017 process to complete was treated as a delivery failure, a timeout while
5018 writing the message to the pipe was logged, but erroneously treated as a
5019 successful delivery. Such timeouts include transport filter timeouts. For
5020 consistency with the overall process timeout, these timeouts are now
5021 treated as errors, giving rise to delivery failures by default. However,
5022 there is now a new Boolean option for the pipe transport called
5023 timeout_defer, which, if set TRUE, converts the failures into defers for
5024 both kinds of timeout. A transport filter timeout is now identified in
5027 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5028 systems where "make" and "gmake" are different, calling "gmake" at top
5029 level broke things. I've arranged for the value of $(MAKE) to be passed
5030 from the Makefile to this script so that it can call the same version of
5034 A note about Exim versions 4.44 and 4.50
5035 ----------------------------------------
5037 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5038 changes of various kinds. As a consequence, a big documentation update was
5039 needed. This delayed the release for rather longer than seemed good, especially
5040 in the light of a couple of (minor) security issues. Therefore, the changes
5041 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5042 release. So 4.44 and 4.50 are in effect two different branches that both start
5045 I have left the 4.50 change log unchanged; it contains all the changes since
5046 4.43. The change log for 4.44 is below; many of its items are identical to
5047 those for 4.50. This seems to be the most sensible way to preserve the
5048 historical information.
5054 1. Minor wording change to the doc/README.SIEVE file.
5056 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5057 computation of the current number of files was incorrect.
5059 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5060 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5063 4. Give more explanation in the error message when the command for a transport
5064 filter fails to execute.
5066 5. There are several places where Exim runs a non-Exim command in a
5067 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5068 was being done only for the command run by the queryprogram router. It is
5069 now done for all such subprocesses. The other cases are: ${run, transport
5070 filters, and the commands run by the lmtp and pipe transports.
5072 6. Added CONFIGURE_GROUP build-time option.
5074 7. Some older OS have a limit of 256 on the maximum number of file
5075 descriptors. Exim was using setrlimit() to set 1000 as a large value
5076 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5077 systems. I've change it so that if it can't get 1000, it tries for 256.
5079 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5080 was an oversight, and furthermore, ever since the addition of extra
5081 controls (e.g. 4.43/32), the checks on when to allow different forms of
5082 "control" were broken. There should now be diagnostics for all cases when a
5083 control that does not make sense is encountered.
5085 9. Added the /retain_sender option to "control=submission".
5087 10. $recipients is now available in the predata ACL (oversight).
5089 11. Tidy the search cache before the fork to do a delivery from a message
5090 received from the command line. Otherwise the child will trigger a lookup
5091 failure and thereby defer the delivery if it tries to use (for example) a
5092 cached ldap connection that the parent has called unbind on.
5094 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5095 of $address_data from the recipient verification was clobbered by the
5096 sender verification.
5098 13. The value of address_data from a sender verification is now available in
5099 $sender_address_data in subsequent conditions in the ACL statement.
5101 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5103 15. Added a new option "connect=<time>" to callout options, to set a different
5106 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5107 was its contents. (It was OK if the option was not defined at all.)
5109 17. A "Completed" log line is now written for messages that are removed from
5110 the spool by the -Mrm option.
5112 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5113 information about exactly what failed.
5115 19. Added -dd to debug only the daemon process.
5117 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5118 handles timeouts, both on the server side and network timeouts. Renamed the
5119 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5121 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5122 It is now set to "smtps".
5124 22. $host_address is now set to the target address during the checking of
5125 ignore_target_hosts.
5127 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5128 being passed; this would have caused $sender_host_name to have been used if
5129 matching the list had actually called for a host name (not very likely,
5130 since this list is usually IP addresses). A host name is now passed as
5133 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5134 code that specifies a non-restarting handler (typically sigaction() in
5135 modern systems) in an attempt to fix a rare and obscure crash bug.
5137 25. Narrowed the window for a race in the daemon that could cause it to ignore
5138 SIGCHLD signals. This is not a major problem, because they are used only to
5139 wake it up if nothing else does.
5141 26. A malformed maildirsize file could cause Exim to calculate negative values
5142 for the mailbox size or file count. Odd effects could occur as a result.
5143 The maildirsize information is now recalculated if the size or filecount
5146 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5147 support for a long time. Removed HAVE_SYS_VFS_H.
5149 28. Installed the latest version of exipick from John Jetmore.
5151 29. In an address list, if the pattern was not a regular expression, an empty
5152 subject address (from a bounce message) matched only if the pattern was an
5153 empty string. Non-empty patterns were not even tested. This was the wrong
5154 because it is perfectly reasonable to use an empty address as part of a
5155 database query. An empty address is now tested by patterns that are
5156 lookups. However, all the other forms of pattern expect the subject to
5157 contain a local part and a domain, and therefore, for them, an empty
5158 address still always fails if the pattern is not itself empty.
5160 30. Exim went into a mad DNS loop when attempting to do a callout where the
5161 host was specified on an smtp transport, and looking it up yielded more
5162 than one IP address.
5164 31. Re-factored the code for checking spool and log partition space into a
5165 function that finds that data and another that does the check. The former
5166 is then used to implement four new variables: $spool_space, $log_space,
5167 $spool_inodes, and $log_inodes.
5169 32. The RFC2047 encoding function was originally intended for short strings
5170 such as real names; it was not keeping to the 75-character limit for
5171 encoded words that the RFC imposes. It now respects the limit, and
5172 generates multiple encoded words if necessary. To be on the safe side, I
5173 have increased the buffer size for the ${rfc2047: expansion operator from
5176 33. It is now permitted to omit both strings after an "if" condition; if the
5177 condition is true, the result is "true". As before, when the second string
5178 is omitted, a false condition yields an empty string. This makes it less
5179 cumbersome to write custom ACL and router conditions.
5181 34. Failure to deliver a bounce message always caused it to be frozen, even if
5182 there was an errors_to setting on the router. The errors_to setting is now
5185 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5186 canonical form (fully expanded) before being placed in
5187 $sender_host_address.
5189 36. The table in the code that translates DNS record types into text (T_A to
5190 "A" for instance) was missing entries for NS and CNAME. It is just possible
5191 that this could have caused confusion if both these types were looked up
5192 for the same domain, because the text type is used as part of Exim's
5193 per-process caching. But the chance of anyone hitting this buglet seems
5196 37. The dnsdb lookup has been extended in a number of ways.
5198 (1) There is a new type, "zns", which walks up the domain tree until it
5199 finds some nameserver records. It should be used with care.
5201 (2) There is a new type, "mxh", which is like "mx" except that it returns
5202 just the host names, not the priorities.
5204 (3) It is now possible to give a list of domains (or IP addresses) to be
5205 looked up. The behaviour when one of the lookups defers can be
5206 controlled by a keyword.
5208 (4) It is now possible to specify the separator character for use when
5209 multiple records are returned.
5211 38. The dnslists ACL condition has been extended: it is now possible to supply
5212 a list of IP addresses and/or domains to be looked up in a particular DNS
5215 39. Added log_selector=+queue_time_overall.
5217 40. When running the queue in the test harness, wait just a tad after forking a
5218 delivery process, to get repeatability of debugging output.
5220 41. Include certificate and key file names in error message when GnuTLS fails
5221 to set them up, because the GnuTLS error message doesn't include the name
5222 of the failing file when there is a problem reading it.
5224 42. Allow both -bf and -bF in the same test run.
5226 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5228 44. The "Exiscan patch" is now merged into the mainline Exim source.
5230 45. Sometimes the final signoff response after QUIT could fail to get
5231 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5232 < 0 before doing a fflush(). This bug looks as though it goes back to the
5233 introduction of TLS in release 3.20, but "sometimes" must have been rare
5234 because the tests only now provoked it.
5236 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5237 (this can affect the format of dates).
5239 47. exim_tidydb, when checking for the continued existence of a message for
5240 which it has found a message-specific retry record, was not finding
5241 messages that were in split spool directories. Consequently, it was
5242 deleting retry records that should have stayed in existence.
5244 48. Steve fixed some bugs in eximstats.
5246 49. The SPA authentication driver was not abandoning authentication and moving
5247 on to the next authenticator when an expansion was forced to fail,
5248 contradicting the general specification for all authenticators. Instead it
5249 was generating a temporary error. It now behaves as specified.
5251 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5252 (the order specifies the preference for clients). The order is now AES256,
5253 AES128, 3DES, ARCFOUR128.
5255 51. Small patch to Sieve code - explicitly set From: when generating an
5258 52. Exim crashed if a remote delivery caused a very long error message to be
5259 recorded - for instance if somebody sent an entire SpamAssassin report back
5260 as a large number of 550 error lines. This bug was coincidentally fixed by
5261 increasing the size of one of Exim's internal buffers (big_buffer) that
5262 happened as part of the Exiscan merge. However, to be on the safe side, I
5263 have made the code more robust (and fixed the comments that describe what
5266 53. Now that there can be additional text after "Completed" in log lines (if
5267 the queue_time_overall log selector is set), a one-byte patch to exigrep
5268 was needed to allow it to recognize "Completed" as not the last thing in
5271 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5272 patch that reportedly fixes this has been added. I am not expert enough to
5273 create a test for it. This is what the patch creator wrote:
5275 "I found a little strange behaviour of ldap code when working with
5276 Windows 2003 AD Domain, where users was placed in more than one
5277 Organization Units. When I tried to give exim partial DN, the exit code
5278 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5279 But simultaneously result of request was absolutely normal ldap result,
5280 so I produce this patch..."
5282 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5283 REFERENCE, so I have modified the code to exclude the patch when that macro
5286 55. Some experimental protocols are using DNS PTR records for new purposes. The
5287 keys for these records are domain names, not reversed IP addresses. The
5288 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5289 leaves it alone. Component reversal etc. now happens only for IP addresses.
5292 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5294 57. Double the size of the debug message buffer (to 2048) so that more of very
5295 long debug lines gets shown.
5297 58. The exicyclog utility now does better if the number of log files to keep
5298 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5300 59. Two changes related to the smtp_active_hostname option:
5302 (1) $smtp_active_hostname is now available as a variable.
5303 (2) The default for smtp_banner uses $smtp_active_hostname instead
5304 of $primary_hostname.
5306 60. The host_aton() function is supposed to be passed a string that is known
5307 to be a valid IP address. However, in the case of IPv6 addresses, it was
5308 not checking this. This is a hostage to fortune. Exim now panics and dies
5309 if the condition is not met. A case was found where this could be provoked
5310 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5311 components; fortuitously, this particular loophole had already been fixed
5312 by change 4.50/55 above.
5314 If there are any other similar loopholes, the new check in host_aton()
5315 itself should stop them being exploited. The report I received stated that
5316 data on the command line could provoke the exploit when Exim was running as
5317 exim, but did not say which command line option was involved. All I could
5318 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5319 running as the user.
5322 61. There was a buffer overflow vulnerability in the SPA authentication code
5323 (which came originally from the Samba project). I have added a test to the
5324 spa_base64_to_bits() function which I hope fixes it.
5327 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5328 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5330 63. The daemon start-up calls getloadavg() while still root for those OS that
5331 need the first call to be done as root, but it missed one case: when
5332 deliver_queue_load_max is set with deliver_drop_privilege. This is
5333 necessary for the benefit of the queue runner, because there is no re-exec
5334 when deliver_drop_privilege is set.
5336 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5337 This has been fixed.
5339 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5340 was in use, was not putting the data itself into the right store pool;
5341 consequently, it could be overwritten for a subsequent message in the same
5342 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5345 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5347 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5348 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5349 the function were treating the return as a boolean value, which happened to
5350 work because 0=false and not-0=true, but is not correct code.
5352 68. The host_aton() function was not handling scoped IPv6 addresses (those
5353 with, for example, "%eth0" on the end) correctly.
5355 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5356 negated items (that is, ~something) in unsigned ints. Some compilers
5357 apparently mutter when there is no cast.
5359 70. If an address verification called from an ACL failed, and did not produce a
5360 user-specific message (i.e. there was only a "system" message), nothing was
5361 put in $acl_verify_message. In this situation, it now puts the system
5364 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5365 session; change 4.31/43 added the unwanted input to the log line - except
5366 that it did not do this in the start of session case. It now does.
5368 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5369 This is wrong and can cause the other end to generate a synchronization
5370 error if it is another Exim or anything else that does the synchronization
5371 check. A QUIT command is no longer sent after a timeout.
5373 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5374 during host lookups.
5376 74. The defer_ok option of callout verification was not working if it was used
5377 when verifying addresses in header lines, that is, for this case:
5379 verify = header_sender/callout=defer_ok
5381 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5382 those file descriptors could be used for SMTP connections. If anything
5383 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5384 could be sent to the SMTP client, causing chaos. The daemon now opens
5385 stdin, stdout, and stderr to /dev/null when it puts itself into the
5388 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5389 log by default. The user can override this with suitable Perl magic.
5391 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5392 the log message when discard triggers, was not working for the DATA ACL or
5393 for the non-SMTP ACL.
5395 78. Error message wording change in sieve.c.
5397 79. If smtp_accept_max_per_host was set, the number of connections could be
5398 restricted to fewer than expected, because the daemon was trying to set up
5399 a new connection before checking whether the processes handling previous
5400 connections had finished. The check for completed processes is now done
5401 earlier. On busy systems, this bug wouldn't be noticed because something
5402 else would have woken the daemon, and it would have reaped the completed
5405 80. If a message was submitted locally by a user whose login name contained one
5406 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5407 It caused a spool format error. I have fixed the spool reading code. A
5408 related problem was that the "from" clause in the Received: line became
5409 illegal because of the space(s). It is now covered by ${quote_local_part.
5411 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5414 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5416 83. Patch from Sieve maintainer to fix unterminated string problem in
5417 "vacation" handling.
5419 84. Some minor changes to the Linux configuration files to help with other
5420 OS variants using glibc.
5422 85. One more patch for Sieve to update vacation handling to latest spec.
5425 ----------------------------------------------------
5426 See the note above about the 4.44 and 4.50 releases.
5427 ----------------------------------------------------
5433 1. Change 4.43/35 introduced a bug that caused file counts to be
5434 incorrectly computed when quota_filecount was set in an appendfile
5437 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5438 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5441 3. Give more explanation in the error message when the command for a transport
5442 filter fails to execute.
5444 4. There are several places where Exim runs a non-Exim command in a
5445 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5446 was being done only for the command run by the queryprogram router. It is
5447 now done for all such subprocesses. The other cases are: ${run, transport
5448 filters, and the commands run by the lmtp and pipe transports.
5450 5. Some older OS have a limit of 256 on the maximum number of file
5451 descriptors. Exim was using setrlimit() to set 1000 as a large value
5452 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5453 systems. I've change it so that if it can't get 1000, it tries for 256.
5455 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5456 was an oversight, and furthermore, ever since the addition of extra
5457 controls (e.g. 4.43/32), the checks on when to allow different forms of
5458 "control" were broken. There should now be diagnostics for all cases when a
5459 control that does not make sense is encountered.
5461 7. $recipients is now available in the predata ACL (oversight).
5463 8. Tidy the search cache before the fork to do a delivery from a message
5464 received from the command line. Otherwise the child will trigger a lookup
5465 failure and thereby defer the delivery if it tries to use (for example) a
5466 cached ldap connection that the parent has called unbind on.
5468 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5469 of $address_data from the recipient verification was clobbered by the
5470 sender verification.
5472 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5473 was its contents. (It was OK if the option was not defined at all.)
5475 11. A "Completed" log line is now written for messages that are removed from
5476 the spool by the -Mrm option.
5478 12. $host_address is now set to the target address during the checking of
5479 ignore_target_hosts.
5481 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5482 being passed; this would have caused $sender_host_name to have been used if
5483 matching the list had actually called for a host name (not very likely,
5484 since this list is usually IP addresses). A host name is now passed as
5487 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5488 code that specifies a non-restarting handler (typically sigaction() in
5489 modern systems) in an attempt to fix a rare and obscure crash bug.
5491 15. Narrowed the window for a race in the daemon that could cause it to ignore
5492 SIGCHLD signals. This is not a major problem, because they are used only to
5493 wake it up if nothing else does.
5495 16. A malformed maildirsize file could cause Exim to calculate negative values
5496 for the mailbox size or file count. Odd effects could occur as a result.
5497 The maildirsize information is now recalculated if the size or filecount
5500 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5501 support for a long time. Removed HAVE_SYS_VFS_H.
5503 18. Updated exipick to current release from John Jetmore.
5505 19. Allow an empty sender to be matched against a lookup in an address list.
5506 Previously the only cases considered were a regular expression, or an
5509 20. Exim went into a mad DNS lookup loop when doing a callout where the
5510 host was specified on the transport, if the DNS lookup yielded more than
5513 21. The RFC2047 encoding function was originally intended for short strings
5514 such as real names; it was not keeping to the 75-character limit for
5515 encoded words that the RFC imposes. It now respects the limit, and
5516 generates multiple encoded words if necessary. To be on the safe side, I
5517 have increased the buffer size for the ${rfc2047: expansion operator from
5520 22. Failure to deliver a bounce message always caused it to be frozen, even if
5521 there was an errors_to setting on the router. The errors_to setting is now
5524 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5525 canonical form (fully expanded) before being placed in
5526 $sender_host_address.
5528 24. Updated eximstats to version 1.33
5530 25. Include certificate and key file names in error message when GnuTLS fails
5531 to set them up, because the GnuTLS error message doesn't include the name
5532 of the failing file when there is a problem reading it.
5534 26. Expand error message when OpenSSL has problems setting up cert/key files.
5537 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5538 (this can affect the format of dates).
5540 28. exim_tidydb, when checking for the continued existence of a message for
5541 which it has found a message-specific retry record, was not finding
5542 messages that were in split spool directories. Consequently, it was
5543 deleting retry records that should have stayed in existence.
5545 29. eximstats updated to version 1.35
5546 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5547 1.35 - bugfix such that pie charts by volume are generated correctly
5549 30. The SPA authentication driver was not abandoning authentication and moving
5550 on to the next authenticator when an expansion was forced to fail,
5551 contradicting the general specification for all authenticators. Instead it
5552 was generating a temporary error. It now behaves as specified.
5554 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5555 (the order specifies the preference for clients). The order is now AES256,
5556 AES128, 3DES, ARCFOUR128.
5558 31. Small patch to Sieve code - explicitly set From: when generating an
5561 32. Exim crashed if a remote delivery caused a very long error message to be
5562 recorded - for instance if somebody sent an entire SpamAssassin report back
5563 as a large number of 550 error lines. This bug was coincidentally fixed by
5564 increasing the size of one of Exim's internal buffers (big_buffer) that
5565 happened as part of the Exiscan merge. However, to be on the safe side, I
5566 have made the code more robust (and fixed the comments that describe what
5569 33. Some experimental protocols are using DNS PTR records for new purposes. The
5570 keys for these records are domain names, not reversed IP addresses. The
5571 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5572 leaves it alone. Component reversal etc. now happens only for IP addresses.
5575 34. The host_aton() function is supposed to be passed a string that is known
5576 to be a valid IP address. However, in the case of IPv6 addresses, it was
5577 not checking this. This is a hostage to fortune. Exim now panics and dies
5578 if the condition is not met. A case was found where this could be provoked
5579 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5580 components; fortuitously, this particular loophole had already been fixed
5581 by change 4.50/55 or 4.44/33 above.
5583 If there are any other similar loopholes, the new check in host_aton()
5584 itself should stop them being exploited. The report I received stated that
5585 data on the command line could provoke the exploit when Exim was running as
5586 exim, but did not say which command line option was involved. All I could
5587 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5588 running as the user.
5591 35. There was a buffer overflow vulnerability in the SPA authentication code
5592 (which came originally from the Samba project). I have added a test to the
5593 spa_base64_to_bits() function which I hope fixes it.
5596 36. The daemon start-up calls getloadavg() while still root for those OS that
5597 need the first call to be done as root, but it missed one case: when
5598 deliver_queue_load_max is set with deliver_drop_privilege. This is
5599 necessary for the benefit of the queue runner, because there is no re-exec
5600 when deliver_drop_privilege is set.
5602 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5603 was in use, was not putting the data itself into the right store pool;
5604 consequently, it could be overwritten for a subsequent message in the same
5605 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5608 38. Sometimes the final signoff response after QUIT could fail to get
5609 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5610 < 0 before doing a fflush(). This bug looks as though it goes back to the
5611 introduction of TLS in release 3.20, but "sometimes" must have been rare
5612 because the tests only now provoked it.
5618 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5619 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5620 Now it returns an integer. A number of calls were still expecting a T/F
5621 return. Fortuitously, in all cases, the tests worked in OK situations,
5622 which is the norm. However, things would have gone wrong on any write
5623 failures on the smtp file descriptor. This function is used when sending
5624 messages over SMTP and also when doing verify callouts.
5626 2. When Exim is called to do synchronous delivery of a locally submitted
5627 message (the -odf or -odi options), it no longer closes stderr before doing
5630 3. Implemented the mua_wrapper option.
5632 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5634 5. Implemented the functions header_remove(), header_testname(),
5635 header_add_at_position(), and receive_remove_recipient(), and exported them
5638 6. If an ACL "warn" statement specified the addition of headers, Exim already
5639 inserted X-ACL-Warn: at the start if there was no header name. However, it
5640 was not making this test for the second and subsequent header lines if
5641 there were newlines in the string. This meant that an invalid header could
5642 be inserted if Exim was badly configured.
5644 7. Allow an ACL "warn" statement to add header lines at the start or after all
5645 the Received: headers, as well as at the end.
5647 8. Added the rcpt_4xx retry error code.
5649 9. Added postmaster_mailfrom=xxx to callout verification option.
5651 10. Added mailfrom=xxxx to the callout verification option, for verify=
5654 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5655 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5657 12. Inserted some casts to stop certain compilers warning when using pointer
5658 differences as field lengths or precisions in printf-type calls (mostly
5659 affecting debugging statements).
5661 13. Added optional readline() support for -be (dynamically loaded).
5663 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5664 same clock tick as a message's arrival, so that its received time was the
5665 same as the "first fail" time on the retry record, and that message
5666 remained on the queue past the ultimate address timeout, every queue runner
5667 would try a delivery (because it was past the ultimate address timeout) but
5668 after another failure, the ultimate address timeout, which should have then
5669 bounced the address, did not kick in. This was a "< instead of <=" error;
5670 in most cases the first failure would have been in the next clock tick
5671 after the received time, and all would be well.
5673 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5674 being recognized when the domain list was tested by the match_domain
5675 condition in an expansion string.
5677 16. Added the ${str2b64: operator.
5679 17. Exim was always calling setrlimit() to set a large limit for the number of
5680 processes, without checking whether the existing limit was already
5681 adequate. (It did check for the limit on file descriptors.) Furthermore,
5682 errors from getrlimit() and setrlimit() were being ignored. Now they are
5683 logged to the main and panic logs, but Exim does carry on, to try to do its
5684 job under whatever limits there are.
5686 18. Imported PCRE 5.0.
5688 19. Trivial typo in log message " temporarily refused connection" (the leading
5691 20. If the log selector return_path_on_delivery was set and an address was
5692 redirected to /dev/null, the delivery process crashed because it assumed
5693 that a return path would always be set for a "successful" delivery. In this
5694 case, the whole delivery is bypassed as an optimization, and therefore no
5697 21. Internal re-arrangement: the function for sending a challenge and reading
5698 a response while authentication was assuming a zero-terminated challenge
5699 string. It's now changed to take a pointer and a length, to allow for
5700 binary data in such strings.
5702 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5704 23. Exim was not respecting finduser_retries when seeking the login of the
5705 uid under which it was called; it was always trying 10 times. (The default
5706 setting of finduser_retries is zero.) Also, it was sleeping after the final
5707 failure, which is pointless.
5709 24. Implemented tls_on_connect_ports.
5711 25. Implemented acl_smtp_predata.
5713 26. If the domain in control=submission is set empty, Exim assumes that the
5714 authenticated id is a complete email address when it generates From: or
5715 Sender: header lines.
5717 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5718 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5719 chown and chgrp in /bin and hostname in /usr/bin.
5721 28. Exim was keeping the "process log" file open after each use, just as it
5722 does for the main log. This opens the possibility of it remaining open for
5723 long periods when the USR1 signal hits a daemon. Occasional processlog
5724 errors were reported, that could have been caused by this. Anyway, it seems
5725 much more sensible not to leave this file open at all, so that is what now
5728 29. The long-running daemon process does not normally write to the log once it
5729 has entered its main loop, and it closes the log before doing so. This is
5730 so that log files can straightforwardly be renamed and moved. However,
5731 there are a couple of unusual error situations where the daemon does write
5732 log entries, and I had neglected to close the log afterwards.
5734 30. The text of an SMTP error response that was received during a remote
5735 delivery was being truncated at 512 bytes. This is too short for some of
5736 the long messages that one sometimes sees. I've increased the limit to
5739 31. It is now possible to make retry rules that apply only when a message has a
5740 specific sender, in particular, an empty sender.
5742 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5743 it possible to be selective about when SMTP synchronization is enforced.
5745 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5747 32. Implemented hosts_connection_nolog.
5749 33. Added an ACL for QUIT.
5751 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5754 35. Added mailbox_size and mailbox_filecount to appendfile.
5756 36. Added control = no_multiline_responses to ACLs.
5758 37. There was a bug in the logic of the code that waits for the clock to tick
5759 in the case where the clock went backwards by a substantial amount such
5760 that the microsecond fraction of "now" was more than the microsecond
5761 fraction of "then" (but the whole seconds number was less).
5763 38. Added support for the libradius Radius client library this is found on
5764 FreeBSD (previously only the radiusclient library was supported).
5770 1. When certain lookups returned multiple values in the form name=value, the
5771 quoting of the values was not always being done properly. Specifically:
5772 (a) If the value started with a double quote, but contained no whitespace,
5774 (b) If the value contained whitespace other than a space character (i.e.
5775 tabs or newlines or carriage returns) it was not quoted.
5776 This fix has been applied to the mysql and pgsql lookups by writing a
5777 separate quoting function and calling it from the lookup code. The fix
5778 should probably also be applied to nisplus, ibase and oracle lookups, but
5779 since I cannot test any of those, I have not disturbed their existing code.
5781 2. A hit in the callout cache for a specific address caused a log line with no
5782 reason for rejecting RCPT. Now it says "Previous (cached) callout
5783 verification failure".
5785 3. There was an off-by-one bug in the queryprogram router. An over-long
5786 return line was truncated at 256 instead of 255 characters, thereby
5787 overflowing its buffer with the terminating zero. As well as fixing this, I
5788 have increased the buffer size to 1024 (and made a note to document this).
5790 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5791 when Exim is waiting for an SMTP response from a remote server, Exim
5792 restarts its select() call on the socket, thereby resetting its timeout.
5793 This is not a problem when such interrupts are rare. Somebody set up a cron
5794 job to run exiwhat every 2 minutes, which is less than the normal select()
5795 timeout (5 or 10 minutes). This meant that the select() timeout never
5796 kicked in because it was always reset. I have fixed this by comparing the
5797 time when an interrupt arrives with the time at the start of the first call
5798 to select(). If more time than the timeout has elapsed, the interrupt is
5799 treated as a timeout.
5801 5. Some internal re-factoring in preparation for the addition of Sieve
5802 extensions (by MH). In particular, the "personal" test is moved to a
5803 separate function, and given an option for scanning Cc: and Bcc: (which is
5804 not set for Exim filters).
5806 6. When Exim created an email address using the login of the caller as the
5807 local part (e.g. when creating a From: or Sender: header line), it was not
5808 quoting the local part when it contained special characters such as @.
5810 7. Installed new OpenBSD configuration files.
5812 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5813 try to make them clearer.
5815 9. Callout options, other than the timeout value, were being ignored when
5816 verifying sender addresses in header lines. For example, when using
5818 verify = header_sender/callout=no_cache
5820 the cache was (incorrectly) being used.
5822 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5823 only the Cygwin environment.
5825 11. When return_path_on_delivery was set as a log selector, if different remote
5826 addresses in the same message used different return paths and parallel
5827 remote delivery occurred, the wrong values would sometimes be logged.
5828 (Whenever a remote delivery process finished, the return path value from
5829 the most recently started remote delivery process was logged.)
5831 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5832 lines when AUTH and/or TLS are in use. This is the "received protocol"
5833 field. Exim used to use "asmtp" for authenticated SMTP, without any
5834 indication (in the protocol name) for TLS use. Now it follows the RFC and
5835 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5836 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5837 names appear in log lines as well as in Received: header lines.
5839 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5840 extensions, and comparison tests, and to fix some bugs.
5842 14. Changes to the "personal" filter test:
5844 (1) The test was buggy in that it was just doing the equivalent of
5845 "contains" tests on header lines. For example, if a user's address was
5846 anne@some.where, the "personal" test would incorrectly be true for
5848 To: susanne@some.where
5850 This test is now done by extracting each address from the header in turn,
5851 and checking the entire address. Other tests that are part of "personal"
5852 are now done using regular expressions (for example, to check local parts
5853 of addresses in From: header lines).
5855 (2) The list of non-personal local parts in From: addresses has been
5856 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5857 taken from the Sieve specification recommendations.
5859 (3) If the message contains any header line starting with "List-" it is
5860 treated as non-personal.
5862 (4) The test for "circular" in the Subject: header line has been removed
5863 because it now seems ill-conceived.
5865 15. Minor typos in src/EDITME comments corrected.
5867 16. Installed latest exipick from John Jetmore.
5869 17. If headers_add on a router specified a text string that was too long for
5870 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5871 of string_sprintf() is now avoided.
5873 18. $message_body_size was not set (it was always zero) when running the DATA
5874 ACL and the local_scan() function.
5876 19. For the "mail" command in an Exim filter, no default was being set for
5877 the once_repeat time, causing a random time value to be used if "once" was
5878 specified. (If the value happened to be <= 0, no repeat happened.) The
5879 default is now 0s, meaning "never repeat". The "vacation" command was OK
5880 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5881 (I found it when inspecting the code).
5883 20. There is now an overall timeout for performing a callout verification. It
5884 defaults to 4 times the callout timeout, which applies to individual SMTP
5885 commands during the callout. The overall timeout applies when there is more
5886 than one host that can be tried. The timeout is checked before trying the
5887 next host. This prevents very long delays if there are a large number of
5888 hosts and all are timing out (e.g. when the network connections are timing
5889 out). The value of the overall timeout can be changed by specifying an
5890 additional sub-option for "callout", called "maxwait". For example:
5892 verify = sender/callout=5s,maxwait=20s
5894 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5895 to the end before writing, but this should make it even safer).
5897 22. Exim was forgetting that it had advertised PIPELINING for the second and
5898 subsequent messages on an SMTP connection. It was also not resetting its
5899 memory on STARTTLS and an internal HELO.
5901 23. When Exim logs an SMTP synchronization error within a session, it now
5902 records whether PIPELINING has been advertised or not.
5904 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5905 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5906 rather than long int.
5908 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5910 26. Added the never_mail option to autoreply.
5916 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5917 crash if the getsockname() call failed; this can happen if a connection is
5918 closed very soon after it is established. The problem was simply in the
5919 order in which certain operations were done, causing Exim to try to write
5920 to the SMTP stream before it had set up the file descriptor. The bug has
5921 been fixed by making things happen in the correct order.
5927 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5928 before the connection was closed, thus losing the rejection response.
5930 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5931 some early Solaris releases, but causes trouble in current releases where
5932 socklen_t is defined.
5934 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5937 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5940 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5941 that releasing the top of it at the end releases what was used for sub-
5942 expansions (unless the block got too big). However, discard this block if
5943 the first thing is a variable or header, so that we can use its block when
5944 it is dynamic (useful for very large $message_headers, for example).
5946 6. Lookups now cache *every* query, not just the most recent. A new, separate
5947 store pool is used for this. It can be recovered when all lookup caches are
5948 flushed. Lookups now release memory at the end of their result strings.
5949 This has involved some general refactoring of the lookup sources.
5951 7. Some code has been added to the store_xxx() functions to reduce the amount
5952 of flapping under certain conditions.
5954 8. log_incoming_interface used to affect only the <= reception log lines. Now
5955 it causes the local interface and port to be added to several more SMTP log
5956 lines, for example "SMTP connection from", and rejection lines.
5958 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5960 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5962 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5963 could be overwritten at the end of the current message (or the start of a
5964 new message if it was set in a HELO ACL). The value is now preserved for
5965 the duration of the SMTP connection.
5967 12. If a transport had a headers_rewrite setting, and a matching header line
5968 contained an unqualified address, that address was qualified, even if it
5969 did not match any rewriting rules. The underlying bug was that the values
5970 of the flags that permit the existence of unqualified sender and recipient
5971 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5972 non-local messages, and by -bnq for local messages) were not being
5973 preserved with the message after it was received.
5975 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5976 "next input=" as part of the text comprising the host identity instead of
5977 the correct text. The code was using the same buffer for two different
5978 strings. However, depending on which order the printing function evaluated
5979 its arguments, the bug did not always show up. Under Linux, for example, my
5980 test suite worked just fine.
5982 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5983 doesn't work with some older versions of Perl. It has been changed to "my",
5984 which in any case is probably the better facility to use.
5986 15. A really picky compiler found some instances of statements for creating
5987 error messages that either had too many or two few arguments for the format
5990 16. The size of the buffer for calls to the DNS resolver has been increased
5991 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5992 for addresses that have a lot of PTR records. This alleviates a problem; it
5993 does not fully solve it.
5995 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5996 buffer now truncates the list and logs the incident, which is the same
5997 action as happens when Exim is looking up a host name and its aliases.
5998 Previously in this situation something unpredictable would happen;
5999 sometimes it was "internal error: store_reset failed".
6001 18. If a server dropped the connection unexpectedly when an Exim client was
6002 using GnuTLS and trying to read a response, the client delivery process
6003 crashed while trying to generate an error log message.
6005 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6006 string, for example:
6008 warn message = H1: something\nH2: something
6010 the text was added as a single header line from Exim's point of view
6011 though it ended up OK in the delivered message. However, searching for the
6012 second and subsequent header lines using $h_h2: did not work. This has been
6013 fixed. Similarly, if a system filter added multiple headers in this way,
6014 the routers could not see them.
6016 20. Expanded the error message when iplsearch is called with an invalid key to
6017 suggest using net-iplsearch in a host list.
6019 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6020 ACLs are no longer actually imposed (and a message to that effect is
6023 22. If a "gecos" field in a passwd entry contained escaped characters, in
6024 particular, if it contained a \" sequence, Exim got it wrong when building
6025 a From: or a Sender: header from that name. A second bug also caused
6026 incorrect handling when an unquoted " was present following a character
6027 that needed quoting.
6029 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6030 was not being matched caselessly.
6032 24. Arranged for all hyphens in the exim.8 source to be escaped with
6035 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6036 properly. Recipient callout cache records were still being keyed to include
6037 the sender, even when use_sender was set false. This led to far more
6038 callouts that were necessary. The sender is no longer included in the key
6039 when use_sender is false.
6041 26. Added "control = submission" modifier to ACLs.
6043 27. Added the ${base62d: operator to decode base 62 numbers.
6045 28. dnsdb lookups can now access SRV records.
6047 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6048 the configuration file.
6050 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6051 (-v) mode. This makes the output for a verbose queue run more intelligible.
6053 31. Added a use_postmaster feature to recipient callouts.
6055 32. Added the $body_zerocount variable, containing the number of binary zero
6056 bytes in the message body.
6058 33. The time of last modification of the "new" subdirectory is now used as the
6059 "mailbox time last read" when there is a quota error for a maildir
6062 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6064 35. Added +ignore_unknown as a special item in host lists.
6066 36. Code for decoding IPv6 addresses in host lists is now included, even if
6067 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6068 address was recognized as an IP address, but was then not correctly decoded
6069 into binary, causing unexpected and incorrect effects when compared with
6076 1. Very minor rewording of debugging text in manualroute to say "list of
6077 hosts" instead of "hostlist".
6079 2. If verify=header_syntax was set, and a header line with an unqualified
6080 address (no domain) and a large number of spaces between the end of the
6081 name and the colon was received, the reception process suffered a buffer
6082 overflow, and (when I tested it) crashed. This was caused by some obsolete
6083 code that should have been removed. The fix is to remove it!
6085 3. When running in the test harness, delay a bit after writing a bounce
6086 message to get a bit more predictability in the log output.
6088 4. Added a call to search_tidyup() just before forking a reception process. In
6089 theory, someone could use a lookup in the expansion of smtp_accept_max_
6090 per_host which, without the tidyup, could leave open a database connection.
6092 5. Added the variables $recipient_data and $sender_data which get set from a
6093 lookup success in an ACL "recipients" or "senders" condition, or a router
6094 "senders" option, similar to $domain_data and $local_part_data.
6096 6. Moved the writing of debug_print from before to after the "senders" test
6099 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6100 problems for message scanning, either using a data ACL, or using
6101 local_scan() because the Received: header was not generated till after they
6102 were called (in order to set the time as the time of reception completion).
6103 I have revised the way this works. The header is now generated after the
6104 body is received, but before the ACL or local_scan() are called. After they
6105 are run, the timestamp in the header is updated.
6111 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6112 before starting a queue runner without re-exec. This happened only when
6113 deliver_drop_privilege was set or when the Exim user was set to root. The
6114 effect of the bug was that timeouts during subsequent deliveries caused
6115 crashes instead of being properly handled. The handler is now left at its
6116 default (and expected) setting.
6118 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6119 message, again when deliver_drop_privilege is set or Exim is run as root.
6120 The bug described in (1) was not present in this case, but the tidying up
6121 of the other signals was missing. I have made the two cases consistent.
6123 3. The ignore_target_hosts setting on a manualroute router was being ignored
6124 for hosts that were looked up using the /MX notation.
6126 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6129 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6130 operated on the sender address. After changing the $sender_address to <>
6131 for the sender address verify, Exim was re-instated it as the original
6132 (before rewriting) address, but remembering that it had rewritten it, so it
6133 wasn't rewriting it again. This bug also had the effect of breaking the
6134 sender address verification caching when the sender address was rewritten.
6136 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6137 This has been changed so that if the ip literal address matches
6138 ignore_target_hosts, the router declines.
6140 7. Added expansion conditions match_domain, match_address, and match_local_
6141 part (NOT match_host).
6143 8. The placeholder for the Received: header didn't have a length field set.
6145 9. Added code to Exim itself and to exim_lock to test for a specific race
6146 condition that could lead to file corruption when using MBX delivery. The
6147 issue is with the lockfile that is created in /tmp. If this file is removed
6148 after a process has opened it but before that process has acquired a lock,
6149 there is the potential for a second process to recreate the file and also
6150 acquire a lock. This could lead to two Exim processes writing to the file
6151 at the same time. The added code performs the same test as UW imapd; it
6152 checks after acquiring the lock that its file descriptor still refers to
6153 the same named file.
6155 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6156 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6159 11. Added the smtp_active_hostname option. If used, this will typically be made
6160 to depend on the incoming interface address. Because $interface_address is
6161 not set up until the daemon has forked a reception process, error responses
6162 that can happen earlier (such as "too many connections") no longer contain
6165 12. If an expansion in a condition on a "warn" statement fails because a lookup
6166 defers, the "warn" statement is abandoned, and the next ACL statement is
6167 processed. Previously this caused the whole ACL to be aborted.
6169 13. Added the iplsearch lookup type.
6171 14. Added ident_timeout as a log selector.
6173 15. Added tls_certificate_verified as a log selector.
6175 16. Added a global option tls_require_ciphers (compare the smtp transport
6176 option of the same name). This controls incoming TLS connections.
6178 17. I finally figured out how to make tls_require_ciphers do a similar thing
6179 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6180 before starting the TLS session.
6182 18. Tabs are now shown as \t in -bP output.
6184 19. If the log selector return_path_on_delivery was set, Exim crashed when
6185 bouncing a message because it had too many Received: header lines.
6187 20. If two routers both had headers_remove settings, and the first one included
6188 a superfluous trailing colon, the final name in the first list and the
6189 first name in the second list were incorrectly joined into one item (with a
6190 colon in the middle).
6196 1. Added -C and -D options to the exinext utility, mainly to make it easier
6197 to include in the automated testing, but these could be helpful when
6198 multiple configurations are in use.
6200 2. The exinext utility was not formatting the output nicely when there was
6201 an alternate port involved in the retry record key, nor when there was a
6202 message id as well (for retries that were specific to a specific message
6203 and a specific host). It was also confused by IPv6 addresses, because of
6204 the additional colons they contain. I have fixed the IPv4 problem, and
6205 patched it up to do a reasonable job for IPv6.
6207 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6208 delivery, the log line now contains "pipelined" if PIPELINING was used.
6210 4. An SMTP transport process used to panic and die if the bind() call to set
6211 an explicit outgoing interface failed. This has been changed; it is now
6212 treated in the same way as a connect() failure.
6214 5. A reference to $sender_host_name in the part of a conditional expansion
6215 that was being skipped was still causing a DNS lookup. This no longer
6218 6. The def: expansion condition was not recognizing references to header lines
6219 that used bh_ and bheader_.
6221 7. Added the _cache feature to named lists.
6223 8. The code for checking quota_filecount in the appendfile transport was
6224 allowing one more file than it should have been.
6226 9. For compatibility with Sendmail, the command line option
6234 and sets the incoming protocol and host name (for trusted callers). The
6235 host name and its colon can be omitted when only the protocol is to be set.
6236 Note the Exim already has two private options, -pd and -ps, that refer to
6237 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6238 "s", but I don't think that's a major issue.
6240 10. A number of refactoring changes to the code, none of which should affect
6243 (a) The number of logging options was getting close to filling up the
6244 32-bit word that was used as a bit map. I have split them into two classes:
6245 those that are passed in the argument to log_write(), and those that are
6246 only ever tested independently outside of that function. These are now in
6247 separate 32-bit words, so there is plenty of room for expansion again.
6248 There is no change in the user interface or the logging behaviour.
6250 (b) When building, for example, log lines, the code previously used a
6251 macro that called string_cat() twice, in order to add two strings. This is
6252 not really sufficiently general. Furthermore, there was one instance where
6253 it was actually wrong because one of the argument was used twice, and in
6254 one call a function was used. (As it happened, calling the function twice
6255 did not affect the overall behaviour.) The macro has been replaced by a
6256 function that can join an arbitrary number of extra strings onto a growing
6259 (c) The code for expansion conditions now uses a table and a binary chop
6260 instead of a serial search (which was left over from when there were very
6261 few conditions). Also, it now recognizes conditions like "pam" even when
6262 the relevant support is not compiled in: a suitably worded error message is
6263 given if an attempt is made to use such a condition.
6265 11. Added ${time_interval:xxxxx}.
6267 12. A bug was causing one of the ddress fields not to be passed back correctly
6268 from remote delivery subprocesses. The field in question was not being
6269 subsequently used, so this caused to problems in practice.
6271 13. Added new log selectors queue_time and deliver_time.
6273 14. Might have fixed a bug in maildirsizefile handling that threw up
6274 "unexpected character" debug warnings, and recalculated the data
6275 unnecessarily. In any case, I expanded the warning message to give more
6278 15. Added the message "Restricted characters in address" to the statements in
6279 the default ACL that block characters like @ and % in local parts.
6281 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6282 Three changes have been made:
6284 (a) There was a serious bug; a negative response to MAIL caused the whole
6285 recipient domain to be cached as invalid, thereby blocking all messages
6286 to all local parts at the same domain, from all senders. This bug has
6287 been fixed. The domain is no longer cached after a negative response to
6288 MAIL if the sender used is not empty.
6290 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6293 (c) A new callout option, "use_sender" has been added for people who want
6294 the modified behaviour.
6300 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6303 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6304 indeed breaks things for older releases.
6306 3. Added additional logging to the case where there is a problem reading data
6307 from a filter that is running in a subprocess using a pipe, in order to
6308 try to track down a specific problem.
6310 4. Testing facility fudge: when running in the test harness and attempting
6311 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6312 getting "No route to host". Convert this to a timeout.
6314 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6317 6. Some OS don't have socklen_t but use size_t instead. This affects the
6318 fifth argument of getsockopt() amongst other things. This is now
6319 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6320 can be set for individual OS. I have set it for SunOS5, OSF1, and
6321 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6322 some earlier ones do not.
6324 7. Change 4.30/15 was not doing the test caselessly.
6326 8. The standard form for an IPv6 address literal was being rejected by address
6327 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6328 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6329 this, as well as the form without the "IPv6" on the front (but only when
6330 address literals are enabled, of course).
6332 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6334 10. Exim crashed if a message with an empty sender address specified by -f
6335 encountered a router with an errors_to setting. This could be provoked only
6336 by a command such as
6340 where an empty string was supplied; "<>" did not hit this bug.
6342 11. Installed PCRE release 4.5.
6344 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6345 remained set. It is now erased.
6347 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6348 times from message ids (which are base 36 rather than the normal 62).
6350 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6351 were being counted as actual protocol errors, and logged if the log
6352 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6353 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6354 and DATA following a set of rejected RCPTs do not count as protocol errors.
6355 In other words, Exim assumes they were pipelined, though this may not
6356 actually be the case. Of course, in all cases the client gets an
6357 appropriate error code.
6359 15. If a lookup fails in an ACL condition, a message about the failure may
6360 be available; it is used if testing the ACL cannot continue, because most
6361 such messages specify what the cause of the deferral is. However, some
6362 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6363 that caused an old message to be retained and used if a later statement
6364 caused a defer, replacing the real cause of the deferral.
6366 16. If an IP address had so many PTR records that the DNS lookup buffer
6367 was not large enough to hold them, Exim could crash while trying to process
6368 the truncated data. It now detects and logs this case.
6370 17. Further to 4.21/58, another change has been made: if (and only if) the
6371 first line of a message (the first header line) ends with CRLF, a bare LF
6372 in a subsequent header line has a space inserted after it, so as not to
6373 terminate the header.
6375 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6376 unnecessarily, used atoi() instead of strtol(), and didn't check the
6377 termination when getting file sizes from file names by regex.
6379 19. Completely re-implemented the support for maildirsize files, in the light
6380 of a number of problems with the previous contributed implementation
6381 (4.30/29). In particular:
6383 . If the quota is zero, the maildirsize file is maintained, but no quota is
6386 . If the maildir directory does not exist, it is created before any attempt
6387 to write a maildirsize file.
6389 . The quota value in the file is just a cache; if the quota is changed in
6390 the transport, the new value overrides.
6392 . A regular expression is available for excluding directories from the
6395 20. The autoreply transport checks the characters in options that define the
6396 message's headers; it allows continued headers, but it was checking with
6397 isspace() after an embedded newline instead of explicitly looking for a
6400 21. If all the "regular" hosts to which an address was routed had passed their
6401 expiry times, and had not reached their retry times, the address was
6402 bounced, even if fallback hosts were defined. Now Exim should go on to try
6405 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6406 equivalent code in the SMTP transport. Some hosts send humungous responses
6407 to HELO/EHLO, more than 1024 it seems.
6409 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6410 gives compiler warnings in some environments. I've now done it "properly",
6413 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6414 (because of problems with the built-in one) was declared to return uschar *
6415 instead of char *, causing compiler failure.
6417 25. Fixed a file descriptor leak when processing alias/forward files.
6419 26. Fixed a minor format string issue in dbfn.c.
6421 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6423 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6424 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6427 29. When a sender address is verified, it is cached, to save repeating the test
6428 when there is more than one recipient in a message. However, when the
6429 verification involves a callout, it is possible for different callout
6430 options to be set for different recipients. It is too complicated to keep
6431 track of this in the cache, so now Exim always runs a verification when a
6432 callout is required, relying on the callout cache for the optimization.
6433 The overhead is duplication of the address routing, but this should not be
6436 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6437 to be verified with callout=postmaster, and the main callout worked but the
6438 postmaster check failed, the verification correctly failed. However, if a
6439 subsequent RCPT command asked for sender verification *without* the
6440 postmaster check, incorrect caching caused this verification also to fail,
6443 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6444 it was not caching the DNS options (qualify_single, search_parents) that
6445 were used when the lookup failed. A subsequent lookup with different
6446 options therefore always gave the same answer, though there were cases
6447 where it should not have. (Example: a "domains = !$mx_any" option on a
6448 dnslookup router: the "domains" option is always processed without any
6449 widening, but the router might have qualify_single set.) Now Exim uses the
6450 cached value only when the same options are set.
6452 32. Added John Jetmore's "exipick" utility to the distribution.
6454 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6455 than a timeout (e.g. a certificate is required, and is not provided), an
6456 Exim server now closes the connection immediately. Previously it waited for
6457 the client to close - but if the client is SSL, it seems that they each
6458 wait for each other, leading to a delay before one of them times out.
6460 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6461 maintained 0.8.x compatibility because I don't think many are using it, and
6462 it is clearly obsolete.
6464 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6467 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6468 client certificate was expired. A simple patch fixes this, though I don't
6469 understand the full logic of why the verify callback is called multiple
6472 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6473 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6474 which causes problems with some clients (such as the Certicom SSL Plus
6475 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6476 disables the coutermeasure allowing Eudora to connect."
6478 38. Exim was not checking that a write() to a log file succeeded. This could
6479 lead to Bad Things if a log got too big, in particular if it hit a file
6480 size limit. Exim now panics and dies if it cannot write to a log file, just
6481 as it does if it cannot open a log file.
6483 39. Modified OS/Makefile-Linux so that it now contains
6485 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6487 The two -D definitions ensure that Exim is compiled with large file
6488 support, which makes it possible to handle log files that are bigger than
6491 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6492 instance) a domain was checked against a named list that involved a lookup,
6493 causing $domain_data to be set, then another domain was checked against the
6494 same list, then the first domain was re-checked, the value of $domain_data
6495 after the final check could be wrong. In particular, if the second check
6496 failed, it could be set empty. This bug probably also applied to
6499 41. The strip_trailing_dot option was not being applied to the address given
6500 with the -f command-line option.
6502 42. The code for reading a message's header from the spool was incrementing
6503 $received_count, but never initializing it. This meant that the value was
6504 incorrect (doubled) while delivering a message in the same process in which
6505 it was received. In the most common configuration of Exim, this never
6506 happens - a fresh exec is done - but it can happen when
6507 deliver_drop_privilege is set.
6509 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6510 it now includes up to 150 characters of the unexpected data in the log
6513 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6514 and building data strings. The size of both of these buffers was 10 000
6515 bytes - far larger than anybody would *ever* want, thought I. Needless to
6516 say, somebody hit the limit. I have increased the maximum line length to
6517 20 000 and the maximum data length of concatenated lines to 100 000. I have
6518 also fixed two bugs, because there was no checking on these buffers. Tsk,
6519 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6520 buffer is too small.
6522 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6523 lsearch lookups. Now it does.
6525 46. When parsing a route_list item in a manualroute router, a fixed-length
6526 buffer was used for the list of hosts. I made this 1024 bytes long,
6527 thinking that nobody would ever have a list of hosts that long. Wrong.
6528 Somebody had a whole pile of complicated expansion conditions, and the
6529 string was silently truncated, leading to an expansion error. It turns out
6530 that it is easier to change to an unlimited length (owing to other changes
6531 that have happened since this code was originally written) than to build
6532 structure for giving a limitation error. The length of the item that
6533 expands into the list of hosts is now unlimited.
6535 47. The lsearch lookup could not handle data where the length of text line was
6536 more than 4095 characters. Such lines were truncated, leading to shortened
6537 data being returned. It should now handle lines of any length.
6539 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6540 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6543 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6544 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6545 no longer happens in comment lines. A list of replacements is now placed
6546 at the head of all of the source files, except those whose only change is
6547 to replace PERL_COMMAND in the very first #! line.
6549 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6550 messages on the queue, with a bottom-up merge sort, using code contributed
6551 by Michael Haardt. This should make operations like -bp somewhat faster on
6552 large queues. It won't affect queue runners, except when queue_run_in_order
6555 51. Installed eximstats 1.31 in the distribution.
6557 52. Added support for SRV lookups to the dnslookup router.
6559 53. If an ACL referred to $message_body or $message_body_end, the value was not
6560 reset for any messages that followed in the same SMTP session.
6562 54. The store-handling optimization for building very long strings was not
6563 differentiating between the different store pools. I don't think this
6564 actually made any difference in practice, but I've tidied it.
6566 55. While running the routers to verify a sender address, $sender_address
6567 was still set to the sender address. This is wrong, because when routing to
6568 send a bounce to the sender, it would be empty. Therefore, I have changed
6569 it so that, while verifying a sender address, $sender_address is set to <>.
6570 (There is no change to what happens when verifying a recipient address.)
6572 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6573 target A or AAAA records (if not already returned) without resetting the
6574 qualify_single or search_parents options of the DNS resolver. These are
6575 inappropriate in this case because the targets of MX and SRV records must
6576 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6577 target that, when qualified, matched something in the local domain. These
6578 two options are now turned off when doing these lookups.
6580 57. It seems that at least some releases of Reiserfs (which does not have the
6581 concept of a fixed number of inodes) returns zero and not -1 for the
6582 number of available inodes. This interacted badly with check_spool_inodes,
6583 which assumed that -1 was the "no such thing" setting. What I have done is
6584 to check that the total number of inodes is greater than zero before doing
6585 the test of how many are available.
6587 58. When a "warn" ACL statement has a log_message modifier, the message is
6588 remembered, and not repeated. This is to avoid a lot of repetition when a
6589 message has many recipients that cause the same warning to be written.
6590 However, Exim was preserving the list of already written lines for an
6591 entire SMTP session, which doesn't seem right. The memory is now reset if a
6592 new message is started.
6594 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6595 debugging output unless log_rewrite was also set.
6597 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6598 of (int)(handle) when we know that handle contains (void *)(-1).
6600 61. The Exim daemon panic-logs an error return when it closes the incoming
6601 connection. However "connection reset by peer" seems to be common, and
6602 isn't really an error worthy of noting specially, so that particular error
6605 62. When Exim is trying to find all the local interfaces, it used to panic and
6606 die if the ioctl to get the interface flags failed. However, it seems that
6607 on at least one OS (Solaris 9) it is possible to have an interface that is
6608 included in the list of interfaces, but for which you get a failure error
6609 for this call. This happens when the interface is not "plumbed" into a
6610 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6611 failure of the "get flags" call assumes that the interface is down.
6613 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6614 makes life easier for people who are doing arithmetic on fields extracted
6615 from dates, where you often get leading zeros that should not be
6616 interpreted as octal.
6618 64. Added qualify_domain to the redirect router, to override the global
6621 65. If a pathologically long header line contained very many addresses (the
6622 report of this problem mentioned 10 000) and each of them was rewritten,
6623 Exim could use up a very large amount of memory. (It kept on making new
6624 copies of the header line as it rewrote, and never released the old ones.)
6625 At the expense of a bit more processing, the header rewriting function has
6626 been changed so that it no longer eats memory in this way.
6628 66. The generation of the Received: header has been moved from the time that a
6629 message starts to be received, to the time that it finishes. The timestamp
6630 in the Received: header should now be very close to that of the <= log
6631 line. There are two side-effects of this change:
6633 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6634 logged header lines no longer include the local Received: line, because
6635 it has not yet been created. The same applies to a copy of the message
6636 that is returned to a non-SMTP sender when a message is rejected.
6638 (b) When a filter file is tested using -bf, no additional Received: header
6639 is added to the test message. After some thought, I decided that this
6642 This change does not affect the value of $received_for. It is still set
6643 after address rewriting, but before local_scan() is called.
6645 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6647 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6648 gave an unhelpful panic error message, and a defer error. I have managed to
6649 change this behaviour so that it now rejects any supplied certificate,
6650 which seems right, as the list of acceptable certificates is empty.
6652 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6653 gave an unhelpful defer error. I have not managed to make this reject any
6654 supplied certificates, but the error message it gives is "no certificate
6655 supplied", which is not helpful.
6657 70. exigrep's output now also includes lines that are not associated with any
6658 message, but which match the given pattern. Implemented by a patch from
6659 Martin Sluka, which also tidied up the Perl a bit.
6661 71. Recipient callout verification, like sender verification, was using <> in
6662 the MAIL FROM command. This isn't really the right thing, since the actual
6663 sender may affect whether the remote host accepts the recipient or not. I
6664 have changed it to use the actual sender in the callout; this means that
6665 the cache record is now keyed on a recipient/sender pair, not just the
6666 recipient address. There doesn't seem to be a real danger of callout loops,
6667 since a callout by the remote host to check the sender would use <>.
6668 [SEE ABOVE: changed after hitting problems.]
6670 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6671 temporary errors. However, in the case of such a code being given after
6672 the end of a data transmission (i.e. after ".") Exim was failing to write
6673 a retry record for the message. (Yes, there was some broken host that was
6674 actually sending 8xx at this point.)
6676 73. An unknown lookup type in a host list could cause Exim to panic-die when
6677 the list was checked. (An example that provoked this was putting <; in the
6678 middle of a list instead of at the start.) If this happened during a DATA
6679 ACL check, a -D file could be left lying around. This kind of configuration
6680 error no longer causes Exim to die; instead it causes a defer error. The
6681 incident is still logged to the main and panic logs.
6683 74. Buglet left over from Exim 3 conversion. The message "too many messages
6684 in one connection" was written to the rejectlog but not the mainlog, except
6685 when address rewriting (yes!) was being logged.
6687 75. Added write_rejectlog option.
6689 76. When a system filter was run not as root (that is, when system_filter_user
6690 was set), the values of the $n variables were not being returned to the
6691 main process; thus, they were not subsequently available in the $sn
6694 77. Added +return_path_on_delivery log selector.
6696 78. A connection timeout was being treated differently from recipients deferred
6697 when testing hosts_max_try with a message that was older than the host's
6698 retry timeout. (The host should not be counted, thus allowing all hosts to
6699 be tried at least once before bouncing.) This may have been the cause of an
6700 occasionally reported bug whereby a message would remain on the queue
6701 longer than the retry timeout, but would be bounced if a delivery was
6702 forced. I say "may" because I never totally pinned down the problem;
6703 setting up timeout/retry tests is difficult. See also the next item.
6705 79. The ultimate address timeout was not being applied to errors that involved
6706 a combination of host plus message (for example, a timeout on a MAIL
6707 command). When an address resolved to a number of possible hosts, and they
6708 were not all tried for each delivery (e.g. because of hosts_max_try), a
6709 message could remain on the queue longer than the retry timeout.
6711 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6714 81. Fixed an obscure SMTP outgoing bug which required at least the following
6715 conditions: (a) there was another message waiting for the same server;
6716 (b) the server returned 5xx to all RCPT commands in the first message so
6717 that the message was not completed; (c) the server dropped the connection
6718 or gave a negative response to the RSET that Exim sends to abort the
6719 transaction. The observed case was a dropped connection after DATA that had
6720 been sent in pipelining mode. That is, the server had advertised PIPELINING
6721 but was not implementing it correctly. The effect of the bug was incorrect
6722 behaviour, such as trying another host, and this could lead to a crash.
6728 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6729 and daemon.c were passed as pointers to ints; they should have been
6730 pointers to socklen_t variables (which are typically unsigned ints).
6732 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6735 3. Fixed a really odd bug that affected only the testing scheme; patching a
6736 certain fixed string in the binary changed the value of another string that
6737 happened to be identical to the end of the original first string.
6739 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6740 name", it returns that address as the IP address. On some operating
6741 systems (e.g. Solaris), it also passes back the IP address string as the
6742 "host name". However, on others (e.g. Linux), it passes back an empty
6743 string. Exim wasn't checking for this, and was changing the host name to an
6744 empty string, assuming it had been canonicalized.
6746 5. Although rare, it is permitted to have more than one PTR record for a given
6747 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6748 all the names associated with an address, because they do in Solaris.
6749 However, it seems that they do not in Linux for data that comes from the
6750 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6751 I found this out when I moved to a new Linux workstation and tried to run
6752 the Exim test suite.
6754 To get round this problem I have changed the code so that it now does its
6755 own call to the DNS to look up PTR records when searching for a host name.
6756 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6757 addresses that are only in /etc/hosts are still found.
6759 This behaviour is, however, controlled by an option called host_lookup_
6760 order, which defaults to "bydns:byaddr". If people want to use the other
6761 order, or indeed, just use one or the other means of lookup, they can
6762 specify it in this variable.
6764 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6765 some operating systems, this comes back from gethostbyaddr() as an empty
6766 string, and this is what Exim used to test for. However, it seems that in
6767 other systems, "." is yielded. Exim now tests for this case too.
6769 7. The values of check_spool_space and check_log_space are now held internally
6770 as a number of kilobytes instead of an absolute number of bytes. If a
6771 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6772 kilobyte. This means that much larger values can be stored.
6774 8. Exim monitor: an attempt to get the action menu when not actually pointing
6775 at a message produces an empty menu entitled "No message selected". This
6776 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6777 no entries in it ("Shell widget menu has zero width and/or height"). So I
6778 have added a single, blank menu entry in this case.
6780 9. Added ${quote_local_part.
6782 10. MIME decoding is now applied to the contents of Subject: header lines when
6785 11. Now that a reference to $sender_host_address automatically causes a reverse
6786 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6787 host lookup before query-style lookups in lists that might use this
6788 variable. This has therefore been abolished, and the "net-" prefix is no
6789 longer necessary for query-style lookups.
6791 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6792 have been a typo for LFLAGS, so it has been changed.
6794 13. The install script calls Exim with "-C /dev/null" in order to find the
6795 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6796 to be output. However, since Exim outputs its version number before the
6797 error, it didn't break the script. It just looked ugly. I fixed this by
6798 always allowing "-C /dev/null" if the caller is root.
6800 14. Ignore overlarge ACL variable number when reading spool file - insurance
6801 against a later release with more variables having written the file.
6803 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6804 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6805 this, as well as the form without the "IPv6" on the front.
6807 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6808 OS/Makefile-Darwin file.
6810 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6811 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6813 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6814 variable. However, some OS use ints for inodes. I've added cast to long int
6815 to get rid of the compiler warning.
6817 19. I had forgotten to lock out "/../" in configuration file names when
6818 ALT_CONFIG_PREFIX was set.
6820 20. Routers used for verification do not need to specify transports. However,
6821 if such a router generated a host list, and callout was configured, Exim
6822 crashed, because it could not find a port number from the (non-existent)
6823 transport. It now assumes port 25 in this circumstance.
6825 21. Added the -t option to exigrep.
6827 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6828 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6829 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6830 them is set, LOOKUP_LSEARCH is forced.
6832 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6834 24. Added sender and host information to the "rejected by local_scan()" log
6835 line; previously there was no indication of these.
6837 25. Added .include_if_exists.
6839 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6840 Linux. It turns out that not all file systems support this. Apparently some
6841 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6842 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6843 not supported on the file descriptor, is now ignored when Exim is trying to
6844 sync a directory. This applies only to Linux.
6846 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6848 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6849 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6852 29. Added support for maildirsize files from supplied patch (modified a bit).
6854 30. The use of :fail: followed by an empty string could lead Exim to respond to
6855 sender verification failures with (e.g.):
6857 550 Verification failed for <xxx>
6858 550 Sender verify failed
6860 where the first response line was missing the '-' that indicates it is not
6861 the final line of the response.
6863 31. The loop for finding the name of the user that called Exim had a hardwired
6864 limit of 10; it now uses the value of finduser_retries, which is used for
6865 all other user lookups.
6867 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6870 33. Exim was neglecting to zero errno before one call of strtol() when
6871 expanding a string and expecting an integer value. On some systems this
6872 resulted in spurious "integer overflow" errors. Also, it was casting the
6873 result into an int without checking.
6875 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6876 did not work. The code looks as if it has *never* worked, though it appears
6877 to have been documented since at least release 1.62. I have made it work.
6879 35. The "timeout_DNS" error in retry rules, also documented since at least
6880 1.62, also never worked. As it isn't clear exactly what this means, and
6881 clearly it isn't a major issue, I have abolished the feature by treating it
6882 as "timeout", and writing a warning to the main and panic logs.
6884 36. The display of retry rules for -brt wasn't always showing the error code
6887 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6888 timeout_connect_A, timeout_connect_MX.
6890 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6891 to the empty sender.
6893 39. The daemon was not analysing the content of -oX till after it had closed
6894 stderr and disconnected from the controlling terminal. This meant that any
6895 syntax errors were only noted on the panic log, and the return code from
6896 the command was 0. By re-arranging the code a little, I've made the
6897 decoding happen first, so such errors now appear on stderr, and the return
6898 code is 1. However, the actual setting up of the sockets still happens in
6899 the disconnected process, so errors there are still only recorded on the
6902 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6903 connections (as happens on some IP stacks) was logged at start up time as
6904 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6905 it from "IPv6 and IPv4", which means that two separate sockets are being
6908 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6909 says whether AF_INET or AF_INET6 was passed as an argument.
6911 42. Exiwhat output was messed up when time zones were included in log
6914 43. Exiwhat now gives more information about the daemon's listening ports,
6915 and whether -tls-on-connect was used.
6917 44. The "port" option of the smtp transport is now expanded.
6919 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6920 silently ignored. Now an error message is written to the main and panic
6923 46. There's a new ACL modifier called "logwrite" which writes to a log file
6924 as soon as it is encountered.
6926 47. Added $local_user_uid and $local_user_gid at routing time.
6928 48. Exim crashed when trying to verify a sender address that was being
6931 49. Exim was recognizing only a space character after ".include". It now also
6932 recognizes a tab character.
6934 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6935 extracting the relevant information from the specification. The man page no
6936 longer contains scrambled data for the -d option, and I've added a section
6937 at the front about calling Exim under different names.
6939 51. Added "extra_headers" argument to the "mail" command in filter files.
6941 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6944 53. Installed eximstats 1.29.
6946 54. Added transport_filter_timeout as a generic transport option.
6948 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6949 Cc: header lines. This was required by RFC 822, but it not required by RFC
6952 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6953 incoming messages that did not have them. Now it does so only if the
6954 message originates locally, that is, if there is no associated remote host
6955 address. When Resent- header lines are present, this applies to the Resent-
6956 lines rather than the non-Resent- lines.
6958 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6959 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6961 58. Messages for configuration errors now include the name of the main
6962 configuration file - useful now that there may be more than one file in a
6963 list (.included file names were always shown).
6965 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6966 for those rare installations that do not start the daemon as root or run it
6967 setuid root. I've cut out the call to initgroups() if the daemon is not
6970 60. The Exim user and group can now be bound into the binary as text strings
6971 that are looked up at the start of Exim's processing.
6973 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6975 62. Added $mailstore_basename variable.
6977 63. Installed patch to sieve.c from Michael Haardt.
6979 64. When Exim failed to open the panic log after failing to open the main log,
6980 the original message it was trying to log was written to stderr and debug
6981 output, but if they were not available (the usual case in production), it
6982 was lost. Now it is written to syslog before the two lines that record the
6983 failures to open the logs.
6985 65. Users' Exim filters run in subprocesses under the user's uid. It is
6986 possible for a "deliver" command or an alias in a "personal" command to
6987 provoke an address rewrite. If logging of address rewriting is configured,
6988 this fails because the process is not running as root or exim. There may be
6989 a better way of dealing with this, but for the moment (because 4.30 needs
6990 to be released), I have disabled address rewrite logging when running a
6991 filter in a non-root, non-exim process.
6997 1. The buildconfig auxiliary program wasn't quoting the value set for
6998 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6999 not defined. This bug was masked in 4.22 by the effect that was fixed in
7002 2. Some messages that were rejected after a message id was allocated were
7003 shown as "incomplete" by exigrep. It no longer does this for messages that
7004 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7006 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7007 have allow_domain_literals set, the ID did not get logged in the <= line.
7008 Domain literals are now always recognized in Message-ID: header lines.
7010 4. The first argument for a ${extract expansion item is the key name or field
7011 number. Leading and trailing spaces in this item were not being ignored,
7012 causing some misleading effects.
7014 5. When deliver_drop_privilege was set, single queue runner processes started
7015 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7016 same command in the process it spins off) were not dropping privilege.
7018 6. When the daemon running as "exim" started a queue runner, it always
7019 re-executed Exim in the spun-off process. This is a waste of effort when
7020 deliver_drop_privilege is set. The new process now just calls the
7021 queue-runner function directly.
7027 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7030 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7031 ignored. Though the use of -oP was forcing the writing of a pid file, it
7032 was always written to the default place.
7034 3. If the message "no IP address found for host xxxx" is generated during
7035 incoming verification, it is now followed by identification of the incoming
7036 connection (so you can more easily find what provoked it).
7038 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7040 5. Added some features to "harden" Exim a bit more against certain attacks:
7042 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7043 be put in Local/Makefile. This is like the never_users runtime option,
7044 but it cannot be overridden. The default setting is "root".
7046 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7047 prefix string with which any file named in a -C command line option
7050 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7051 is retained for -C and -D only if the caller of Exim is root. Without
7052 it, the exim user may also use -C and -D and retain privilege.
7054 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7055 command line option is disabled.
7057 6. Macro names set by the -D option must start with an upper case letter, just
7058 like macro names defined in the configuration file.
7060 7. Added "dereference=" facility to LDAP.
7062 8. Two instances of the typo "uknown" in the source files are fixed.
7064 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7065 the Configure-Makefile script screwed up while processing it.
7067 10. Incorporated PCRE 4.4.
7069 11. The SMTP synchronization check was not operating right at the start of an
7070 SMTP session. For example, it could not catch a HELO sent before the client
7071 waited for the greeting. There is now a check for outstanding input at the
7072 point when the greeting is written. Because of the duplex, asynchronous
7073 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7074 way, but not yet received, when the check is performed.
7076 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7077 on TCP/IP sockets, because this apparently causes some broken clients to
7080 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7081 unchanged) from the Cygwin maintainer.
7083 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7084 was supported instead of testing for mbx. Effectively a typo.
7086 15. The spa authenticator server code was not checking that the input it
7087 received was valid base64.
7089 16. The debug output line for the "set" modifier in ACLs was not showing the
7090 name of the variable that was being set.
7092 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7094 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7095 lookup on its own. Something else was needed to trigger the lookup. For
7096 example, a match in host_lookup or the need for a host name in a host list.
7097 Now, if $sender_host_name is referenced and the host name has not yet been
7098 looked up, a lookup is performed. If the lookup fails, the variable remains
7099 empty, and $host_lookup_failed is set to "1".
7101 19. Added "eqi" as a case-independent comparison operator.
7103 20. The saslauthd authentication condition could segfault if neither service
7104 nor realm was specified.
7106 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7107 error message that was logged was misleading, and incoming SMTP
7108 connections were dropped. The message is now more accurate, and temporary
7109 errors are given to SMTP connections.
7111 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7112 (except RSET) with a 421 error, until QUIT is received. However, it was
7113 failing to send a response to QUIT.
7115 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7116 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7117 ACL is now run after the helo_try_verify_hosts code.
7119 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7120 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7121 case-independent, so other case variants are also recognized). Apparently
7122 some systems use these upper case variants.
7124 25. If more than two messages were waiting for the same host, and a transport
7125 filter was specified for the transport, Exim sent two messages over the
7126 same TCP/IP connection, and then failed with "socket operation on non-
7127 socket" when it tried to send the third.
7129 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7131 27. The extern definition of crypt16() in expand.c was not being excluded when
7132 the OS had its own crypt16() function.
7134 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7135 as a preferred synonym for return_size_limit, both as an option and as an
7138 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7140 30. Changed the default configuration ACL to relax the local part checking rule
7141 for addresses that are not in any local domains. For these addresses,
7142 slashes and pipe symbols are allowed within local parts, but the sequence
7143 /../ is explicitly forbidden.
7145 31. SPA server authentication was not clearing the challenge buffer before
7148 32. log_message in a "warn" ACL statement was writing to the reject log as
7149 well as to the main log, which contradicts the documentation and doesn't
7150 seem right (because no rejection is happening). So I have stopped it.
7152 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7153 However, I am unable to do any testing of this.
7155 34. Fixed an infelicity in the appendfile transport. When checking directories
7156 for a mailbox, to see if any needed to be created, it was accidentally
7157 using path names with one or more superfluous leading slashes; tracing
7158 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7160 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7161 contents are added to the log line that is written for every discarded
7162 recipient. (Previously a log_message setting was ignored.)
7164 36. The ${quote: operator now quotes the string if it is empty.
7166 37. The install script runs exim in order to find its version number. If for
7167 some reason other than non-existence or emptiness, which it checks, it
7168 could not run './exim', it was installing it with an empty version number,
7169 i.e. as "exim-". This error state is now caught, and the installation is
7172 38. An argument was missing from the function that creates an error message
7173 when Exim fails to connect to the socket for saslauthd authentication.
7174 This could cause Exim to crash, or give a corrupted message.
7176 39. Added isip, isip4, and isip6 to ${if conditions.
7178 40. The ACL variables $acl_xx are now saved with the message, and can be
7179 accessed later in routers, transports, and filters.
7181 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7182 strings in the file are not string-expanded.
7184 42. If a MAIL command specified a SIZE value that was too large to fit into an
7185 int variable, the check against message_size_limit failed. Such values are
7186 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7187 day this will have to be increased, but I don't think I want to be around
7188 when emails are that large.
7195 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7196 iconv() is not standard in FreeBSD.
7198 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7199 IPv6 enabled. The observed symptom was a segmentation fault on return from
7200 the function os_common_find_running_interfaces() in src/os.c.
7202 3. In the check_special_case() function in daemon.c I had used "errno" as an
7203 argument name, which causes warnings on some systems. This was basically a
7204 typo, since it was named "eno" in the comments!
7206 4. The code that waits for the clock to tick (at a resolution of some fraction
7207 of a second) so as to ensure message-id uniqueness was always waiting for
7208 at least one whole tick, when it could have waited for less. [This is
7209 almost certainly not relevant at current processor speeds, where it is
7210 unlikely to ever wait at all. But we try to future-proof.]
7212 5. The function that sleeps for a time interval that includes fractions of a
7213 second contained a race. It did not block SIGALRM between setting the
7214 timer, and suspending (a couple of lines later). If the interval was short
7215 and the sigsuspend() was delayed until after it had expired, the suspension
7216 never ended. On busy systems this could lead to processes getting stuck for
7219 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7220 process, before it forks any delivery processes. The open lookup caching
7221 mechanism meant that the open file or database connection was passed into
7222 the delivery process. The problem was that delivery processes always tidy
7223 up cached lookup data. This could cause a problem for the next delivery
7224 process started by the queue runner, because the external queue runner
7225 process does not know about the closure. So the next delivery process
7226 still has data in the lookup cache. In the case of a file lookup, there was
7227 no problem because closing a file descriptor in a subprocess doesn't affect
7228 the parent. However, if the lookup was caching a connection to a database,
7229 the connection was closed, and the second delivery process was likely to
7230 see errors such as "PGSQL: query failed: server closed the connection
7231 unexpectedly". The problem has been fixed by closing all cached lookups
7232 in a queue runner before running a delivery process.
7234 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7235 seem to have the "const" qualifier which it has on other OS. I've
7238 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7239 *of the same type* (client or server) with the same public name that an
7240 error should be diagnosed.
7242 9. When Exim looked up a host name for an IP address, but failed to find the
7243 original IP address when looking up the host name (a safety check), it
7244 output the message "<ip address> does not match any IP for NULL", which was
7245 confusing, to say the least. The bug was that the host name should have
7246 appeared instead of "NULL".
7248 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7249 user that is built into the binary, and the -C or -D options is used, root
7250 privilege is dropped before the configuration file is read. In addition,
7251 logging is switched to stderr instead of the normal log files. If the
7252 configuration then re-defines the Exim user, the unprivileged environment
7253 is probably not what is expected, so Exim logs a panic warning message (but
7256 However, if deliver_drop_privilege is set, the unprivileged state may well
7257 be exactly what is intended, so the warning has been cut out in that case,
7258 and Exim is allowed to try to write to its normal log files.
7264 1. smtp_return_error_details was not giving details for temporary sender
7265 or receiver verification errors.
7267 2. Diagnose a configuration error if two authenticators have the same public
7270 3. Exim used not to create the message log file for a message until the first
7271 delivery attempt. This could be confusing when incoming messages were held
7272 for policy or load reasons. The message log file is now created at the time
7273 the message is received, and an initial "Received" line is written to it.
7275 4. The automatically generated man page for command line options had a minor
7276 bug that caused no ill effects; however, a more serious problem was that
7277 the procedure for building the man page automatically didn't always
7278 operate. Consequently, release 4.20 contains an out-of-date version. This
7279 shouldn't happen again.
7281 5. When building Exim with embedded Perl support, the script that builds the
7282 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7283 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7285 6. The freeze_tell option was not being used for messages that were frozen on
7286 arrival, either by an ACL or by local_scan().
7288 7. Added the smtp_incomplete_transaction log selector.
7290 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7291 was accepting AUTH without a new EHLO.
7293 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7294 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7297 10. Logging of TCP/IP connections (when configured) now happens in the main
7298 daemon process instead of the child process, so that the TCP/IP connection
7299 count is more accurate (but it can never be perfect).
7301 11. The use of "drop" in a nested ACL was not being handled correctly in the
7302 outer ACL. Now, if condition failure induced by the nested "drop" causes
7303 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7304 or "require"), the connection is dropped.
7306 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7307 that yield "discard" can now be used with an "accept" or a "discard" verb,
7308 but an error is generated for any others (because I can't see a useful way
7309 to define what should happen).
7311 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7312 are now processed in the same way as lines in the Exim configuration file.
7313 In particular, continuation lines are supported.
7315 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7317 15. Added -ti meaning -t -i.
7319 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7320 domains, and warn by logging if others are found.
7322 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7323 returned from GIFCONF when Exim is trying to find the list of interfaces on
7324 a host. The code in os.c has been modified to copy each ifreq to an aligned
7325 structure in all cases.
7327 Also, in some cases, the returned ifreq's were being copied to a 'struct
7328 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7329 means the last couple of bytes of an IPv6 address could be chopped if the
7330 ifreq contained only a normal sockaddr (14 bytes storage).
7332 18. Named domain lists were not supported in the hosts_treat_as_local option.
7333 An entry such as +xxxx was not recognized, and was treated as a literal
7336 19. Ensure that header lines added by a DATA ACL are included in the reject log
7337 if the ACL subsequently rejects the message.
7339 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7340 MD5 (which is deprecated).
7342 21. When testing a filter file using -bf, Exim was writing a message when it
7343 took the sender from a "From " line in the message, but it was not doing so
7344 when it took $return_path from a Return-Path: header line. It now does.
7346 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7347 with a valid header line field name (a series of printing characters
7348 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7350 23. Changed "disc" in the source to "disk" to conform to the documentation and
7351 the book and for uniformity.
7353 24. Ignore Sendmail's -Ooption=value command line item.
7355 25. When execve() failed while trying to run a command in a pipe transport,
7356 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7357 could be confused with a return value of 69 from the command itself. This
7358 has been changed to 127, the value the shell returns if it is asked to run
7359 a non-existent command. The wording for the related log line suggests a
7360 non-existent command as the problem.
7362 26. If received_header_text expands to an empty string, do not add a Received:
7363 header line to the message. (Well, it adds a token one on the spool, but
7364 marks it "old" so that it doesn't get used or transmitted.)
7366 27. Installed eximstats 1.28 (addition of -nt option).
7368 28. There was no check for failure on the call to getsockname() in the daemon
7369 code. This can fail if there is a shortage of resources on the system, with
7370 ENOMEM, for example. A temporary error is now given on failure.
7372 29. Contrary to the C standard, it seems that in some environments, the
7373 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7374 program. Exim now does this explicitly; it affects the formatting of
7375 timestamps using strftime().
7377 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7378 complaints. I've now initialized all the variables, to avoid this.
7380 32. Header lines added by a system filter were not being "seen" during
7381 transport-time rewrites.
7383 33. The info_callback() function passed to OpenSSL is set up with type void
7384 (*)(SSL *, int, int), as described somewhere. However, when calling the
7385 function (actually a macro) that sets it up, the type void(*)() is
7386 expected. I've put in a cast to prevent warnings from picky compilers.
7388 34. If a DNS black list lookup found a CNAME record, but there were no A
7389 records associated with the domain it pointed at, Exim crashed.
7391 35. If a DNS black list lookup returned more than one A record, Exim ignored
7392 all but the first. It now scans all returned addresses if a particular IP
7393 value is being sought. In this situation, the contents of the
7394 $dnslist_value variable are a list of all the addresses, separated by a
7397 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7398 to accept a host name and all its aliases if the forward lookup for any of
7399 them yielded the IP address of the incoming connection. Now it accepts only
7400 those names whose forward lookup yields the correct IP address. Any other
7401 names are discarded. This closes a loophole whereby a rogue DNS
7402 administrator could create reverse DNS records to break through a
7403 wildcarded host restriction in an ACL.
7405 37. If a user filter or a system filter that ran in a subprocess used any of
7406 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7407 the wrong values were passed to the pipe command ($thisaddress had the
7408 value of $0, $0 had the value of $1, etc). This bug was introduced by
7409 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7411 38. Improved the line breaking for long SMTP error messages from ACLs.
7412 Previously, if there was no break point between 40 and 75 characters, Exim
7413 left the rest of the message alone. Two changes have been made: (a) I've
7414 reduced the minimum length to 35 characters; (b) if it can't find a break
7415 point between 35 and 75 characters, it looks ahead and uses the first one
7416 that it finds. This may give the occasional overlong line, but at least the
7417 remaining text gets split now.
7419 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7420 file descriptors might be low, and that setting 1000 would always raise it.
7421 It turns out that in some environments, the limit is already over 1000 and
7422 that lowering it causes trouble. So now Exim takes care not to decrease it.
7424 40. When delivering a message, the value of $return_path is set to $sender_
7425 address at the start of routing (routers may change the value). By an
7426 oversight, this default was not being set up when an address was tested by
7427 -bt or -bv, which affected the outcome if any router or filter referred to
7430 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7431 log only when "message" or "log_message" are set. However, if one of the
7432 conditions was an address verification, or a call to a nested ACL, the
7433 messages generated by the underlying test were being passed through. This
7434 no longer happens. The underlying message is available in $acl_verify_
7435 message for both "message" and "log_message" expansions, so it can be
7436 passed through if needed.
7438 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7439 new expansion $bh_ to give the encoded byte string without charset
7440 translation. Translation happens only if iconv() is available; HAVE_ICONV
7441 indicates this at build time. HEADERS_CHARSET gives the charset to
7442 translate to; headers_charset can change it in the configuration, and
7443 "headers charset" can change it in an individual filter file.
7445 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7446 that creates RFC 2047 encoded "words" labels them as that charset instead
7447 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7448 expansion operator; (ii) when Exim creates a From: line for a local
7449 message; (iii) when a header line is rewritten to include a "phrase" part.
7451 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7452 buggy, causing it to skip the first lines of messages whose message ID
7453 ended in 'D'. This would not have bitten before Exim release 4.14, because
7454 message IDs were unlikely to end in 'D' before then. The effect was to have
7455 incorrect size information for certain domains.
7457 45. #include "config.h" was missing at the start of the crypt16.c module. This
7458 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7461 46. If there was a timeout during a "random" callout check, Exim treated it as
7462 a failure of the random address, and carried on sending RSET and the real
7463 address. If the delay was just some slowness somewhere, the response to the
7464 original RCPT would be taken as a response to RSET and so on, causing
7465 mayhem of various kinds.
7467 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7468 when I implemented it. It didn't allow for the fact that some option values
7469 may legitimately be negative (e.g. size_addition), and it didn't even do
7470 the right test for positive values.
7472 48. Domain names in DNS records are case-independent. Exim always looks them up
7473 in lower case. Some resolvers return domain names in exactly the case they
7474 appear in the zone file, that is, they may contain uppercase letters. Not
7475 all resolvers do this - some return always lower case. Exim was treating a
7476 change of case by a resolver as a change of domain, similar to a widening
7477 of a domain abbreviation. This triggered its re-routing code and so it was
7478 trying to route what was effectively the same domain again. This normally
7479 caused routing to fail (because the router wouldn't handle the domain
7480 twice). Now Exim checks for this case specially, and just changes the
7481 casing of the domain that it ultimately uses when it transmits the message
7484 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7487 50. If a filter generated a file delivery with a non-absolute name (possible if
7488 no home directory exists for the router), the forbid_file option was not
7491 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7492 the existing equality matching.
7494 52. Exim was using ints instead of ino_t variables in some places where it was
7495 dealing with inode numbers.
7497 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7498 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7499 called TMPDIR, and if it finds it is different, it changes its value.
7501 54. The smtp_printf() function is now made available to local_scan() so
7502 additional output lines can be written before returning. There is also an
7503 smtp_fflush() function to enable the detection of a dropped connection.
7504 The variables smtp_input and smtp_batched_input are exported to
7507 55. Changed the default runtime configuration: the message "Unknown user"
7508 has been removed from the ACL, and instead placed on the localuser router,
7509 using the cannot_route_message feature. This means that any verification
7510 failures that generate their own messages won't get overridden. Similarly,
7511 the "Unrouteable address" message that was in the ACL for unverifiable
7512 relay addresses has also been removed.
7514 56. Added hosts_avoid_esmtp to the smtp transport.
7516 57. The exicyclog script was not checking for the esoteric option
7517 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7518 will work only if exicyclog is run under the appropriate euid.
7520 58. Following a discussion on the list, the rules by which Exim recognises line
7521 endings on incoming messages have been changed. The -dropcr and drop_cr
7522 options are now no-ops, retained only for backwards compatibility. The
7523 following line terminators are recognized: LF CRLF CR. However, special
7524 processing applies to CR:
7526 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7527 nor a local message in the state where . is a terminator.
7529 (ii) If a bare CR is encountered in a header line, an extra space is added
7530 after the line terminator so as not to end the header. The reasoning
7531 behind this is that bare CRs in header lines are most likely either
7532 to be mistakes, or people trying to play silly games.
7534 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7535 was being incorrectly given as 18 bytes larger than it should have been.
7536 This is a VOB (very old bug).
7538 60. This may never have affected anything current, but just in case it has:
7539 When the local host is found other than at the start of a list of hosts,
7540 the local host, those with the same MX, and any that follow, are discarded.
7541 When the list in question was part of a longer list of hosts, the following
7542 hosts (not currently being processed) were also being discarded. This no
7543 longer happens. I'm not sure if this situation could ever has previously
7546 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7549 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7550 header. This is something that is recommended in a new Internet Draft, and
7551 is something that is documented as being done by Sendmail. There are two
7552 possible values. For messages generated by the autoreply transport, Exim
7555 Auto-Submitted: auto-replied
7557 whereas for all other generated messages (e.g. bounces) it adds
7559 Auto-Submitted: auto-generated
7561 63. The "personal" condition in filters now includes a test for the
7562 Auto-Submitted: header. If it contains the string "auto-" the message it
7563 not considered personal.
7565 64. Added rcpt_include_affixes as a generic transport option.
7567 65. Added queue_only_override (default true).
7569 66. Added the syslog_duplication option.
7571 67. If what should have been the first header line of a message consisted of
7572 a space followed by a colon, Exim was mis-interpreting it as a header line.
7573 It isn't of course - it is syntactically invalid and should therefore be
7574 treated as the start of the message body. The misbehaviour could have
7575 caused a number of strange effects, including loss of data in subsequent
7576 header lines, and spool format errors.
7578 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7579 client host had authenticated. This control can now be exercised by an ACL
7580 for more flexibility.
7582 69. By default, callouts do not happen when testing with -bh. There is now a
7583 variant, -bhc, which does actually run the callout code, including
7584 consulting and updating the callout cache.
7586 70. Added support for saslauthd authentication, courtesy of Alexander
7589 71. If statvfs() failed on the spool or log directories while checking their
7590 size for availability, Exim confusingly gave the error "space shortage".
7591 Furthermore, in debugging mode it crashed with a floating point exception.
7592 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7593 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7594 serious problem, Exim now writes to the main and panic logs when this
7595 happens, with details of the failure. It then refuses to accept the
7596 incoming message, giving the message "spool directory problem" or "log
7597 directory problem" with a 421 code for SMTP messages.
7599 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7600 0, 1, and 2 exist, because some OS complain for execs without them (see
7601 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7602 descriptors. However, the code omitted to check that the open succeeded,
7603 causing mysterious errors if for some reason the permissions on /dev/null
7604 got screwed. Now Exim writes a message to the main and panic logs, and
7605 bombs out if it can't open /dev/null.
7607 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7608 interact so that it is all more flexible. It is supposed to remain
7609 backwards compatible. Also added extra_local_interfaces.
7611 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7612 to bomb out with an assertion failure - to the client this appears as a
7613 connection drop. This problem occurs in the part of the code that was taken
7614 from the Samba project. Fortunately, the assertion is in a very simple
7615 function, so I have fixed this by reproducing the function inline in the
7616 one place where it is called, and arranging for authentication to fail
7617 instead of killing the process with assert().
7619 75. The SPA client code was not working when the server requested OEM rather
7620 than Unicode encoding.
7622 76. Added code to make require_files with a specific uid setting more usable in
7623 the case where statting the file as root fails - usually a non-root-mounted
7624 NFS file system. When this happens and the failure is EACCES, Exim now
7625 forks a subprocess and does the per-uid checking as the relevant uid.
7627 77. Added process_log_path.
7629 78. If log_file_path was not explicitly set, a setting of check_log_space or
7630 check_log_inodes was ignored.
7632 79. If a space check for the spool or log partitions fails, the incident is now
7633 logged. Of course, in the latter case the data may get lost...
7635 80. Added the %p formatting code to string_format() so that it can be used to
7636 print addresses in debug_print(). Adjusted all the address printing in the
7637 debugging in store.c to use %p rather than %d.
7639 81. There was a concern that a line of code in smtp_in.c could overflow a
7640 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7641 initially expressed, the concern was not well-founded, because trailing
7642 spaces are removed early. However, if the trailing spaces were followed by
7643 a NULL, they did not get removed, so the overflow was possible. Two fixes
7646 (a) I re-wrote the offending code in a cleaner fashion.
7647 (b) If an incoming SMTP command contains a NULL character, it is rejected
7650 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7651 runs initgroups(), so that if the Exim user is in any additional groups,
7652 they will be used during message reception.
7658 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob