1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
77 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
78 for the expansion of the server_scram_iter and server_scram_salt options.
80 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
81 are now specifically given a HOST_NOT_FOUND response without hitting the
82 system resolver. The library goes on to do the now-standard TXT lookup.
83 Use of dnsdb lookups is not affected.
90 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
91 side, to decode a TLS 1.3 packet capture.
93 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
94 Previously the default library behaviour applied, sending two, each in
97 JH/03 Debug output for ACL now gives the config file name and line number for
100 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
102 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
104 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
105 buffer overrun for (non-chunking) other transports.
107 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
108 TLS1.3, means that a server rejecting a client certificate is not visible
109 to the client until the first read of encrypted data (typically the
110 response to EHLO). Add detection for that case and treat it as a failed
111 TLS connection attempt, so that the normal retry-in-clear can work (if
112 suitably configured).
114 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
115 and/or domain. Found and fixed by Jason Betts.
117 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
118 configuration). If a CNAME target was not a wellformed name pattern, a
121 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
122 the OS reports them interleaved with other addresses.
124 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
125 used both for input and for a verify callout, both encrypted, SMTP
126 responses being sent by the server could be lost. This resulted in
127 dropped connections and sometimes bounces generated by a peer sending
130 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
131 string. Previously it was possible to cause undefined behaviour in a
132 library routine (usually a crash). Found by "zerons".
134 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
137 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
138 API was removed, so update to use the newer ones.
140 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
141 any timeout set, is taking a long time. Previously we would hang on to a
142 rotated logfile "forever" if the input was arriving with long gaps
143 (a previous attempt to fix addressed lack, for a long time, of initial
146 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
147 shared (NFS) environment. The length of the tempfile name is now
148 4 + 16 ("hdr.$message_exim_id") which might break on file
149 systems which restrict the file name length to lower values.
152 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
153 shared (NFS) environment.
155 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
156 did for all versions <4.90). Notably -M, -m, --invert, -I may be
159 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
160 on some platforms for bit 31.
162 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
163 to changes apparently associated with TLS1.3 handling some of the APIs
164 previously used were either nonfunctional or inappropriate. Strings
165 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
166 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
167 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
168 This affects log line X= elements, the $tls_{in,out}_cipher variables,
169 and the use of specific cipher names in the encrypted= ACL condition.
171 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
173 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
174 verification result was not updated unless hosts_require_ocsp applied.
176 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
177 queue_list_requires_admin set to false, non-admin users were denied the
180 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
181 directory-of-certs mode. Previously they were advertised despite the
184 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
185 A single TCP connection by a client will now hold a TLS connection open
186 for multiple message deliveries, by default. Previoud the default was to
189 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
190 default. If built with the facility, DANE will be used. The facility
191 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
193 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
194 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
195 must be defined and you must still, unless you define DISABLE_TLS, manage
196 the the include-dir and library-file requirements that go with that
197 choice. Non-TLS builds are still supported.
199 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
202 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
203 default. If the platform supports and has the facility enabled, it will
204 be requested on all coneections.
206 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
207 controlled by the build-time option SUPPORT_PIPE_CONNECT.
209 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
211 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
212 success-DSN messages. Previously the From: header was always the default
213 one for these; the option was ignored.
215 JH/28 Fix the timeout on smtp response to apply to the whole response.
216 Previously it was reset for every read, so a teergrubing peer sending
217 single bytes within the time limit could extend the connection for a
218 long time. Credit to Qualsys Security Advisory Team for the discovery.
220 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
221 delivery address, which leaked information of the results of local
222 forwarding. Change to the original envelope recipient address, per
225 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
226 requested. Previously not bounce was generated and a log entry of
227 error ignored was made.
229 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
231 JH/32 Introduce a general tainting mechanism for values read from the input
232 channel, and values derived from them. Refuse to expand any tainted
233 values, to catch one form of exploit.
235 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
236 was unused and the unexpanded text used for the test. Found and
237 fixed by Ruben Jenster.
239 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
240 an attempt to use a TLS library read routine dereffed a nul pointer,
243 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
244 them in our smtp response.
246 JH/36 Have the general router option retry_use_local_part default to true when
247 any of the restrictive preconditions are set (to anything). Previously it
248 was only for check_local user. The change removes one item of manual
249 configuration which is required for proper retries when a remote router
250 handles a subset of addresses for a domain.
252 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
253 link count into consideration.
255 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
256 caused the extension of big_buffer, the following lines were ignored.
258 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
259 accordance with RFC 2308. Previously there was no expiry, so a longlived
260 receive process (eg. due to ACL delays) versus a short SOA value could
263 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
265 JH/39 Promote DMARC support to mainline.
267 JH/40 Bug 2452: Add a References: header to DSNs.
269 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
270 parameters. The relevant library call is documented as "Deprecated: This
271 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
272 3.6.0, DH parameters are negotiated following RFC7919."
274 HS/06 Change the default of dnssec_request_domains to "*"
276 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
277 carried on and emitted a BDAT command, even when PIPELINING was not
280 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
281 buffer was used for the filename, resulting in a trap when tainted
282 arguments (eg. $domain) were used.
284 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
285 recommended to avoid a possible server-load attack. The feature can be
286 re-enabled via the openssl_options main cofiguration option.
288 JH/45 local_scan API: documented the current smtp_printf() call. This changed
289 for version 4.90 - adding a "more data" boolean to the arguments.
290 Bumped the ABI version number also, this having been missed previously;
291 release versions 4.90 to 4.92.3 inclusive were effectively broken in
292 respect of usage of smtp_printf() by either local_scan code or libraries
293 accessed via the ${dlfunc } expansion item. Both will need coding
294 adjustment for any calls to smtp_printf() to match the new function
295 signature; a FALSE value for the new argument is always safe.
297 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
298 the file-offset (which the Linux syscall does, and exim expects); this
299 resulted in an indefinite loop.
301 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
302 to do ARC verification. The Authentication-Results: header line added
303 by the configuration then had no ARC item.
309 JH/01 Remove code calling the customisable local_scan function, unless a new
310 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
312 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
313 non-signal-safe functions being used.
315 JH/03 Bug 2269: When presented with a received message having a stupidly large
316 number of DKIM-Signature headers, disable DKIM verification to avoid
317 a resource-consumption attack. The limit is set at twenty.
319 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
320 report of oldest_pass in ${authres } in consequence, and separate out
321 some descriptions of reasons for verification fail.
323 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
324 files in the spool were present and unlocked. A queue-runner could spot
325 them, resulting in a duplicate delivery. Fix that by doing the unlock
326 after the unlink. Investigation by Tim Stewart. Take the opportunity to
327 add more error-checking on spoolfile handling while that code is being
330 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
331 No known attacks, no CVE, this is defensive hardening.
333 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
334 a queue-runner could start a delivery while other operations were ongoing.
335 Cutthrough delivery was a common victim, resulting in duplicate delivery.
336 Found and investigated by Tim Stewart. Fix by using the open message data
337 file handle rather than opening another, and not locally closing it (which
338 releases a lock) for that case, while creating the temporary .eml format
339 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
341 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
342 $sender_verify_failure/$recipient_verify_failure to "random".
344 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
347 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
348 Previously this would segfault.
350 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
353 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
354 like zero, since the resolver should be doing this for us, But we need one
355 as a CNAME but no MX presence gets the CNAME returned; we need to check
356 that doesn't point to an MX to declare it "no MX returned" rather than
357 "error, loop". A new main option is added so the older capability of
358 following some limited number of chain links is maintained.
360 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
362 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
363 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
364 now-deprecated earlier definitions used only specified the range up to TLS
365 1.2 (in the older-version library docs).
367 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
369 JH/15 Rework TLS client-side context management. Stop using a global, and
370 explicitly pass a context around. This enables future use of TLS for
371 connections to service-daemons (eg. malware scanning) while a client smtp
372 connection is using TLS; with cutthrough connections this is quite likely.
374 JH/16 Fix ARC verification to do AS checks in reverse order.
376 JH/17 Support a "tls" option on the ${readsocket } expansion item.
378 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
379 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
380 Previously the "utf8" would be re-prepended for every additional message.
382 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
383 Previously thery were accepted, resulting in issues when attempting to
384 forward messages to a non-supporting MTA.
386 PP/02 Let -n work with printing macros too, not just options.
388 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
389 one parent address was copied, and bogus data was used at delivery-logging
390 time. Either a crash (after delivery) or bogus log data could result.
391 Discovery and analysis by Tim Stewart.
393 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
394 Previously if the string ended mid-character, we did not insert the
395 promised '?' replacement.
397 PP/04 Documentation: current string operators work on bytes, not codepoints.
399 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
400 should pack well giving a smaller memory footprint so better caching and
401 therefore performance. Group the declarations where this can't be done so
402 that the byte-sized flag variables are not interspersed among pointer
403 variables, giving a better chance of good packing by the compiler.
405 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
406 non-null, to avoid issues with sites running BATV. Previously reports were
407 sent with an empty envelope sender so looked like bounces.
409 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
410 The ignore_error flag wasn't being returned from the filter subprocess so
411 was not set for later routers. Investigation and fix by Matthias Kurz.
413 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
414 and a msg:complete for the whole, when a message is manually removed using
415 -Mrm. Developement by Matthias Kurz, hacked on by JH.
417 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
418 a "Gnu special" function, asprintf() in the DB utility binary builds; I
419 hope that is portable enough.
421 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
422 requiring a known-CA anchor certificate; make it now rely entirely on the
423 TLSA as an anchor. Checking the name on the leaf cert against the name
424 on the A-record for the host is still done for TA (but not for EE mode).
426 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
427 included in delivery lines for non-proxied connections, when compiled with
428 SUPPORT_SOCKS and running with proxy logging enabled.
430 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
431 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
432 move the existing event to fire before the normal logging of message
433 failure so that custom logging is bracketed by normal logging.
435 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
436 msg:fail:internal event. Developement by Matthias Kurz.
438 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
439 far too small for todays use of crypto signatures stored there. Go all
440 the way to the max DNS message size of 64kB, even though this might be
441 overmuch for IOT constrained device use.
443 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
444 copy a string over itself. The library routine is documented as not
445 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
447 JH/32 For main options check_spool_space and check_inode_space, where the
448 platform supports 64b integers, support more than the previous 2^31 kB
449 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
450 the previous G, M, k.
452 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
453 $authenticated_fail_id variable on authentication failure. Previously
456 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
457 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
458 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
459 GNUTLS_SEC_PARAM_MEDIUM.
461 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
462 side. Previously we would continue as if no SNI had been received.
464 JH/36 Harden the handling of string-lists. When a list consisted of a sole
465 "<" character, which should be a list-separator specification, we walked
466 off past the nul-terimation.
468 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
469 causes) even when the retry time is not yet met. Previously they were
470 not, meaning that when (say) an account was over-quota and temp-rejecting,
471 and multiple senders' messages were queued, only one sender would get
472 notified on each configured delay_warning cycle.
474 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
476 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
477 cipher-suites, an error can be left on the stack even for a succeeding
478 accept; this results in impossible error messages when a later operation
481 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
482 return error codes indicating retry. Under TLS1.3 this becomes required.
484 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
485 it only wrote the new authenticators, resulting in a lack of tracking of
486 peer changes of ESMTP extensions until the next cache flush.
488 JH/41 Fix the loop reading a message header line to check for integer overflow,
489 and more-often against header_maxsize. Previously a crafted message could
490 induce a crash of the recive process; now the message is cleanly rejected.
492 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
493 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
499 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
500 When redis_servers is set to a list of > 1 element, and the Redis servers
501 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
502 case of MOVED into a DEFER case instead, thus moving the query onto the
503 next server in the list. For a cluster of N elements, all N servers must
504 be defined in redis_servers.
506 GF/02 Catch and remove uninitialized value warning in exiqsumm
507 Check for existence of @ARGV before looking at $ARGV[0]
509 JH/01 Replace the store_release() internal interface with store_newblock(),
510 which internalises the check required to safely use the old one, plus
511 the allocate and data copy operations duplicated in both (!) of the
512 extant use locations.
514 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
515 modifier. This matches the restriction on the commandline.
517 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
518 Previously only the last row was returned.
520 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
521 we assumed that tags in the header were well-formed, and parsed the
522 element content after inspecting only the first char of the tag.
523 Assumptions at that stage could crash the receive process on malformed
526 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
527 While running the DKIM ACL we operate on the Permanent memory pool so that
528 variables created with "set" persist to the DATA ACL. Also (at any time)
529 DNS lookups that fail create cache records using the Permanent pool. But
530 expansions release any allocations made on the current pool - so a dnsdb
531 lookup expansion done in the DKIM ACL releases the memory used for the
532 DNS negative-cache, and bad things result. Solution is to switch to the
533 Main pool for expansions.
534 While we're in that code, add checks on the DNS cache during store_reset,
535 active in the testsuite.
536 Problem spotted, and debugging aided, by Wolfgang Breyha.
538 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
539 When none of the hosts presented to a transport match an already-open
540 connection, close it and proceed with the list. Previously we would
541 queue the message. Spotted by Lena with Yahoo, probably involving
544 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
545 Previously a spurious "250 OK id=" response was appended to the proper
548 JH/08 The "support for" informational output now, which built with Content
549 Scanning support, has a line for the malware scanner interfaces compiled
550 in. Interface can be individually included or not at build time.
552 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
553 by the template makefile "src/EDITME". The "STREAM" support for an older
554 ClamAV interface method is removed.
556 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
557 rows affected is given instead).
559 JH/11 The runtime Berkeley DB library version is now additionally output by
560 "exim -d -bV". Previously only the compile-time version was shown.
562 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
563 SMTP connection. Previously, when one had more receipients than the
564 first, an abortive onward connection was made. Move to full support for
565 multiple onward connections in sequence, handling cutthrough connection
566 for all multi-message initiating connections.
568 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
569 routers. Previously, a multi-recipient message would fail to match the
570 onward-connection opened for the first recipient, and cause its closure.
572 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
573 a timeout on read on a GnuTLS initiating connection, resulting in the
574 initiating connection being dropped. This mattered most when the callout
575 was marked defer_ok. Fix to keep the two timeout-detection methods
578 JH/15 Relax results from ACL control request to enable cutthrough, in
579 unsupported situations, from error to silently (except under debug)
580 ignoring. This covers use with PRDR, frozen messages, queue-only and
583 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
585 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
586 metadata, resulting in a crash in free().
588 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
589 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
590 Broken also in d185889f4, with init system revamp.
592 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
593 Previously we abruptly closed the connection after reading a malware-
594 found indication; now we go on to read the "scan ok" response line,
597 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
598 ACL. Previously, a crash would result.
600 JH/19 Speed up macro lookups during configuration file read, by skipping non-
601 macro text after a replacement (previously it was only once per line) and
602 by skipping builtin macros when searching for an uppercase lead character.
604 JH/20 DANE support moved from Experimental to mainline. The Makefile control
605 for the build is renamed.
607 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
608 was allocated for every new TLS startup, meaning one per message. Fix
609 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
611 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
612 reported the original. Fix to report (as far as possible) the ACL
613 result replacing the original.
615 JH/23 Fix memory leak during multi-message connections using STARTTLS under
616 OpenSSL. Certificate information is loaded for every new TLS startup,
617 and the resources needed to be freed.
619 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
621 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
624 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
625 DATA response info to the (existing) per-recipient response info for
626 the "C=" log element. It can have useful tracking info from the
627 destination system. Patch from Simon Arlott.
629 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
630 length value. Previously this would segfault.
632 HS/02 Support Avast multiline protoocol, this allows passing flags to
633 newer versions of the scanner.
635 JH/28 Ensure that variables possibly set during message acceptance are marked
636 dead before release of memory in the daemon loop. This stops complaints
637 about them when the debug_store option is enabled. Discovered specifically
638 for sender_rate_period, but applies to a whole set of variables.
639 Do the same for the queue-runner and queue-list loops, for variables set
640 from spool message files. Do the same for the SMTP per-message loop, for
641 certain variables indirectly set in ACL operations.
643 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
644 as a multi-recipient message from a mailinglist manager). The coding had
645 an arbitrary cutoff number of characters while checking for more input;
646 enforced by writing a NUL into the buffer. This corrupted long / fast
647 input. The problem was exposed more widely when more pipelineing of SMTP
648 responses was introduced, and one Exim system was feeding another.
649 The symptom is log complaints of SMTP syntax error (NUL chars) on the
650 receiving system, and refused recipients seen by the sending system
651 (propating to people being dropped from mailing lists).
652 Discovered and pinpointed by David Carter.
654 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
655 replaced by the ${authresults } expansion.
657 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
659 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
660 allows proper process termination in container environments.
662 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
663 Previously the "final dot" had a newline after it; ensure it is CR,LF.
665 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
666 and "err_perm", deprecated since 4.83 when the RFC-defined words
667 "temperror" and "permerror" were introduced.
669 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
670 transport-filters or DKIM-signing. The restriction was lost in the
671 consolidation of verify-callout and delivery SMTP handling.
672 Extend the restriction to also cover ARC-signing.
674 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
675 in defer=pass mode supply a 450 to the initiator. Previously the message
678 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
679 tls_require_ciphers is used as before.
681 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
682 "pass_unscanned". Only tmpfails from the scanner are written to
683 the paniclog, as they may require admin intervention (permission
684 denied, license issues). Other scanner errors (like decompression
685 bombs) do not cause a paniclog entry.
687 JH/36 Fix reinitialisation of DKIM logging variable between messages.
688 Previously it was possible to log spurious information in receive log
691 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
692 triggered odd behaviour from Outlook Express clients.
694 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
697 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
698 since the IETF WG has not yet settled on that versus the original
699 "bare" representation.
701 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
702 Previously the millisecond value corrupted the output.
703 Fix also for syslog_pid=no and log_selector +pid, for which the pid
704 corrupted the output.
710 JH/01 Rework error string handling in TLS interface so that the caller in
711 more cases is responsible for logging. This permits library-sourced
712 string to be attached to addresses during delivery, and collapses
713 pairs of long lines into single ones.
715 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
716 during configuration. Wildcards are allowed and expanded.
718 JH/02 Rework error string handling in DKIM to pass more info back to callers.
719 This permits better logging.
721 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
722 do not close it down and have the child transport start it up again on
723 the passed-on TCP connection. Instead, proxy the child (and any
724 subsequent ones) for TLS via a unix-domain socket channel. Logging is
725 affected: the continued delivery log lines do not have any DNSSEC, TLS
726 Certificate or OCSP information. TLS cipher information is still logged.
728 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
729 identical IP addresses on different listening ports. Will also affect
732 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
733 add noisy ifdef guards to special-case this sillyness.
734 Patch from Bernd Kuhls.
736 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
737 than 255 are no longer allowed.
739 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
740 Disable the session-cache too, which might reduce our load. Since we
741 currrectly use a new context for every connection, both as server and
742 client, there is no benefit for these.
743 GnuTLS appears to not support tickets server-side by default (we don't
744 call gnutls_session_ticket_enable_server()) but client side is enabled
745 by default on recent versions (3.1.3 +) unless the PFS priority string
748 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
749 <https://reproducible-builds.org/specs/source-date-epoch/>.
751 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
752 the check for any unsuccessful recipients did not notice the limit, and
753 erroneously found still-pending ones.
755 JH/08 Pipeline CHUNKING command and data together, on kernels that support
756 MSG_MORE. Only in-clear (not on TLS connections).
758 JH/09 Avoid using a temporary file during transport using dkim. Unless a
759 transport-filter is involved we can buffer the headers in memory for
760 creating the signature, and read the spool data file once for the
761 signature and again for transmission.
763 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
764 4.77 as the kernel support then wasn't solid, having issues in 64bit
765 mode. Now, it's been long enough. Add support for FreeBSD also.
767 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
768 case where the routing stage had gathered several addresses to send to
769 a host before calling the transport for the first, we previously failed
770 to close down TLS in the old transport process before passing the TCP
771 connection to the new process. The new one sent a STARTTLS command
772 which naturally failed, giving a failed delivery and bloating the retry
773 database. Investigation and fix prototype from Wolfgang Breyha.
775 JH/12 Fix check on SMTP command input synchronisation. Previously there were
776 false-negatives in the check that the sender had not preempted a response
777 or prompt from Exim (running as a server), due to that code's lack of
778 awareness of the SMTP input buffering.
780 PP/04 Add commandline_checks_require_admin option.
781 Exim drops privileges sanely, various checks such as -be aren't a
782 security problem, as long as you trust local users with access to their
783 own account. When invoked by services which pass untrusted data to
784 Exim, this might be an issue. Set this option in main configuration
785 AND make fixes to the calling application, such as using `--` to stop
788 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
789 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
790 responses to those, into a single TLS record each way (this usually means
791 a single packet). As a side issue, smtp_enforce_sync now works on TLS
794 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
795 affects you only if you're dancing at the edge of the param size limits.
796 If you are, and this message makes sense to you, then: raise the
797 configured limit or use OpenSSL 1.1. Nothing we can do for older
800 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
801 cmdline element to get the documented default one. Previously it was
804 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
805 are unsuable for non-message-specific reasons - eg. connection timeout,
806 banner-time rejection.
808 JH/16 Fix logging of delivery remote port, when specified by router, under
811 PP/06 Repair manualroute's ability to take options in any order, even if one
812 is the name of a transport.
815 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
817 JH/17 Change the list-building routines interface to use the expanding-string
818 triplet model, for better allocation and copying behaviour.
820 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
821 Previously it was constructed the first time a possibly-matching string
822 was met in the configuration file input during startup; now it is done
825 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
826 compatible one, to avoid the (poorly documented) possibility of a config
827 file in the working directory redirecting the DB files, possibly correpting
828 some existing file. CVE-2017-10140 assigned for BDB.
830 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
831 cache-hot. Previously, although the result was properly cached, the
832 initial verify call returned a defer.
834 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
835 the main verify for receipient in uncached-mode.
837 JH/22 Retire historical build files to an "unsupported" subdir. These are
838 defined as "ones for which we have no current evidence of testing".
840 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
841 if present. Previously it was ignored.
843 JH/24 Start using specified-initialisers in C structure init coding. This is
844 a C99 feature (it's 2017, so now considered safe).
846 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
847 if was a fixed-sized field and bitmask ops via macros; it is now more
850 PP/07 GitHub PR 56: Apply MariaDB build fix.
851 Patch provided by Jaroslav Škarvada.
853 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
854 during Coverity cleanups [4.87 JH/47]
855 Diagnosis and fix provided by Michael Fischer v. Mollard.
857 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
858 the right size to place the terminating semicolon on its own folded
859 line, the header hash was calculated to an incorrect value thanks to
860 the (relaxed) space the fold became.
862 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
863 and confused the parent.
865 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
866 which could crash as a result. This could lead to undeliverable messages.
868 JH/28 Logging: "next input sent too soon" now shows where input was truncated
871 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
872 matters on fast-turnover and PID-randomising systems, which were getting
873 out-of-order delivery.
875 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
876 a possibly-overlapping copy. The symptom was that "Remote host closed
877 connection in response to HELO" was logged instead of the actual 4xx
880 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
881 Previously only that bufferd was discarded, resulting in SYMTP command
884 JH/32 DKIM: when a message has multiple signatures matching an identity given
885 in dkim_verify_signers, run the dkim acl once for each. Previously only
886 one run was done. Bug 2189.
888 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
889 "panic the current process" to "deliberately defer". The panic log is
890 still written with the problem list name; the mail and reject logs now
891 get a temp-reject line for the message that was being handled, saying
892 something like "domains check lookup or other defer". The SMTP 451
893 message is still "Temporary local problem".
895 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
896 A crafted sequence of BDAT commands could result in in-use memory beeing
897 freed. CVE-2017-16943.
899 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
900 from SMTP input. Previously it was always done; now only done for DATA
901 and not BDAT commands. CVE-2017-16944.
903 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
904 to the message (such as an overlong header line). Previously this was
905 not done and we did not exit BDAT mode. Followon from the previous item
906 though a different problem.
912 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
913 than -2003 did; needs libidn2 in addition to libidn.
915 JH/02 The path option on a pipe transport is now expanded before use.
917 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
918 Patch provided by "Björn", documentation fix added too.
920 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
921 missing a wire-to-host endian conversion.
923 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
924 close after a BDAT command line could be taken as a following command,
925 giving a synch failure. Fix by only checking for synch immediately
926 before acknowledging the chunk.
928 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
929 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
930 macro. Patches provided by Josh Soref.
932 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
933 Previously we did not; the RFC seems ambiguous and VRFY is not listed
934 by IANA as a service extension. However, John Klensin suggests that we
937 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
938 the dkim code may be unix-mode line endings rather than smtp wire-format
939 CRLF, so prepend a CR to any bare LF.
941 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
942 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
944 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
945 the first were themselves being wrongly included in the feed into dkim
946 processing; with most chunk sizes in use this resulted in an incorrect
947 body hash calculated value.
949 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
950 DKIM signature block, for verification. Although advised against by
951 standards it is specifically not ruled illegal.
953 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
955 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
956 missing a body hash (the bh= tag).
958 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
959 It seems that HAProxy sends the Proxy Protocol information in clear and
960 only then does a TLS startup, so do the same.
962 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
963 TCP connections (such as for Spamd) unless the daemon successfully set
964 Fast Open mode on its listening sockets. This fixes breakage seen on
965 too-old kernels or those not configured for Fast Open, at the cost of
966 requiring both directions being enabled for TFO, and TFO never being used
967 by non-daemon-related Exim processes.
969 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
970 endings, at least on the first header line. Try to canonify any that get
971 past that check, despite the cost.
973 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
974 now limited to an arbitrary five deep, while parsing addresses with the
975 strip_excess_angle_brackets option enabled.
977 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
978 instead leave the unprompted TLS handshake in socket buffer for the
979 TLS library to consume.
981 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
983 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
985 JH/16 Drop variables when they go out of scope. Memory management drops a whole
986 region in one operation, for speed, and this leaves assigned pointers
987 dangling. Add checks run only under the testsuite which checks all
988 variables at a store-reset and panics on a dangling pointer; add code
989 explicitly nulling out all the variables discovered. Fixes one known
990 bug: a transport crash, where a dangling pointer for $sending_ip_address
991 originally assigned in a verify callout, is re-used.
993 PP/06 Drop '.' from @INC in various Perl scripts.
995 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
997 PP/08 Reduce a number of compilation warnings under clang; building with
998 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
999 should be warning-free.
1001 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1003 HS/01 Fix portability problems introduced by PP/08 for platforms where
1004 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1006 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1007 chunk. This allows us to accept broken chunked messages. We need a more
1008 general solution here.
1010 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1011 already-broken messages in the queue.
1013 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1015 JH/19 Fix reference counting bug in routing-generated-address tracking.
1021 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1022 supports it and a size is available (ie. the sending peer gave us one).
1024 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1025 years of being deprecated). The replacements are the ACLs
1026 acl_smtp_mime and acl_not_smtp_mime.
1028 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1029 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1030 or even an in-clear connection were permitted. Now, if the host lookup
1031 was dnssec and dane was requested then the host is only used if the
1032 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1033 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1034 if one fails this test.
1035 This means that a poorly-configured remote DNS will make it incommunicado;
1036 but it protects against a DNS-interception attack on it.
1038 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1039 non-noisy when a race steals the message being considered.
1041 JH/05 If main configuration option tls_certificate is unset, generate a
1042 self-signed certificate for inbound TLS connections.
1044 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1045 in rewrites and routers.
1047 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1048 and logged a warning sing 4.83; now they are a configuration file error.
1050 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1051 (lacking @domain). Apply the same qualification processing as RCPT.
1053 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1055 JH/10 Support ${sha256:} applied to a string (as well as the previous
1058 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1059 a cutthrough deliver is pending, as we always want to make a connection.
1060 This also avoids re-routing the message when later placing the cutthrough
1061 connection after a verify cache hit.
1062 Do not update it with the verify result either.
1064 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1065 when routing results in more than one destination address.
1067 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1068 signing (which inhibits the cutthrough capability). Previously only
1069 the presence of an option was tested; now an expansion evaluating as
1070 empty is permissible (obviously it should depend only on data available
1071 when the cutthrough connection is made).
1073 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1074 the relevant preceding SMTP command did not note the pipelining mode.
1076 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1077 Previously they were not counted.
1079 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1080 as one having no matching records. Previously we deferred the message
1081 that needed the lookup.
1083 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1084 distinguished as "(=".
1086 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1087 for missing MX records. Previously it only worked for missing A records.
1089 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1091 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1092 after the data-go-ahead and data-ack. Patch from Jason Betts.
1094 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1095 even for a "none" policy. Patch from Tony Meyer.
1097 JH/22 Fix continued use of a connection for further deliveries. If a port was
1098 specified by a router, it must also match for the delivery to be
1101 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1102 When one of the recipients of a message was unsuitable for the connection
1103 (has no matching addresses), we lost track of needing to mark it
1104 deferred. As a result mail would be lost.
1106 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1108 JH/25 Decoding ACL controls is now done using a binary search; the source code
1109 takes up less space and should be simpler to maintain. Merge the ACL
1110 condition decode tables also, with similar effect.
1112 JH/26 Fix problem with one_time used on a redirect router which returned the
1113 parent address unchanged. A retry would see the parent address marked as
1114 delivered, so not attempt the (identical) child. As a result mail would
1117 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1118 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1119 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1122 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1123 Main config options check_{log,spool}_{inodes,space} are now
1124 100 inodes, 10MB unless set otherwise in the configuration.
1126 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1127 Previously it only applied to the main-section connection policy
1130 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1132 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1133 by me. Added RFC7919 DH primes as an alternative.
1135 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1136 are not in the system include path.
1138 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1139 GnuTLS, when a session startup failed (eg because the client disconnected)
1140 Exim did stdio operations after fclose. This was exposed by a recent
1141 change which nulled out the file handle after the fclose.
1143 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1144 signed directly by the cert-signing cert, rather than an intermediate
1145 OCSP-signing cert. This is the model used by LetsEncrypt.
1147 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1149 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1150 an incoming connection.
1152 HS/02 Bug 1802: Do not half-close the connection after sending a request
1155 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1156 fallback to "prime256v1".
1158 JH/34 SECURITY: Use proper copy of DATA command in error message.
1159 Could leak key material. Remotely exploitable. CVE-2016-9963.
1165 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1166 and 3.4.4 - once the server is enabled to respond to an OCSP request
1167 it does even when not requested, resulting in a stapling non-aware
1168 client dropping the TLS connection.
1170 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1171 support variable-length bit vectors. No functional change.
1173 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1174 The I= interface field on outgoing lines is now after the H= remote
1175 host field, same as incoming lines. There is a separate
1176 outgoing_interface log selector which allows you to disable the
1179 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1180 If not running log_selector +smtp_connection the mainlog would be held
1181 open indefinitely after a "too many connections" event, including to a
1182 deleted file after a log rotate. Leave the per net connection logging
1183 leaving it open for efficiency as that will be quickly detected by the
1184 check on the next write.
1186 HS/01 Bug 1671: Fix post transport crash.
1187 Processing the wait-<transport> messages could crash the delivery
1188 process if the message IDs didn't exist for some reason. When
1189 using 'split_spool_directory=yes' the construction of the spool
1190 file name failed already, exposing the same netto behaviour.
1192 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1193 mime_regex ACL conditions.
1195 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1196 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1197 to HELO, local diagnostic string.
1199 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1200 log line to debug. Even when configured with a tls authenticator many
1201 client connections are expected to not authenticate in this way, so
1202 an authenticate fail is not an error.
1204 HS/02 Add the Exim version string to the process info. This way exiwhat
1205 gives some more detail about the running daemon.
1207 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1208 matter for fast-change records such as DNSBLs.
1210 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1211 retry record, even if constant. There may be multiple transports with
1212 different interface settings and the retry behaviour needs to be kept
1215 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1217 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1219 JH/10 Bug 840: fix log_defer_output option of pipe transport
1221 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1222 hosts_randomize. This matters a lot when combined with mua_wrapper.
1224 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1225 ${quote_pgsql:<string>} operator.
1227 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1229 JH/14 Change header file name for freeradius-client. Relevant if compiling
1230 with Radius support; from the Gentoo tree and checked under Fedora.
1232 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1234 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1235 option transport_filter, meaning no filtering.
1237 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1239 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1240 defaults to "*" (all hosts). The variable is now available when not built
1241 with TLS, default unset, mainly to enable keeping the testsuite sane.
1242 If a server certificate is not supplied (via tls_certificate) an error is
1243 logged, and clients will find TLS connections fail on startup. Presumably
1244 they will retry in-clear.
1245 Packagers of Exim are strongly encouraged to create a server certificate
1246 at installation time.
1248 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1249 with the $config_file variable.
1251 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1252 in transport context, after the attempt, and per-recipient. The latter type
1253 is per host attempted. The event data is the error message, and the errno
1254 information encodes the lookup type (A vs. MX) used for the (first) host,
1255 and the trailing two digits of the smtp 4xx response.
1257 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1258 to write to mainlog (or rejectlog, paniclog) in the window between file
1259 creation and permissions/ownership being changed. Particularly affects
1260 installations where exicyclog is run as root, rather than exim user;
1261 result is that the running daemon panics and dies.
1263 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1265 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1266 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1267 "pri" and "weight". Note that the previous implicit priority given by the
1268 list order is no longer honoured.
1270 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1271 for DKIM processing.
1273 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1274 by defining SUPPORT_SOCKS.
1276 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1277 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1278 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1279 variables are renamed to proxy_{local,external}_{address,port}.
1281 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1282 for a build by defining SUPPORT_I18N
1284 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1285 of the query string, and make ${quote_redis:} do that quoting.
1287 JH/29 Move Events support from Experimental to mainline, enabled by default
1288 and removable for a build by defining DISABLE_EVENT.
1290 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1292 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1293 cached by the daemon.
1295 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1296 by defining LOOKUP_REDIS. The libhiredis library is required.
1298 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1299 keys are given for lookup.
1301 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1302 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1303 only supported when built with TLS support. The PolarSSL SHA routines
1304 are still used when the TLS library is too old for convenient support.
1306 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1307 openssl_options), for security. OpenSSL forces this from version 1.1.0
1308 server-side so match that on older versions.
1310 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1311 allocation for $value could be released as the expansion processing
1312 concluded, but leaving the global pointer active for it.
1314 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1315 and to use the domains and local_parts ACL conditions.
1317 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1318 incorrectly not doubled on cutthrough transmission, hence seen as a
1319 body-termination at the receiving system - resulting in truncated mails.
1320 Commonly the sender saw a TCP-level error, and retransmitted the message
1321 via the normal store-and-forward channel. This could result in duplicates
1322 received - but deduplicating mailstores were liable to retain only the
1323 initial truncated version.
1325 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1327 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1329 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1330 we're in there, support oversigning also; bug 1309.
1332 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1334 HS/04 Add support for keep_environment and add_environment options.
1336 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1337 either intentional arithmetic overflow during PRNG, or testing config-
1340 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1341 delivery resulted in actual delivery. Cancel cutthrough before DATA
1344 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1345 rejects a recipient: pass the reject to the originator.
1347 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1348 Many were false-positives and ignorable, but it's worth fixing the
1351 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1352 for the new environment-manipulation done at startup. Move the routines
1353 from being local to tls.c to being global via the os.c file.
1355 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1356 an extract embedded as result-arg for a map, the first arg for extract
1357 is unavailable so we cannot tell if this is a numbered or keyed
1358 extraction. Accept either.
1364 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1367 JH/02 The smtp transport option "multi_domain" is now expanded.
1369 JH/03 The smtp transport now requests PRDR by default, if the server offers
1372 JH/04 Certificate name checking on server certificates, when exim is a client,
1373 is now done by default. The transport option tls_verify_cert_hostnames
1374 can be used to disable this per-host. The build option
1375 EXPERIMENTAL_CERTNAMES is withdrawn.
1377 JH/05 The value of the tls_verify_certificates smtp transport and main options
1378 default to the word "system" to access the system default CA bundle.
1379 For GnuTLS, only version 3.0.20 or later.
1381 JH/06 Verification of the server certificate for a TLS connection is now tried
1382 (but not required) by default. The verification status is now logged by
1383 default, for both outbound TLS and client-certificate supplying inbound
1386 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1389 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1390 Status Notification (bounce) messages are now MIME format per RFC 3464.
1391 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1392 under the control of the dsn_advertise_hosts option, and routers may
1393 have a dsn_lasthop option.
1395 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1396 default, modifiable by a malware= option. The list separator for
1397 the options can now be changed in the usual way. Bug 68.
1399 JH/10 The smtp_receive_timeout main option is now expanded before use.
1401 JH/11 The incoming_interface log option now also enables logging of the
1402 local interface on delivery outgoing connections.
1404 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1405 if the interface and destination host and port all match.
1407 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1410 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1411 Patch from Andrew Lewis.
1413 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1414 now supports optional time-restrictions, weighting, and priority
1415 modifiers per server. Patch originally by <rommer@active.by>.
1417 JH/16 The spamd_address main option now supports a mixed list of local
1418 and remote servers. Remote servers can be IPv6 addresses, and
1419 specify a port-range.
1421 JH/17 Bug 68: The spamd_address main option now supports an optional
1422 timeout value per server.
1424 JH/18 Bug 1581: Router and transport options headers_add/remove can
1425 now have the list separator specified.
1427 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1430 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1433 JH/21 Support for the A6 type of dns record is withdrawn.
1435 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1436 rather than the verbs used.
1438 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1439 from 255 to 1024 chars.
1441 JH/24 Verification callouts now attempt to use TLS by default.
1443 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1444 are generic router options now. The defaults didn't change.
1446 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1447 Original patch from Alexander Shikoff, worked over by JH.
1449 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1450 files if ZCAT_COMMAND is not executable.
1452 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1454 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1456 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1457 Normally benign, it bites when the pair was led to by a CNAME;
1458 modern usage is to not canonicalize the domain to a CNAME target
1459 (and we were inconsistent anyway for A-only vs AAAA+A).
1461 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1463 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1464 when evaluating $sender_host_dnssec.
1466 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1467 $sender_helo_dnssec variable.
1469 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1471 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1473 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1475 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1476 documented as working, but never had. Support all but $spam_report.
1478 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1479 added for tls authenticator.
1481 HS/03 Add perl_taintmode main config option
1487 TL/01 When running the test suite, the README says that variables such as
1488 no_msglog_check are global and can be placed anywhere in a specific
1489 test's script, however it was observed that placement needed to be near
1490 the beginning for it to behave that way. Changed the runtest perl
1491 script to read through the entire script once to detect and set these
1492 variables, reset to the beginning of the script, and then run through
1493 the script parsing/test process like normal.
1495 TL/02 The BSD's have an arc4random API. One of the functions to induce
1496 adding randomness was arc4random_stir(), but it has been removed in
1497 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1498 function when detected.
1500 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1501 cause callback expansion.
1503 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1504 syntax errors in an expansion can be treated as a string instead of
1505 logging or causing an error, due to the internal use of bool_lax
1506 instead of bool when processing it.
1508 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1509 server certificates when making smtp deliveries.
1511 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1513 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1515 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1517 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1518 Merged patch from Sebastian Wiedenroth.
1520 JH/05 Fix results-pipe from transport process. Several recipients, combined
1521 with certificate use, exposed issues where response data items split
1522 over buffer boundaries were not parsed properly. This eventually
1523 resulted in duplicates being sent. This issue only became common enough
1524 to notice due to the introduction of connection certificate information,
1525 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1527 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1528 size buffer was used, resulting in syntax errors when an expansion
1531 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1532 version 3.3.6 or later.
1534 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1535 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1536 both become "event_action", the variables become $event_name, $event_data
1537 and $event_defer_errno. There is a new variable $verify_mode, usable in
1538 routers, transports and related events. The tls:cert event is now also
1539 raised for inbound connections, if the main configuration event_action
1542 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1543 early OCSP support, but no stapling (appears to be less than 1.0.0).
1545 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1546 server certificate names available under the smtp transport option
1547 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1550 JH/10 Time-related extraction expansions from certificates now use the main
1551 option "timezone" setting for output formatting, and are consistent
1552 between OpenSSL and GnuTLS compilations. Bug 1541.
1554 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1555 encoded parameter in the incoming message. Bug 1558.
1557 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1558 include certificate info, eximon was claiming there were spoolfile
1561 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1563 JH/14 Log delivery-related information more consistently, using the sequence
1564 "H=<name> [<ip>]" wherever possible.
1566 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1567 are problematic for Debian distribution, omit them from the release
1570 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1572 JH/16 Fix string representation of time values on 64bit time_t architectures.
1575 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1576 output list separator was used.
1581 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1582 checkers that were complaining about end of non-void function with no
1585 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1586 This was a regression introduced in 4.83 by another bugfix.
1588 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1590 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1591 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1597 TF/01 Correctly close the server side of TLS when forking for delivery.
1599 When a message was received over SMTP with TLS, Exim failed to clear up
1600 the incoming connection properly after forking off the child process to
1601 deliver the message. In some situations the subsequent outgoing
1602 delivery connection happened to have the same fd number as the incoming
1603 connection previously had. Exim would try to use TLS and fail, logging
1604 a "Bad file descriptor" error.
1606 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1607 utilities have not been installed.
1609 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1610 temporary space as the ACL may create new global variables.
1612 TL/01 LDAP support uses per connection or global context settings, depending
1613 upon the detected version of the libraries at build time.
1615 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1616 to extract and use the src ip:port in logging and expansions as if it
1617 were a direct connection from the outside internet. PPv2 support was
1618 updated based on HAProxy spec change in May 2014.
1620 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1622 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1623 Properly escape header and check for NULL return.
1625 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1628 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1630 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1631 characters in header names, implemented as a verify condition.
1632 Contributed by Michael Fischer v. Mollard.
1634 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1635 results permerror and temperror. Previous values are deprecated but
1636 still accepted. In a future release, err_perm and err_temp will be
1637 completely removed, which will be a backward incompatibility if the
1638 ACL tests for either of these two old results. Patch contributed by
1639 user bes-internal on the mailing list.
1641 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1643 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1644 selectors, in both main and reject logs.
1646 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1649 JH/07 Add malware type "sock" for talking to simple daemon.
1651 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1653 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1654 routers/transports under cutthrough routing.
1656 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1657 numbers. Touch up "bool" conditional to keep the same definition.
1659 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1661 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1663 JH/12 Expand items in router/transport headers_add or headers_remove lists
1664 individually rather than the list as a whole. Bug 1452.
1666 Required for reasonable handling of multiple headers_ options when
1667 they may be empty; requires that headers_remove items with embedded
1668 colons must have them doubled (or the list-separator changed).
1670 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1671 view the policy declared in the DMARC record. Currently, $dmarc_status
1672 is a combined value of both the record presence and the result of the
1675 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1677 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1678 dnslookup router and the smtp transport (applying to the forward
1681 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1682 of ldap servers used for a specific lookup. Patch provided by Heiko
1685 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1686 New variable $lookup_dnssec_authenticated for observability.
1688 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1689 Patch submitted by Lars Timman.
1691 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1693 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1694 Requires trusted mode and valid format message id, aborts otherwise.
1695 Patch contributed by Heiko Schlichting.
1697 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1698 certextract with support for various fields. Bug 1358.
1700 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1701 is requested by default, modifiable by smtp transport option
1704 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1705 operate on certificate variables to give certificate fingerprints
1706 Also new ${sha256:cert_variable}.
1708 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1710 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1711 Christian Aistleitner.
1713 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1715 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1716 file. Patch from Wolfgang Breyha.
1718 JH/25 Expand the coverage of the delivery $host and $host_address to
1719 client authenticators run in verify callout. Bug 1476.
1721 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1722 align with daemon_smtp_ports. Bug 72.
1724 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1725 support and error reporting did not work properly.
1727 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1728 and is readable. Patch from Andrew Colin Kissa.
1730 TL/14 Enhance documentation of ${run expansion and how it parses the
1731 commandline after expansion, particularly in the case when an
1732 unquoted variable expansion results in an empty value.
1734 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1736 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1737 by renaming away from it. Addresses GH issue 15, reported by
1740 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1742 TL/15 SECURITY: prevent double expansion in math comparison functions
1743 (can expand unsanitized data). Not remotely exploitable.
1750 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1752 PP/02 Make -n do something, by making it not do something.
1753 When combined with -bP, the name of an option is not output.
1755 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1758 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1759 $sender_host_name and config options to manage this, and basic check
1762 PP/05 DSCP support for outbound connections and control modifier for inbound.
1764 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1765 (Only plugin which currently uses this is kerberos4, which nobody should
1766 be using, but we should make it available and other future plugins might
1767 conceivably use it, even though it would break NAT; stuff *should* be
1768 using channel bindings instead).
1770 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1771 name; added for Sendmail compatibility; requires admin caller.
1772 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1773 just ignore it); requires trusted caller.
1774 Also parse but ignore: -Ac -Am -X<logfile>
1777 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1779 TL/02 Add +smtp_confirmation as a default logging option.
1781 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1782 Patch by Magnus Holmgren from 2007-02-20.
1784 TL/04 Bugzilla 1281 - Spec typo.
1785 Bugzilla 1283 - Spec typo.
1786 Bugzilla 1290 - Spec grammar fixes.
1788 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1790 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1792 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1793 mailing list by by Dmitry Isaikin.
1795 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1797 JH/02 Support "G" suffix to numbers in ${if comparisons.
1799 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1801 NM/01 Bugzilla 1197 - Spec typo
1802 Bugzilla 1196 - Spec examples corrections
1804 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1806 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1807 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1810 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1811 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1813 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1814 "acl {{name}{arg}...}", and optional args on acl condition
1817 JH/05 Permit multiple router/transport headers_add/remove lines.
1819 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1821 JH/07 Avoid using a waiting database for a single-message-only transport.
1822 Performance patch from Paul Fisher. Bugzilla 1262.
1824 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1827 JH/09 Add $headers_added variable, with content from use of ACL modifier
1828 add_header (but not yet added to the message). Bugzilla 199.
1830 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1831 Pulled from Bugzilla 817 by Wolfgang Breyha.
1833 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1835 (nb: this is the same fix as in Exim 4.80.1)
1837 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1840 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1843 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1845 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1846 advertise SMTP AUTH mechanism to us, instead of a generic
1847 protocol violation error. Also, make Exim more robust to bad
1848 data from the Dovecot auth socket.
1850 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1852 When a queue runner is handling a message, Exim first routes the
1853 recipient addresses, during which it prunes them based on the retry
1854 hints database. After that it attempts to deliver the message to
1855 any remaining recipients. It then updates the hints database using
1858 So if a recipient address works intermittently, it can get repeatedly
1859 deferred at routing time. The retry hints record remains fresh so the
1860 address never reaches the final cutoff time.
1862 This is a fairly common occurrence when a user is bumping up against
1863 their storage quota. Exim had some logic in its local delivery code
1864 to deal with this. However it did not apply to per-recipient defers
1865 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1867 This change adds a proper retry rule check during routing so that the
1868 final cutoff time is checked against the message's age. We only do
1869 this check if there is an address retry record and there is not a
1870 domain retry record; this implies that previous attempts to handle
1871 the address had the retry_use_local_parts option turned on. We use
1872 this as an approximation for the destination being like a local
1873 delivery, as in LMTP.
1875 I suspect this new check makes the old local delivery cutoff check
1876 redundant, but I have not verified this so I left the code in place.
1878 TF/02 Correct gecos expansion when From: is a prefix of the username.
1880 Test 0254 submits a message to Exim with the header
1884 When I ran the test suite under the user fanf2, Exim expanded
1885 the header to contain my full name, whereas it should have added
1886 a Resent-Sender: header. It erroneously treats any prefix of the
1887 username as equal to the username.
1889 This change corrects that bug.
1891 GF/01 DCC debug and logging tidyup
1892 Error conditions log to paniclog rather than rejectlog.
1893 Debug lines prefixed by "DCC: " to remove any ambiguity.
1895 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1897 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1898 Bug spotted by Jeremy Harris; was flawed since initial commit.
1899 Would have resulted in OCSP responses post-SNI triggering an Exim
1900 NULL dereference and crash.
1902 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1904 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1905 Bug detection, analysis and fix by Samuel Thibault.
1906 Bugzilla 1331, Debian bug #698092.
1908 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1910 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1911 Server implementation by Todd Lyons, client by JH.
1912 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1913 config variable "prdr_enable" controls whether the server
1914 advertises the facility. If the client requests PRDR a new
1915 acl_data_smtp_prdr ACL is called once for each recipient, after
1916 the body content is received and before the acl_smtp_data ACL.
1917 The client is controlled by both of: a hosts_try_prdr option
1918 on the smtp transport, and the server advertisement.
1919 Default client logging of deliveries and rejections involving
1920 PRDR are flagged with the string "PRDR".
1922 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1923 fclose(). Diagnosis by Todd Lyons.
1925 PP/17 Update configure.default to handle IPv6 localhost better.
1926 Patch by Alain Williams (plus minor tweaks).
1929 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1930 This is now consistent with GnuTLS, and is now documented: the
1931 previous undocumented portable approach to treating the option as
1932 unset was to force an expansion failure. That still works, and
1933 an empty string is now equivalent.
1935 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1936 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1937 not performing validation itself.
1939 PP/20 Added force_command boolean option to pipe transport.
1940 Patch from Nick Koston, of cPanel Inc.
1942 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1945 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1947 PP/21 Fix eximon continuous updating with timestamped log-files.
1948 Broken in a format-string cleanup in 4.80, missed when I repaired the
1949 other false fix of the same issue.
1950 Report and fix from Heiko Schlichting.
1953 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1954 Report from Prashanth Katuri.
1956 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1957 It's SecureTransport, so affects any MacOS clients which use the
1958 system-integrated TLS libraries, including email clients.
1960 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1961 using a MIME ACL for non-SMTP local injection.
1962 Report and assistance in diagnosis by Warren Baker.
1964 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1966 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1968 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1969 last id that failed so it may be referenced in subsequent ACL's.
1971 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1974 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1975 ldap library initialization, allowing self-signed CA's to be
1976 used. Also properly sets require_cert option later in code by
1977 using NULL (global ldap config) instead of ldap handle (per
1978 session). Bug diagnosis and testing by alxgomz.
1980 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1981 the src/util/ subdirectory.
1983 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1984 renamed to Transport Post Delivery Action by Jeremy Harris, as
1987 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1988 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1989 redis_servers = needs to be configured which will be used by the redis
1990 lookup. Patch from Warren Baker, of The Packet Hub.
1992 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1994 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1995 hostname or reverse DNS when processing a host list. Used suggestions
1996 from multiple comments on this bug.
1998 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2000 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2001 Missed a few lines, added it to make the runtest require no keyboard
2004 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2005 contains upper case chars. Make router use caseful_local_part.
2007 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2008 support when GnuTLS has been built with p11-kit.
2014 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2016 This, or similar/improved, will also be change PP/11 of 4.82.
2022 PP/01 Handle short writes when writing local log-files.
2023 In practice, only affects FreeBSD (8 onwards).
2024 Bugzilla 1053, with thanks to Dmitry Isaikin.
2026 NM/01 Bugzilla 949 - Documentation tweak
2028 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2031 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2033 PP/02 Implemented gsasl authenticator.
2035 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2037 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2038 `pkg-config foo` for cflags/libs.
2040 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2041 with rest of GSASL and with heimdal_gssapi.
2043 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2044 `pkg-config foo` for cflags/libs for the TLS implementation.
2046 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2047 properties get this fed in as external SSF. A number of robustness
2048 and debugging improvements to the cyrus_sasl authenticator.
2050 PP/08 cyrus_sasl server now expands the server_realm option.
2052 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2053 Patch by Jeremy Harris.
2055 PP/10 Added dbmjz lookup type.
2057 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2059 PP/12 MAIL args handles TAB as well as SP, for better interop with
2060 non-compliant senders.
2061 Analysis and variant patch by Todd Lyons.
2063 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2064 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2065 Patch from Dirk Mueller <dmueller@suse.com>
2067 PP/13 tls_peerdn now print-escaped for spool files.
2068 Observed some $tls_peerdn in wild which contained \n, which resulted
2069 in spool file corruption.
2071 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2072 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2073 or write after TLS renegotiation, which otherwise led to messages
2076 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2077 as a tracking header (ie: a signed header comes before the signature).
2078 Patch from Wolfgang Breyha.
2080 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2081 comma-sep list; embedded commas doubled.
2083 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2085 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2087 Report and patch from Dmitry Banschikov.
2089 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2090 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2091 failures. We appear to now support TLS1.1+ with Exim.
2093 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2094 lets Exim select keys and certificates based upon TLS SNI from client.
2095 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2096 before an outbound SMTP session. New log_selector, +tls_sni.
2098 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2099 NULL dereference. Report and patch from Alun Jones.
2101 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2102 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2103 Not seeing resolver debug output on NetBSD, but suspect this is a
2104 resolver implementation change.
2106 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2107 Left warnings. Added "eximon gdb" invocation mode.
2109 PP/21 Defaulting "accept_8bitmime" to true, not false.
2111 PP/22 Added -bw for inetd wait mode support.
2113 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2114 locate the relevant includes and libraries. Made this the default.
2116 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2117 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2119 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2120 This may cause build issues on older platforms.
2122 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2123 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2124 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2125 Added SNI support via GnuTLS too.
2126 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2128 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2130 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2131 Applied second patch from Janne, implementing suggestion to default
2132 multiple-strings-in-record handling to match SPF spec.
2134 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2136 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2137 read-only, out of scope).
2138 Patch from Wolfgang Breyha, report from Stuart Northfield.
2140 PP/29 Fix three issues highlighted by clang analyser static analysis.
2141 Only crash-plausible issue would require the Cambridge-specific
2142 iplookup router and a misconfiguration.
2143 Report from Marcin Mirosław.
2145 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2147 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2148 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2149 As part of this, removing so much warning spew let me fix some minor
2150 real issues in debug logging.
2152 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2153 assignment on my part. Fixed.
2155 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2156 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2157 Janne Snabb (who went above and beyond: thank you).
2159 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2160 string otherwise requires a connection and a bunch more work and it's
2161 relatively easy to get wrong. Should also expose TLS library linkage
2164 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2165 64-bit ${eval} (JH/03).
2167 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2168 GNU libc to support some of the 64-bit stuff, should not lead to
2169 conflicts. Defined before os.h is pulled in, so if a given platform
2170 needs to override this, it can.
2172 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2173 protection layer was required, which is not implemented.
2174 Bugzilla 1254, patch from Wolfgang Breyha.
2176 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2177 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2178 tls_dhparam take prime identifiers. Also unbreak combination of
2179 OpenSSL+DH_params+TLSSNI.
2181 PP/39 Disable SSLv2 by default in OpenSSL support.
2187 PP/01 Solaris build fix for Oracle's LDAP libraries.
2188 Bugzilla 1109, patch from Stephen Usher.
2190 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2192 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2195 TF/02 Fix a couple more cases where we did not log the error message
2196 when unlink() failed. See also change 4.74-TF/03.
2198 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2199 lock up or crash if it happened to be inside a call to libc when it
2200 got a SIGUSR1 from exiwhat.
2202 The SIGUSR1 handler appends the current process status to the process
2203 log which is later printed by exiwhat. It used to use the general
2204 purpose logging code to do this, but several functions it calls are
2205 not safe for signals.
2207 The new output code in the SIGUSR1 handler is specific to the process
2208 log, and simple enough that it's easy to inspect for signal safety.
2209 Removing some special cases also simplifies the general logging code.
2210 Removing the spurious timestamps from the process log simplifies
2213 TF/04 Improved ratelimit ACL condition.
2215 The /noupdate option has been deprecated in favour of /readonly which
2216 has clearer semantics. The /leaky, /strict, and /readonly update modes
2217 are mutually exclusive. The update mode is no longer included in the
2218 database key; it just determines when the database is updated. (This
2219 means that when you upgrade Exim will forget old rate measurements.)
2221 Exim now checks that the per_* options are used with an update mode that
2222 makes sense for the current ACL. For example, when Exim is processing a
2223 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2224 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2225 must specify per_mail/readonly. If you omit the update mode it defaults to
2226 /leaky where that makes sense (as before) or /readonly where required.
2228 The /noupdate option is now undocumented but still supported for
2229 backwards compatibility. It is equivalent to /readonly except that in
2230 ACLs where /readonly is required you may specify /leaky/noupdate or
2231 /strict/noupdate which are treated the same as /readonly.
2233 A useful new feature is the /count= option. This is a generalization
2234 of the per_byte option, so that you can measure the throughput of other
2235 aggregate values. For example, the per_byte option is now equivalent
2236 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2238 The per_rcpt option has been generalized using the /count= mechanism
2239 (though it's more complicated than the per_byte equivalence). When it is
2240 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2241 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2242 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2243 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2244 Note that using per_rcpt with a non-readonly update mode in more than
2245 one ACL will cause the recipients to be double-counted. (The per_mail
2246 and per_byte options don't have this problem.)
2248 The handling of very low rates has changed slightly. If the computed rate
2249 is less than the event's count (usually one) then this event is the first
2250 after a long gap. In this case the rate is set to the same as this event's
2251 count, so that the first message of a spam run is counted properly.
2253 The major new feature is a mechanism for counting the rate of unique
2254 events. The new per_addr option counts the number of different
2255 recipients that someone has sent messages to in the last time period. It
2256 behaves like per_rcpt if all the recipient addresses are different, but
2257 duplicate recipient addresses do not increase the measured rate. Like
2258 the /count= option this is a general mechanism, so the per_addr option
2259 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2260 example, measure the rate that a client uses different sender addresses
2261 with the options per_mail/unique=$sender_address. There are further
2262 details in the main documentation.
2264 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2266 TF/06 Removed a few PCRE remnants.
2268 TF/07 Automatically extract Exim's version number from tags in the git
2269 repository when doing development or release builds.
2271 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2272 Bugzilla 879. Patch from Paul Fisher.
2274 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2275 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2278 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2280 PP/05 Variable $av_failed, true if the AV scanner deferred.
2281 Bugzilla 1078. Patch from John Horne.
2283 PP/06 Stop make process more reliably on build failure.
2284 Bugzilla 1087. Patch from Heiko Schlittermann.
2286 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2287 Bugzilla 1089. Patch from Heiko Schlittermann.
2289 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2290 Bugzilla 1131. Patch from Holger Weiß.
2292 PP/09 Handle IPv6 addresses with SPF.
2293 Bugzilla 860. Patch from Wolfgang Breyha.
2295 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2297 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2300 PP/11 match_* no longer expand right-hand-side by default.
2301 New compile-time build option, EXPAND_LISTMATCH_RHS.
2302 New expansion conditions, "inlist", "inlisti".
2304 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2306 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2308 PP/14 fix log_write() format string regression from TF/03.
2309 Bugzilla 1152. Patch from Dmitry Isaikin.
2315 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2317 PP/02 Harmonised TLS library version reporting; only show if debugging.
2318 Layout now matches that introduced for other libraries in 4.74 PP/03.
2320 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2322 PP/04 New "dns_use_edns0" global option.
2324 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2327 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2328 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2330 TK/01 Updated PolarSSL code to 0.14.2.
2331 Bugzilla 1097. Patch from Andreas Metzler.
2333 PP/07 Catch divide-by-zero in ${eval:...}.
2334 Fixes bugzilla 1102.
2336 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2339 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2340 format-string attack -- SECURITY: remote arbitrary code execution.
2342 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2343 time unintentionally subject to list matching rules, letting the header
2344 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2345 arbitrary string expansion). This allowed for information disclosure.
2347 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2348 INT_MIN/-1 -- value coerced to INT_MAX.
2354 NM/01 Workaround for PCRE version dependency in version reporting
2357 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2358 This fixes portability to compilers other than gcc, notably
2359 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2361 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2362 makefiles for portability to HP-UX and POSIX correctness.
2364 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2365 Also via indented variable definition in the Makefile.
2366 (Debugging by Oliver Heesakkers).
2368 PP/02 Restore caching of spamd results with expanded spamd_address.
2369 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2371 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2372 Improves build reliability. Fix from: Frank Elsner
2374 NM/02 Fix wide character breakage in the rfc2047 coding
2375 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2377 NM/03 Allow underscore in dnslist lookups
2378 Fixes bug 1026. Patch from Graeme Fowler
2380 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2381 Code patches from Adam Ciarcinski of NetBSD.
2383 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2386 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2387 is logged, to avoid truncation. Patch from John Horne.
2389 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2390 Patch from Jakob Hirsch.
2392 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2393 SQL string expansion failure details.
2394 Patch from Andrey Oktyabrski.
2396 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2397 Patch from Simon Arlott.
2399 PP/09 New lookups functionality failed to compile on old gcc which rejects
2400 extern declarations in function scope.
2401 Patch from Oliver Fleischmann
2403 PP/10 Use sig_atomic_t for flags set from signal handlers.
2404 Check getgroups() return and improve debugging.
2405 Fixed developed for diagnosis in bug 927 (which turned out to be
2408 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2409 Patch from Mark Zealey.
2411 PP/12 Bugzilla 1056: Improved spamd server selection.
2412 Patch from Mark Zealey.
2414 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2415 Based on patch from Heiko Schlittermann.
2417 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2418 Patch from Uwe Doering, sign-off by Michael Haardt.
2420 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2421 variable declaration deep within a block. Bug and patch from
2424 PP/15 lookups-Makefile IRIX compatibility coercion.
2426 PP/16 Make DISABLE_DKIM build knob functional.
2428 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2429 Patch by Simon Arlott
2431 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2432 variable argument macros. Our copy now differs from upstream.
2438 TF/01 Failure to get a lock on a hints database can have serious
2439 consequences so log it to the panic log.
2441 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2442 controlled using the smtp_confirmation log selector.
2444 TF/03 Include the error message when we fail to unlink a spool file.
2446 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2447 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2448 for maintaining out-of-tree patches for some time.
2450 PP/01 Bugzilla 139: Documentation and portability issues.
2451 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2452 Handle per-OS dynamic-module compilation flags.
2454 PP/02 Let /dev/null have normal permissions.
2455 The 4.73 fixes were a little too stringent and complained about the
2456 permissions on /dev/null. Exempt it from some checks.
2457 Reported by Andreas M. Kirchwitz.
2459 PP/03 Report version information for many libraries, including
2460 Exim version information for dynamically loaded libraries. Created
2461 version.h, now support a version extension string for distributors
2462 who patch heavily. Dynamic module ABI change.
2464 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2465 privilege escalation vulnerability whereby the Exim run-time user
2466 can cause root to append content of the attacker's choosing to
2469 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2472 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2473 If dropping privileges for untrusted macros, we disabled normal logging
2474 on the basis that it would fail; for the Exim run-time user, this is not
2475 the case, and it resulted in successful deliveries going unlogged.
2476 Fixed. Reported by Andreas Metzler.
2482 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2483 only prepend for the Resent-* case. Fixes regression introduced in
2484 Exim 4.70 by NM/22 for Bugzilla 607.
2486 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2487 increasing numbers of administrators be bitten by this.
2489 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2491 PP/03 Bugzilla 994: added openssl_options main configuration option.
2493 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2495 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2497 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2499 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2500 without a peer certificate, leading to a segfault because of an
2501 assumption that peers always have certificates. Be a little more
2502 paranoid. Problem reported by Martin Tscholak.
2504 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2505 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2506 NB: ClamAV planning to remove STREAM in "middle of 2010".
2507 CL also introduces -bmalware, various -d+acl logging additions and
2508 more caution in buffer sizes.
2510 PP/09 Implemented reverse_ip expansion operator.
2512 PP/10 Bugzilla 937: provide a "debug" ACL control.
2514 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2516 PP/12 Bugzilla 973: Implement --version.
2518 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2520 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2522 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2524 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2525 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2526 ignore trailing whitespace.
2528 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2530 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2533 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2534 Notification from Dr Andrew Aitchison.
2536 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2537 ExtendedDetectionInfo response format.
2538 Notification from John Horne.
2540 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2543 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2544 XSL and documented dependency on system catalogs, with examples of how
2547 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2550 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2551 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2552 configuration file which is writeable by the Exim user or group.
2554 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2555 of configuration files to cover files specified with the -C option if
2556 they are going to be used with root privileges, not just the default
2559 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2560 option (effectively making it always true).
2562 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2563 files to be used while preserving root privileges.
2565 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2566 that rogue child processes cannot use them.
2568 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2569 run-time user, instead of root.
2571 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2572 Exim run-time user without dropping privileges.
2574 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2575 result string, instead of calling string_vformat() twice with the same
2578 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2579 for other users. Others should always drop root privileges if they use
2580 -C on the command line, even for a whitelisted configure file.
2582 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2584 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2590 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2591 $data_path, and $header_path variables; fixed documentation bugs and
2594 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2595 exipick to access non-standard spools, including the "frozen" queue
2598 NM/01 Bugzilla 965: Support mysql stored procedures.
2599 Patch from Alain Williams
2601 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2603 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2604 Patch from Andreas Metzler
2606 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2607 Patch from Kirill Miazine
2609 NM/05 Bugzilla 671: Added umask to procmail example.
2611 JJ/03 installed exipick 20100323.0, fixing doc bug
2613 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2614 directory. Notification and patch from Dan Rosenberg.
2616 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2618 TK/02 Improve log output when DKIM signing operation fails.
2620 MH/01 Treat the transport option dkim_domain as a colon separated
2621 list, not as a single string, and sign the message with each element,
2622 omitting multiple occurences of the same signer.
2624 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2625 Bugzilla 985, 986. Patch by Simon Arlott
2627 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2628 Patch by Simon Arlott
2630 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2631 MBX locking. Notification from Dan Rosenberg.
2637 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2639 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2641 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2643 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2645 NM/04 Bugzilla 915: Flag broken perl installation during build.
2651 TK/01 Added patch by Johannes Berg that expands the main option
2652 "spamd_address" if it starts with a dollar sign.
2654 TK/02 Write list of recipients to X-Envelope-Sender header when building
2655 the mbox-format spool file for content scanning (suggested by Jakob
2658 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2659 (http://www.dcc-servers.net/) support via dccifd. Activated by
2660 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2662 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2663 by Mark Daniel Reidel <mr@df.eu>.
2665 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2666 When building exim an external PCRE library is now needed -
2667 PCRE is a system library on the majority of modern systems.
2668 See entry on PCRE_LIBS in EDITME file.
2670 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2671 conversation. Added nologin parameter to request.
2672 Patch contributed by Kirill Miazine.
2674 TF/01 Do not log submission mode rewrites if they do not change the address.
2676 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2678 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2679 log files in place. Contributed by Roberto Lima.
2681 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2683 TF/03 Bugzilla 615: When checking the local_parts router precondition
2684 after a local_part_suffix or local_part_prefix option, Exim now
2685 does not use the address's named list lookup cache, since this
2686 contains cached lookups for the whole local part.
2688 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2689 Robert Millan. Documentation is in experimental-spec.txt.
2691 TF/04 Bugzilla 668: Fix parallel build (make -j).
2693 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2695 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2696 Patch provided by Jan Srzednicki.
2698 TF/05 Leading white space used to be stripped from $spam_report which
2699 wrecked the formatting. Now it is preserved.
2701 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2702 that they are available at delivery time.
2704 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2706 TF/08 TLS error reporting now respects the incoming_interface and
2707 incoming_port log selectors.
2709 TF/09 Produce a more useful error message if an SMTP transport's hosts
2710 setting expands to an empty string.
2712 NM/06 Bugzilla 744: EXPN did not work under TLS.
2713 Patch provided by Phil Pennock.
2715 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2716 Patch provided by Richard Godbee.
2718 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2719 acl_smtp_notquit, added index entry.
2721 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2722 Patch provided by Eugene Bujak.
2724 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2725 accept(). Patch provided by Maxim Dounin.
2727 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2728 Patch provided by Phil Pennock.
2730 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2732 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2733 Patch provided by Brad "anomie" Jorsch.
2735 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2736 Patch provided by Dean Brooks.
2738 TK/05 Add native DKIM support (does not depend on external libraries).
2740 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2741 Patch provided by Graeme Fowler.
2743 NM/16 Bugzilla 851: Documentation example syntax fix.
2745 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2747 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2750 NM/19 Bugzilla 745: TLS version reporting.
2751 Patch provided by Phil Pennock.
2753 NM/20 Bugzilla 167: bool: condition support.
2754 Patch provided by Phil Pennock.
2756 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2757 clients. Patch provided by Phil Pennock.
2759 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2760 Patch provided by Brad "anomie" Jorsch.
2762 NM/23 Bugzilla 687: Fix misparses in eximstats.
2763 Patch provided by Heiko Schlittermann.
2765 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2766 Patch provided by Heiko Schlittermann.
2768 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2769 plus update to original patch.
2771 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2773 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2774 Patch provided by David Brownlee.
2776 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2778 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2780 NM/30 Bugzilla 888: TLS documentation bugfixes.
2782 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2784 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2785 Unlike the original bugzilla I have changed all shell scripts in src tree.
2787 NM/33 Bugzilla 898: Transport filter timeout fix.
2788 Patch by Todd Rinaldo.
2790 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2791 Patch by Serge Demonchaux.
2793 NM/35 Bugzilla 39: Base64 decode bug fixes.
2794 Patch by Jakob Hirsch.
2796 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2798 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2800 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2802 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2808 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2809 ALT-N's libdkim that I have put here:
2810 http://duncanthrax.net/exim-experimental/
2812 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2813 were called 'true' and it seems that C99 defines that as a reserved
2814 keyword to be used with 'bool' variable types. That means you could
2815 not include C99-style headers which use bools without triggering
2816 build errors in sieve.c.
2818 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2819 as mailq or other aliases. Changed the --help handling significantly
2820 to do whats expected. exim_usage() emits usage/help information.
2822 SC/01 Added the -bylocaldomain option to eximstats.
2824 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2826 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2828 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2834 PH/01 Another patch from the Sieve maintainer.
2836 PH/02 When an IPv6 address is converted to a string for single-key lookup
2837 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2838 dots are used instead of colons so that keys in lsearch files need not
2839 contain colons. This was done some time before quoting was made available
2840 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2841 (notated using the quote facility) so as to distinguish them from IPv4
2842 keys. This meant that lookups for IP addresses in host lists did not work
2843 for iplsearch lookups.
2845 This has been fixed by arranging for IPv6 addresses to be expressed with
2846 colons if the lookup type is iplsearch. This is not incompatible, because
2847 previously such lookups could never work.
2849 The situation is now rather anomalous, since one *can* have colons in
2850 ordinary lsearch keys. However, making the change in all cases is
2851 incompatible and would probably break a number of configurations.
2853 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2856 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2857 conversion specification without a maximum field width, thereby enabling
2858 a rogue spamd server to cause a buffer overflow. While nobody in their
2859 right mind would setup Exim to query an untrusted spamd server, an
2860 attacker that gains access to a server running spamd could potentially
2861 exploit this vulnerability to run arbitrary code as the Exim user.
2863 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2864 $primary_hostname instead of what libspf2 thinks the hosts name is.
2866 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2867 a directory entry by the name of the lookup key. Previously, if a
2868 symlink pointed to a non-existing file or a file in a directory that
2869 Exim lacked permissions to read, a lookup for a key matching that
2870 symlink would fail. Now it is enough that a matching directory entry
2871 exists, symlink or not. (Bugzilla 503.)
2873 PH/03 The body_linecount and body_zerocount variables are now exported in the
2876 PH/04 Added the $dnslist_matched variable.
2878 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2879 This means they are set thereafter only if the connection becomes
2882 PH/06 Added the client_condition to authenticators so that some can be skipped
2883 by clients under certain conditions.
2885 PH/07 The error message for a badly-placed control=no_multiline_responses left
2886 "_responses" off the end of the name.
2888 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2890 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2891 (without spaces) instead of just copying the configuration text.
2893 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2895 PH/11 Added $max_received_linelength.
2897 PH/12 Added +ignore_defer and +include_defer to host lists.
2899 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2900 way in which PCRE > 7.0 is built.
2902 PH/14 Implemented queue_only_load_latch.
2904 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2905 MAIL command. The effect was to mangle the value on 64-bit systems.
2907 PH/16 Another patch from the Sieve maintainer.
2909 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2911 PH/18 If a system quota error occurred while trying to create the file for
2912 a maildir delivery, the message "Mailbox is full" was not appended to the
2913 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2914 only to a quota excession during the actual writing of the file.
2916 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2917 characters?) which causes problems in log lines. The DN values are now
2918 passed through string_printing() before being added to log lines.
2920 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2921 and InterBase are left for another time.)
2923 PH/21 Added message_body_newlines option.
2925 PH/22 Guard against possible overflow in moan_check_errorcopy().
2927 PH/23 POSIX allows open() to be a macro; guard against that.
2929 PH/24 If the recipient of an error message contained an @ in the local part
2930 (suitably quoted, of course), incorrect values were put in $domain and
2931 $local_part during the evaluation of errors_copy.
2937 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2938 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2941 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2942 issue a MAIL command.
2944 PH/02 In an ACL statement such as
2946 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2948 if a client was not listed at all, or was listed with a value other than
2949 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2950 the condition was not true (as it should be), so access was not denied.
2951 The bug was that the ! inversion was incorrectly passed on to the second
2952 item. This has been fixed.
2954 PH/03 Added additional dnslists conditions == and =& which are different from
2955 = and & when the dns lookup returns more than one IP address.
2957 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2958 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2960 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2961 FSYNC, which compiles an option called disable_fsync that allows for
2962 bypassing fsync(). The documentation is heavily laced with warnings.
2964 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2966 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2967 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2968 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2969 including adding "make clean"; (3) Added -fPIC when compiling the test
2970 dynamically loaded module, to get rid of a warning.
2972 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2973 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2974 = 0s. The bug is otherwise harmless.
2976 PH/07 There was a bug in the dovecot authenticator such that the value of
2977 $auth1 could be overwritten, and so not correctly preserved, after a
2978 successful authentication. This usually meant that the value preserved by
2979 the server_setid option was incorrect.
2981 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2983 PH/09 Installed PCRE release 7.0.
2985 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2986 run for batched SMTP input. It is now run at the start of every message
2987 in the batch. While fixing this I discovered that the process information
2988 (output by running exiwhat) was not always getting set for -bs and -bS
2989 input. This is fixed, and it now also says "batched" for BSMTP.
2991 PH/11 Added control=no_pipelining.
2993 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2994 patch, slightly modified), and move the expansion of helo_data till after
2995 the connection is made in the smtp transport (so it can use these
2998 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3000 PH/14 Added log_selector = +pid.
3002 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3004 PH/16 Add ${if forany and ${if forall.
3006 PH/17 Added dsn_from option to vary the From: line in DSNs.
3008 PH/18 Flush SMTP output before performing a callout, unless control =
3009 no_callout_flush is set.
3011 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3012 was true (the default) a successful delivery failed to delete the retry
3013 item, thus causing premature timeout of the address. The bug is now
3016 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3018 PH/21 Long custom messages for fakedefer and fakereject are now split up
3019 into multiline responses in the same way that messages for "deny" and
3020 other ACL rejections are.
3022 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3023 with slight modification.
3025 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3026 draft, changing the syntax and factoring some duplicate code".
3028 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3029 for deliveries of the second and subsequent messages over the same SMTP
3032 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3033 ${reduce, with only minor "tidies".
3035 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3037 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3038 expansion side effects.
3040 PH/27 When a message times out after an over-quota error from an Exim-imposed
3041 quota, the bounce message says "mailbox is full". This message was not
3042 being given when it was a system quota that was exceeded. It now should
3045 MH/03 Made $recipients available in local_scan(). local_scan() already has
3046 better access to the recipient list through recipients_list[], but
3047 $recipients can be useful in postmaster-provided expansion strings.
3049 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3050 in the case of a MAIL command with additional options following the
3051 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3052 were accidentally chopped off.
3054 PH/29 SMTP synchronization checks are implemented when a command is read -
3055 there is a check that no more input is waiting when there shouldn't be
3056 any. However, for some commands, a delay in an ACL can mean that it is
3057 some time before the response is written. In this time, more input might
3058 arrive, invalidly. So now there are extra checks after an ACL has run for
3059 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3060 pipelining has not been advertised.
3062 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3064 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3065 returned in the SMTP response when smtp_return_error_details was false.
3066 This has been fixed.
3068 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3069 instead of the C I/O that was originally supplied, because problems were
3070 reported on Solaris.
3072 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3073 Exim which did not show up earlier: it was assuming that a call to
3074 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3075 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3076 was a macro that became an assignment, so it seemed to work. This has
3077 changed to a proper function call with a void return, hence the compile
3078 error. Exim's code has been fixed.
3080 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3083 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3085 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3087 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3088 because it turns out to be too big or there is a timeout) there is no
3089 "Completed" line in the log. When some messages of this type were
3090 selected by exigrep, they were listed as "not completed". Others were
3091 picked up by some special patterns. I have improved the selection
3092 criteria to be more general.
3094 PH/38 The host_find_failed option in the manualroute router can now be set
3095 to "ignore", to completely ignore a host whose IP address cannot be
3096 found. If all hosts are ignored, the behaviour is controlled by the new
3097 host_all_ignored option.
3099 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3100 homing or because of multiple MX records with /mx) generated more than
3101 one IP address, and the following item turned out to be the local host,
3102 all the secondary addresses of the first item were incorrectly removed
3103 from the list, along with the local host and any following hosts (which
3104 is what is supposed to happen).
3106 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3107 whoever called Exim into the -H file. In the case of the daemon it was
3108 behaving confusingly. When first started, it used values for whoever
3109 started the daemon, but after a SIGHUP it used the Exim user (because it
3110 calls itself on a restart). I have changed the code so that it now always
3113 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3114 message are rejected with the same error (e.g. no authentication or bad
3115 sender address), and a DATA command is nevertheless sent (as can happen
3116 with PIPELINING or a stupid MUA), the error message that was given to the
3117 RCPT commands is included in the rejection of the DATA command. This is
3118 intended to be helpful for MUAs that show only the final error to their
3121 PH/42 Another patch from the Sieve maintainer.
3123 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3124 Eximstats - Fixed some broken HTML links and added missing column headers
3126 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3127 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3129 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3131 PH/43 Yet another patch from the Sieve maintainer.
3133 PH/44 I found a way to check for a TCP/IP connection going away before sending
3134 the response to the final '.' that terminates a message, but only in the
3135 case where the client has not sent further data following the '.'
3136 (unfortunately, this is allowed). However, in many cases there won't be
3137 any further data because there won't be any more messages to send. A call
3138 to select() can be used: if it shows that the input is "ready", there is
3139 either input waiting, or the socket has been closed. An attempt to read
3140 the next input character can distinguish the two cases. Previously, Exim
3141 would have sent an OK response which the client would never have see.
3142 This could lead to message repetition. This fix should cure that, at
3143 least in a lot of common cases.
3145 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3146 advertised in response to EHLO.
3152 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3153 fixed by 4.65/MH/01 (is this a record?) are fixed:
3155 (i) An empty string was always treated as zero by the numeric comparison
3156 operators. This behaviour has been restored.
3158 (ii) It is documented that the numeric comparison operators always treat
3159 their arguments as decimal numbers. This was broken in that numbers
3160 starting with 0 were being interpreted as octal.
3162 While fixing these problems I realized that there was another issue that
3163 hadn't been noticed. Values of message_size_limit (both the global option
3164 and the transport option) were treated as octal if they started with 0.
3165 The documentation was vague. These values are now always treated as
3166 decimal, and I will make that clear in the documentation.
3172 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3173 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3176 MH/01 Don't check that the operands of numeric comparison operators are
3177 integers when their expansion is in "skipping" mode (fixes bug
3178 introduced by 4.64-PH/07).
3180 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3181 child addresses, Exim now panics and dies. Previously, because the count
3182 is held in a short int, deliveries were likely to be lost. As such a
3183 large number of recipients for a single message is ridiculous
3184 (performance will be very, very poor), I have chosen to impose a limit
3185 rather than extend the field.
3191 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3192 leftover -K file (the existence of which was triggered by #402).
3193 While we were at it, introduced process PID as part of the -K
3194 filename. This should rule out race conditions when creating
3197 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3198 processing considerably. Previous code took too long for large mails,
3199 triggering a timeout which in turn triggers #401.
3201 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3202 in the DK code in transports.c. sendfile() is not really portable,
3203 hence the _LINUX specificness.
3205 TF/01 In the add_headers option to the mail command in an Exim filter,
3206 there was a bug that Exim would claim a syntax error in any
3207 header after the first one which had an odd number of characters
3210 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3211 callout verification, Exim cached a "reject" for the entire domain. This
3212 is correct for most verifications, but it is not correct for a recipient
3213 verification with use_sender or use_postmaster set, because in that case
3214 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3215 case of MAIL FROM:<> rejection from other early rejections (e.g.
3216 rejection of HELO). When verifying a recipient using a non-null MAIL
3217 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3218 Whatever the result of the callout, the value of the domain cache is
3219 left unchanged (for any other kind of callout, getting as far as trying
3220 RCPT means that the domain itself is ok).
3222 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3225 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3226 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3227 session, the daemon ignores ECONNECT errors and logs others; it now
3228 ignores EPIPE as well.
3230 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3231 (quoted-printable decoding).
3233 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3234 later the small subsequent patch to fix an introduced bug.
3236 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3238 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3240 PH/08 An error is now given if message_size_limit is specified negative.
3242 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3243 to be given (somewhat) arbitrary names.
3245 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3248 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3249 miscellaneous code fixes
3251 PH/10 Added the log_reject_target ACL modifier to specify where to log
3254 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3255 hostname. This is wrong, because it relates to the incoming message (and
3256 probably the interface on which it is arriving) and not to the outgoing
3257 callout (which could be using a different interface). This has been
3258 changed to use the value of the helo_data option from the smtp transport
3259 instead - this is what is used when a message is actually being sent. If
3260 there is no remote transport (possible with a router that sets up host
3261 addresses), $smtp_active_hostname is used.
3263 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3264 tweaks were necessary in order to get it to work (see also 21 below):
3265 (a) The code assumed that strncpy() returns a negative number on buffer
3266 overflow, which isn't the case. Replaced with Exim's string_format()
3268 (b) There were several signed/unsigned issues. I just did the minimum
3269 hacking in of casts. There is scope for a larger refactoring.
3270 (c) The code used strcasecmp() which is not a standard C function.
3271 Replaced with Exim's strcmpic() function.
3272 (d) The code set only $1; it now sets $auth1 as well.
3273 (e) A simple test gave the error "authentication client didn't specify
3274 service in request". It would seem that Dovecot has changed its
3275 interface. Fortunately there's a specification; I followed it and
3276 changed what the client sends and it appears to be working now.
3278 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3281 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3282 address A is aliased to B and C, where B exists and C does not. Without
3283 -v the output is "A verified" because verification stops after a
3284 successful redirection if more than one address is generated. However,
3285 with -v the child addresses are also verified. Exim was outputting "A
3286 failed to verify" and then showing the successful verification for C,
3287 with its parentage. It now outputs "B failed to verify", showing B's
3288 parentage before showing the successful verification of C.
3290 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3291 look up a TXT record in a specific list after matching in a combined
3294 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3295 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3296 they consult the DNS. I had assumed they would set it the way they
3297 wanted; and indeed my experiments on Linux seem to show that in some
3298 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3299 To be on the safe side, however, I have now made the interface to
3300 host_find_byname() similar to host_find_bydns(), with an argument
3301 containing the DNS resolver options. The host_find_byname() function now
3302 sets these options at its start, just as host_find_bydns() does. The smtp
3303 transport options dns_qualify_single and dns_search_parents are passed to
3304 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3305 of host_find_byname() use the default settings of RES_DEFNAMES
3306 (qualify_single) but not RES_DNSRCH (search_parents).
3308 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3309 spool_read_header() do less string testing, by means of a preliminary
3310 switch on the second character of optional "-foo" lines. (This is
3311 overdue, caused by the large number of possibilities that now exist.
3312 Originally there were few.) While I was there, I also converted the
3313 str(n)cmp tests so they don't re-test the leading "-" and the first
3314 character, in the hope this might squeeze out yet more improvement.
3316 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3317 flag allowing group syntax was set by the header_syntax check but not
3318 turned off, possible causing trouble later; (2) The flag was not being
3319 set at all for the header_verify test, causing "group"-style headers to
3320 be rejected. I have now set it in this case, and also caused header_
3321 verify to ignore an empty address taken from a group. While doing this, I
3322 came across some other cases where the code for allowing group syntax
3323 while scanning a header line wasn't quite right (mostly, not resetting
3324 the flag correctly in the right place). These bugs could have caused
3325 trouble for malformed header lines. I hope it is now all correct.
3327 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3328 with the "reply" argument non-NULL. The code, however (which originally
3329 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3330 but it didn't always do it. This confused somebody who was copying the
3331 code for some other use. I have removed all the tests.
3333 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3334 feature that was used to support insecure browsers during the U.S. crypto
3335 embargo. It requires special client support, and Exim is probably the
3336 only MTA that supported it -- and would never use it because real RSA is
3337 always available. This code has been removed, because it had the bad
3338 effect of slowing Exim down by computing (never used) parameters for the
3339 RSA_EXPORT functionality.
3341 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3342 authenticator to fail if there's a tab character in the incoming data
3343 (there should never be unless someone is messing about, as it's supposed
3344 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3345 if the connection is using TLS or if the remote IP is the same as the
3346 local IP, and the "valid-client-cert option" if a client certificate has
3349 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3350 authenticators. This can be used for authorization after authentication
3351 succeeds. (In the case of plaintext, it servers for both authentication
3354 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3355 if any retry times were supplied.
3357 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3358 connection, where there is no client IP address to check. In this
3359 situation, the verify now always succeeds.
3361 PH/25 Applied John Jetmore's -Mset patch.
3363 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3365 PH/27 In a string expansion for a processed (not raw) header when multiple
3366 headers of the same name were present, leading whitespace was being
3367 removed from all of them, but trailing whitespace was being removed only
3368 from the last one. Now trailing whitespace is removed from each header
3369 before concatenation. Completely empty headers in a concatenation (as
3370 before) are ignored.
3372 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3373 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3375 PH/29 [Removed. This was a change that I later backed out, and forgot to
3376 correct the ChangeLog entry (that I had efficiently created) before
3377 committing the later change.]
3379 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3380 address errors (4xx response to RCPT) over the same connection as other
3381 messages routed to the same hosts. Such deliveries are always "forced",
3382 so retry times are not inspected. This resulted in far too many retries
3383 for the affected addresses. The effect occurred only when there were more
3384 hosts than the hosts_max_try setting in the smtp transport when it had
3385 the 4xx errors. Those hosts that it had tried were not added to the list
3386 of hosts for which the message was waiting, so if all were tried, there
3387 was no problem. Two fixes have been applied:
3389 (i) If there are any address or message errors in an SMTP delivery, none
3390 of the hosts (tried or untried) are now added to the list of hosts
3391 for which the message is waiting, so the message should not be a
3392 candidate for sending over the same connection that was used for a
3393 successful delivery of some other message. This seems entirely
3394 reasonable: after all the message is NOT "waiting for some host".
3395 This is so "obvious" that I'm not sure why it wasn't done
3396 previously. Hope I haven't missed anything, but it can't do any
3397 harm, as the worst effect is to miss an optimization.
3399 (ii) If, despite (i), such a delivery is accidentally attempted, the
3400 routing retry time is respected, so at least it doesn't keep
3401 hammering the server.
3403 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3404 in ${readsocket because some servers need this prod.
3406 PH/32 Added some extra debug output when updating a wait-xxx database.
3408 PH/33 The hint "could be header name not terminated by colon", which has been
3409 given for certain expansion errors for a long time, was not being given
3410 for the ${if def:h_colon_omitted{... case.
3412 PH/34 The spec says: "With one important exception, whenever a domain list is
3413 being scanned, $domain contains the subject domain." There was at least
3414 one case where this was not true.
3416 PH/35 The error "getsockname() failed: connection reset by peer" was being
3417 written to the panic log as well as the main log, but it isn't really
3418 panic-worthy as it just means the connection died rather early on. I have
3419 removed the panic log writing for the ECONNRESET error when getsockname()
3422 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3423 runs only) independently of the message's sender address. This meant
3424 that, if the 4xx error was in fact related to the sender, a different
3425 message to the same recipient with a different sender could confuse
3426 things. In particular, this can happen when sending to a greylisting
3427 server, but other circumstances could also provoke similar problems.
3428 I have changed the default so that the retry time for these errors is now
3429 based a combination of the sender and recipient addresses. This change
3430 can be overridden by setting address_retry_include_sender=false in the
3433 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3434 remote server are returned as part of bounce messages. This was not
3435 happening for LMTP over a pipe (the lmtp transport), but now it is the
3436 same for both kinds of LMTP.
3438 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3439 in header lines that were in fact CNAMEs. This is no longer the case.
3441 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3442 and queue runs started by the daemon processed all messages. This has
3443 been fixed so that -R and -S can now usefully be given with -q<time>.
3445 PH/40 Import PCRE release 6.7 (fixes some bugs).
3447 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3449 PH/42 Give an error if -q is specified more than once.
3451 PH/43 Renamed the variables $interface_address and $interface_port as
3452 $received_ip_address and $received_port, to make it clear that these
3453 values apply to message reception, and not to the outgoing interface when
3454 a message is delivered. (The old names remain recognized, of course.)
3456 PH/44 There was no timeout on the connect() call when using a Unix domain
3457 socket in the ${readsocket expansion. There now is.
3459 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3460 be meaningful with "accept".
3462 SC/01 Eximstats V1.43
3463 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3465 SC/02 Eximstats V1.44
3466 Use a glob alias rather than an array ref in the generated
3467 parser. This improves both readability and performance.
3469 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3470 Collect SpamAssassin and rejection statistics.
3471 Don't display local sender or destination tables unless
3472 there is data to show.
3473 Added average volumes into the top table text output.
3475 SC/04 Eximstats V1.46
3476 Collect data on the number of addresses (recipients)
3477 as well as the number of messages.
3479 SC/05 Eximstats V1.47
3480 Added 'Message too big' to the list of mail rejection
3481 reasons (thanks to Marco Gaiarin).
3483 SC/06 Eximstats V1.48
3484 Mainlog lines which have GMT offsets and are too short to
3485 have a flag are now skipped.
3487 SC/07 Eximstats V1.49 (Alain Williams)
3488 Added the -emptyok flag.
3490 SC/08 Eximstats V1.50
3491 Fixes for obtaining the IP address from reject messages.
3493 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3494 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3495 whitespace changes from 4.64-PH/27
3497 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3500 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3501 are found, allow negative numbers in numeric criteria)
3503 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3505 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3506 to match changes made in 4.64-PH/43
3508 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3510 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3511 "domain" is actually an IP address, and force a failure. This locks out
3512 those revolvers/nameservers that support "A-for-A" lookups, in
3513 contravention of the specifications.
3515 PH/48 When a host name was looked up from an IP address, and the subsequent
3516 forward lookup of the name timed out, the host name was left in
3517 $sender_host_name, contrary to the specification.
3519 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3520 restricted to single-key lookups, Exim was not diagnosing an error if
3521 * or *@ was used with a query-style lookup.
3523 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3525 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3526 long ago, but noone interested enough thought of it. Let's just say that
3527 the "1.1" means that there are some new functions that weren't there at
3528 some point in the past.
3530 PH/51 Error processing for expansion failure of helo_data from an smtp
3531 transport during callout processing was broken.
3533 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3534 tested/used via the -bh/-bhc/-bs options.
3536 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3537 bug, fixed in subsequent PCRE releases).
3539 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3540 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3542 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3548 SC/01 Use a glob alias rather than an array ref in eximstats generated
3549 parser. This improves both readability and performance.
3551 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3552 Don't display local sender or destination tables in eximstats unless
3553 there is data to show.
3554 Added average volumes into the eximstats top table text output.
3556 SC/03 Collect data on the number of addresses (recipients) as well
3557 as the number of messages in eximstats.
3559 TF/01 Correct an error in the documentation for the redirect router. Exim
3560 does (usually) call initgroups() when daemonizing.
3562 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3563 with consistent privilege compared to when running as a daemon.
3565 TF/03 Note in the spec that $authenticated_id is not set for local
3566 submissions from trusted users.
3568 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3569 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3571 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3572 by adding some example configuration directives to the default
3573 configuration file. A little bit of work is required to uncomment the
3574 directives and define how usernames and passwords are checked, but
3575 there is now a framework to start from.
3577 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3578 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3579 without this. I don't know how relevant this is to other LDAP libraries.
3581 PH/02 Add the verb name to the "unknown ACL verb" error.
3583 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3585 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3587 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3588 directory not expanded when it should be if an expanded home directory
3589 was set for the address (which is overridden by the transport).
3591 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3594 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3595 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3596 because it is too late at that time, and has no effect.
3598 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3599 security issue with \' (bugzilla #107). I could not use the
3600 PQescapeStringConn() function, because it needs a PGconn value as one of
3603 PH/08 When testing addresses using -bt, indicate those final addresses that
3604 are duplicates that would not cause an additional delivery. At least one
3605 person was confused, thinking that -bt output corresponded to deliveries.
3606 (Suppressing duplicates isn't a good idea as you lose the information
3607 about possibly different redirections that led to the duplicates.)
3609 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3610 systems where poll() doesn't work, in particular OS X.
3612 PH/10 Added more information to debugging output for retry time not reached.
3614 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3615 operations in malware.c.
3617 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3620 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3621 syslog_duplication set false, log lines that would normally be written
3622 both the the main log and to the reject log were not written to syslog at
3625 PH/14 In the default configuration, change the use of "message" in ACL warn
3626 statements to "add_header".
3628 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3629 not followed by a command (e.g. "seen endif").
3631 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3632 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3635 PH/17 Added extra conditions to the default value of delay_warning_condition
3639 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3640 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3641 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3644 The Auto-Submitted: and various List- headers are standardised, whereas I
3645 don't think Precedence: ever was.
3647 PH/18 Refactored debugging code in route_finduser() to show more information,
3648 in particular, the error code if getpwnam() issues one.
3650 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3651 This is apparently needed in addition to the PH/07 change above to avoid
3652 any possible encoding problems.
3654 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3655 but not after initializing Perl.
3657 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3658 output them only if debugging. By default they are written stderr,
3659 apparently, which is not desirable.
3661 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3664 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3667 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3669 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3670 authenticated or an ident call has been made. Suppress the default
3671 values for $authenticated_id and $authenticated_sender (but permit -oMai
3672 and -oMas) when testing with -bh.
3674 PH/24 Re-jigged the order of the tests in the default configuration so that the
3675 tests for valid domains and recipients precede the DNS black list and CSA
3676 tests, on the grounds that those ones are more expensive.
3678 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3679 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3680 This bug exists in every version of Exim that I still have, right back to
3683 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3684 However, an attempt to turn on case-sensitivity in a regex key by
3685 including (?-i) didn't work because the subject string was already
3686 lowercased, and the effects were non-intuitive. It turns out that a
3687 one-line patch can be used to allow (?-i) to work as expected.
3693 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3694 other effects) broke the use of negated acl sub-conditions.
3696 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3699 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3700 "Deny" causes Exim to reject the incoming connection with a 554 error.
3701 Unfortunately, if there is a major crisis, such as a disk failure,
3702 tcp-wrappers gives "deny", whereas what one would like would be some
3703 kind of temporary error. A kludge has been added to help with this.
3704 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3705 554 error is used if errno is still zero or contains ENOENT (which occurs
3706 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3709 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3711 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3712 errors. Otherwise a message that provokes a temporary error (when other
3713 messages do not) can cause a whole host to time out.
3715 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3716 addresses were routed directly to files or pipes from a redirect router.
3717 File deliveries just didn't batch; pipe deliveries might have suffered
3720 PH/06 A failure to get a lock for a hints database would erroneously always say
3721 "Failed to get write lock", even when it was really a read lock.
3723 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3724 of 0600. This has been changed to use the value of the lockfile_mode
3725 option (which defaults to 0600).
3727 PH/08 Applied small patch from the Sieve maintainer.
3729 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3730 folder from quota calculations, a direct delivery into this folder messed
3731 up the contents of the maildirsize file. This was because the regex was
3732 used only to exclude .Trash (or whatever) when the size of the mailbox
3733 was calculated. There was no check that a delivery was happening into an
3734 excluded directory. This bug has been fixed by ignoring all quota
3735 processing for deliveries into excluded directories.
3737 PH/10 Added the maildirfolder_create_regex option to appendfile.
3743 PH/01 The code for finding all the local interface addresses on a FreeBSD
3744 system running IPv6 was broken. This may well have applied to all BSD
3745 systems, as well as to others that have similar system calls. The broken
3746 code found IPv4 interfaces correctly, but gave incorrect values for the
3747 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3748 that it would not match correctly against @[] and not recognize the IPv6
3751 PH/02 The ipliteral router was not recognizing addresses of the form user@
3752 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3754 PH/03 Added disable_ipv6.
3756 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3757 decoded form, because it is most often used to construct To: headers
3758 lines in autoreplies, and the decoded form may well be syntactically
3759 invalid. However, $reply_address has leading white space removed, and all
3760 newlines turned into spaces so that the autoreply transport does not
3763 PH/05 If group was specified without a user on a router, and no group or user
3764 was specified on a transport, the group from the router was ignored.
3766 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3767 for visible compile-time settings that can be used to change these
3768 numbers, for those that want even more. Backwards compatibility with old
3769 spool files has been maintained. However, going back to a previous Exim
3770 release will lost any variables that are in spool files.
3772 PH/07 Two small changes when running in the test harness: increase delay when
3773 passing a TCP/IP connection to a new process, in case the original
3774 process has to generate a bounce, and remove special handling of
3775 127.0.0.2 (sic), which is no longer necessary.
3777 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3778 be the same on different OS.
3780 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3783 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3784 whether --show-vars was specified or not
3786 JJ/02 exipick: Added support for new ACL variable spool format introduced
3789 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3790 syntactically invalid From: or Reply-to: line, and a filter used this to
3791 generate an autoreply, and therefore failed to obtain an address for the
3792 autoreply, Exim could try to deliver to a non-existent relative file
3793 name, causing unrelated and misleading errors. What now happens is that
3794 it logs this as a hard delivery error, but does not attempt to create a
3797 PH/11 The exinext utility has a -C option for testing purposes, but although
3798 the given file was scanned by exinext itself; it wasn't being passed on
3799 when Exim was called.
3801 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3802 an end-of-file indication when reading a command response.
3804 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3805 compiled. In many other places in Exim, IPv6 addresses are always
3806 recognized, so I have changed this. It also means that IPv4 domain
3807 literals of the form [IPV4:n.n.n.n] are now always recognized.
3809 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3810 used if the router is not running as root, for example, when verifying at
3811 ACL time, or when using -bh. The debugging output from this situation was
3812 non-existent - all you got was a failure to exec. I have made two
3815 (a) Failures to set uid/gid, the current directory, or a process leader
3816 in a subprocess such as that created by queryprogram now generate
3817 suitable debugging output when -d is set.
3819 (b) The queryprogram router detects when it is not running as root,
3820 outputs suitable debugging information if -d is set, and then runs
3821 the subprocess without attempting to change uid/gid.
3823 PH/15 Minor change to Makefile for building test_host (undocumented testing
3826 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3827 additional section of a DNS packet that returns MX or SRV records.
3828 Instead, it always explicitly searches for A/AAAA records. This avoids
3829 major problems that occur when a DNS server includes only records of one
3830 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3831 fixed another bug: if SRV records were looked up and the corresponding
3832 address records were *not* found in the additional section, the port
3833 values from the SRV records were lost.
3835 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3836 using the correct key (the original address) when searching the retry
3837 rules in order to find which one to use for generating the retry hint.
3839 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3840 adding the default one. Similarly, if it contains a Reply-To: header, the
3841 errors_reply_to option, if set, is not used.
3843 PH/19 When calculating a retry time, Exim used to measure the "time since
3844 failure" by looking at the "first failed" field in the retry record. Now
3845 it does not use this if it is later than than the arrival time of the
3846 message. Instead it uses the arrival time. This makes for better
3847 behaviour in cases where some deliveries succeed, thus re-setting the
3848 "first failed" field. An example is a quota failure for a huge message
3849 when small messages continue to be delivered. Without this change, the
3850 "time since failure" will always be short, possible causing more frequent
3851 delivery attempts for the huge message than are intended.
3852 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3854 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3855 $1, $2, $3) because the numerical variables can be reset during some
3856 expansion items (e.g. "match"), thereby losing the authentication data.
3858 PH/21 Make -bV show the size of off_t variables so that the test suite can
3859 decide whether to run tests for quotas > 2G.
3861 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3862 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3863 filecount value is greater than 2G or if a quota value is greater than 2G
3864 on a system where the size of off_t is not greater than 4, a panic error
3867 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3868 never match. The debug and -bh output now contains an explicit error
3869 message indicating a malformed IPv4 address or mask.
3871 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3872 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3873 PH/23 above applies.
3875 PH/25 Do not write to syslog when running in the test harness. The only
3876 occasion when this arises is a failure to open the main or panic logs
3877 (for which there is an explicit test).
3879 PH/26 Added the /no_tell option to "control=freeze".
3881 PH/27 If a host name lookup failed very early in a connection, for example, if
3882 the IP address matched host_lookup and the reverse lookup yielded a name
3883 that did not have a forward lookup, an error message of the form "no IP
3884 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3885 could be logged. Now it outputs the IP address instead of "NULL".
3887 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3888 allows the sender and the authenticated sender to be set when
3889 submitting a message from within Exim. Since child_open_exim() is
3890 documented for local_scan(), the new function should be too.
3892 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3893 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3894 results in an empty string is now treated as unset.
3896 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3898 PH/31 Added sender_verify_fail logging option.
3900 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3901 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3902 tidied the source and removed it altogether.
3904 PH/33 When a queue run was abandoned because the load average was too high, a
3905 log line was always written; now it is written only if the queue_run log
3906 selector is set. In addition, the log line for abandonment now contains
3907 information about the queue run such as the pid. This is always present
3908 in "start" and "stop" lines but was omitted from the "abandon" line.
3910 PH/34 Omit spaces between a header name and the colon in the error message that
3911 is given when verify = headers_syntax fails (if there are lots of them,
3912 the message gets confusing).
3914 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3915 names, as there are now some PTR records that contain slashes. This check
3916 is only to protect against broken name servers that fall over on strange
3917 characters, so the fact that it applies to all lookups doesn't matter.
3919 PH/36 Now that the new test suite is complete, we can remove some of the
3920 special code in Exim that was needed for the old test suite. For example,
3921 sorting DNS records because real resolvers return them in an arbitrary
3922 order. The new test suite's fake resolver always returns records in the
3925 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3926 bounces) except when queue_only is set, to avoid logging races between
3927 the different processes.
3929 PH/38 Panic-die if .include specifies a non-absolute path.
3931 PH/39 A tweak to the "H" retry rule from its user.
3933 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3934 a label. They prevented compilation on older perls.
3936 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3937 a warning to be raised on newish perls.
3939 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3940 on queue. Changes to match documented behaviour of showing count of
3941 messages matching specified criteria.
3943 PH/40 Changed the default ident timeout from 30s to 5s.
3945 PH/41 Added support for the use of login_cap features, on those BSD systems
3946 that have them, for controlling the resources used by pipe deliveries.
3948 PH/42 The content-scanning code uses fopen() to create files in which to put
3949 message data. Previously it was not paying any attention to the mode of
3950 the files. Exim runs with umask(0) because the rest of the code creates
3951 files with open(), and sets the required mode explicitly. Thus, these
3952 files were ending up world-writeable. This was not a big issue, because,
3953 being within the spool directory, they were not world-accessible. I have
3954 created a function called modefopen, which takes an additional mode
3955 argument. It sets umask(777), creates the file, chmods it to the required
3956 mode, then resets the umask. All the relevant calls to fopen() in the
3957 content scanning code have been changed to use this function.
3959 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3960 to 24 hours. This avoids potential overflow problems when processing G
3961 and H retry rules. I suspect nobody ever tinkers with this value.
3963 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3965 PH/45 When the plaintext authenticator is running as a client, the server's
3966 challenges are checked to ensure they are valid base64 strings. By
3967 default, the authentication attempt is cancelled if an invalid string is
3968 received. Setting client_ignore_invalid_base64 true ignores these errors.
3969 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3970 they are received. Thus, the responses can be made to depend on the
3971 challenges. If an invalid string is ignored, an empty string is placed in
3974 PH/46 Messages that are created by the autoreply transport now contains a
3975 References: header, in accordance with RFCs 2822 and 3834.
3977 PH/47 Added authenticated_sender_force to the smtp transport.
3979 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3981 PH/49 Installed latest patch from the Sieve maintainer.
3983 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3984 was also set, the appendfile transport was unnecessarily scanning a
3985 directory of message files (e.g. for maildir delivery) to find the count
3986 of files (along with the size), even though it did not need this
3987 information. It now does the scan only if it needs to find either the
3988 size of the count of files.
3990 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3992 PH/52 Two bugs concerned with error handling when the smtp transport is
3995 (i) Exim was not creating retry information for temporary errors given
3996 for individual recipients after the DATA command when the smtp transport
3997 was used in LMTP mode. This meant that they could be retried too
3998 frequently, and not timed out correctly.
4000 (ii) Exim was setting the flag that allows error details to be returned
4001 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4002 recipients that were returned after the DATA command.
4004 PH/53 This is related to PH/52, but is more general: for any failing address,
4005 when detailed error information was permitted to be returned to the
4006 sender, but the error was temporary, then after the final timeout, only
4007 "retry timeout exceeded" was returned. Now it returns the full error as
4008 well as "retry timeout exceeded".
4010 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4011 do this, and (what is worse) MTAs that accept it.
4013 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4014 will now be deprecated.
4016 PH/56 New os.c-cygwin from the Cygwin maintainer.
4018 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4019 formats (previously only available in exim formats via -bpr, -bpru,
4020 and -bpra. Now also available in native and exiqgrep formats)
4022 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4023 with very large, slow to parse queues
4025 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4027 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4029 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4030 responses to authentication challenges, though it was showing the
4031 challenges; (ii) I've removed the CR characters from the debug output for
4034 PH/58 Allow for the insertion of a newline as well as a space when a string
4035 is turned into more than one encoded-word during RFC 2047 encoding. The
4036 Sieve code now uses this.
4038 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4039 data_4xx, lost_connection, tls_required.
4041 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4042 message was being sent as an SMTP response.
4044 PH/61 Add -l and -k options to exicyclog.
4046 PH/62 When verifying, if an address was redirected to one new address, so that
4047 verification continued, and the new address failed or deferred after
4048 having set something in $address_data, the value of $address_data was not
4049 passed back to the ACL. This was different to the case when no
4050 redirection occurred. The value is now passed back in both cases.
4052 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4053 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4054 use login_cap.h, so on its own it isn't the distinguishing feature. The
4055 new name refers directly to the setclassresources() function.
4057 PH/65 Added configuration files for NetBSD3.
4059 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4061 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4062 is preferred over IPv4.
4064 PH/68 The bounce_return_message and bounce_return_body options were not being
4065 honoured for bounces generated during the reception of non-SMTP messages.
4066 In particular, this applied to messages rejected by the ACL. This bug has
4067 been fixed. However, if bounce_return_message is true and bounce_return_
4068 body is false, the headers that are returned for a non-SMTP message
4069 include only those that have been read before the error was detected.
4070 (In the case of an ACL rejection, they have all been read.)
4072 PH/69 The HTML version of the specification is now built in a directory called
4073 spec_html instead of spec.html, because the latter looks like a path with
4074 a MIME-type, and this confuses some software.
4076 PH/70 Catch two compiler warnings in sieve.c.
4078 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4079 function verify_get_ident() calls ip_connect() to connect a socket, but
4080 if the "connect()" function timed out, ip_connect() used to close the
4081 socket. However, verify_get_ident() also closes the socket later, and in
4082 between Exim writes to the log, which may get opened at this point. When
4083 the socket was closed in ip_connect(), the log could get the same file
4084 descriptor number as the socket. This naturally causes chaos. The fix is
4085 not to close the socket in ip_connect(); the socket should be closed by
4086 the function that creates it. There was only one place in the code where
4087 this was missing, in the iplookup router, which I don't think anybody now
4088 uses, but I've fixed it anyway.
4090 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4091 well as to direct DNS lookups. Otherwise the handling of names in host
4092 lists is inconsistent and therefore confusing.
4098 PH/01 Two changes to the default runtime configuration:
4100 (1) Move the checks for relay_from_hosts and authenticated clients from
4101 after to before the (commented out) DNS black list checks.
4103 (2) Add control=submission to the relay_from_hosts and authenticated
4104 clients checks, on the grounds that messages accepted by these
4105 statements are most likely to be submissions.
4107 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4109 (1) Generate an error if the third argument for the ${prvs expansion is
4112 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4115 (3) Reset the variables that are obtained from the first argument of
4116 ${prvscheck and used in the second argument before leaving the code,
4117 because their memory is reclaimed, so using them afterwards may do
4120 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4121 one (it's much easier than Tom thought :-).
4123 (5) Because of (4), we can now allow for the use of $prvscheck_result
4124 inside the third argument.
4126 PH/03 For some reason, the default setting of PATH when running a command from
4127 a pipe transport was just "/usr/bin". I have changed it to
4130 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4131 anything to be listed in the output from -bV.
4133 PH/05 When a filter generated an autoreply, the entire To: header line was
4134 quoted in the delivery log line, like this:
4136 => >A.N.Other <ano@some.domain> <original@ddress> ...
4138 This has been changed so that it extracts the operative address. There
4139 may be more than one such address. If so, they are comma-separated, like
4142 => >ano@some.domain,ona@other.domain <original@ddress> ...
4144 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4145 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4146 not being looked up in the rDNS to get a host name, Exim was showing the
4147 IP address twice in Received: lines, even though the IP addresses were
4148 identical. For example:
4150 Received: from [1.2.3.4] (helo=[1.2.3.4])
4152 However, if the real host name was known, it was omitting the HELO data
4153 if it matched the actual IP address. This has been tidied up so that it
4154 doesn't show the same IP address twice.
4156 PH/07 When both +timestamp and +memory debugging was on, the value given by
4157 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4158 called by the debug printing, thereby overwriting the timestamp buffer.
4159 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4161 PH/08 When the original message was included in an autoreply transport, it
4162 always said "this is a copy of the message, including all the headers",
4163 even if body_only or headers_only was set. It now gives an appropriate
4166 PH/09 Applied a patch from the Sieve maintainer which:
4168 o fixes some comments
4169 o adds the (disabled) notify extension core
4170 o adds some debug output for the result of if/elsif tests
4171 o points to the current vacation draft in the documentation
4172 and documents the missing references header update
4176 o fixes a bug in processing the envelope test (when testing
4177 multiple envelope elements, the last element determined the
4180 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4181 Electronic Mail") by including:
4183 Auto-submitted: auto-generated
4185 in the messages that it generates (bounce messages and others, such as
4186 warnings). In the case of bounce messages for non-SMTP messages, there was
4187 also a typo: it was using "Auto_submitted" (underscore instead of
4188 hyphen). Since every message generated by Exim is necessarily in response
4189 to another message, thes have all been changed to:
4191 Auto-Submitted: auto-replied
4193 in accordance with these statements in the RFC:
4195 The auto-replied keyword:
4197 - SHOULD be used on messages sent in direct response to another
4198 message by an automatic process,
4200 - MUST NOT be used on manually-generated messages,
4202 - MAY be used on Delivery Status Notifications (DSNs) and Message
4203 Disposition Notifications (MDNs),
4205 - MUST NOT be used on messages generated by automatic or periodic
4206 processes, except for messages which are automatic responses to
4209 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4210 to the default Received: header definition.
4212 PH/12 Added log selector acl_warn_skipped (default on).
4214 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4215 variables because (a) they are in the wrong storage pool and (b) even if
4216 they were copied, it wouldn't work properly because of the caching.
4218 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4219 checking when decoding. Apparently there are clients that generate
4220 overlong encoded strings. Why am I not surprised?
4222 PH/15 If the first argument of "${if match_address" was not empty, but did not
4223 contain an "@" character, Exim crashed. Now it writes a panic log message
4224 and treats the condition as false.
4226 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4228 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4229 extension "envelope-auth". The code is finished and in agreement with
4230 other implementations, but there is no documentation so far and in fact,
4231 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4232 not changing the active code.
4234 Print executed "if" and "elsif" statements when debugging is used. This
4235 helps a great deal to understand what a filter does.
4237 Document more things not specified clearly in RFC3028. I had all this
4238 sorted out, when out of a sudden new issues came to my mind. Oops."
4240 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4243 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4244 canonical form (as documented). However, after a host name lookup from
4245 the IP address, check_host() was doing a simple string comparison with
4246 addresses acquired from the DNS when checking that the found name did
4247 have the original IP as one of its addresses. Since any found IPv6
4248 addresses are likely to be in abbreviated form, the comparison could
4249 fail. Luckily, there already exists a function for doing the comparison
4250 by converting both addresses to binary, so now that is used instead of
4251 the text comparison.
4253 PH/20 There was another similar case to PH/19, when a complete host name was
4254 given in a host list; looking up its IP address could give an abbreviated
4255 form, whereas the current host's name might or might not be abbreviated.
4256 The same fix has been applied.
4262 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4263 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4266 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4267 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4269 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4271 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4272 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4273 Also, for "Auto-Submitted", treat anything other than "no" as
4274 non-personal, in accordance with RFC 3834. (Previously it treated
4275 anything starting "auto-" as non-personal.)
4277 TF/01 The control=submission/name=... option had a problem with syntax
4278 errors if the name included a slash character. The /name= option
4279 now slurps the rest of the string, so it can include any characters
4280 but it must come last in the list of options (after /sender_retain
4283 PH/05 Some modifications to the interface to the fake nameserver for the new
4291 TK/01 Added the "success_on_redirect" address verification option. See
4292 NewStuff for rationale and an example.
4294 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4296 PH/02 Patch to exigrep to allow it to work on syslog lines.
4298 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4299 fread() to skip over the body file's header line, because in Cygwin the
4300 header line is locked and is inaccessible.
4302 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4303 co-exist for some time) to make it clear that it is the Exim ID that is
4304 referenced, not the Message-ID: header line.
4306 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4307 string_format() function, because snprintf() does not exist on all
4310 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4311 ${stat: expansion item.
4313 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4314 protocol synchronization error", to keep the pedants happy.
4316 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4317 well as for IRIX systems, when gcc is being used. See the host.c source
4320 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4322 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4325 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4326 transport and to the smtp transport in LMTP mode.
4328 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4330 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4331 run a filter in a subprocess. This could lead to confusion in subsequent
4332 lookups in the parent process. There should also be a search_tidyup() at
4333 the end of the subprocess.
4335 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4336 only if the host matched helo_try_verify_hosts, which caused the
4337 verification to occur when the EHLO/HELO command was issued. The ACL just
4338 tested the remembered result. Now, if a previous verification attempt has
4339 not happened, "verify = helo" does it there and then.
4341 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4343 TK/03 Fix log output including CR from clamd.
4345 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4346 exist provoked a memory error which could cause a segfault.
4348 PH/15 Installed PCRE 6.2
4350 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4352 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4353 of the problem. Specifically, suggested +O2 rather than +O1 for the
4356 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4358 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4359 record, the port was not correctly logged when the outgoing_port log
4360 selector was set (it logged the transort's default port).
4362 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4363 fallback_hosts, and "hosts" in the smtp transport.
4365 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4366 host errors such as "Connection refused".
4368 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4369 authentication with radiusclient 0.4.9:
4371 - Error returned from rc_read_config was caught wrongly
4372 - Username/password not passed on to radius server due to wrong length.
4374 The presumption is that some radiusclient API changes for 4.51/PH/17
4375 were not taken care of correctly. The code is still untested by me (my
4376 Linux distribution still has 0.3.2 of radiusclient), but it was
4377 contributed by a Radius user.
4379 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4380 expanding the "port" option of the smtp transport.
4382 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4383 while reading a MIME header. Thanks to Tom Hughes for a patch.
4385 PH/24 Include config.h inside local_scan.h so that configuration settings are
4388 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4389 that in an ACL for RCPT (for example), you can examine exactly what was
4392 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4393 commands, but it was not correctly comparing the address with the actual
4394 client host address. Thus, it would show the EHLO address in Received:
4395 header lines when this was not necessary.
4397 PH/27 Added the % operator to ${eval:}.
4399 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4400 it should be ignoring failures (because with -C, for example, it has lost
4401 privilege). It wasn't ignoring creation failures other than "already
4404 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4407 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4408 because an input line was too long, either on its own, or by virtue of
4409 too many continuations, the temporary file was not being removed, and the
4410 return code was incorrect.
4412 PH/31 Missing "BOOL" in function definition in filtertest.c.
4414 PH/32 Applied Sieve patches from the maintainer.
4416 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4418 PH/33 Added "verify = not_blind".
4420 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4421 Local/Makefile (with some defaults set). These are used in built scripts
4422 such as exicyclog, but they have never been used in the exim_install
4423 script (though there are many overriding facilities there). I have
4424 arranged that the exim_install script now takes note of these two
4427 PH/35 Installed configuration files for Dragonfly.
4429 PH/36 When a locally submitted message by a trusted user did not contain a
4430 From: header, and the sender address was obtained from -f or from an SMTP
4431 MAIL command, and the trusted user did not use -F to supply a sender
4432 name, $originator_name was incorrectly used when constructing a From:
4433 header. Furthermore, $originator_name was used for submission mode
4434 messages from external hosts without From: headers in a similar way,
4435 which is clearly wrong.
4437 PH/37 Added control=suppress_local_fixups.
4439 PH/38 When log_selector = +received_sender was set, and the addition of the
4440 sender made the log line's construction buffer exactly full, or one byte
4441 less than full, an overflow happened when the terminating "\n" was
4444 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4445 when the result of a list match is failure because a DNS lookup failed.
4447 PH/40 RM_COMMAND is now used in the building process.
4449 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4450 the "build-* directories that it finds.
4452 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4453 address in a domain literal was a prefix of an interface address.
4455 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4456 when verifying a sender address, unless rewrite_headers is false.
4458 PH/44 Wrote a long comment about why errors_to addresses are verified as
4459 recipients, not senders.
4461 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4462 the ratelimit ACL was added.
4464 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4466 PH/46 Added extra information about PostgreSQL errors to the error string.
4468 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4469 avoiding the need to install special zones in a real server. This is
4470 backwards compatible; if it can't find the fake resolver, it drops back.
4471 Thus, both old and new test suites can be run.
4473 TF/02 Added util/ratelimit.pl
4475 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4476 clock is set back in time.
4478 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4479 Candler <B.Candler@pobox.com>.
4481 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4482 OK for addresses that are the result of redirections.
4484 PH/48 A number of further additions for the benefit of the new test suite,
4485 including a fake gethostbyname() that interfaces to the fake DNS resolver
4488 TF/06 The fix for widen_domains has also been applied to qualify_single and
4489 search_parents which are the other dnslookup options that can cause
4492 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4495 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4497 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4498 service" if it encountered an MX record with an empty target hostname.
4499 The message is now "an MX or SRV record indicated no SMTP service".
4501 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4502 if the DNS of the sending site is misconfigured. This is quite a
4503 common situation. This change restores the behaviour of treating a
4504 helo verification defer as a failure.
4506 PH/51 If self=fail was set on a router, the bounce message did not include the
4507 actual error message.
4513 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4515 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4516 command itself ended in error, the underlying message about the transport
4517 filter timeout was being overwritten with the pipe command error. Now the
4518 underlying error message should be appended to the second error message.
4520 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4522 PH/02 Reduce the amount of output that "make" produces by default. Full output
4523 can still be requested.
4525 PH/03 The warning log line about a condition test deferring for a "warn" verb
4526 was being output only once per connection, rather than after each
4527 occurrence (because it was using the same function as for successful
4528 "warn" verbs). This seems wrong, so I have changed it.
4530 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4531 it should not have, which might have caused a crash in the right
4532 circumstances, but probably never did.
4534 PH/04 Installed a modified version of Tony Finch's patch to make submission
4535 mode fix the return path as well as the Sender: header line, and to
4536 add a /name= option so that you can make the user's friendly name appear
4539 TF/03 Added the control = fakedefer ACL modifier.
4541 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4542 Mark Lowes for thorough testing.
4544 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4546 TK/03 Merged latest SRS patch from Miles Wilton.
4548 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4549 with the definition in sysexits.h (which is #included earlier).
4550 Fortunately, Exim does not actually use EX_OK. The code used to try to
4551 preserve the sysexits.h value, by assuming that macro definitions were
4552 scanned for macro replacements. I have been disabused of this notion,
4553 so now the code just undefines EX_OK before #including unistd.h.
4555 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4556 in the smtp transport. When a block could not be written in a single
4557 write() function, the timeout was being re-applied to each part-write.
4558 This seems wrong - if the receiver was accepting one byte at a time it
4559 would take for ever. The timeout is now adjusted when this happens. It
4560 doesn't have to be particularly precise.
4562 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4563 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4565 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4566 without a domain if the check to <postmaster@domain> fails.
4568 SC/01 Eximstats: added -xls and the ability to specify output files
4569 (patch written by Frank Heydlauf).
4571 SC/02 Eximstats: use FileHandles for outputting results.
4573 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4575 SC/04 Eximstats: fixed display of large numbers with -nvr option
4577 SC/05 Eximstats: fixed merging of reports with empty tables.
4579 SC/06 Eximstats: added the -include_original_destination flag
4581 SC/07 Eximstats: removed tabs and trailing whitespace.
4583 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4585 TK/06 MBOX spool code: Add real "From " MBOX separator line
4586 so the .eml file is really in mbox format (even though
4587 most programs do not really care). Patch from Alex Miller.
4589 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4590 The latter is generated from $received_to and is only set if the
4591 message has one envelope recipient. SA can use these headers,
4592 obviously out-of-the-box. Patch from Alex Miller.
4594 PH/08 The ${def test on a variable was returning false if the variable's
4595 value was "0", contrary to what the specification has always said!
4596 The result should be true unless the variable is empty.
4598 PH/09 The syntax error of a character other than { following "${if
4599 def:variable_name" (after optional whitespace) was not being diagnosed.
4600 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4601 accidental colon was present, for example, could give incorrect results.
4603 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4604 result is used (not including appendfile, where other changes are about
4607 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4608 This involved changing a lot of size variables from int to off_t. It
4609 should work with maildirs and everything.
4611 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4612 spamd dying while we are connected to it.
4614 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4617 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4618 function for BDB 4.3.
4620 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4622 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4623 the sources for PCRE have been re-arranged and more files are now
4626 PH/15 The code I had for printing potentially long long variables in PH/11
4627 above was not the best (it lost precision). The length of off_t variables
4628 is now inspected at build time, and an appropriate printing format (%ld
4629 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4630 to be "long long int" or "long int". This is needed for the internal
4631 formatting function string_vformat().
4633 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4634 the configuration file to be ":syslog", then the script "guesses" where
4635 the logs files are, rather than using the compiled in default. In our
4636 case the guess is not the same as the compiled default, so the script
4637 suddenly stopped working when I started to use syslog. The patch checks
4638 to see if log_file_path is "". If so, it attempts to read it from exim
4639 with no configuration file to get the compiled in version, before it
4640 falls back to the previous guessing code."
4642 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4643 implementing BATV in an Exim configuration. See NewStuff for the gory
4646 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4647 Makefile that are specific to HP-UX.
4649 PH/18 If the "use_postmaster" option was set for a recipient callout together
4650 with the "random" option, the postmaster address was used as the MAIL
4651 FROM address for the random test, but not for the subsequent recipient
4652 test. It is now used for both.
4654 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4655 patch removes a few documentation additions to RFC 3028, because the
4656 latest draft now contains them. It adds the new en;ascii-case comparator
4657 and a new error check for 8bit text in MIME parts. Comparator and
4658 require names are now matched exactly. I enabled the subaddress
4659 extension, but it is not well tested yet (read: it works for me)."
4661 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4662 rework some of the code of TK/09 above to avoid the hardwired use of
4663 "%lld" and "long long". Replaced the call to snprintf() with a call to
4666 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4667 records point to non-existent hosts", "retry timeout exceeded", and
4668 "retry time not reached for any host after a long failure period".
4670 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4671 experimental DomainKeys support:
4673 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4674 (2) On an error such as an illegally used "control", the wrong name for
4675 the control was given.
4677 These problems did NOT occur unless DomainKeys support was compiled.
4679 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4681 PH/24 Added ${if match_ip condition.
4683 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4684 databases so that it will be absolutely obvious if a crash occurs in the
4685 DB library. This is a regular occurrence (often caused by mis-matched
4688 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4689 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4690 picked up on a user's system that detects such things. There doesn't seem
4691 to be a gcc warning option for this - only an attribute that has to be
4692 put on the function's prototype. It seems that in Fedora Core 4 they have
4693 set this on a number of new functions. No doubt there will be more in due
4696 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4697 specify a transport. However, if an address that was verified by such a
4698 router was the subject of a callout, Exim crashed because it tried to
4699 read the rcpt_include_affixes from the non-existent transport. Now it
4700 just assumes that the setting of that option is false. This bug was
4701 introduced by 4.51/PH/31.
4703 PH/28 Changed -d+all to exclude +memory, because that information is very
4704 rarely of interest, but it makes the output a lot bigger. People tend to
4705 do -d+all out of habit.
4707 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4708 code in os-type was giving problems when libc.so lives in lib64, like on
4711 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4712 aren't the modern standard, and it seems that some systems' include files
4713 don't always have them. Exim was already checking for some of the newer
4714 ones like T_AAAA, and defining it itself. I've added checks for all the
4715 record types that Exim uses.
4717 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4718 not automatically generating a new one, as it is supposed to. This
4719 prevented TLS from working. If the file did exist, but contained invalid
4720 data, a new version was generated, as expected. It was only the case of a
4721 non-existent file that was broken.
4723 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4724 with a change in libdomainkeys > 0.64.
4726 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4727 from DNS. If the selector record carries the flag, it now has
4728 precedence over the domain-wide flag.
4730 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4732 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4733 the use of a port name that isn't defined in /etc/services) occurred, the
4734 message was deferred as in a normal delivery, and thus remained on the
4735 spool, instead of being failed because of the mua_wrapper setting. This
4736 is now fixed, and I tidied up some of the mua_wrapper messages at the
4739 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4740 the messages in a hash of arrays rather than using individual hashes.
4741 This is a bit cleaner and results in dramatic memory savings, albeit
4742 at a slight CPU cost.
4744 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4745 as requested by Marc Sherman.
4747 SC/10 Eximstats: added histograms for user specified patterns as requested
4750 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4752 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4753 fopen() in the content-scanning modules that did not already have it.
4759 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4760 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4762 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4764 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4766 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4767 to test Sieve filters that use "vacation".
4769 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4770 that changes the way the GnuTLS parameters are stored in the cache file.
4771 The new format can be generated externally. For backward compatibility,
4772 if the data in the cache doesn't make sense, Exim assumes it has read an
4773 old-format file, and it generates new data and writes a new file. This
4774 means that you can't go back to an older release without removing the
4777 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4778 work if there are any delivery delays because "one_time" forces the
4779 parent to be marked "delivered", so its unseen clone is never tried
4780 again. For this reason, Exim now forbids the simultaneous setting of
4783 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4784 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4785 entry if you want to know the details. The fix, however, neglected to
4786 consider the case when local delivery batching is involved. The test for
4787 "previously delivered" was not happening when checking to see if an
4788 address could be batched with a previous (undelivered) one; under
4789 certain circumstances this could lead to multiple deliveries to the same
4792 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4793 in its include files, and this causes problems building Exim.
4795 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4796 header_syntax) but Exim was just ignoring anything given after a slash.
4797 In particular, this caused confusion with an attempt to use "verify =
4798 reverse_host_lookup/defer_ok". An error is now given when options are
4799 supplied for verify items that do not have them. (Maybe reverse_host_
4800 lookup should have a defer_ok option, but that's a different point.)
4802 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4803 defined by RFC 821) to 2048, because there were problems with some AUTH
4804 commands, and RFC 1869 says the size should be increased for extended
4805 SMTP commands that take arguments.
4807 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4810 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4811 "unknown" error; now it says that the functionality isn't in the binary.
4813 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4814 an address' error message when a string expansion fails (syntax or
4815 whatever). Otherwise the password may appear in the log. Following change
4816 PH/42 below, there is no longer a chance of it appearing in a bounce
4819 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4821 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4822 of its addresses was ever tried. (Bugzilla bug #2.)
4824 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4825 the result incorrectly in the debug output. (It correctly added a newline
4826 to what was transported.)
4828 TF/01 Added $received_time.
4830 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4831 commented out examples of how to interface to a virus scanner and to
4832 SpamAssassin. Also added commented examples of av_scanner and
4833 spamd_address settings.
4835 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4836 and controls are allowed in which ACLs. There were a couple of minor
4837 errors. Some of the entries in the conditions table (which is a table of
4838 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4839 negation of where the condition IS allowed.
4841 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4843 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4844 header file does not have a version number, so I've had to invent a new
4845 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4846 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4847 radiusclient), but it was contributed by a Radius user.
4849 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4850 files or directories, for OpenSSL.
4852 PH/19 When an Exim process that is running as root has to create an Exim log
4853 file, it does so in a subprocess that runs as exim:exim so as to get the
4854 ownership right at creation (otherwise, other Exim processes might see
4855 the file with the wrong ownership). There was no test for failure of this
4856 fork() call, which would lead to the process getting stuck as it waited
4857 for a non-existent subprocess. Forks do occasionally fail when resources
4858 run out. I reviewed all the other calls to fork(); they all seem to check
4861 PH/20 When checking for unexpected SMTP input at connect time (before writing
4862 the banner), Exim was not dealing correctly with a non-positive return
4863 from the read() function. If the client had disconnected by this time,
4864 the result was a log entry for a synchronization error with an empty
4865 string after "input=" when read() returned zero. If read() returned -1
4866 (an event I could not check), uninitialized data bytes were printed.
4867 There were reports of junk text (parts of files, etc) appearing after
4870 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4872 PH/22 Added support for macro redefinition, and (re)definition in between
4873 driver and ACL definitions.
4875 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4876 forgetting to use the resulting value; it was using the unexpanded value.
4878 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4879 hadn't been configured. The fix is from Juergen Kreileder, who
4880 understands it better than I do:
4882 "Here's what I see happening with three configured cyrus_sasl
4883 authenticators configured (plain, login, cram-md5):
4885 On startup auth_cyrus_sasl_init() gets called for each of these.
4886 This means three calls to sasl_listmech() without a specified mech_list.
4887 => SASL tests which mechs of all available mechs actually work
4888 => three warnings about OTP not working
4889 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4891 With the patch, sasl_listmech() also gets called three times. But now
4892 SASL's mech_list option is set to the server_mech specified in the the
4893 authenticator. Or in other words, the answer from sasl_listmech()
4894 gets limited to just the mech you're testing for (which is different
4896 => the return list contains just 'plain' or 'login', 'cram-md5' or
4897 nothing depending on the value of ob->server_mech.
4899 I've just tested the patch: Authentication still works fine,
4900 unavailable mechs specified in the exim configuration are still
4901 caught, and the auth.log warnings about OTP are gone."
4903 PH/25 When debugging is enabled, the contents of the command line are added
4904 to the debugging output, even when log_selector=+arguments is not
4907 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4908 answer is "GNU", and only if the return is "GNU/something" is the answer
4911 PH/27 $acl_verify_message is now set immediately after the failure of a
4912 verification in an ACL, and so is available in subsequent modifiers. In
4913 particular, the message can be preserved by coding like this:
4915 warn !verify = sender
4916 set acl_m0 = $acl_verify_message
4918 Previously, $acl_verify_message was set only while expanding "message"
4919 and "log_message" when a very denied access.
4921 PH/28 Modified OS/os.c-Linux with
4923 -#ifndef OS_LOAD_AVERAGE
4924 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4926 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4927 nomenclature these days.)
4929 PH/29 Installed patch from the Sieve maintainer that adds the options
4930 sieve_useraddress and sieve_subaddress to the redirect router.
4932 PH/30 In these circumstances:
4933 . Two addresses routed to the same list of hosts;
4934 . First host does not offer TLS;
4935 . First host accepts first address;
4936 . First host gives temporary error to second address;
4937 . Second host offers TLS and a TLS session is established;
4938 . Second host accepts second address.
4939 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4940 and peerdn, if requested) that were in fact used only for the second
4943 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4944 attention to any local part prefix or suffix that was matched by the
4945 router that accepted the address. It now behaves in the same way as it
4946 does for delivery: the affixes are removed from the local part unless
4947 rcpt_include_affixes is set on the transport.
4949 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4950 timeout during the DATA phase of an incoming message.
4952 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4953 applied a patch sanctioned by the Sieve maintainer.
4955 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4956 the uid or gid is negative. A case of a negative gid caused this to be
4957 noticed. The fix allows for either to be negative.
4959 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4960 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4963 PH/36 The change PH/12 above was broken. Fixed it.
4965 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4966 the grounds that routing the same address twice would always produce the
4967 same answer. This might have been true once, but it is certainly no
4968 longer true now. Routing a child address may depend on the previous
4969 routing that produced that child. Some complicated redirection strategies
4970 went wrong when messages had multiple recipients, and made Exim's
4971 behaviour dependent on the order in which the addresses were given.
4973 I have moved the duplicate checking until after the routing is complete.
4974 Exim scans the addresses that are assigned to local and remote
4975 transports, and removes any duplicates. This means that more work will be
4976 done, as duplicates will always all be routed, but duplicates are
4977 presumably rare, so I don't expect this is of any significance.
4979 For deliveries to pipes, files, and autoreplies, the duplicate checking
4980 still happens during the routing process, since they are not going to be
4983 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4984 It corrects a timeout issue with spamd. This is Ian's comment: "The
4985 background is that sometimes spamd either never reads data from a
4986 connection it has accepted, or it never writes response data. The exiscan
4987 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4988 blindly assumes that writes won't block so it may never time out."
4990 PH/39 Allow G after quota size as well as K and M.
4992 PH/40 The value set for $authenticated_id in an authenticator may not contain
4993 binary zeroes or newlines because the value is written to log lines and
4994 to spool files. There was no check on this. Now the value is run through
4995 the string_printing() function so that such characters are converted to
4996 printable escape sequences.
4998 PH/41 $message_linecount is a new variable that contains the total number of
4999 lines in the message. Compare $body_linecount, which is the count for the
5002 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5003 bounce and delay warning messages, except in certain special cases, which
5006 (a) An SMTP error message from a remote host;
5007 (b) A message specified in a :fail: redirection;
5008 (c) A message specified in a "fail" command in a system filter;
5009 (d) A message specified in a FAIL return from the queryprogram router;
5010 (e) A message specified by the cannot_route_message router option.
5012 In these cases only, Exim does include the error details in bounce and
5013 warning messages. There are also a few cases where bland messages such
5014 as "unrouteable address" or "local delivery error" are given.
5016 PH/43 $value is now also set for the "else" part of a ${run expansion.
5018 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5019 being worked on, but at least Exim now implements the latest version to
5022 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5023 process to complete was treated as a delivery failure, a timeout while
5024 writing the message to the pipe was logged, but erroneously treated as a
5025 successful delivery. Such timeouts include transport filter timeouts. For
5026 consistency with the overall process timeout, these timeouts are now
5027 treated as errors, giving rise to delivery failures by default. However,
5028 there is now a new Boolean option for the pipe transport called
5029 timeout_defer, which, if set TRUE, converts the failures into defers for
5030 both kinds of timeout. A transport filter timeout is now identified in
5033 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5034 systems where "make" and "gmake" are different, calling "gmake" at top
5035 level broke things. I've arranged for the value of $(MAKE) to be passed
5036 from the Makefile to this script so that it can call the same version of
5040 A note about Exim versions 4.44 and 4.50
5041 ----------------------------------------
5043 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5044 changes of various kinds. As a consequence, a big documentation update was
5045 needed. This delayed the release for rather longer than seemed good, especially
5046 in the light of a couple of (minor) security issues. Therefore, the changes
5047 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5048 release. So 4.44 and 4.50 are in effect two different branches that both start
5051 I have left the 4.50 change log unchanged; it contains all the changes since
5052 4.43. The change log for 4.44 is below; many of its items are identical to
5053 those for 4.50. This seems to be the most sensible way to preserve the
5054 historical information.
5060 1. Minor wording change to the doc/README.SIEVE file.
5062 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5063 computation of the current number of files was incorrect.
5065 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5066 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5069 4. Give more explanation in the error message when the command for a transport
5070 filter fails to execute.
5072 5. There are several places where Exim runs a non-Exim command in a
5073 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5074 was being done only for the command run by the queryprogram router. It is
5075 now done for all such subprocesses. The other cases are: ${run, transport
5076 filters, and the commands run by the lmtp and pipe transports.
5078 6. Added CONFIGURE_GROUP build-time option.
5080 7. Some older OS have a limit of 256 on the maximum number of file
5081 descriptors. Exim was using setrlimit() to set 1000 as a large value
5082 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5083 systems. I've change it so that if it can't get 1000, it tries for 256.
5085 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5086 was an oversight, and furthermore, ever since the addition of extra
5087 controls (e.g. 4.43/32), the checks on when to allow different forms of
5088 "control" were broken. There should now be diagnostics for all cases when a
5089 control that does not make sense is encountered.
5091 9. Added the /retain_sender option to "control=submission".
5093 10. $recipients is now available in the predata ACL (oversight).
5095 11. Tidy the search cache before the fork to do a delivery from a message
5096 received from the command line. Otherwise the child will trigger a lookup
5097 failure and thereby defer the delivery if it tries to use (for example) a
5098 cached ldap connection that the parent has called unbind on.
5100 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5101 of $address_data from the recipient verification was clobbered by the
5102 sender verification.
5104 13. The value of address_data from a sender verification is now available in
5105 $sender_address_data in subsequent conditions in the ACL statement.
5107 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5109 15. Added a new option "connect=<time>" to callout options, to set a different
5112 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5113 was its contents. (It was OK if the option was not defined at all.)
5115 17. A "Completed" log line is now written for messages that are removed from
5116 the spool by the -Mrm option.
5118 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5119 information about exactly what failed.
5121 19. Added -dd to debug only the daemon process.
5123 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5124 handles timeouts, both on the server side and network timeouts. Renamed the
5125 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5127 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5128 It is now set to "smtps".
5130 22. $host_address is now set to the target address during the checking of
5131 ignore_target_hosts.
5133 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5134 being passed; this would have caused $sender_host_name to have been used if
5135 matching the list had actually called for a host name (not very likely,
5136 since this list is usually IP addresses). A host name is now passed as
5139 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5140 code that specifies a non-restarting handler (typically sigaction() in
5141 modern systems) in an attempt to fix a rare and obscure crash bug.
5143 25. Narrowed the window for a race in the daemon that could cause it to ignore
5144 SIGCHLD signals. This is not a major problem, because they are used only to
5145 wake it up if nothing else does.
5147 26. A malformed maildirsize file could cause Exim to calculate negative values
5148 for the mailbox size or file count. Odd effects could occur as a result.
5149 The maildirsize information is now recalculated if the size or filecount
5152 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5153 support for a long time. Removed HAVE_SYS_VFS_H.
5155 28. Installed the latest version of exipick from John Jetmore.
5157 29. In an address list, if the pattern was not a regular expression, an empty
5158 subject address (from a bounce message) matched only if the pattern was an
5159 empty string. Non-empty patterns were not even tested. This was the wrong
5160 because it is perfectly reasonable to use an empty address as part of a
5161 database query. An empty address is now tested by patterns that are
5162 lookups. However, all the other forms of pattern expect the subject to
5163 contain a local part and a domain, and therefore, for them, an empty
5164 address still always fails if the pattern is not itself empty.
5166 30. Exim went into a mad DNS loop when attempting to do a callout where the
5167 host was specified on an smtp transport, and looking it up yielded more
5168 than one IP address.
5170 31. Re-factored the code for checking spool and log partition space into a
5171 function that finds that data and another that does the check. The former
5172 is then used to implement four new variables: $spool_space, $log_space,
5173 $spool_inodes, and $log_inodes.
5175 32. The RFC2047 encoding function was originally intended for short strings
5176 such as real names; it was not keeping to the 75-character limit for
5177 encoded words that the RFC imposes. It now respects the limit, and
5178 generates multiple encoded words if necessary. To be on the safe side, I
5179 have increased the buffer size for the ${rfc2047: expansion operator from
5182 33. It is now permitted to omit both strings after an "if" condition; if the
5183 condition is true, the result is "true". As before, when the second string
5184 is omitted, a false condition yields an empty string. This makes it less
5185 cumbersome to write custom ACL and router conditions.
5187 34. Failure to deliver a bounce message always caused it to be frozen, even if
5188 there was an errors_to setting on the router. The errors_to setting is now
5191 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5192 canonical form (fully expanded) before being placed in
5193 $sender_host_address.
5195 36. The table in the code that translates DNS record types into text (T_A to
5196 "A" for instance) was missing entries for NS and CNAME. It is just possible
5197 that this could have caused confusion if both these types were looked up
5198 for the same domain, because the text type is used as part of Exim's
5199 per-process caching. But the chance of anyone hitting this buglet seems
5202 37. The dnsdb lookup has been extended in a number of ways.
5204 (1) There is a new type, "zns", which walks up the domain tree until it
5205 finds some nameserver records. It should be used with care.
5207 (2) There is a new type, "mxh", which is like "mx" except that it returns
5208 just the host names, not the priorities.
5210 (3) It is now possible to give a list of domains (or IP addresses) to be
5211 looked up. The behaviour when one of the lookups defers can be
5212 controlled by a keyword.
5214 (4) It is now possible to specify the separator character for use when
5215 multiple records are returned.
5217 38. The dnslists ACL condition has been extended: it is now possible to supply
5218 a list of IP addresses and/or domains to be looked up in a particular DNS
5221 39. Added log_selector=+queue_time_overall.
5223 40. When running the queue in the test harness, wait just a tad after forking a
5224 delivery process, to get repeatability of debugging output.
5226 41. Include certificate and key file names in error message when GnuTLS fails
5227 to set them up, because the GnuTLS error message doesn't include the name
5228 of the failing file when there is a problem reading it.
5230 42. Allow both -bf and -bF in the same test run.
5232 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5234 44. The "Exiscan patch" is now merged into the mainline Exim source.
5236 45. Sometimes the final signoff response after QUIT could fail to get
5237 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5238 < 0 before doing a fflush(). This bug looks as though it goes back to the
5239 introduction of TLS in release 3.20, but "sometimes" must have been rare
5240 because the tests only now provoked it.
5242 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5243 (this can affect the format of dates).
5245 47. exim_tidydb, when checking for the continued existence of a message for
5246 which it has found a message-specific retry record, was not finding
5247 messages that were in split spool directories. Consequently, it was
5248 deleting retry records that should have stayed in existence.
5250 48. Steve fixed some bugs in eximstats.
5252 49. The SPA authentication driver was not abandoning authentication and moving
5253 on to the next authenticator when an expansion was forced to fail,
5254 contradicting the general specification for all authenticators. Instead it
5255 was generating a temporary error. It now behaves as specified.
5257 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5258 (the order specifies the preference for clients). The order is now AES256,
5259 AES128, 3DES, ARCFOUR128.
5261 51. Small patch to Sieve code - explicitly set From: when generating an
5264 52. Exim crashed if a remote delivery caused a very long error message to be
5265 recorded - for instance if somebody sent an entire SpamAssassin report back
5266 as a large number of 550 error lines. This bug was coincidentally fixed by
5267 increasing the size of one of Exim's internal buffers (big_buffer) that
5268 happened as part of the Exiscan merge. However, to be on the safe side, I
5269 have made the code more robust (and fixed the comments that describe what
5272 53. Now that there can be additional text after "Completed" in log lines (if
5273 the queue_time_overall log selector is set), a one-byte patch to exigrep
5274 was needed to allow it to recognize "Completed" as not the last thing in
5277 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5278 patch that reportedly fixes this has been added. I am not expert enough to
5279 create a test for it. This is what the patch creator wrote:
5281 "I found a little strange behaviour of ldap code when working with
5282 Windows 2003 AD Domain, where users was placed in more than one
5283 Organization Units. When I tried to give exim partial DN, the exit code
5284 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5285 But simultaneously result of request was absolutely normal ldap result,
5286 so I produce this patch..."
5288 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5289 REFERENCE, so I have modified the code to exclude the patch when that macro
5292 55. Some experimental protocols are using DNS PTR records for new purposes. The
5293 keys for these records are domain names, not reversed IP addresses. The
5294 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5295 leaves it alone. Component reversal etc. now happens only for IP addresses.
5298 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5300 57. Double the size of the debug message buffer (to 2048) so that more of very
5301 long debug lines gets shown.
5303 58. The exicyclog utility now does better if the number of log files to keep
5304 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5306 59. Two changes related to the smtp_active_hostname option:
5308 (1) $smtp_active_hostname is now available as a variable.
5309 (2) The default for smtp_banner uses $smtp_active_hostname instead
5310 of $primary_hostname.
5312 60. The host_aton() function is supposed to be passed a string that is known
5313 to be a valid IP address. However, in the case of IPv6 addresses, it was
5314 not checking this. This is a hostage to fortune. Exim now panics and dies
5315 if the condition is not met. A case was found where this could be provoked
5316 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5317 components; fortuitously, this particular loophole had already been fixed
5318 by change 4.50/55 above.
5320 If there are any other similar loopholes, the new check in host_aton()
5321 itself should stop them being exploited. The report I received stated that
5322 data on the command line could provoke the exploit when Exim was running as
5323 exim, but did not say which command line option was involved. All I could
5324 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5325 running as the user.
5328 61. There was a buffer overflow vulnerability in the SPA authentication code
5329 (which came originally from the Samba project). I have added a test to the
5330 spa_base64_to_bits() function which I hope fixes it.
5333 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5334 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5336 63. The daemon start-up calls getloadavg() while still root for those OS that
5337 need the first call to be done as root, but it missed one case: when
5338 deliver_queue_load_max is set with deliver_drop_privilege. This is
5339 necessary for the benefit of the queue runner, because there is no re-exec
5340 when deliver_drop_privilege is set.
5342 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5343 This has been fixed.
5345 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5346 was in use, was not putting the data itself into the right store pool;
5347 consequently, it could be overwritten for a subsequent message in the same
5348 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5351 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5353 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5354 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5355 the function were treating the return as a boolean value, which happened to
5356 work because 0=false and not-0=true, but is not correct code.
5358 68. The host_aton() function was not handling scoped IPv6 addresses (those
5359 with, for example, "%eth0" on the end) correctly.
5361 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5362 negated items (that is, ~something) in unsigned ints. Some compilers
5363 apparently mutter when there is no cast.
5365 70. If an address verification called from an ACL failed, and did not produce a
5366 user-specific message (i.e. there was only a "system" message), nothing was
5367 put in $acl_verify_message. In this situation, it now puts the system
5370 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5371 session; change 4.31/43 added the unwanted input to the log line - except
5372 that it did not do this in the start of session case. It now does.
5374 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5375 This is wrong and can cause the other end to generate a synchronization
5376 error if it is another Exim or anything else that does the synchronization
5377 check. A QUIT command is no longer sent after a timeout.
5379 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5380 during host lookups.
5382 74. The defer_ok option of callout verification was not working if it was used
5383 when verifying addresses in header lines, that is, for this case:
5385 verify = header_sender/callout=defer_ok
5387 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5388 those file descriptors could be used for SMTP connections. If anything
5389 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5390 could be sent to the SMTP client, causing chaos. The daemon now opens
5391 stdin, stdout, and stderr to /dev/null when it puts itself into the
5394 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5395 log by default. The user can override this with suitable Perl magic.
5397 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5398 the log message when discard triggers, was not working for the DATA ACL or
5399 for the non-SMTP ACL.
5401 78. Error message wording change in sieve.c.
5403 79. If smtp_accept_max_per_host was set, the number of connections could be
5404 restricted to fewer than expected, because the daemon was trying to set up
5405 a new connection before checking whether the processes handling previous
5406 connections had finished. The check for completed processes is now done
5407 earlier. On busy systems, this bug wouldn't be noticed because something
5408 else would have woken the daemon, and it would have reaped the completed
5411 80. If a message was submitted locally by a user whose login name contained one
5412 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5413 It caused a spool format error. I have fixed the spool reading code. A
5414 related problem was that the "from" clause in the Received: line became
5415 illegal because of the space(s). It is now covered by ${quote_local_part.
5417 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5420 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5422 83. Patch from Sieve maintainer to fix unterminated string problem in
5423 "vacation" handling.
5425 84. Some minor changes to the Linux configuration files to help with other
5426 OS variants using glibc.
5428 85. One more patch for Sieve to update vacation handling to latest spec.
5431 ----------------------------------------------------
5432 See the note above about the 4.44 and 4.50 releases.
5433 ----------------------------------------------------
5439 1. Change 4.43/35 introduced a bug that caused file counts to be
5440 incorrectly computed when quota_filecount was set in an appendfile
5443 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5444 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5447 3. Give more explanation in the error message when the command for a transport
5448 filter fails to execute.
5450 4. There are several places where Exim runs a non-Exim command in a
5451 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5452 was being done only for the command run by the queryprogram router. It is
5453 now done for all such subprocesses. The other cases are: ${run, transport
5454 filters, and the commands run by the lmtp and pipe transports.
5456 5. Some older OS have a limit of 256 on the maximum number of file
5457 descriptors. Exim was using setrlimit() to set 1000 as a large value
5458 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5459 systems. I've change it so that if it can't get 1000, it tries for 256.
5461 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5462 was an oversight, and furthermore, ever since the addition of extra
5463 controls (e.g. 4.43/32), the checks on when to allow different forms of
5464 "control" were broken. There should now be diagnostics for all cases when a
5465 control that does not make sense is encountered.
5467 7. $recipients is now available in the predata ACL (oversight).
5469 8. Tidy the search cache before the fork to do a delivery from a message
5470 received from the command line. Otherwise the child will trigger a lookup
5471 failure and thereby defer the delivery if it tries to use (for example) a
5472 cached ldap connection that the parent has called unbind on.
5474 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5475 of $address_data from the recipient verification was clobbered by the
5476 sender verification.
5478 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5479 was its contents. (It was OK if the option was not defined at all.)
5481 11. A "Completed" log line is now written for messages that are removed from
5482 the spool by the -Mrm option.
5484 12. $host_address is now set to the target address during the checking of
5485 ignore_target_hosts.
5487 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5488 being passed; this would have caused $sender_host_name to have been used if
5489 matching the list had actually called for a host name (not very likely,
5490 since this list is usually IP addresses). A host name is now passed as
5493 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5494 code that specifies a non-restarting handler (typically sigaction() in
5495 modern systems) in an attempt to fix a rare and obscure crash bug.
5497 15. Narrowed the window for a race in the daemon that could cause it to ignore
5498 SIGCHLD signals. This is not a major problem, because they are used only to
5499 wake it up if nothing else does.
5501 16. A malformed maildirsize file could cause Exim to calculate negative values
5502 for the mailbox size or file count. Odd effects could occur as a result.
5503 The maildirsize information is now recalculated if the size or filecount
5506 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5507 support for a long time. Removed HAVE_SYS_VFS_H.
5509 18. Updated exipick to current release from John Jetmore.
5511 19. Allow an empty sender to be matched against a lookup in an address list.
5512 Previously the only cases considered were a regular expression, or an
5515 20. Exim went into a mad DNS lookup loop when doing a callout where the
5516 host was specified on the transport, if the DNS lookup yielded more than
5519 21. The RFC2047 encoding function was originally intended for short strings
5520 such as real names; it was not keeping to the 75-character limit for
5521 encoded words that the RFC imposes. It now respects the limit, and
5522 generates multiple encoded words if necessary. To be on the safe side, I
5523 have increased the buffer size for the ${rfc2047: expansion operator from
5526 22. Failure to deliver a bounce message always caused it to be frozen, even if
5527 there was an errors_to setting on the router. The errors_to setting is now
5530 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5531 canonical form (fully expanded) before being placed in
5532 $sender_host_address.
5534 24. Updated eximstats to version 1.33
5536 25. Include certificate and key file names in error message when GnuTLS fails
5537 to set them up, because the GnuTLS error message doesn't include the name
5538 of the failing file when there is a problem reading it.
5540 26. Expand error message when OpenSSL has problems setting up cert/key files.
5543 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5544 (this can affect the format of dates).
5546 28. exim_tidydb, when checking for the continued existence of a message for
5547 which it has found a message-specific retry record, was not finding
5548 messages that were in split spool directories. Consequently, it was
5549 deleting retry records that should have stayed in existence.
5551 29. eximstats updated to version 1.35
5552 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5553 1.35 - bugfix such that pie charts by volume are generated correctly
5555 30. The SPA authentication driver was not abandoning authentication and moving
5556 on to the next authenticator when an expansion was forced to fail,
5557 contradicting the general specification for all authenticators. Instead it
5558 was generating a temporary error. It now behaves as specified.
5560 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5561 (the order specifies the preference for clients). The order is now AES256,
5562 AES128, 3DES, ARCFOUR128.
5564 31. Small patch to Sieve code - explicitly set From: when generating an
5567 32. Exim crashed if a remote delivery caused a very long error message to be
5568 recorded - for instance if somebody sent an entire SpamAssassin report back
5569 as a large number of 550 error lines. This bug was coincidentally fixed by
5570 increasing the size of one of Exim's internal buffers (big_buffer) that
5571 happened as part of the Exiscan merge. However, to be on the safe side, I
5572 have made the code more robust (and fixed the comments that describe what
5575 33. Some experimental protocols are using DNS PTR records for new purposes. The
5576 keys for these records are domain names, not reversed IP addresses. The
5577 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5578 leaves it alone. Component reversal etc. now happens only for IP addresses.
5581 34. The host_aton() function is supposed to be passed a string that is known
5582 to be a valid IP address. However, in the case of IPv6 addresses, it was
5583 not checking this. This is a hostage to fortune. Exim now panics and dies
5584 if the condition is not met. A case was found where this could be provoked
5585 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5586 components; fortuitously, this particular loophole had already been fixed
5587 by change 4.50/55 or 4.44/33 above.
5589 If there are any other similar loopholes, the new check in host_aton()
5590 itself should stop them being exploited. The report I received stated that
5591 data on the command line could provoke the exploit when Exim was running as
5592 exim, but did not say which command line option was involved. All I could
5593 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5594 running as the user.
5597 35. There was a buffer overflow vulnerability in the SPA authentication code
5598 (which came originally from the Samba project). I have added a test to the
5599 spa_base64_to_bits() function which I hope fixes it.
5602 36. The daemon start-up calls getloadavg() while still root for those OS that
5603 need the first call to be done as root, but it missed one case: when
5604 deliver_queue_load_max is set with deliver_drop_privilege. This is
5605 necessary for the benefit of the queue runner, because there is no re-exec
5606 when deliver_drop_privilege is set.
5608 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5609 was in use, was not putting the data itself into the right store pool;
5610 consequently, it could be overwritten for a subsequent message in the same
5611 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5614 38. Sometimes the final signoff response after QUIT could fail to get
5615 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5616 < 0 before doing a fflush(). This bug looks as though it goes back to the
5617 introduction of TLS in release 3.20, but "sometimes" must have been rare
5618 because the tests only now provoked it.
5624 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5625 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5626 Now it returns an integer. A number of calls were still expecting a T/F
5627 return. Fortuitously, in all cases, the tests worked in OK situations,
5628 which is the norm. However, things would have gone wrong on any write
5629 failures on the smtp file descriptor. This function is used when sending
5630 messages over SMTP and also when doing verify callouts.
5632 2. When Exim is called to do synchronous delivery of a locally submitted
5633 message (the -odf or -odi options), it no longer closes stderr before doing
5636 3. Implemented the mua_wrapper option.
5638 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5640 5. Implemented the functions header_remove(), header_testname(),
5641 header_add_at_position(), and receive_remove_recipient(), and exported them
5644 6. If an ACL "warn" statement specified the addition of headers, Exim already
5645 inserted X-ACL-Warn: at the start if there was no header name. However, it
5646 was not making this test for the second and subsequent header lines if
5647 there were newlines in the string. This meant that an invalid header could
5648 be inserted if Exim was badly configured.
5650 7. Allow an ACL "warn" statement to add header lines at the start or after all
5651 the Received: headers, as well as at the end.
5653 8. Added the rcpt_4xx retry error code.
5655 9. Added postmaster_mailfrom=xxx to callout verification option.
5657 10. Added mailfrom=xxxx to the callout verification option, for verify=
5660 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5661 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5663 12. Inserted some casts to stop certain compilers warning when using pointer
5664 differences as field lengths or precisions in printf-type calls (mostly
5665 affecting debugging statements).
5667 13. Added optional readline() support for -be (dynamically loaded).
5669 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5670 same clock tick as a message's arrival, so that its received time was the
5671 same as the "first fail" time on the retry record, and that message
5672 remained on the queue past the ultimate address timeout, every queue runner
5673 would try a delivery (because it was past the ultimate address timeout) but
5674 after another failure, the ultimate address timeout, which should have then
5675 bounced the address, did not kick in. This was a "< instead of <=" error;
5676 in most cases the first failure would have been in the next clock tick
5677 after the received time, and all would be well.
5679 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5680 being recognized when the domain list was tested by the match_domain
5681 condition in an expansion string.
5683 16. Added the ${str2b64: operator.
5685 17. Exim was always calling setrlimit() to set a large limit for the number of
5686 processes, without checking whether the existing limit was already
5687 adequate. (It did check for the limit on file descriptors.) Furthermore,
5688 errors from getrlimit() and setrlimit() were being ignored. Now they are
5689 logged to the main and panic logs, but Exim does carry on, to try to do its
5690 job under whatever limits there are.
5692 18. Imported PCRE 5.0.
5694 19. Trivial typo in log message " temporarily refused connection" (the leading
5697 20. If the log selector return_path_on_delivery was set and an address was
5698 redirected to /dev/null, the delivery process crashed because it assumed
5699 that a return path would always be set for a "successful" delivery. In this
5700 case, the whole delivery is bypassed as an optimization, and therefore no
5703 21. Internal re-arrangement: the function for sending a challenge and reading
5704 a response while authentication was assuming a zero-terminated challenge
5705 string. It's now changed to take a pointer and a length, to allow for
5706 binary data in such strings.
5708 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5710 23. Exim was not respecting finduser_retries when seeking the login of the
5711 uid under which it was called; it was always trying 10 times. (The default
5712 setting of finduser_retries is zero.) Also, it was sleeping after the final
5713 failure, which is pointless.
5715 24. Implemented tls_on_connect_ports.
5717 25. Implemented acl_smtp_predata.
5719 26. If the domain in control=submission is set empty, Exim assumes that the
5720 authenticated id is a complete email address when it generates From: or
5721 Sender: header lines.
5723 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5724 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5725 chown and chgrp in /bin and hostname in /usr/bin.
5727 28. Exim was keeping the "process log" file open after each use, just as it
5728 does for the main log. This opens the possibility of it remaining open for
5729 long periods when the USR1 signal hits a daemon. Occasional processlog
5730 errors were reported, that could have been caused by this. Anyway, it seems
5731 much more sensible not to leave this file open at all, so that is what now
5734 29. The long-running daemon process does not normally write to the log once it
5735 has entered its main loop, and it closes the log before doing so. This is
5736 so that log files can straightforwardly be renamed and moved. However,
5737 there are a couple of unusual error situations where the daemon does write
5738 log entries, and I had neglected to close the log afterwards.
5740 30. The text of an SMTP error response that was received during a remote
5741 delivery was being truncated at 512 bytes. This is too short for some of
5742 the long messages that one sometimes sees. I've increased the limit to
5745 31. It is now possible to make retry rules that apply only when a message has a
5746 specific sender, in particular, an empty sender.
5748 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5749 it possible to be selective about when SMTP synchronization is enforced.
5751 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5753 32. Implemented hosts_connection_nolog.
5755 33. Added an ACL for QUIT.
5757 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5760 35. Added mailbox_size and mailbox_filecount to appendfile.
5762 36. Added control = no_multiline_responses to ACLs.
5764 37. There was a bug in the logic of the code that waits for the clock to tick
5765 in the case where the clock went backwards by a substantial amount such
5766 that the microsecond fraction of "now" was more than the microsecond
5767 fraction of "then" (but the whole seconds number was less).
5769 38. Added support for the libradius Radius client library this is found on
5770 FreeBSD (previously only the radiusclient library was supported).
5776 1. When certain lookups returned multiple values in the form name=value, the
5777 quoting of the values was not always being done properly. Specifically:
5778 (a) If the value started with a double quote, but contained no whitespace,
5780 (b) If the value contained whitespace other than a space character (i.e.
5781 tabs or newlines or carriage returns) it was not quoted.
5782 This fix has been applied to the mysql and pgsql lookups by writing a
5783 separate quoting function and calling it from the lookup code. The fix
5784 should probably also be applied to nisplus, ibase and oracle lookups, but
5785 since I cannot test any of those, I have not disturbed their existing code.
5787 2. A hit in the callout cache for a specific address caused a log line with no
5788 reason for rejecting RCPT. Now it says "Previous (cached) callout
5789 verification failure".
5791 3. There was an off-by-one bug in the queryprogram router. An over-long
5792 return line was truncated at 256 instead of 255 characters, thereby
5793 overflowing its buffer with the terminating zero. As well as fixing this, I
5794 have increased the buffer size to 1024 (and made a note to document this).
5796 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5797 when Exim is waiting for an SMTP response from a remote server, Exim
5798 restarts its select() call on the socket, thereby resetting its timeout.
5799 This is not a problem when such interrupts are rare. Somebody set up a cron
5800 job to run exiwhat every 2 minutes, which is less than the normal select()
5801 timeout (5 or 10 minutes). This meant that the select() timeout never
5802 kicked in because it was always reset. I have fixed this by comparing the
5803 time when an interrupt arrives with the time at the start of the first call
5804 to select(). If more time than the timeout has elapsed, the interrupt is
5805 treated as a timeout.
5807 5. Some internal re-factoring in preparation for the addition of Sieve
5808 extensions (by MH). In particular, the "personal" test is moved to a
5809 separate function, and given an option for scanning Cc: and Bcc: (which is
5810 not set for Exim filters).
5812 6. When Exim created an email address using the login of the caller as the
5813 local part (e.g. when creating a From: or Sender: header line), it was not
5814 quoting the local part when it contained special characters such as @.
5816 7. Installed new OpenBSD configuration files.
5818 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5819 try to make them clearer.
5821 9. Callout options, other than the timeout value, were being ignored when
5822 verifying sender addresses in header lines. For example, when using
5824 verify = header_sender/callout=no_cache
5826 the cache was (incorrectly) being used.
5828 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5829 only the Cygwin environment.
5831 11. When return_path_on_delivery was set as a log selector, if different remote
5832 addresses in the same message used different return paths and parallel
5833 remote delivery occurred, the wrong values would sometimes be logged.
5834 (Whenever a remote delivery process finished, the return path value from
5835 the most recently started remote delivery process was logged.)
5837 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5838 lines when AUTH and/or TLS are in use. This is the "received protocol"
5839 field. Exim used to use "asmtp" for authenticated SMTP, without any
5840 indication (in the protocol name) for TLS use. Now it follows the RFC and
5841 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5842 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5843 names appear in log lines as well as in Received: header lines.
5845 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5846 extensions, and comparison tests, and to fix some bugs.
5848 14. Changes to the "personal" filter test:
5850 (1) The test was buggy in that it was just doing the equivalent of
5851 "contains" tests on header lines. For example, if a user's address was
5852 anne@some.where, the "personal" test would incorrectly be true for
5854 To: susanne@some.where
5856 This test is now done by extracting each address from the header in turn,
5857 and checking the entire address. Other tests that are part of "personal"
5858 are now done using regular expressions (for example, to check local parts
5859 of addresses in From: header lines).
5861 (2) The list of non-personal local parts in From: addresses has been
5862 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5863 taken from the Sieve specification recommendations.
5865 (3) If the message contains any header line starting with "List-" it is
5866 treated as non-personal.
5868 (4) The test for "circular" in the Subject: header line has been removed
5869 because it now seems ill-conceived.
5871 15. Minor typos in src/EDITME comments corrected.
5873 16. Installed latest exipick from John Jetmore.
5875 17. If headers_add on a router specified a text string that was too long for
5876 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5877 of string_sprintf() is now avoided.
5879 18. $message_body_size was not set (it was always zero) when running the DATA
5880 ACL and the local_scan() function.
5882 19. For the "mail" command in an Exim filter, no default was being set for
5883 the once_repeat time, causing a random time value to be used if "once" was
5884 specified. (If the value happened to be <= 0, no repeat happened.) The
5885 default is now 0s, meaning "never repeat". The "vacation" command was OK
5886 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5887 (I found it when inspecting the code).
5889 20. There is now an overall timeout for performing a callout verification. It
5890 defaults to 4 times the callout timeout, which applies to individual SMTP
5891 commands during the callout. The overall timeout applies when there is more
5892 than one host that can be tried. The timeout is checked before trying the
5893 next host. This prevents very long delays if there are a large number of
5894 hosts and all are timing out (e.g. when the network connections are timing
5895 out). The value of the overall timeout can be changed by specifying an
5896 additional sub-option for "callout", called "maxwait". For example:
5898 verify = sender/callout=5s,maxwait=20s
5900 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5901 to the end before writing, but this should make it even safer).
5903 22. Exim was forgetting that it had advertised PIPELINING for the second and
5904 subsequent messages on an SMTP connection. It was also not resetting its
5905 memory on STARTTLS and an internal HELO.
5907 23. When Exim logs an SMTP synchronization error within a session, it now
5908 records whether PIPELINING has been advertised or not.
5910 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5911 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5912 rather than long int.
5914 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5916 26. Added the never_mail option to autoreply.
5922 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5923 crash if the getsockname() call failed; this can happen if a connection is
5924 closed very soon after it is established. The problem was simply in the
5925 order in which certain operations were done, causing Exim to try to write
5926 to the SMTP stream before it had set up the file descriptor. The bug has
5927 been fixed by making things happen in the correct order.
5933 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5934 before the connection was closed, thus losing the rejection response.
5936 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5937 some early Solaris releases, but causes trouble in current releases where
5938 socklen_t is defined.
5940 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5943 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5946 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5947 that releasing the top of it at the end releases what was used for sub-
5948 expansions (unless the block got too big). However, discard this block if
5949 the first thing is a variable or header, so that we can use its block when
5950 it is dynamic (useful for very large $message_headers, for example).
5952 6. Lookups now cache *every* query, not just the most recent. A new, separate
5953 store pool is used for this. It can be recovered when all lookup caches are
5954 flushed. Lookups now release memory at the end of their result strings.
5955 This has involved some general refactoring of the lookup sources.
5957 7. Some code has been added to the store_xxx() functions to reduce the amount
5958 of flapping under certain conditions.
5960 8. log_incoming_interface used to affect only the <= reception log lines. Now
5961 it causes the local interface and port to be added to several more SMTP log
5962 lines, for example "SMTP connection from", and rejection lines.
5964 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5966 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5968 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5969 could be overwritten at the end of the current message (or the start of a
5970 new message if it was set in a HELO ACL). The value is now preserved for
5971 the duration of the SMTP connection.
5973 12. If a transport had a headers_rewrite setting, and a matching header line
5974 contained an unqualified address, that address was qualified, even if it
5975 did not match any rewriting rules. The underlying bug was that the values
5976 of the flags that permit the existence of unqualified sender and recipient
5977 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5978 non-local messages, and by -bnq for local messages) were not being
5979 preserved with the message after it was received.
5981 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5982 "next input=" as part of the text comprising the host identity instead of
5983 the correct text. The code was using the same buffer for two different
5984 strings. However, depending on which order the printing function evaluated
5985 its arguments, the bug did not always show up. Under Linux, for example, my
5986 test suite worked just fine.
5988 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5989 doesn't work with some older versions of Perl. It has been changed to "my",
5990 which in any case is probably the better facility to use.
5992 15. A really picky compiler found some instances of statements for creating
5993 error messages that either had too many or two few arguments for the format
5996 16. The size of the buffer for calls to the DNS resolver has been increased
5997 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5998 for addresses that have a lot of PTR records. This alleviates a problem; it
5999 does not fully solve it.
6001 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6002 buffer now truncates the list and logs the incident, which is the same
6003 action as happens when Exim is looking up a host name and its aliases.
6004 Previously in this situation something unpredictable would happen;
6005 sometimes it was "internal error: store_reset failed".
6007 18. If a server dropped the connection unexpectedly when an Exim client was
6008 using GnuTLS and trying to read a response, the client delivery process
6009 crashed while trying to generate an error log message.
6011 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6012 string, for example:
6014 warn message = H1: something\nH2: something
6016 the text was added as a single header line from Exim's point of view
6017 though it ended up OK in the delivered message. However, searching for the
6018 second and subsequent header lines using $h_h2: did not work. This has been
6019 fixed. Similarly, if a system filter added multiple headers in this way,
6020 the routers could not see them.
6022 20. Expanded the error message when iplsearch is called with an invalid key to
6023 suggest using net-iplsearch in a host list.
6025 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6026 ACLs are no longer actually imposed (and a message to that effect is
6029 22. If a "gecos" field in a passwd entry contained escaped characters, in
6030 particular, if it contained a \" sequence, Exim got it wrong when building
6031 a From: or a Sender: header from that name. A second bug also caused
6032 incorrect handling when an unquoted " was present following a character
6033 that needed quoting.
6035 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6036 was not being matched caselessly.
6038 24. Arranged for all hyphens in the exim.8 source to be escaped with
6041 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6042 properly. Recipient callout cache records were still being keyed to include
6043 the sender, even when use_sender was set false. This led to far more
6044 callouts that were necessary. The sender is no longer included in the key
6045 when use_sender is false.
6047 26. Added "control = submission" modifier to ACLs.
6049 27. Added the ${base62d: operator to decode base 62 numbers.
6051 28. dnsdb lookups can now access SRV records.
6053 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6054 the configuration file.
6056 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6057 (-v) mode. This makes the output for a verbose queue run more intelligible.
6059 31. Added a use_postmaster feature to recipient callouts.
6061 32. Added the $body_zerocount variable, containing the number of binary zero
6062 bytes in the message body.
6064 33. The time of last modification of the "new" subdirectory is now used as the
6065 "mailbox time last read" when there is a quota error for a maildir
6068 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6070 35. Added +ignore_unknown as a special item in host lists.
6072 36. Code for decoding IPv6 addresses in host lists is now included, even if
6073 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6074 address was recognized as an IP address, but was then not correctly decoded
6075 into binary, causing unexpected and incorrect effects when compared with
6082 1. Very minor rewording of debugging text in manualroute to say "list of
6083 hosts" instead of "hostlist".
6085 2. If verify=header_syntax was set, and a header line with an unqualified
6086 address (no domain) and a large number of spaces between the end of the
6087 name and the colon was received, the reception process suffered a buffer
6088 overflow, and (when I tested it) crashed. This was caused by some obsolete
6089 code that should have been removed. The fix is to remove it!
6091 3. When running in the test harness, delay a bit after writing a bounce
6092 message to get a bit more predictability in the log output.
6094 4. Added a call to search_tidyup() just before forking a reception process. In
6095 theory, someone could use a lookup in the expansion of smtp_accept_max_
6096 per_host which, without the tidyup, could leave open a database connection.
6098 5. Added the variables $recipient_data and $sender_data which get set from a
6099 lookup success in an ACL "recipients" or "senders" condition, or a router
6100 "senders" option, similar to $domain_data and $local_part_data.
6102 6. Moved the writing of debug_print from before to after the "senders" test
6105 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6106 problems for message scanning, either using a data ACL, or using
6107 local_scan() because the Received: header was not generated till after they
6108 were called (in order to set the time as the time of reception completion).
6109 I have revised the way this works. The header is now generated after the
6110 body is received, but before the ACL or local_scan() are called. After they
6111 are run, the timestamp in the header is updated.
6117 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6118 before starting a queue runner without re-exec. This happened only when
6119 deliver_drop_privilege was set or when the Exim user was set to root. The
6120 effect of the bug was that timeouts during subsequent deliveries caused
6121 crashes instead of being properly handled. The handler is now left at its
6122 default (and expected) setting.
6124 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6125 message, again when deliver_drop_privilege is set or Exim is run as root.
6126 The bug described in (1) was not present in this case, but the tidying up
6127 of the other signals was missing. I have made the two cases consistent.
6129 3. The ignore_target_hosts setting on a manualroute router was being ignored
6130 for hosts that were looked up using the /MX notation.
6132 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6135 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6136 operated on the sender address. After changing the $sender_address to <>
6137 for the sender address verify, Exim was re-instated it as the original
6138 (before rewriting) address, but remembering that it had rewritten it, so it
6139 wasn't rewriting it again. This bug also had the effect of breaking the
6140 sender address verification caching when the sender address was rewritten.
6142 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6143 This has been changed so that if the ip literal address matches
6144 ignore_target_hosts, the router declines.
6146 7. Added expansion conditions match_domain, match_address, and match_local_
6147 part (NOT match_host).
6149 8. The placeholder for the Received: header didn't have a length field set.
6151 9. Added code to Exim itself and to exim_lock to test for a specific race
6152 condition that could lead to file corruption when using MBX delivery. The
6153 issue is with the lockfile that is created in /tmp. If this file is removed
6154 after a process has opened it but before that process has acquired a lock,
6155 there is the potential for a second process to recreate the file and also
6156 acquire a lock. This could lead to two Exim processes writing to the file
6157 at the same time. The added code performs the same test as UW imapd; it
6158 checks after acquiring the lock that its file descriptor still refers to
6159 the same named file.
6161 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6162 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6165 11. Added the smtp_active_hostname option. If used, this will typically be made
6166 to depend on the incoming interface address. Because $interface_address is
6167 not set up until the daemon has forked a reception process, error responses
6168 that can happen earlier (such as "too many connections") no longer contain
6171 12. If an expansion in a condition on a "warn" statement fails because a lookup
6172 defers, the "warn" statement is abandoned, and the next ACL statement is
6173 processed. Previously this caused the whole ACL to be aborted.
6175 13. Added the iplsearch lookup type.
6177 14. Added ident_timeout as a log selector.
6179 15. Added tls_certificate_verified as a log selector.
6181 16. Added a global option tls_require_ciphers (compare the smtp transport
6182 option of the same name). This controls incoming TLS connections.
6184 17. I finally figured out how to make tls_require_ciphers do a similar thing
6185 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6186 before starting the TLS session.
6188 18. Tabs are now shown as \t in -bP output.
6190 19. If the log selector return_path_on_delivery was set, Exim crashed when
6191 bouncing a message because it had too many Received: header lines.
6193 20. If two routers both had headers_remove settings, and the first one included
6194 a superfluous trailing colon, the final name in the first list and the
6195 first name in the second list were incorrectly joined into one item (with a
6196 colon in the middle).
6202 1. Added -C and -D options to the exinext utility, mainly to make it easier
6203 to include in the automated testing, but these could be helpful when
6204 multiple configurations are in use.
6206 2. The exinext utility was not formatting the output nicely when there was
6207 an alternate port involved in the retry record key, nor when there was a
6208 message id as well (for retries that were specific to a specific message
6209 and a specific host). It was also confused by IPv6 addresses, because of
6210 the additional colons they contain. I have fixed the IPv4 problem, and
6211 patched it up to do a reasonable job for IPv6.
6213 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6214 delivery, the log line now contains "pipelined" if PIPELINING was used.
6216 4. An SMTP transport process used to panic and die if the bind() call to set
6217 an explicit outgoing interface failed. This has been changed; it is now
6218 treated in the same way as a connect() failure.
6220 5. A reference to $sender_host_name in the part of a conditional expansion
6221 that was being skipped was still causing a DNS lookup. This no longer
6224 6. The def: expansion condition was not recognizing references to header lines
6225 that used bh_ and bheader_.
6227 7. Added the _cache feature to named lists.
6229 8. The code for checking quota_filecount in the appendfile transport was
6230 allowing one more file than it should have been.
6232 9. For compatibility with Sendmail, the command line option
6240 and sets the incoming protocol and host name (for trusted callers). The
6241 host name and its colon can be omitted when only the protocol is to be set.
6242 Note the Exim already has two private options, -pd and -ps, that refer to
6243 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6244 "s", but I don't think that's a major issue.
6246 10. A number of refactoring changes to the code, none of which should affect
6249 (a) The number of logging options was getting close to filling up the
6250 32-bit word that was used as a bit map. I have split them into two classes:
6251 those that are passed in the argument to log_write(), and those that are
6252 only ever tested independently outside of that function. These are now in
6253 separate 32-bit words, so there is plenty of room for expansion again.
6254 There is no change in the user interface or the logging behaviour.
6256 (b) When building, for example, log lines, the code previously used a
6257 macro that called string_cat() twice, in order to add two strings. This is
6258 not really sufficiently general. Furthermore, there was one instance where
6259 it was actually wrong because one of the argument was used twice, and in
6260 one call a function was used. (As it happened, calling the function twice
6261 did not affect the overall behaviour.) The macro has been replaced by a
6262 function that can join an arbitrary number of extra strings onto a growing
6265 (c) The code for expansion conditions now uses a table and a binary chop
6266 instead of a serial search (which was left over from when there were very
6267 few conditions). Also, it now recognizes conditions like "pam" even when
6268 the relevant support is not compiled in: a suitably worded error message is
6269 given if an attempt is made to use such a condition.
6271 11. Added ${time_interval:xxxxx}.
6273 12. A bug was causing one of the ddress fields not to be passed back correctly
6274 from remote delivery subprocesses. The field in question was not being
6275 subsequently used, so this caused to problems in practice.
6277 13. Added new log selectors queue_time and deliver_time.
6279 14. Might have fixed a bug in maildirsizefile handling that threw up
6280 "unexpected character" debug warnings, and recalculated the data
6281 unnecessarily. In any case, I expanded the warning message to give more
6284 15. Added the message "Restricted characters in address" to the statements in
6285 the default ACL that block characters like @ and % in local parts.
6287 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6288 Three changes have been made:
6290 (a) There was a serious bug; a negative response to MAIL caused the whole
6291 recipient domain to be cached as invalid, thereby blocking all messages
6292 to all local parts at the same domain, from all senders. This bug has
6293 been fixed. The domain is no longer cached after a negative response to
6294 MAIL if the sender used is not empty.
6296 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6299 (c) A new callout option, "use_sender" has been added for people who want
6300 the modified behaviour.
6306 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6309 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6310 indeed breaks things for older releases.
6312 3. Added additional logging to the case where there is a problem reading data
6313 from a filter that is running in a subprocess using a pipe, in order to
6314 try to track down a specific problem.
6316 4. Testing facility fudge: when running in the test harness and attempting
6317 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6318 getting "No route to host". Convert this to a timeout.
6320 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6323 6. Some OS don't have socklen_t but use size_t instead. This affects the
6324 fifth argument of getsockopt() amongst other things. This is now
6325 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6326 can be set for individual OS. I have set it for SunOS5, OSF1, and
6327 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6328 some earlier ones do not.
6330 7. Change 4.30/15 was not doing the test caselessly.
6332 8. The standard form for an IPv6 address literal was being rejected by address
6333 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6334 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6335 this, as well as the form without the "IPv6" on the front (but only when
6336 address literals are enabled, of course).
6338 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6340 10. Exim crashed if a message with an empty sender address specified by -f
6341 encountered a router with an errors_to setting. This could be provoked only
6342 by a command such as
6346 where an empty string was supplied; "<>" did not hit this bug.
6348 11. Installed PCRE release 4.5.
6350 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6351 remained set. It is now erased.
6353 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6354 times from message ids (which are base 36 rather than the normal 62).
6356 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6357 were being counted as actual protocol errors, and logged if the log
6358 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6359 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6360 and DATA following a set of rejected RCPTs do not count as protocol errors.
6361 In other words, Exim assumes they were pipelined, though this may not
6362 actually be the case. Of course, in all cases the client gets an
6363 appropriate error code.
6365 15. If a lookup fails in an ACL condition, a message about the failure may
6366 be available; it is used if testing the ACL cannot continue, because most
6367 such messages specify what the cause of the deferral is. However, some
6368 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6369 that caused an old message to be retained and used if a later statement
6370 caused a defer, replacing the real cause of the deferral.
6372 16. If an IP address had so many PTR records that the DNS lookup buffer
6373 was not large enough to hold them, Exim could crash while trying to process
6374 the truncated data. It now detects and logs this case.
6376 17. Further to 4.21/58, another change has been made: if (and only if) the
6377 first line of a message (the first header line) ends with CRLF, a bare LF
6378 in a subsequent header line has a space inserted after it, so as not to
6379 terminate the header.
6381 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6382 unnecessarily, used atoi() instead of strtol(), and didn't check the
6383 termination when getting file sizes from file names by regex.
6385 19. Completely re-implemented the support for maildirsize files, in the light
6386 of a number of problems with the previous contributed implementation
6387 (4.30/29). In particular:
6389 . If the quota is zero, the maildirsize file is maintained, but no quota is
6392 . If the maildir directory does not exist, it is created before any attempt
6393 to write a maildirsize file.
6395 . The quota value in the file is just a cache; if the quota is changed in
6396 the transport, the new value overrides.
6398 . A regular expression is available for excluding directories from the
6401 20. The autoreply transport checks the characters in options that define the
6402 message's headers; it allows continued headers, but it was checking with
6403 isspace() after an embedded newline instead of explicitly looking for a
6406 21. If all the "regular" hosts to which an address was routed had passed their
6407 expiry times, and had not reached their retry times, the address was
6408 bounced, even if fallback hosts were defined. Now Exim should go on to try
6411 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6412 equivalent code in the SMTP transport. Some hosts send humungous responses
6413 to HELO/EHLO, more than 1024 it seems.
6415 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6416 gives compiler warnings in some environments. I've now done it "properly",
6419 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6420 (because of problems with the built-in one) was declared to return uschar *
6421 instead of char *, causing compiler failure.
6423 25. Fixed a file descriptor leak when processing alias/forward files.
6425 26. Fixed a minor format string issue in dbfn.c.
6427 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6429 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6430 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6433 29. When a sender address is verified, it is cached, to save repeating the test
6434 when there is more than one recipient in a message. However, when the
6435 verification involves a callout, it is possible for different callout
6436 options to be set for different recipients. It is too complicated to keep
6437 track of this in the cache, so now Exim always runs a verification when a
6438 callout is required, relying on the callout cache for the optimization.
6439 The overhead is duplication of the address routing, but this should not be
6442 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6443 to be verified with callout=postmaster, and the main callout worked but the
6444 postmaster check failed, the verification correctly failed. However, if a
6445 subsequent RCPT command asked for sender verification *without* the
6446 postmaster check, incorrect caching caused this verification also to fail,
6449 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6450 it was not caching the DNS options (qualify_single, search_parents) that
6451 were used when the lookup failed. A subsequent lookup with different
6452 options therefore always gave the same answer, though there were cases
6453 where it should not have. (Example: a "domains = !$mx_any" option on a
6454 dnslookup router: the "domains" option is always processed without any
6455 widening, but the router might have qualify_single set.) Now Exim uses the
6456 cached value only when the same options are set.
6458 32. Added John Jetmore's "exipick" utility to the distribution.
6460 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6461 than a timeout (e.g. a certificate is required, and is not provided), an
6462 Exim server now closes the connection immediately. Previously it waited for
6463 the client to close - but if the client is SSL, it seems that they each
6464 wait for each other, leading to a delay before one of them times out.
6466 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6467 maintained 0.8.x compatibility because I don't think many are using it, and
6468 it is clearly obsolete.
6470 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6473 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6474 client certificate was expired. A simple patch fixes this, though I don't
6475 understand the full logic of why the verify callback is called multiple
6478 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6479 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6480 which causes problems with some clients (such as the Certicom SSL Plus
6481 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6482 disables the coutermeasure allowing Eudora to connect."
6484 38. Exim was not checking that a write() to a log file succeeded. This could
6485 lead to Bad Things if a log got too big, in particular if it hit a file
6486 size limit. Exim now panics and dies if it cannot write to a log file, just
6487 as it does if it cannot open a log file.
6489 39. Modified OS/Makefile-Linux so that it now contains
6491 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6493 The two -D definitions ensure that Exim is compiled with large file
6494 support, which makes it possible to handle log files that are bigger than
6497 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6498 instance) a domain was checked against a named list that involved a lookup,
6499 causing $domain_data to be set, then another domain was checked against the
6500 same list, then the first domain was re-checked, the value of $domain_data
6501 after the final check could be wrong. In particular, if the second check
6502 failed, it could be set empty. This bug probably also applied to
6505 41. The strip_trailing_dot option was not being applied to the address given
6506 with the -f command-line option.
6508 42. The code for reading a message's header from the spool was incrementing
6509 $received_count, but never initializing it. This meant that the value was
6510 incorrect (doubled) while delivering a message in the same process in which
6511 it was received. In the most common configuration of Exim, this never
6512 happens - a fresh exec is done - but it can happen when
6513 deliver_drop_privilege is set.
6515 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6516 it now includes up to 150 characters of the unexpected data in the log
6519 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6520 and building data strings. The size of both of these buffers was 10 000
6521 bytes - far larger than anybody would *ever* want, thought I. Needless to
6522 say, somebody hit the limit. I have increased the maximum line length to
6523 20 000 and the maximum data length of concatenated lines to 100 000. I have
6524 also fixed two bugs, because there was no checking on these buffers. Tsk,
6525 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6526 buffer is too small.
6528 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6529 lsearch lookups. Now it does.
6531 46. When parsing a route_list item in a manualroute router, a fixed-length
6532 buffer was used for the list of hosts. I made this 1024 bytes long,
6533 thinking that nobody would ever have a list of hosts that long. Wrong.
6534 Somebody had a whole pile of complicated expansion conditions, and the
6535 string was silently truncated, leading to an expansion error. It turns out
6536 that it is easier to change to an unlimited length (owing to other changes
6537 that have happened since this code was originally written) than to build
6538 structure for giving a limitation error. The length of the item that
6539 expands into the list of hosts is now unlimited.
6541 47. The lsearch lookup could not handle data where the length of text line was
6542 more than 4095 characters. Such lines were truncated, leading to shortened
6543 data being returned. It should now handle lines of any length.
6545 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6546 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6549 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6550 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6551 no longer happens in comment lines. A list of replacements is now placed
6552 at the head of all of the source files, except those whose only change is
6553 to replace PERL_COMMAND in the very first #! line.
6555 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6556 messages on the queue, with a bottom-up merge sort, using code contributed
6557 by Michael Haardt. This should make operations like -bp somewhat faster on
6558 large queues. It won't affect queue runners, except when queue_run_in_order
6561 51. Installed eximstats 1.31 in the distribution.
6563 52. Added support for SRV lookups to the dnslookup router.
6565 53. If an ACL referred to $message_body or $message_body_end, the value was not
6566 reset for any messages that followed in the same SMTP session.
6568 54. The store-handling optimization for building very long strings was not
6569 differentiating between the different store pools. I don't think this
6570 actually made any difference in practice, but I've tidied it.
6572 55. While running the routers to verify a sender address, $sender_address
6573 was still set to the sender address. This is wrong, because when routing to
6574 send a bounce to the sender, it would be empty. Therefore, I have changed
6575 it so that, while verifying a sender address, $sender_address is set to <>.
6576 (There is no change to what happens when verifying a recipient address.)
6578 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6579 target A or AAAA records (if not already returned) without resetting the
6580 qualify_single or search_parents options of the DNS resolver. These are
6581 inappropriate in this case because the targets of MX and SRV records must
6582 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6583 target that, when qualified, matched something in the local domain. These
6584 two options are now turned off when doing these lookups.
6586 57. It seems that at least some releases of Reiserfs (which does not have the
6587 concept of a fixed number of inodes) returns zero and not -1 for the
6588 number of available inodes. This interacted badly with check_spool_inodes,
6589 which assumed that -1 was the "no such thing" setting. What I have done is
6590 to check that the total number of inodes is greater than zero before doing
6591 the test of how many are available.
6593 58. When a "warn" ACL statement has a log_message modifier, the message is
6594 remembered, and not repeated. This is to avoid a lot of repetition when a
6595 message has many recipients that cause the same warning to be written.
6596 However, Exim was preserving the list of already written lines for an
6597 entire SMTP session, which doesn't seem right. The memory is now reset if a
6598 new message is started.
6600 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6601 debugging output unless log_rewrite was also set.
6603 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6604 of (int)(handle) when we know that handle contains (void *)(-1).
6606 61. The Exim daemon panic-logs an error return when it closes the incoming
6607 connection. However "connection reset by peer" seems to be common, and
6608 isn't really an error worthy of noting specially, so that particular error
6611 62. When Exim is trying to find all the local interfaces, it used to panic and
6612 die if the ioctl to get the interface flags failed. However, it seems that
6613 on at least one OS (Solaris 9) it is possible to have an interface that is
6614 included in the list of interfaces, but for which you get a failure error
6615 for this call. This happens when the interface is not "plumbed" into a
6616 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6617 failure of the "get flags" call assumes that the interface is down.
6619 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6620 makes life easier for people who are doing arithmetic on fields extracted
6621 from dates, where you often get leading zeros that should not be
6622 interpreted as octal.
6624 64. Added qualify_domain to the redirect router, to override the global
6627 65. If a pathologically long header line contained very many addresses (the
6628 report of this problem mentioned 10 000) and each of them was rewritten,
6629 Exim could use up a very large amount of memory. (It kept on making new
6630 copies of the header line as it rewrote, and never released the old ones.)
6631 At the expense of a bit more processing, the header rewriting function has
6632 been changed so that it no longer eats memory in this way.
6634 66. The generation of the Received: header has been moved from the time that a
6635 message starts to be received, to the time that it finishes. The timestamp
6636 in the Received: header should now be very close to that of the <= log
6637 line. There are two side-effects of this change:
6639 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6640 logged header lines no longer include the local Received: line, because
6641 it has not yet been created. The same applies to a copy of the message
6642 that is returned to a non-SMTP sender when a message is rejected.
6644 (b) When a filter file is tested using -bf, no additional Received: header
6645 is added to the test message. After some thought, I decided that this
6648 This change does not affect the value of $received_for. It is still set
6649 after address rewriting, but before local_scan() is called.
6651 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6653 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6654 gave an unhelpful panic error message, and a defer error. I have managed to
6655 change this behaviour so that it now rejects any supplied certificate,
6656 which seems right, as the list of acceptable certificates is empty.
6658 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6659 gave an unhelpful defer error. I have not managed to make this reject any
6660 supplied certificates, but the error message it gives is "no certificate
6661 supplied", which is not helpful.
6663 70. exigrep's output now also includes lines that are not associated with any
6664 message, but which match the given pattern. Implemented by a patch from
6665 Martin Sluka, which also tidied up the Perl a bit.
6667 71. Recipient callout verification, like sender verification, was using <> in
6668 the MAIL FROM command. This isn't really the right thing, since the actual
6669 sender may affect whether the remote host accepts the recipient or not. I
6670 have changed it to use the actual sender in the callout; this means that
6671 the cache record is now keyed on a recipient/sender pair, not just the
6672 recipient address. There doesn't seem to be a real danger of callout loops,
6673 since a callout by the remote host to check the sender would use <>.
6674 [SEE ABOVE: changed after hitting problems.]
6676 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6677 temporary errors. However, in the case of such a code being given after
6678 the end of a data transmission (i.e. after ".") Exim was failing to write
6679 a retry record for the message. (Yes, there was some broken host that was
6680 actually sending 8xx at this point.)
6682 73. An unknown lookup type in a host list could cause Exim to panic-die when
6683 the list was checked. (An example that provoked this was putting <; in the
6684 middle of a list instead of at the start.) If this happened during a DATA
6685 ACL check, a -D file could be left lying around. This kind of configuration
6686 error no longer causes Exim to die; instead it causes a defer error. The
6687 incident is still logged to the main and panic logs.
6689 74. Buglet left over from Exim 3 conversion. The message "too many messages
6690 in one connection" was written to the rejectlog but not the mainlog, except
6691 when address rewriting (yes!) was being logged.
6693 75. Added write_rejectlog option.
6695 76. When a system filter was run not as root (that is, when system_filter_user
6696 was set), the values of the $n variables were not being returned to the
6697 main process; thus, they were not subsequently available in the $sn
6700 77. Added +return_path_on_delivery log selector.
6702 78. A connection timeout was being treated differently from recipients deferred
6703 when testing hosts_max_try with a message that was older than the host's
6704 retry timeout. (The host should not be counted, thus allowing all hosts to
6705 be tried at least once before bouncing.) This may have been the cause of an
6706 occasionally reported bug whereby a message would remain on the queue
6707 longer than the retry timeout, but would be bounced if a delivery was
6708 forced. I say "may" because I never totally pinned down the problem;
6709 setting up timeout/retry tests is difficult. See also the next item.
6711 79. The ultimate address timeout was not being applied to errors that involved
6712 a combination of host plus message (for example, a timeout on a MAIL
6713 command). When an address resolved to a number of possible hosts, and they
6714 were not all tried for each delivery (e.g. because of hosts_max_try), a
6715 message could remain on the queue longer than the retry timeout.
6717 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6720 81. Fixed an obscure SMTP outgoing bug which required at least the following
6721 conditions: (a) there was another message waiting for the same server;
6722 (b) the server returned 5xx to all RCPT commands in the first message so
6723 that the message was not completed; (c) the server dropped the connection
6724 or gave a negative response to the RSET that Exim sends to abort the
6725 transaction. The observed case was a dropped connection after DATA that had
6726 been sent in pipelining mode. That is, the server had advertised PIPELINING
6727 but was not implementing it correctly. The effect of the bug was incorrect
6728 behaviour, such as trying another host, and this could lead to a crash.
6734 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6735 and daemon.c were passed as pointers to ints; they should have been
6736 pointers to socklen_t variables (which are typically unsigned ints).
6738 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6741 3. Fixed a really odd bug that affected only the testing scheme; patching a
6742 certain fixed string in the binary changed the value of another string that
6743 happened to be identical to the end of the original first string.
6745 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6746 name", it returns that address as the IP address. On some operating
6747 systems (e.g. Solaris), it also passes back the IP address string as the
6748 "host name". However, on others (e.g. Linux), it passes back an empty
6749 string. Exim wasn't checking for this, and was changing the host name to an
6750 empty string, assuming it had been canonicalized.
6752 5. Although rare, it is permitted to have more than one PTR record for a given
6753 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6754 all the names associated with an address, because they do in Solaris.
6755 However, it seems that they do not in Linux for data that comes from the
6756 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6757 I found this out when I moved to a new Linux workstation and tried to run
6758 the Exim test suite.
6760 To get round this problem I have changed the code so that it now does its
6761 own call to the DNS to look up PTR records when searching for a host name.
6762 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6763 addresses that are only in /etc/hosts are still found.
6765 This behaviour is, however, controlled by an option called host_lookup_
6766 order, which defaults to "bydns:byaddr". If people want to use the other
6767 order, or indeed, just use one or the other means of lookup, they can
6768 specify it in this variable.
6770 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6771 some operating systems, this comes back from gethostbyaddr() as an empty
6772 string, and this is what Exim used to test for. However, it seems that in
6773 other systems, "." is yielded. Exim now tests for this case too.
6775 7. The values of check_spool_space and check_log_space are now held internally
6776 as a number of kilobytes instead of an absolute number of bytes. If a
6777 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6778 kilobyte. This means that much larger values can be stored.
6780 8. Exim monitor: an attempt to get the action menu when not actually pointing
6781 at a message produces an empty menu entitled "No message selected". This
6782 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6783 no entries in it ("Shell widget menu has zero width and/or height"). So I
6784 have added a single, blank menu entry in this case.
6786 9. Added ${quote_local_part.
6788 10. MIME decoding is now applied to the contents of Subject: header lines when
6791 11. Now that a reference to $sender_host_address automatically causes a reverse
6792 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6793 host lookup before query-style lookups in lists that might use this
6794 variable. This has therefore been abolished, and the "net-" prefix is no
6795 longer necessary for query-style lookups.
6797 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6798 have been a typo for LFLAGS, so it has been changed.
6800 13. The install script calls Exim with "-C /dev/null" in order to find the
6801 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6802 to be output. However, since Exim outputs its version number before the
6803 error, it didn't break the script. It just looked ugly. I fixed this by
6804 always allowing "-C /dev/null" if the caller is root.
6806 14. Ignore overlarge ACL variable number when reading spool file - insurance
6807 against a later release with more variables having written the file.
6809 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6810 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6811 this, as well as the form without the "IPv6" on the front.
6813 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6814 OS/Makefile-Darwin file.
6816 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6817 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6819 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6820 variable. However, some OS use ints for inodes. I've added cast to long int
6821 to get rid of the compiler warning.
6823 19. I had forgotten to lock out "/../" in configuration file names when
6824 ALT_CONFIG_PREFIX was set.
6826 20. Routers used for verification do not need to specify transports. However,
6827 if such a router generated a host list, and callout was configured, Exim
6828 crashed, because it could not find a port number from the (non-existent)
6829 transport. It now assumes port 25 in this circumstance.
6831 21. Added the -t option to exigrep.
6833 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6834 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6835 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6836 them is set, LOOKUP_LSEARCH is forced.
6838 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6840 24. Added sender and host information to the "rejected by local_scan()" log
6841 line; previously there was no indication of these.
6843 25. Added .include_if_exists.
6845 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6846 Linux. It turns out that not all file systems support this. Apparently some
6847 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6848 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6849 not supported on the file descriptor, is now ignored when Exim is trying to
6850 sync a directory. This applies only to Linux.
6852 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6854 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6855 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6858 29. Added support for maildirsize files from supplied patch (modified a bit).
6860 30. The use of :fail: followed by an empty string could lead Exim to respond to
6861 sender verification failures with (e.g.):
6863 550 Verification failed for <xxx>
6864 550 Sender verify failed
6866 where the first response line was missing the '-' that indicates it is not
6867 the final line of the response.
6869 31. The loop for finding the name of the user that called Exim had a hardwired
6870 limit of 10; it now uses the value of finduser_retries, which is used for
6871 all other user lookups.
6873 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6876 33. Exim was neglecting to zero errno before one call of strtol() when
6877 expanding a string and expecting an integer value. On some systems this
6878 resulted in spurious "integer overflow" errors. Also, it was casting the
6879 result into an int without checking.
6881 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6882 did not work. The code looks as if it has *never* worked, though it appears
6883 to have been documented since at least release 1.62. I have made it work.
6885 35. The "timeout_DNS" error in retry rules, also documented since at least
6886 1.62, also never worked. As it isn't clear exactly what this means, and
6887 clearly it isn't a major issue, I have abolished the feature by treating it
6888 as "timeout", and writing a warning to the main and panic logs.
6890 36. The display of retry rules for -brt wasn't always showing the error code
6893 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6894 timeout_connect_A, timeout_connect_MX.
6896 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6897 to the empty sender.
6899 39. The daemon was not analysing the content of -oX till after it had closed
6900 stderr and disconnected from the controlling terminal. This meant that any
6901 syntax errors were only noted on the panic log, and the return code from
6902 the command was 0. By re-arranging the code a little, I've made the
6903 decoding happen first, so such errors now appear on stderr, and the return
6904 code is 1. However, the actual setting up of the sockets still happens in
6905 the disconnected process, so errors there are still only recorded on the
6908 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6909 connections (as happens on some IP stacks) was logged at start up time as
6910 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6911 it from "IPv6 and IPv4", which means that two separate sockets are being
6914 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6915 says whether AF_INET or AF_INET6 was passed as an argument.
6917 42. Exiwhat output was messed up when time zones were included in log
6920 43. Exiwhat now gives more information about the daemon's listening ports,
6921 and whether -tls-on-connect was used.
6923 44. The "port" option of the smtp transport is now expanded.
6925 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6926 silently ignored. Now an error message is written to the main and panic
6929 46. There's a new ACL modifier called "logwrite" which writes to a log file
6930 as soon as it is encountered.
6932 47. Added $local_user_uid and $local_user_gid at routing time.
6934 48. Exim crashed when trying to verify a sender address that was being
6937 49. Exim was recognizing only a space character after ".include". It now also
6938 recognizes a tab character.
6940 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6941 extracting the relevant information from the specification. The man page no
6942 longer contains scrambled data for the -d option, and I've added a section
6943 at the front about calling Exim under different names.
6945 51. Added "extra_headers" argument to the "mail" command in filter files.
6947 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6950 53. Installed eximstats 1.29.
6952 54. Added transport_filter_timeout as a generic transport option.
6954 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6955 Cc: header lines. This was required by RFC 822, but it not required by RFC
6958 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6959 incoming messages that did not have them. Now it does so only if the
6960 message originates locally, that is, if there is no associated remote host
6961 address. When Resent- header lines are present, this applies to the Resent-
6962 lines rather than the non-Resent- lines.
6964 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6965 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6967 58. Messages for configuration errors now include the name of the main
6968 configuration file - useful now that there may be more than one file in a
6969 list (.included file names were always shown).
6971 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6972 for those rare installations that do not start the daemon as root or run it
6973 setuid root. I've cut out the call to initgroups() if the daemon is not
6976 60. The Exim user and group can now be bound into the binary as text strings
6977 that are looked up at the start of Exim's processing.
6979 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6981 62. Added $mailstore_basename variable.
6983 63. Installed patch to sieve.c from Michael Haardt.
6985 64. When Exim failed to open the panic log after failing to open the main log,
6986 the original message it was trying to log was written to stderr and debug
6987 output, but if they were not available (the usual case in production), it
6988 was lost. Now it is written to syslog before the two lines that record the
6989 failures to open the logs.
6991 65. Users' Exim filters run in subprocesses under the user's uid. It is
6992 possible for a "deliver" command or an alias in a "personal" command to
6993 provoke an address rewrite. If logging of address rewriting is configured,
6994 this fails because the process is not running as root or exim. There may be
6995 a better way of dealing with this, but for the moment (because 4.30 needs
6996 to be released), I have disabled address rewrite logging when running a
6997 filter in a non-root, non-exim process.
7003 1. The buildconfig auxiliary program wasn't quoting the value set for
7004 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7005 not defined. This bug was masked in 4.22 by the effect that was fixed in
7008 2. Some messages that were rejected after a message id was allocated were
7009 shown as "incomplete" by exigrep. It no longer does this for messages that
7010 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7012 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7013 have allow_domain_literals set, the ID did not get logged in the <= line.
7014 Domain literals are now always recognized in Message-ID: header lines.
7016 4. The first argument for a ${extract expansion item is the key name or field
7017 number. Leading and trailing spaces in this item were not being ignored,
7018 causing some misleading effects.
7020 5. When deliver_drop_privilege was set, single queue runner processes started
7021 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7022 same command in the process it spins off) were not dropping privilege.
7024 6. When the daemon running as "exim" started a queue runner, it always
7025 re-executed Exim in the spun-off process. This is a waste of effort when
7026 deliver_drop_privilege is set. The new process now just calls the
7027 queue-runner function directly.
7033 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7036 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7037 ignored. Though the use of -oP was forcing the writing of a pid file, it
7038 was always written to the default place.
7040 3. If the message "no IP address found for host xxxx" is generated during
7041 incoming verification, it is now followed by identification of the incoming
7042 connection (so you can more easily find what provoked it).
7044 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7046 5. Added some features to "harden" Exim a bit more against certain attacks:
7048 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7049 be put in Local/Makefile. This is like the never_users runtime option,
7050 but it cannot be overridden. The default setting is "root".
7052 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7053 prefix string with which any file named in a -C command line option
7056 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7057 is retained for -C and -D only if the caller of Exim is root. Without
7058 it, the exim user may also use -C and -D and retain privilege.
7060 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7061 command line option is disabled.
7063 6. Macro names set by the -D option must start with an upper case letter, just
7064 like macro names defined in the configuration file.
7066 7. Added "dereference=" facility to LDAP.
7068 8. Two instances of the typo "uknown" in the source files are fixed.
7070 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7071 the Configure-Makefile script screwed up while processing it.
7073 10. Incorporated PCRE 4.4.
7075 11. The SMTP synchronization check was not operating right at the start of an
7076 SMTP session. For example, it could not catch a HELO sent before the client
7077 waited for the greeting. There is now a check for outstanding input at the
7078 point when the greeting is written. Because of the duplex, asynchronous
7079 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7080 way, but not yet received, when the check is performed.
7082 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7083 on TCP/IP sockets, because this apparently causes some broken clients to
7086 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7087 unchanged) from the Cygwin maintainer.
7089 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7090 was supported instead of testing for mbx. Effectively a typo.
7092 15. The spa authenticator server code was not checking that the input it
7093 received was valid base64.
7095 16. The debug output line for the "set" modifier in ACLs was not showing the
7096 name of the variable that was being set.
7098 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7100 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7101 lookup on its own. Something else was needed to trigger the lookup. For
7102 example, a match in host_lookup or the need for a host name in a host list.
7103 Now, if $sender_host_name is referenced and the host name has not yet been
7104 looked up, a lookup is performed. If the lookup fails, the variable remains
7105 empty, and $host_lookup_failed is set to "1".
7107 19. Added "eqi" as a case-independent comparison operator.
7109 20. The saslauthd authentication condition could segfault if neither service
7110 nor realm was specified.
7112 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7113 error message that was logged was misleading, and incoming SMTP
7114 connections were dropped. The message is now more accurate, and temporary
7115 errors are given to SMTP connections.
7117 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7118 (except RSET) with a 421 error, until QUIT is received. However, it was
7119 failing to send a response to QUIT.
7121 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7122 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7123 ACL is now run after the helo_try_verify_hosts code.
7125 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7126 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7127 case-independent, so other case variants are also recognized). Apparently
7128 some systems use these upper case variants.
7130 25. If more than two messages were waiting for the same host, and a transport
7131 filter was specified for the transport, Exim sent two messages over the
7132 same TCP/IP connection, and then failed with "socket operation on non-
7133 socket" when it tried to send the third.
7135 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7137 27. The extern definition of crypt16() in expand.c was not being excluded when
7138 the OS had its own crypt16() function.
7140 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7141 as a preferred synonym for return_size_limit, both as an option and as an
7144 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7146 30. Changed the default configuration ACL to relax the local part checking rule
7147 for addresses that are not in any local domains. For these addresses,
7148 slashes and pipe symbols are allowed within local parts, but the sequence
7149 /../ is explicitly forbidden.
7151 31. SPA server authentication was not clearing the challenge buffer before
7154 32. log_message in a "warn" ACL statement was writing to the reject log as
7155 well as to the main log, which contradicts the documentation and doesn't
7156 seem right (because no rejection is happening). So I have stopped it.
7158 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7159 However, I am unable to do any testing of this.
7161 34. Fixed an infelicity in the appendfile transport. When checking directories
7162 for a mailbox, to see if any needed to be created, it was accidentally
7163 using path names with one or more superfluous leading slashes; tracing
7164 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7166 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7167 contents are added to the log line that is written for every discarded
7168 recipient. (Previously a log_message setting was ignored.)
7170 36. The ${quote: operator now quotes the string if it is empty.
7172 37. The install script runs exim in order to find its version number. If for
7173 some reason other than non-existence or emptiness, which it checks, it
7174 could not run './exim', it was installing it with an empty version number,
7175 i.e. as "exim-". This error state is now caught, and the installation is
7178 38. An argument was missing from the function that creates an error message
7179 when Exim fails to connect to the socket for saslauthd authentication.
7180 This could cause Exim to crash, or give a corrupted message.
7182 39. Added isip, isip4, and isip6 to ${if conditions.
7184 40. The ACL variables $acl_xx are now saved with the message, and can be
7185 accessed later in routers, transports, and filters.
7187 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7188 strings in the file are not string-expanded.
7190 42. If a MAIL command specified a SIZE value that was too large to fit into an
7191 int variable, the check against message_size_limit failed. Such values are
7192 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7193 day this will have to be increased, but I don't think I want to be around
7194 when emails are that large.
7201 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7202 iconv() is not standard in FreeBSD.
7204 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7205 IPv6 enabled. The observed symptom was a segmentation fault on return from
7206 the function os_common_find_running_interfaces() in src/os.c.
7208 3. In the check_special_case() function in daemon.c I had used "errno" as an
7209 argument name, which causes warnings on some systems. This was basically a
7210 typo, since it was named "eno" in the comments!
7212 4. The code that waits for the clock to tick (at a resolution of some fraction
7213 of a second) so as to ensure message-id uniqueness was always waiting for
7214 at least one whole tick, when it could have waited for less. [This is
7215 almost certainly not relevant at current processor speeds, where it is
7216 unlikely to ever wait at all. But we try to future-proof.]
7218 5. The function that sleeps for a time interval that includes fractions of a
7219 second contained a race. It did not block SIGALRM between setting the
7220 timer, and suspending (a couple of lines later). If the interval was short
7221 and the sigsuspend() was delayed until after it had expired, the suspension
7222 never ended. On busy systems this could lead to processes getting stuck for
7225 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7226 process, before it forks any delivery processes. The open lookup caching
7227 mechanism meant that the open file or database connection was passed into
7228 the delivery process. The problem was that delivery processes always tidy
7229 up cached lookup data. This could cause a problem for the next delivery
7230 process started by the queue runner, because the external queue runner
7231 process does not know about the closure. So the next delivery process
7232 still has data in the lookup cache. In the case of a file lookup, there was
7233 no problem because closing a file descriptor in a subprocess doesn't affect
7234 the parent. However, if the lookup was caching a connection to a database,
7235 the connection was closed, and the second delivery process was likely to
7236 see errors such as "PGSQL: query failed: server closed the connection
7237 unexpectedly". The problem has been fixed by closing all cached lookups
7238 in a queue runner before running a delivery process.
7240 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7241 seem to have the "const" qualifier which it has on other OS. I've
7244 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7245 *of the same type* (client or server) with the same public name that an
7246 error should be diagnosed.
7248 9. When Exim looked up a host name for an IP address, but failed to find the
7249 original IP address when looking up the host name (a safety check), it
7250 output the message "<ip address> does not match any IP for NULL", which was
7251 confusing, to say the least. The bug was that the host name should have
7252 appeared instead of "NULL".
7254 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7255 user that is built into the binary, and the -C or -D options is used, root
7256 privilege is dropped before the configuration file is read. In addition,
7257 logging is switched to stderr instead of the normal log files. If the
7258 configuration then re-defines the Exim user, the unprivileged environment
7259 is probably not what is expected, so Exim logs a panic warning message (but
7262 However, if deliver_drop_privilege is set, the unprivileged state may well
7263 be exactly what is intended, so the warning has been cut out in that case,
7264 and Exim is allowed to try to write to its normal log files.
7270 1. smtp_return_error_details was not giving details for temporary sender
7271 or receiver verification errors.
7273 2. Diagnose a configuration error if two authenticators have the same public
7276 3. Exim used not to create the message log file for a message until the first
7277 delivery attempt. This could be confusing when incoming messages were held
7278 for policy or load reasons. The message log file is now created at the time
7279 the message is received, and an initial "Received" line is written to it.
7281 4. The automatically generated man page for command line options had a minor
7282 bug that caused no ill effects; however, a more serious problem was that
7283 the procedure for building the man page automatically didn't always
7284 operate. Consequently, release 4.20 contains an out-of-date version. This
7285 shouldn't happen again.
7287 5. When building Exim with embedded Perl support, the script that builds the
7288 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7289 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7291 6. The freeze_tell option was not being used for messages that were frozen on
7292 arrival, either by an ACL or by local_scan().
7294 7. Added the smtp_incomplete_transaction log selector.
7296 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7297 was accepting AUTH without a new EHLO.
7299 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7300 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7303 10. Logging of TCP/IP connections (when configured) now happens in the main
7304 daemon process instead of the child process, so that the TCP/IP connection
7305 count is more accurate (but it can never be perfect).
7307 11. The use of "drop" in a nested ACL was not being handled correctly in the
7308 outer ACL. Now, if condition failure induced by the nested "drop" causes
7309 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7310 or "require"), the connection is dropped.
7312 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7313 that yield "discard" can now be used with an "accept" or a "discard" verb,
7314 but an error is generated for any others (because I can't see a useful way
7315 to define what should happen).
7317 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7318 are now processed in the same way as lines in the Exim configuration file.
7319 In particular, continuation lines are supported.
7321 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7323 15. Added -ti meaning -t -i.
7325 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7326 domains, and warn by logging if others are found.
7328 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7329 returned from GIFCONF when Exim is trying to find the list of interfaces on
7330 a host. The code in os.c has been modified to copy each ifreq to an aligned
7331 structure in all cases.
7333 Also, in some cases, the returned ifreq's were being copied to a 'struct
7334 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7335 means the last couple of bytes of an IPv6 address could be chopped if the
7336 ifreq contained only a normal sockaddr (14 bytes storage).
7338 18. Named domain lists were not supported in the hosts_treat_as_local option.
7339 An entry such as +xxxx was not recognized, and was treated as a literal
7342 19. Ensure that header lines added by a DATA ACL are included in the reject log
7343 if the ACL subsequently rejects the message.
7345 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7346 MD5 (which is deprecated).
7348 21. When testing a filter file using -bf, Exim was writing a message when it
7349 took the sender from a "From " line in the message, but it was not doing so
7350 when it took $return_path from a Return-Path: header line. It now does.
7352 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7353 with a valid header line field name (a series of printing characters
7354 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7356 23. Changed "disc" in the source to "disk" to conform to the documentation and
7357 the book and for uniformity.
7359 24. Ignore Sendmail's -Ooption=value command line item.
7361 25. When execve() failed while trying to run a command in a pipe transport,
7362 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7363 could be confused with a return value of 69 from the command itself. This
7364 has been changed to 127, the value the shell returns if it is asked to run
7365 a non-existent command. The wording for the related log line suggests a
7366 non-existent command as the problem.
7368 26. If received_header_text expands to an empty string, do not add a Received:
7369 header line to the message. (Well, it adds a token one on the spool, but
7370 marks it "old" so that it doesn't get used or transmitted.)
7372 27. Installed eximstats 1.28 (addition of -nt option).
7374 28. There was no check for failure on the call to getsockname() in the daemon
7375 code. This can fail if there is a shortage of resources on the system, with
7376 ENOMEM, for example. A temporary error is now given on failure.
7378 29. Contrary to the C standard, it seems that in some environments, the
7379 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7380 program. Exim now does this explicitly; it affects the formatting of
7381 timestamps using strftime().
7383 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7384 complaints. I've now initialized all the variables, to avoid this.
7386 32. Header lines added by a system filter were not being "seen" during
7387 transport-time rewrites.
7389 33. The info_callback() function passed to OpenSSL is set up with type void
7390 (*)(SSL *, int, int), as described somewhere. However, when calling the
7391 function (actually a macro) that sets it up, the type void(*)() is
7392 expected. I've put in a cast to prevent warnings from picky compilers.
7394 34. If a DNS black list lookup found a CNAME record, but there were no A
7395 records associated with the domain it pointed at, Exim crashed.
7397 35. If a DNS black list lookup returned more than one A record, Exim ignored
7398 all but the first. It now scans all returned addresses if a particular IP
7399 value is being sought. In this situation, the contents of the
7400 $dnslist_value variable are a list of all the addresses, separated by a
7403 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7404 to accept a host name and all its aliases if the forward lookup for any of
7405 them yielded the IP address of the incoming connection. Now it accepts only
7406 those names whose forward lookup yields the correct IP address. Any other
7407 names are discarded. This closes a loophole whereby a rogue DNS
7408 administrator could create reverse DNS records to break through a
7409 wildcarded host restriction in an ACL.
7411 37. If a user filter or a system filter that ran in a subprocess used any of
7412 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7413 the wrong values were passed to the pipe command ($thisaddress had the
7414 value of $0, $0 had the value of $1, etc). This bug was introduced by
7415 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7417 38. Improved the line breaking for long SMTP error messages from ACLs.
7418 Previously, if there was no break point between 40 and 75 characters, Exim
7419 left the rest of the message alone. Two changes have been made: (a) I've
7420 reduced the minimum length to 35 characters; (b) if it can't find a break
7421 point between 35 and 75 characters, it looks ahead and uses the first one
7422 that it finds. This may give the occasional overlong line, but at least the
7423 remaining text gets split now.
7425 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7426 file descriptors might be low, and that setting 1000 would always raise it.
7427 It turns out that in some environments, the limit is already over 1000 and
7428 that lowering it causes trouble. So now Exim takes care not to decrease it.
7430 40. When delivering a message, the value of $return_path is set to $sender_
7431 address at the start of routing (routers may change the value). By an
7432 oversight, this default was not being set up when an address was tested by
7433 -bt or -bv, which affected the outcome if any router or filter referred to
7436 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7437 log only when "message" or "log_message" are set. However, if one of the
7438 conditions was an address verification, or a call to a nested ACL, the
7439 messages generated by the underlying test were being passed through. This
7440 no longer happens. The underlying message is available in $acl_verify_
7441 message for both "message" and "log_message" expansions, so it can be
7442 passed through if needed.
7444 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7445 new expansion $bh_ to give the encoded byte string without charset
7446 translation. Translation happens only if iconv() is available; HAVE_ICONV
7447 indicates this at build time. HEADERS_CHARSET gives the charset to
7448 translate to; headers_charset can change it in the configuration, and
7449 "headers charset" can change it in an individual filter file.
7451 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7452 that creates RFC 2047 encoded "words" labels them as that charset instead
7453 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7454 expansion operator; (ii) when Exim creates a From: line for a local
7455 message; (iii) when a header line is rewritten to include a "phrase" part.
7457 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7458 buggy, causing it to skip the first lines of messages whose message ID
7459 ended in 'D'. This would not have bitten before Exim release 4.14, because
7460 message IDs were unlikely to end in 'D' before then. The effect was to have
7461 incorrect size information for certain domains.
7463 45. #include "config.h" was missing at the start of the crypt16.c module. This
7464 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7467 46. If there was a timeout during a "random" callout check, Exim treated it as
7468 a failure of the random address, and carried on sending RSET and the real
7469 address. If the delay was just some slowness somewhere, the response to the
7470 original RCPT would be taken as a response to RSET and so on, causing
7471 mayhem of various kinds.
7473 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7474 when I implemented it. It didn't allow for the fact that some option values
7475 may legitimately be negative (e.g. size_addition), and it didn't even do
7476 the right test for positive values.
7478 48. Domain names in DNS records are case-independent. Exim always looks them up
7479 in lower case. Some resolvers return domain names in exactly the case they
7480 appear in the zone file, that is, they may contain uppercase letters. Not
7481 all resolvers do this - some return always lower case. Exim was treating a
7482 change of case by a resolver as a change of domain, similar to a widening
7483 of a domain abbreviation. This triggered its re-routing code and so it was
7484 trying to route what was effectively the same domain again. This normally
7485 caused routing to fail (because the router wouldn't handle the domain
7486 twice). Now Exim checks for this case specially, and just changes the
7487 casing of the domain that it ultimately uses when it transmits the message
7490 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7493 50. If a filter generated a file delivery with a non-absolute name (possible if
7494 no home directory exists for the router), the forbid_file option was not
7497 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7498 the existing equality matching.
7500 52. Exim was using ints instead of ino_t variables in some places where it was
7501 dealing with inode numbers.
7503 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7504 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7505 called TMPDIR, and if it finds it is different, it changes its value.
7507 54. The smtp_printf() function is now made available to local_scan() so
7508 additional output lines can be written before returning. There is also an
7509 smtp_fflush() function to enable the detection of a dropped connection.
7510 The variables smtp_input and smtp_batched_input are exported to
7513 55. Changed the default runtime configuration: the message "Unknown user"
7514 has been removed from the ACL, and instead placed on the localuser router,
7515 using the cannot_route_message feature. This means that any verification
7516 failures that generate their own messages won't get overridden. Similarly,
7517 the "Unrouteable address" message that was in the ACL for unverifiable
7518 relay addresses has also been removed.
7520 56. Added hosts_avoid_esmtp to the smtp transport.
7522 57. The exicyclog script was not checking for the esoteric option
7523 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7524 will work only if exicyclog is run under the appropriate euid.
7526 58. Following a discussion on the list, the rules by which Exim recognises line
7527 endings on incoming messages have been changed. The -dropcr and drop_cr
7528 options are now no-ops, retained only for backwards compatibility. The
7529 following line terminators are recognized: LF CRLF CR. However, special
7530 processing applies to CR:
7532 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7533 nor a local message in the state where . is a terminator.
7535 (ii) If a bare CR is encountered in a header line, an extra space is added
7536 after the line terminator so as not to end the header. The reasoning
7537 behind this is that bare CRs in header lines are most likely either
7538 to be mistakes, or people trying to play silly games.
7540 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7541 was being incorrectly given as 18 bytes larger than it should have been.
7542 This is a VOB (very old bug).
7544 60. This may never have affected anything current, but just in case it has:
7545 When the local host is found other than at the start of a list of hosts,
7546 the local host, those with the same MX, and any that follow, are discarded.
7547 When the list in question was part of a longer list of hosts, the following
7548 hosts (not currently being processed) were also being discarded. This no
7549 longer happens. I'm not sure if this situation could ever has previously
7552 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7555 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7556 header. This is something that is recommended in a new Internet Draft, and
7557 is something that is documented as being done by Sendmail. There are two
7558 possible values. For messages generated by the autoreply transport, Exim
7561 Auto-Submitted: auto-replied
7563 whereas for all other generated messages (e.g. bounces) it adds
7565 Auto-Submitted: auto-generated
7567 63. The "personal" condition in filters now includes a test for the
7568 Auto-Submitted: header. If it contains the string "auto-" the message it
7569 not considered personal.
7571 64. Added rcpt_include_affixes as a generic transport option.
7573 65. Added queue_only_override (default true).
7575 66. Added the syslog_duplication option.
7577 67. If what should have been the first header line of a message consisted of
7578 a space followed by a colon, Exim was mis-interpreting it as a header line.
7579 It isn't of course - it is syntactically invalid and should therefore be
7580 treated as the start of the message body. The misbehaviour could have
7581 caused a number of strange effects, including loss of data in subsequent
7582 header lines, and spool format errors.
7584 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7585 client host had authenticated. This control can now be exercised by an ACL
7586 for more flexibility.
7588 69. By default, callouts do not happen when testing with -bh. There is now a
7589 variant, -bhc, which does actually run the callout code, including
7590 consulting and updating the callout cache.
7592 70. Added support for saslauthd authentication, courtesy of Alexander
7595 71. If statvfs() failed on the spool or log directories while checking their
7596 size for availability, Exim confusingly gave the error "space shortage".
7597 Furthermore, in debugging mode it crashed with a floating point exception.
7598 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7599 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7600 serious problem, Exim now writes to the main and panic logs when this
7601 happens, with details of the failure. It then refuses to accept the
7602 incoming message, giving the message "spool directory problem" or "log
7603 directory problem" with a 421 code for SMTP messages.
7605 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7606 0, 1, and 2 exist, because some OS complain for execs without them (see
7607 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7608 descriptors. However, the code omitted to check that the open succeeded,
7609 causing mysterious errors if for some reason the permissions on /dev/null
7610 got screwed. Now Exim writes a message to the main and panic logs, and
7611 bombs out if it can't open /dev/null.
7613 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7614 interact so that it is all more flexible. It is supposed to remain
7615 backwards compatible. Also added extra_local_interfaces.
7617 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7618 to bomb out with an assertion failure - to the client this appears as a
7619 connection drop. This problem occurs in the part of the code that was taken
7620 from the Samba project. Fortunately, the assertion is in a very simple
7621 function, so I have fixed this by reproducing the function inline in the
7622 one place where it is called, and arranging for authentication to fail
7623 instead of killing the process with assert().
7625 75. The SPA client code was not working when the server requested OEM rather
7626 than Unicode encoding.
7628 76. Added code to make require_files with a specific uid setting more usable in
7629 the case where statting the file as root fails - usually a non-root-mounted
7630 NFS file system. When this happens and the failure is EACCES, Exim now
7631 forks a subprocess and does the per-uid checking as the relevant uid.
7633 77. Added process_log_path.
7635 78. If log_file_path was not explicitly set, a setting of check_log_space or
7636 check_log_inodes was ignored.
7638 79. If a space check for the spool or log partitions fails, the incident is now
7639 logged. Of course, in the latter case the data may get lost...
7641 80. Added the %p formatting code to string_format() so that it can be used to
7642 print addresses in debug_print(). Adjusted all the address printing in the
7643 debugging in store.c to use %p rather than %d.
7645 81. There was a concern that a line of code in smtp_in.c could overflow a
7646 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7647 initially expressed, the concern was not well-founded, because trailing
7648 spaces are removed early. However, if the trailing spaces were followed by
7649 a NULL, they did not get removed, so the overflow was possible. Two fixes
7652 (a) I re-wrote the offending code in a cleaner fashion.
7653 (b) If an incoming SMTP command contains a NULL character, it is rejected
7656 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7657 runs initgroups(), so that if the Exim user is in any additional groups,
7658 they will be used during message reception.
7664 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob