1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
77 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
78 for the expansion of the server_scram_iter and server_scram_salt options.
80 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
81 are now specifically given a NO_DATA response without hitting the system
82 resolver. The library goes on to do the now-standard TXT lookup.
83 Use of dnsdb lookups is not affected.
85 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
86 only retrieve the errormessage once. Previously two calls to dlerror()
87 were used, and the second one (for mainlog/paniclog) retrieved null
90 JH/20 Taint checking: disallow use of tainted data for
91 - the appendfile transport file and directory options
92 - the pipe transport command
93 - the autoreply transport file, log and once options
94 - file names used by the redirect router (including filter files)
96 Previously this was permitted.
98 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
99 adjusted the size of a major service buffer; this failed because the
100 buffer was in use at the time. Change to a compile-time increase in the
101 buffer size, when this authenticator is compiled into exim.
103 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
104 previous fast-mode was untenable in the face of glibs using mmap to
105 support larger malloc requests.
107 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
108 New values supported, if defined on system where compiled:
109 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
110 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
112 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
113 running a limited number of proceses in parallel, a benefit is gained. The
114 amount varies with the platform hardware and load. The use of the option
115 queue_run_in_order means we cannot do this, as ordering becomes
118 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
119 had introduced a string-copy (for ensuring NUL-termination) which was not
120 appropriate for that case, which can include embedded NUL bytes in the
121 block of data. Investigation showed the copy to actually be needless, the
122 data being length-specified.
124 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
125 done during a receiving connection, and both used TLS, global info was
126 used rather than per-connection info for tracking the state of data
127 queued for transmission. This could result in a connection hang.
129 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
130 Previously, when delivering serveral messages down a single connection
131 only the first would provide a SIZE. This was due to the size information
132 not being properly tracked.
134 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
135 TAI (at 37 seconds currently), pretend to be in UTC for time-related
136 expansion and logging. Previously, spurious values such as a future
137 minute could be seen.
139 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
140 it could crash from a null-deref. This could also affect the
141 ${addresses: } operator and ${readsock } item.
147 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
148 side, to decode a TLS 1.3 packet capture.
150 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
151 Previously the default library behaviour applied, sending two, each in
154 JH/03 Debug output for ACL now gives the config file name and line number for
157 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
159 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
161 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
162 buffer overrun for (non-chunking) other transports.
164 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
165 TLS1.3, means that a server rejecting a client certificate is not visible
166 to the client until the first read of encrypted data (typically the
167 response to EHLO). Add detection for that case and treat it as a failed
168 TLS connection attempt, so that the normal retry-in-clear can work (if
169 suitably configured).
171 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
172 and/or domain. Found and fixed by Jason Betts.
174 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
175 configuration). If a CNAME target was not a wellformed name pattern, a
178 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
179 the OS reports them interleaved with other addresses.
181 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
182 used both for input and for a verify callout, both encrypted, SMTP
183 responses being sent by the server could be lost. This resulted in
184 dropped connections and sometimes bounces generated by a peer sending
187 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
188 string. Previously it was possible to cause undefined behaviour in a
189 library routine (usually a crash). Found by "zerons".
191 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
194 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
195 API was removed, so update to use the newer ones.
197 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
198 any timeout set, is taking a long time. Previously we would hang on to a
199 rotated logfile "forever" if the input was arriving with long gaps
200 (a previous attempt to fix addressed lack, for a long time, of initial
203 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
204 shared (NFS) environment. The length of the tempfile name is now
205 4 + 16 ("hdr.$message_exim_id") which might break on file
206 systems which restrict the file name length to lower values.
209 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
210 shared (NFS) environment.
212 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
213 did for all versions <4.90). Notably -M, -m, --invert, -I may be
216 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
217 on some platforms for bit 31.
219 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
220 to changes apparently associated with TLS1.3 handling some of the APIs
221 previously used were either nonfunctional or inappropriate. Strings
222 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
223 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
224 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
225 This affects log line X= elements, the $tls_{in,out}_cipher variables,
226 and the use of specific cipher names in the encrypted= ACL condition.
228 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
230 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
231 verification result was not updated unless hosts_require_ocsp applied.
233 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
234 queue_list_requires_admin set to false, non-admin users were denied the
237 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
238 directory-of-certs mode. Previously they were advertised despite the
241 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
242 A single TCP connection by a client will now hold a TLS connection open
243 for multiple message deliveries, by default. Previoud the default was to
246 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
247 default. If built with the facility, DANE will be used. The facility
248 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
250 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
251 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
252 must be defined and you must still, unless you define DISABLE_TLS, manage
253 the the include-dir and library-file requirements that go with that
254 choice. Non-TLS builds are still supported.
256 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
259 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
260 default. If the platform supports and has the facility enabled, it will
261 be requested on all coneections.
263 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
264 controlled by the build-time option SUPPORT_PIPE_CONNECT.
266 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
268 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
269 success-DSN messages. Previously the From: header was always the default
270 one for these; the option was ignored.
272 JH/28 Fix the timeout on smtp response to apply to the whole response.
273 Previously it was reset for every read, so a teergrubing peer sending
274 single bytes within the time limit could extend the connection for a
275 long time. Credit to Qualsys Security Advisory Team for the discovery.
277 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
278 delivery address, which leaked information of the results of local
279 forwarding. Change to the original envelope recipient address, per
282 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
283 requested. Previously not bounce was generated and a log entry of
284 error ignored was made.
286 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
288 JH/32 Introduce a general tainting mechanism for values read from the input
289 channel, and values derived from them. Refuse to expand any tainted
290 values, to catch one form of exploit.
292 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
293 was unused and the unexpanded text used for the test. Found and
294 fixed by Ruben Jenster.
296 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
297 an attempt to use a TLS library read routine dereffed a nul pointer,
300 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
301 them in our smtp response.
303 JH/36 Have the general router option retry_use_local_part default to true when
304 any of the restrictive preconditions are set (to anything). Previously it
305 was only for check_local user. The change removes one item of manual
306 configuration which is required for proper retries when a remote router
307 handles a subset of addresses for a domain.
309 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
310 link count into consideration.
312 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
313 caused the extension of big_buffer, the following lines were ignored.
315 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
316 accordance with RFC 2308. Previously there was no expiry, so a longlived
317 receive process (eg. due to ACL delays) versus a short SOA value could
320 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
322 JH/39 Promote DMARC support to mainline.
324 JH/40 Bug 2452: Add a References: header to DSNs.
326 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
327 parameters. The relevant library call is documented as "Deprecated: This
328 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
329 3.6.0, DH parameters are negotiated following RFC7919."
331 HS/06 Change the default of dnssec_request_domains to "*"
333 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
334 carried on and emitted a BDAT command, even when PIPELINING was not
337 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
338 buffer was used for the filename, resulting in a trap when tainted
339 arguments (eg. $domain) were used.
341 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
342 recommended to avoid a possible server-load attack. The feature can be
343 re-enabled via the openssl_options main cofiguration option.
345 JH/45 local_scan API: documented the current smtp_printf() call. This changed
346 for version 4.90 - adding a "more data" boolean to the arguments.
347 Bumped the ABI version number also, this having been missed previously;
348 release versions 4.90 to 4.92.3 inclusive were effectively broken in
349 respect of usage of smtp_printf() by either local_scan code or libraries
350 accessed via the ${dlfunc } expansion item. Both will need coding
351 adjustment for any calls to smtp_printf() to match the new function
352 signature; a FALSE value for the new argument is always safe.
354 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
355 the file-offset (which the Linux syscall does, and exim expects); this
356 resulted in an indefinite loop.
358 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
359 to do ARC verification. The Authentication-Results: header line added
360 by the configuration then had no ARC item.
366 JH/01 Remove code calling the customisable local_scan function, unless a new
367 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
369 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
370 non-signal-safe functions being used.
372 JH/03 Bug 2269: When presented with a received message having a stupidly large
373 number of DKIM-Signature headers, disable DKIM verification to avoid
374 a resource-consumption attack. The limit is set at twenty.
376 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
377 report of oldest_pass in ${authres } in consequence, and separate out
378 some descriptions of reasons for verification fail.
380 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
381 files in the spool were present and unlocked. A queue-runner could spot
382 them, resulting in a duplicate delivery. Fix that by doing the unlock
383 after the unlink. Investigation by Tim Stewart. Take the opportunity to
384 add more error-checking on spoolfile handling while that code is being
387 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
388 No known attacks, no CVE, this is defensive hardening.
390 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
391 a queue-runner could start a delivery while other operations were ongoing.
392 Cutthrough delivery was a common victim, resulting in duplicate delivery.
393 Found and investigated by Tim Stewart. Fix by using the open message data
394 file handle rather than opening another, and not locally closing it (which
395 releases a lock) for that case, while creating the temporary .eml format
396 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
398 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
399 $sender_verify_failure/$recipient_verify_failure to "random".
401 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
404 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
405 Previously this would segfault.
407 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
410 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
411 like zero, since the resolver should be doing this for us, But we need one
412 as a CNAME but no MX presence gets the CNAME returned; we need to check
413 that doesn't point to an MX to declare it "no MX returned" rather than
414 "error, loop". A new main option is added so the older capability of
415 following some limited number of chain links is maintained.
417 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
419 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
420 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
421 now-deprecated earlier definitions used only specified the range up to TLS
422 1.2 (in the older-version library docs).
424 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
426 JH/15 Rework TLS client-side context management. Stop using a global, and
427 explicitly pass a context around. This enables future use of TLS for
428 connections to service-daemons (eg. malware scanning) while a client smtp
429 connection is using TLS; with cutthrough connections this is quite likely.
431 JH/16 Fix ARC verification to do AS checks in reverse order.
433 JH/17 Support a "tls" option on the ${readsocket } expansion item.
435 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
436 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
437 Previously the "utf8" would be re-prepended for every additional message.
439 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
440 Previously thery were accepted, resulting in issues when attempting to
441 forward messages to a non-supporting MTA.
443 PP/02 Let -n work with printing macros too, not just options.
445 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
446 one parent address was copied, and bogus data was used at delivery-logging
447 time. Either a crash (after delivery) or bogus log data could result.
448 Discovery and analysis by Tim Stewart.
450 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
451 Previously if the string ended mid-character, we did not insert the
452 promised '?' replacement.
454 PP/04 Documentation: current string operators work on bytes, not codepoints.
456 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
457 should pack well giving a smaller memory footprint so better caching and
458 therefore performance. Group the declarations where this can't be done so
459 that the byte-sized flag variables are not interspersed among pointer
460 variables, giving a better chance of good packing by the compiler.
462 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
463 non-null, to avoid issues with sites running BATV. Previously reports were
464 sent with an empty envelope sender so looked like bounces.
466 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
467 The ignore_error flag wasn't being returned from the filter subprocess so
468 was not set for later routers. Investigation and fix by Matthias Kurz.
470 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
471 and a msg:complete for the whole, when a message is manually removed using
472 -Mrm. Developement by Matthias Kurz, hacked on by JH.
474 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
475 a "Gnu special" function, asprintf() in the DB utility binary builds; I
476 hope that is portable enough.
478 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
479 requiring a known-CA anchor certificate; make it now rely entirely on the
480 TLSA as an anchor. Checking the name on the leaf cert against the name
481 on the A-record for the host is still done for TA (but not for EE mode).
483 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
484 included in delivery lines for non-proxied connections, when compiled with
485 SUPPORT_SOCKS and running with proxy logging enabled.
487 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
488 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
489 move the existing event to fire before the normal logging of message
490 failure so that custom logging is bracketed by normal logging.
492 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
493 msg:fail:internal event. Developement by Matthias Kurz.
495 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
496 far too small for todays use of crypto signatures stored there. Go all
497 the way to the max DNS message size of 64kB, even though this might be
498 overmuch for IOT constrained device use.
500 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
501 copy a string over itself. The library routine is documented as not
502 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
504 JH/32 For main options check_spool_space and check_inode_space, where the
505 platform supports 64b integers, support more than the previous 2^31 kB
506 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
507 the previous G, M, k.
509 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
510 $authenticated_fail_id variable on authentication failure. Previously
513 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
514 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
515 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
516 GNUTLS_SEC_PARAM_MEDIUM.
518 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
519 side. Previously we would continue as if no SNI had been received.
521 JH/36 Harden the handling of string-lists. When a list consisted of a sole
522 "<" character, which should be a list-separator specification, we walked
523 off past the nul-terimation.
525 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
526 causes) even when the retry time is not yet met. Previously they were
527 not, meaning that when (say) an account was over-quota and temp-rejecting,
528 and multiple senders' messages were queued, only one sender would get
529 notified on each configured delay_warning cycle.
531 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
533 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
534 cipher-suites, an error can be left on the stack even for a succeeding
535 accept; this results in impossible error messages when a later operation
538 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
539 return error codes indicating retry. Under TLS1.3 this becomes required.
541 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
542 it only wrote the new authenticators, resulting in a lack of tracking of
543 peer changes of ESMTP extensions until the next cache flush.
545 JH/41 Fix the loop reading a message header line to check for integer overflow,
546 and more-often against header_maxsize. Previously a crafted message could
547 induce a crash of the recive process; now the message is cleanly rejected.
549 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
550 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
556 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
557 When redis_servers is set to a list of > 1 element, and the Redis servers
558 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
559 case of MOVED into a DEFER case instead, thus moving the query onto the
560 next server in the list. For a cluster of N elements, all N servers must
561 be defined in redis_servers.
563 GF/02 Catch and remove uninitialized value warning in exiqsumm
564 Check for existence of @ARGV before looking at $ARGV[0]
566 JH/01 Replace the store_release() internal interface with store_newblock(),
567 which internalises the check required to safely use the old one, plus
568 the allocate and data copy operations duplicated in both (!) of the
569 extant use locations.
571 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
572 modifier. This matches the restriction on the commandline.
574 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
575 Previously only the last row was returned.
577 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
578 we assumed that tags in the header were well-formed, and parsed the
579 element content after inspecting only the first char of the tag.
580 Assumptions at that stage could crash the receive process on malformed
583 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
584 While running the DKIM ACL we operate on the Permanent memory pool so that
585 variables created with "set" persist to the DATA ACL. Also (at any time)
586 DNS lookups that fail create cache records using the Permanent pool. But
587 expansions release any allocations made on the current pool - so a dnsdb
588 lookup expansion done in the DKIM ACL releases the memory used for the
589 DNS negative-cache, and bad things result. Solution is to switch to the
590 Main pool for expansions.
591 While we're in that code, add checks on the DNS cache during store_reset,
592 active in the testsuite.
593 Problem spotted, and debugging aided, by Wolfgang Breyha.
595 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
596 When none of the hosts presented to a transport match an already-open
597 connection, close it and proceed with the list. Previously we would
598 queue the message. Spotted by Lena with Yahoo, probably involving
601 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
602 Previously a spurious "250 OK id=" response was appended to the proper
605 JH/08 The "support for" informational output now, which built with Content
606 Scanning support, has a line for the malware scanner interfaces compiled
607 in. Interface can be individually included or not at build time.
609 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
610 by the template makefile "src/EDITME". The "STREAM" support for an older
611 ClamAV interface method is removed.
613 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
614 rows affected is given instead).
616 JH/11 The runtime Berkeley DB library version is now additionally output by
617 "exim -d -bV". Previously only the compile-time version was shown.
619 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
620 SMTP connection. Previously, when one had more receipients than the
621 first, an abortive onward connection was made. Move to full support for
622 multiple onward connections in sequence, handling cutthrough connection
623 for all multi-message initiating connections.
625 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
626 routers. Previously, a multi-recipient message would fail to match the
627 onward-connection opened for the first recipient, and cause its closure.
629 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
630 a timeout on read on a GnuTLS initiating connection, resulting in the
631 initiating connection being dropped. This mattered most when the callout
632 was marked defer_ok. Fix to keep the two timeout-detection methods
635 JH/15 Relax results from ACL control request to enable cutthrough, in
636 unsupported situations, from error to silently (except under debug)
637 ignoring. This covers use with PRDR, frozen messages, queue-only and
640 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
642 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
643 metadata, resulting in a crash in free().
645 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
646 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
647 Broken also in d185889f4, with init system revamp.
649 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
650 Previously we abruptly closed the connection after reading a malware-
651 found indication; now we go on to read the "scan ok" response line,
654 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
655 ACL. Previously, a crash would result.
657 JH/19 Speed up macro lookups during configuration file read, by skipping non-
658 macro text after a replacement (previously it was only once per line) and
659 by skipping builtin macros when searching for an uppercase lead character.
661 JH/20 DANE support moved from Experimental to mainline. The Makefile control
662 for the build is renamed.
664 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
665 was allocated for every new TLS startup, meaning one per message. Fix
666 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
668 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
669 reported the original. Fix to report (as far as possible) the ACL
670 result replacing the original.
672 JH/23 Fix memory leak during multi-message connections using STARTTLS under
673 OpenSSL. Certificate information is loaded for every new TLS startup,
674 and the resources needed to be freed.
676 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
678 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
681 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
682 DATA response info to the (existing) per-recipient response info for
683 the "C=" log element. It can have useful tracking info from the
684 destination system. Patch from Simon Arlott.
686 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
687 length value. Previously this would segfault.
689 HS/02 Support Avast multiline protoocol, this allows passing flags to
690 newer versions of the scanner.
692 JH/28 Ensure that variables possibly set during message acceptance are marked
693 dead before release of memory in the daemon loop. This stops complaints
694 about them when the debug_store option is enabled. Discovered specifically
695 for sender_rate_period, but applies to a whole set of variables.
696 Do the same for the queue-runner and queue-list loops, for variables set
697 from spool message files. Do the same for the SMTP per-message loop, for
698 certain variables indirectly set in ACL operations.
700 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
701 as a multi-recipient message from a mailinglist manager). The coding had
702 an arbitrary cutoff number of characters while checking for more input;
703 enforced by writing a NUL into the buffer. This corrupted long / fast
704 input. The problem was exposed more widely when more pipelineing of SMTP
705 responses was introduced, and one Exim system was feeding another.
706 The symptom is log complaints of SMTP syntax error (NUL chars) on the
707 receiving system, and refused recipients seen by the sending system
708 (propating to people being dropped from mailing lists).
709 Discovered and pinpointed by David Carter.
711 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
712 replaced by the ${authresults } expansion.
714 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
716 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
717 allows proper process termination in container environments.
719 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
720 Previously the "final dot" had a newline after it; ensure it is CR,LF.
722 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
723 and "err_perm", deprecated since 4.83 when the RFC-defined words
724 "temperror" and "permerror" were introduced.
726 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
727 transport-filters or DKIM-signing. The restriction was lost in the
728 consolidation of verify-callout and delivery SMTP handling.
729 Extend the restriction to also cover ARC-signing.
731 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
732 in defer=pass mode supply a 450 to the initiator. Previously the message
735 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
736 tls_require_ciphers is used as before.
738 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
739 "pass_unscanned". Only tmpfails from the scanner are written to
740 the paniclog, as they may require admin intervention (permission
741 denied, license issues). Other scanner errors (like decompression
742 bombs) do not cause a paniclog entry.
744 JH/36 Fix reinitialisation of DKIM logging variable between messages.
745 Previously it was possible to log spurious information in receive log
748 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
749 triggered odd behaviour from Outlook Express clients.
751 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
754 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
755 since the IETF WG has not yet settled on that versus the original
756 "bare" representation.
758 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
759 Previously the millisecond value corrupted the output.
760 Fix also for syslog_pid=no and log_selector +pid, for which the pid
761 corrupted the output.
767 JH/01 Rework error string handling in TLS interface so that the caller in
768 more cases is responsible for logging. This permits library-sourced
769 string to be attached to addresses during delivery, and collapses
770 pairs of long lines into single ones.
772 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
773 during configuration. Wildcards are allowed and expanded.
775 JH/02 Rework error string handling in DKIM to pass more info back to callers.
776 This permits better logging.
778 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
779 do not close it down and have the child transport start it up again on
780 the passed-on TCP connection. Instead, proxy the child (and any
781 subsequent ones) for TLS via a unix-domain socket channel. Logging is
782 affected: the continued delivery log lines do not have any DNSSEC, TLS
783 Certificate or OCSP information. TLS cipher information is still logged.
785 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
786 identical IP addresses on different listening ports. Will also affect
789 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
790 add noisy ifdef guards to special-case this sillyness.
791 Patch from Bernd Kuhls.
793 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
794 than 255 are no longer allowed.
796 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
797 Disable the session-cache too, which might reduce our load. Since we
798 currrectly use a new context for every connection, both as server and
799 client, there is no benefit for these.
800 GnuTLS appears to not support tickets server-side by default (we don't
801 call gnutls_session_ticket_enable_server()) but client side is enabled
802 by default on recent versions (3.1.3 +) unless the PFS priority string
805 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
806 <https://reproducible-builds.org/specs/source-date-epoch/>.
808 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
809 the check for any unsuccessful recipients did not notice the limit, and
810 erroneously found still-pending ones.
812 JH/08 Pipeline CHUNKING command and data together, on kernels that support
813 MSG_MORE. Only in-clear (not on TLS connections).
815 JH/09 Avoid using a temporary file during transport using dkim. Unless a
816 transport-filter is involved we can buffer the headers in memory for
817 creating the signature, and read the spool data file once for the
818 signature and again for transmission.
820 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
821 4.77 as the kernel support then wasn't solid, having issues in 64bit
822 mode. Now, it's been long enough. Add support for FreeBSD also.
824 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
825 case where the routing stage had gathered several addresses to send to
826 a host before calling the transport for the first, we previously failed
827 to close down TLS in the old transport process before passing the TCP
828 connection to the new process. The new one sent a STARTTLS command
829 which naturally failed, giving a failed delivery and bloating the retry
830 database. Investigation and fix prototype from Wolfgang Breyha.
832 JH/12 Fix check on SMTP command input synchronisation. Previously there were
833 false-negatives in the check that the sender had not preempted a response
834 or prompt from Exim (running as a server), due to that code's lack of
835 awareness of the SMTP input buffering.
837 PP/04 Add commandline_checks_require_admin option.
838 Exim drops privileges sanely, various checks such as -be aren't a
839 security problem, as long as you trust local users with access to their
840 own account. When invoked by services which pass untrusted data to
841 Exim, this might be an issue. Set this option in main configuration
842 AND make fixes to the calling application, such as using `--` to stop
845 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
846 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
847 responses to those, into a single TLS record each way (this usually means
848 a single packet). As a side issue, smtp_enforce_sync now works on TLS
851 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
852 affects you only if you're dancing at the edge of the param size limits.
853 If you are, and this message makes sense to you, then: raise the
854 configured limit or use OpenSSL 1.1. Nothing we can do for older
857 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
858 cmdline element to get the documented default one. Previously it was
861 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
862 are unsuable for non-message-specific reasons - eg. connection timeout,
863 banner-time rejection.
865 JH/16 Fix logging of delivery remote port, when specified by router, under
868 PP/06 Repair manualroute's ability to take options in any order, even if one
869 is the name of a transport.
872 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
874 JH/17 Change the list-building routines interface to use the expanding-string
875 triplet model, for better allocation and copying behaviour.
877 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
878 Previously it was constructed the first time a possibly-matching string
879 was met in the configuration file input during startup; now it is done
882 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
883 compatible one, to avoid the (poorly documented) possibility of a config
884 file in the working directory redirecting the DB files, possibly correpting
885 some existing file. CVE-2017-10140 assigned for BDB.
887 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
888 cache-hot. Previously, although the result was properly cached, the
889 initial verify call returned a defer.
891 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
892 the main verify for receipient in uncached-mode.
894 JH/22 Retire historical build files to an "unsupported" subdir. These are
895 defined as "ones for which we have no current evidence of testing".
897 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
898 if present. Previously it was ignored.
900 JH/24 Start using specified-initialisers in C structure init coding. This is
901 a C99 feature (it's 2017, so now considered safe).
903 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
904 if was a fixed-sized field and bitmask ops via macros; it is now more
907 PP/07 GitHub PR 56: Apply MariaDB build fix.
908 Patch provided by Jaroslav Škarvada.
910 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
911 during Coverity cleanups [4.87 JH/47]
912 Diagnosis and fix provided by Michael Fischer v. Mollard.
914 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
915 the right size to place the terminating semicolon on its own folded
916 line, the header hash was calculated to an incorrect value thanks to
917 the (relaxed) space the fold became.
919 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
920 and confused the parent.
922 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
923 which could crash as a result. This could lead to undeliverable messages.
925 JH/28 Logging: "next input sent too soon" now shows where input was truncated
928 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
929 matters on fast-turnover and PID-randomising systems, which were getting
930 out-of-order delivery.
932 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
933 a possibly-overlapping copy. The symptom was that "Remote host closed
934 connection in response to HELO" was logged instead of the actual 4xx
937 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
938 Previously only that bufferd was discarded, resulting in SYMTP command
941 JH/32 DKIM: when a message has multiple signatures matching an identity given
942 in dkim_verify_signers, run the dkim acl once for each. Previously only
943 one run was done. Bug 2189.
945 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
946 "panic the current process" to "deliberately defer". The panic log is
947 still written with the problem list name; the mail and reject logs now
948 get a temp-reject line for the message that was being handled, saying
949 something like "domains check lookup or other defer". The SMTP 451
950 message is still "Temporary local problem".
952 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
953 A crafted sequence of BDAT commands could result in in-use memory beeing
954 freed. CVE-2017-16943.
956 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
957 from SMTP input. Previously it was always done; now only done for DATA
958 and not BDAT commands. CVE-2017-16944.
960 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
961 to the message (such as an overlong header line). Previously this was
962 not done and we did not exit BDAT mode. Followon from the previous item
963 though a different problem.
969 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
970 than -2003 did; needs libidn2 in addition to libidn.
972 JH/02 The path option on a pipe transport is now expanded before use.
974 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
975 Patch provided by "Björn", documentation fix added too.
977 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
978 missing a wire-to-host endian conversion.
980 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
981 close after a BDAT command line could be taken as a following command,
982 giving a synch failure. Fix by only checking for synch immediately
983 before acknowledging the chunk.
985 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
986 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
987 macro. Patches provided by Josh Soref.
989 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
990 Previously we did not; the RFC seems ambiguous and VRFY is not listed
991 by IANA as a service extension. However, John Klensin suggests that we
994 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
995 the dkim code may be unix-mode line endings rather than smtp wire-format
996 CRLF, so prepend a CR to any bare LF.
998 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
999 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1001 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1002 the first were themselves being wrongly included in the feed into dkim
1003 processing; with most chunk sizes in use this resulted in an incorrect
1004 body hash calculated value.
1006 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1007 DKIM signature block, for verification. Although advised against by
1008 standards it is specifically not ruled illegal.
1010 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1012 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1013 missing a body hash (the bh= tag).
1015 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1016 It seems that HAProxy sends the Proxy Protocol information in clear and
1017 only then does a TLS startup, so do the same.
1019 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1020 TCP connections (such as for Spamd) unless the daemon successfully set
1021 Fast Open mode on its listening sockets. This fixes breakage seen on
1022 too-old kernels or those not configured for Fast Open, at the cost of
1023 requiring both directions being enabled for TFO, and TFO never being used
1024 by non-daemon-related Exim processes.
1026 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1027 endings, at least on the first header line. Try to canonify any that get
1028 past that check, despite the cost.
1030 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1031 now limited to an arbitrary five deep, while parsing addresses with the
1032 strip_excess_angle_brackets option enabled.
1034 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1035 instead leave the unprompted TLS handshake in socket buffer for the
1036 TLS library to consume.
1038 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1040 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1042 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1043 region in one operation, for speed, and this leaves assigned pointers
1044 dangling. Add checks run only under the testsuite which checks all
1045 variables at a store-reset and panics on a dangling pointer; add code
1046 explicitly nulling out all the variables discovered. Fixes one known
1047 bug: a transport crash, where a dangling pointer for $sending_ip_address
1048 originally assigned in a verify callout, is re-used.
1050 PP/06 Drop '.' from @INC in various Perl scripts.
1052 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1054 PP/08 Reduce a number of compilation warnings under clang; building with
1055 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1056 should be warning-free.
1058 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1060 HS/01 Fix portability problems introduced by PP/08 for platforms where
1061 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1063 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1064 chunk. This allows us to accept broken chunked messages. We need a more
1065 general solution here.
1067 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1068 already-broken messages in the queue.
1070 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1072 JH/19 Fix reference counting bug in routing-generated-address tracking.
1078 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1079 supports it and a size is available (ie. the sending peer gave us one).
1081 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1082 years of being deprecated). The replacements are the ACLs
1083 acl_smtp_mime and acl_not_smtp_mime.
1085 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1086 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1087 or even an in-clear connection were permitted. Now, if the host lookup
1088 was dnssec and dane was requested then the host is only used if the
1089 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1090 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1091 if one fails this test.
1092 This means that a poorly-configured remote DNS will make it incommunicado;
1093 but it protects against a DNS-interception attack on it.
1095 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1096 non-noisy when a race steals the message being considered.
1098 JH/05 If main configuration option tls_certificate is unset, generate a
1099 self-signed certificate for inbound TLS connections.
1101 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1102 in rewrites and routers.
1104 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1105 and logged a warning sing 4.83; now they are a configuration file error.
1107 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1108 (lacking @domain). Apply the same qualification processing as RCPT.
1110 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1112 JH/10 Support ${sha256:} applied to a string (as well as the previous
1115 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1116 a cutthrough deliver is pending, as we always want to make a connection.
1117 This also avoids re-routing the message when later placing the cutthrough
1118 connection after a verify cache hit.
1119 Do not update it with the verify result either.
1121 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1122 when routing results in more than one destination address.
1124 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1125 signing (which inhibits the cutthrough capability). Previously only
1126 the presence of an option was tested; now an expansion evaluating as
1127 empty is permissible (obviously it should depend only on data available
1128 when the cutthrough connection is made).
1130 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1131 the relevant preceding SMTP command did not note the pipelining mode.
1133 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1134 Previously they were not counted.
1136 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1137 as one having no matching records. Previously we deferred the message
1138 that needed the lookup.
1140 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1141 distinguished as "(=".
1143 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1144 for missing MX records. Previously it only worked for missing A records.
1146 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1148 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1149 after the data-go-ahead and data-ack. Patch from Jason Betts.
1151 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1152 even for a "none" policy. Patch from Tony Meyer.
1154 JH/22 Fix continued use of a connection for further deliveries. If a port was
1155 specified by a router, it must also match for the delivery to be
1158 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1159 When one of the recipients of a message was unsuitable for the connection
1160 (has no matching addresses), we lost track of needing to mark it
1161 deferred. As a result mail would be lost.
1163 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1165 JH/25 Decoding ACL controls is now done using a binary search; the source code
1166 takes up less space and should be simpler to maintain. Merge the ACL
1167 condition decode tables also, with similar effect.
1169 JH/26 Fix problem with one_time used on a redirect router which returned the
1170 parent address unchanged. A retry would see the parent address marked as
1171 delivered, so not attempt the (identical) child. As a result mail would
1174 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1175 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1176 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1179 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1180 Main config options check_{log,spool}_{inodes,space} are now
1181 100 inodes, 10MB unless set otherwise in the configuration.
1183 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1184 Previously it only applied to the main-section connection policy
1187 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1189 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1190 by me. Added RFC7919 DH primes as an alternative.
1192 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1193 are not in the system include path.
1195 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1196 GnuTLS, when a session startup failed (eg because the client disconnected)
1197 Exim did stdio operations after fclose. This was exposed by a recent
1198 change which nulled out the file handle after the fclose.
1200 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1201 signed directly by the cert-signing cert, rather than an intermediate
1202 OCSP-signing cert. This is the model used by LetsEncrypt.
1204 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1206 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1207 an incoming connection.
1209 HS/02 Bug 1802: Do not half-close the connection after sending a request
1212 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1213 fallback to "prime256v1".
1215 JH/34 SECURITY: Use proper copy of DATA command in error message.
1216 Could leak key material. Remotely exploitable. CVE-2016-9963.
1222 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1223 and 3.4.4 - once the server is enabled to respond to an OCSP request
1224 it does even when not requested, resulting in a stapling non-aware
1225 client dropping the TLS connection.
1227 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1228 support variable-length bit vectors. No functional change.
1230 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1231 The I= interface field on outgoing lines is now after the H= remote
1232 host field, same as incoming lines. There is a separate
1233 outgoing_interface log selector which allows you to disable the
1236 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1237 If not running log_selector +smtp_connection the mainlog would be held
1238 open indefinitely after a "too many connections" event, including to a
1239 deleted file after a log rotate. Leave the per net connection logging
1240 leaving it open for efficiency as that will be quickly detected by the
1241 check on the next write.
1243 HS/01 Bug 1671: Fix post transport crash.
1244 Processing the wait-<transport> messages could crash the delivery
1245 process if the message IDs didn't exist for some reason. When
1246 using 'split_spool_directory=yes' the construction of the spool
1247 file name failed already, exposing the same netto behaviour.
1249 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1250 mime_regex ACL conditions.
1252 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1253 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1254 to HELO, local diagnostic string.
1256 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1257 log line to debug. Even when configured with a tls authenticator many
1258 client connections are expected to not authenticate in this way, so
1259 an authenticate fail is not an error.
1261 HS/02 Add the Exim version string to the process info. This way exiwhat
1262 gives some more detail about the running daemon.
1264 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1265 matter for fast-change records such as DNSBLs.
1267 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1268 retry record, even if constant. There may be multiple transports with
1269 different interface settings and the retry behaviour needs to be kept
1272 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1274 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1276 JH/10 Bug 840: fix log_defer_output option of pipe transport
1278 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1279 hosts_randomize. This matters a lot when combined with mua_wrapper.
1281 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1282 ${quote_pgsql:<string>} operator.
1284 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1286 JH/14 Change header file name for freeradius-client. Relevant if compiling
1287 with Radius support; from the Gentoo tree and checked under Fedora.
1289 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1291 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1292 option transport_filter, meaning no filtering.
1294 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1296 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1297 defaults to "*" (all hosts). The variable is now available when not built
1298 with TLS, default unset, mainly to enable keeping the testsuite sane.
1299 If a server certificate is not supplied (via tls_certificate) an error is
1300 logged, and clients will find TLS connections fail on startup. Presumably
1301 they will retry in-clear.
1302 Packagers of Exim are strongly encouraged to create a server certificate
1303 at installation time.
1305 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1306 with the $config_file variable.
1308 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1309 in transport context, after the attempt, and per-recipient. The latter type
1310 is per host attempted. The event data is the error message, and the errno
1311 information encodes the lookup type (A vs. MX) used for the (first) host,
1312 and the trailing two digits of the smtp 4xx response.
1314 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1315 to write to mainlog (or rejectlog, paniclog) in the window between file
1316 creation and permissions/ownership being changed. Particularly affects
1317 installations where exicyclog is run as root, rather than exim user;
1318 result is that the running daemon panics and dies.
1320 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1322 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1323 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1324 "pri" and "weight". Note that the previous implicit priority given by the
1325 list order is no longer honoured.
1327 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1328 for DKIM processing.
1330 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1331 by defining SUPPORT_SOCKS.
1333 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1334 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1335 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1336 variables are renamed to proxy_{local,external}_{address,port}.
1338 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1339 for a build by defining SUPPORT_I18N
1341 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1342 of the query string, and make ${quote_redis:} do that quoting.
1344 JH/29 Move Events support from Experimental to mainline, enabled by default
1345 and removable for a build by defining DISABLE_EVENT.
1347 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1349 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1350 cached by the daemon.
1352 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1353 by defining LOOKUP_REDIS. The libhiredis library is required.
1355 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1356 keys are given for lookup.
1358 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1359 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1360 only supported when built with TLS support. The PolarSSL SHA routines
1361 are still used when the TLS library is too old for convenient support.
1363 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1364 openssl_options), for security. OpenSSL forces this from version 1.1.0
1365 server-side so match that on older versions.
1367 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1368 allocation for $value could be released as the expansion processing
1369 concluded, but leaving the global pointer active for it.
1371 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1372 and to use the domains and local_parts ACL conditions.
1374 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1375 incorrectly not doubled on cutthrough transmission, hence seen as a
1376 body-termination at the receiving system - resulting in truncated mails.
1377 Commonly the sender saw a TCP-level error, and retransmitted the message
1378 via the normal store-and-forward channel. This could result in duplicates
1379 received - but deduplicating mailstores were liable to retain only the
1380 initial truncated version.
1382 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1384 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1386 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1387 we're in there, support oversigning also; bug 1309.
1389 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1391 HS/04 Add support for keep_environment and add_environment options.
1393 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1394 either intentional arithmetic overflow during PRNG, or testing config-
1397 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1398 delivery resulted in actual delivery. Cancel cutthrough before DATA
1401 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1402 rejects a recipient: pass the reject to the originator.
1404 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1405 Many were false-positives and ignorable, but it's worth fixing the
1408 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1409 for the new environment-manipulation done at startup. Move the routines
1410 from being local to tls.c to being global via the os.c file.
1412 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1413 an extract embedded as result-arg for a map, the first arg for extract
1414 is unavailable so we cannot tell if this is a numbered or keyed
1415 extraction. Accept either.
1421 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1424 JH/02 The smtp transport option "multi_domain" is now expanded.
1426 JH/03 The smtp transport now requests PRDR by default, if the server offers
1429 JH/04 Certificate name checking on server certificates, when exim is a client,
1430 is now done by default. The transport option tls_verify_cert_hostnames
1431 can be used to disable this per-host. The build option
1432 EXPERIMENTAL_CERTNAMES is withdrawn.
1434 JH/05 The value of the tls_verify_certificates smtp transport and main options
1435 default to the word "system" to access the system default CA bundle.
1436 For GnuTLS, only version 3.0.20 or later.
1438 JH/06 Verification of the server certificate for a TLS connection is now tried
1439 (but not required) by default. The verification status is now logged by
1440 default, for both outbound TLS and client-certificate supplying inbound
1443 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1446 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1447 Status Notification (bounce) messages are now MIME format per RFC 3464.
1448 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1449 under the control of the dsn_advertise_hosts option, and routers may
1450 have a dsn_lasthop option.
1452 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1453 default, modifiable by a malware= option. The list separator for
1454 the options can now be changed in the usual way. Bug 68.
1456 JH/10 The smtp_receive_timeout main option is now expanded before use.
1458 JH/11 The incoming_interface log option now also enables logging of the
1459 local interface on delivery outgoing connections.
1461 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1462 if the interface and destination host and port all match.
1464 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1467 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1468 Patch from Andrew Lewis.
1470 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1471 now supports optional time-restrictions, weighting, and priority
1472 modifiers per server. Patch originally by <rommer@active.by>.
1474 JH/16 The spamd_address main option now supports a mixed list of local
1475 and remote servers. Remote servers can be IPv6 addresses, and
1476 specify a port-range.
1478 JH/17 Bug 68: The spamd_address main option now supports an optional
1479 timeout value per server.
1481 JH/18 Bug 1581: Router and transport options headers_add/remove can
1482 now have the list separator specified.
1484 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1487 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1490 JH/21 Support for the A6 type of dns record is withdrawn.
1492 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1493 rather than the verbs used.
1495 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1496 from 255 to 1024 chars.
1498 JH/24 Verification callouts now attempt to use TLS by default.
1500 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1501 are generic router options now. The defaults didn't change.
1503 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1504 Original patch from Alexander Shikoff, worked over by JH.
1506 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1507 files if ZCAT_COMMAND is not executable.
1509 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1511 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1513 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1514 Normally benign, it bites when the pair was led to by a CNAME;
1515 modern usage is to not canonicalize the domain to a CNAME target
1516 (and we were inconsistent anyway for A-only vs AAAA+A).
1518 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1520 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1521 when evaluating $sender_host_dnssec.
1523 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1524 $sender_helo_dnssec variable.
1526 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1528 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1530 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1532 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1533 documented as working, but never had. Support all but $spam_report.
1535 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1536 added for tls authenticator.
1538 HS/03 Add perl_taintmode main config option
1544 TL/01 When running the test suite, the README says that variables such as
1545 no_msglog_check are global and can be placed anywhere in a specific
1546 test's script, however it was observed that placement needed to be near
1547 the beginning for it to behave that way. Changed the runtest perl
1548 script to read through the entire script once to detect and set these
1549 variables, reset to the beginning of the script, and then run through
1550 the script parsing/test process like normal.
1552 TL/02 The BSD's have an arc4random API. One of the functions to induce
1553 adding randomness was arc4random_stir(), but it has been removed in
1554 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1555 function when detected.
1557 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1558 cause callback expansion.
1560 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1561 syntax errors in an expansion can be treated as a string instead of
1562 logging or causing an error, due to the internal use of bool_lax
1563 instead of bool when processing it.
1565 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1566 server certificates when making smtp deliveries.
1568 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1570 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1572 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1574 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1575 Merged patch from Sebastian Wiedenroth.
1577 JH/05 Fix results-pipe from transport process. Several recipients, combined
1578 with certificate use, exposed issues where response data items split
1579 over buffer boundaries were not parsed properly. This eventually
1580 resulted in duplicates being sent. This issue only became common enough
1581 to notice due to the introduction of connection certificate information,
1582 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1584 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1585 size buffer was used, resulting in syntax errors when an expansion
1588 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1589 version 3.3.6 or later.
1591 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1592 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1593 both become "event_action", the variables become $event_name, $event_data
1594 and $event_defer_errno. There is a new variable $verify_mode, usable in
1595 routers, transports and related events. The tls:cert event is now also
1596 raised for inbound connections, if the main configuration event_action
1599 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1600 early OCSP support, but no stapling (appears to be less than 1.0.0).
1602 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1603 server certificate names available under the smtp transport option
1604 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1607 JH/10 Time-related extraction expansions from certificates now use the main
1608 option "timezone" setting for output formatting, and are consistent
1609 between OpenSSL and GnuTLS compilations. Bug 1541.
1611 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1612 encoded parameter in the incoming message. Bug 1558.
1614 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1615 include certificate info, eximon was claiming there were spoolfile
1618 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1620 JH/14 Log delivery-related information more consistently, using the sequence
1621 "H=<name> [<ip>]" wherever possible.
1623 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1624 are problematic for Debian distribution, omit them from the release
1627 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1629 JH/16 Fix string representation of time values on 64bit time_t architectures.
1632 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1633 output list separator was used.
1638 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1639 checkers that were complaining about end of non-void function with no
1642 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1643 This was a regression introduced in 4.83 by another bugfix.
1645 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1647 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1648 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1654 TF/01 Correctly close the server side of TLS when forking for delivery.
1656 When a message was received over SMTP with TLS, Exim failed to clear up
1657 the incoming connection properly after forking off the child process to
1658 deliver the message. In some situations the subsequent outgoing
1659 delivery connection happened to have the same fd number as the incoming
1660 connection previously had. Exim would try to use TLS and fail, logging
1661 a "Bad file descriptor" error.
1663 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1664 utilities have not been installed.
1666 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1667 temporary space as the ACL may create new global variables.
1669 TL/01 LDAP support uses per connection or global context settings, depending
1670 upon the detected version of the libraries at build time.
1672 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1673 to extract and use the src ip:port in logging and expansions as if it
1674 were a direct connection from the outside internet. PPv2 support was
1675 updated based on HAProxy spec change in May 2014.
1677 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1679 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1680 Properly escape header and check for NULL return.
1682 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1685 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1687 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1688 characters in header names, implemented as a verify condition.
1689 Contributed by Michael Fischer v. Mollard.
1691 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1692 results permerror and temperror. Previous values are deprecated but
1693 still accepted. In a future release, err_perm and err_temp will be
1694 completely removed, which will be a backward incompatibility if the
1695 ACL tests for either of these two old results. Patch contributed by
1696 user bes-internal on the mailing list.
1698 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1700 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1701 selectors, in both main and reject logs.
1703 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1706 JH/07 Add malware type "sock" for talking to simple daemon.
1708 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1710 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1711 routers/transports under cutthrough routing.
1713 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1714 numbers. Touch up "bool" conditional to keep the same definition.
1716 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1718 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1720 JH/12 Expand items in router/transport headers_add or headers_remove lists
1721 individually rather than the list as a whole. Bug 1452.
1723 Required for reasonable handling of multiple headers_ options when
1724 they may be empty; requires that headers_remove items with embedded
1725 colons must have them doubled (or the list-separator changed).
1727 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1728 view the policy declared in the DMARC record. Currently, $dmarc_status
1729 is a combined value of both the record presence and the result of the
1732 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1734 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1735 dnslookup router and the smtp transport (applying to the forward
1738 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1739 of ldap servers used for a specific lookup. Patch provided by Heiko
1742 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1743 New variable $lookup_dnssec_authenticated for observability.
1745 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1746 Patch submitted by Lars Timman.
1748 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1750 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1751 Requires trusted mode and valid format message id, aborts otherwise.
1752 Patch contributed by Heiko Schlichting.
1754 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1755 certextract with support for various fields. Bug 1358.
1757 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1758 is requested by default, modifiable by smtp transport option
1761 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1762 operate on certificate variables to give certificate fingerprints
1763 Also new ${sha256:cert_variable}.
1765 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1767 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1768 Christian Aistleitner.
1770 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1772 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1773 file. Patch from Wolfgang Breyha.
1775 JH/25 Expand the coverage of the delivery $host and $host_address to
1776 client authenticators run in verify callout. Bug 1476.
1778 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1779 align with daemon_smtp_ports. Bug 72.
1781 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1782 support and error reporting did not work properly.
1784 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1785 and is readable. Patch from Andrew Colin Kissa.
1787 TL/14 Enhance documentation of ${run expansion and how it parses the
1788 commandline after expansion, particularly in the case when an
1789 unquoted variable expansion results in an empty value.
1791 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1793 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1794 by renaming away from it. Addresses GH issue 15, reported by
1797 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1799 TL/15 SECURITY: prevent double expansion in math comparison functions
1800 (can expand unsanitized data). Not remotely exploitable.
1807 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1809 PP/02 Make -n do something, by making it not do something.
1810 When combined with -bP, the name of an option is not output.
1812 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1815 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1816 $sender_host_name and config options to manage this, and basic check
1819 PP/05 DSCP support for outbound connections and control modifier for inbound.
1821 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1822 (Only plugin which currently uses this is kerberos4, which nobody should
1823 be using, but we should make it available and other future plugins might
1824 conceivably use it, even though it would break NAT; stuff *should* be
1825 using channel bindings instead).
1827 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1828 name; added for Sendmail compatibility; requires admin caller.
1829 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1830 just ignore it); requires trusted caller.
1831 Also parse but ignore: -Ac -Am -X<logfile>
1834 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1836 TL/02 Add +smtp_confirmation as a default logging option.
1838 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1839 Patch by Magnus Holmgren from 2007-02-20.
1841 TL/04 Bugzilla 1281 - Spec typo.
1842 Bugzilla 1283 - Spec typo.
1843 Bugzilla 1290 - Spec grammar fixes.
1845 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1847 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1849 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1850 mailing list by by Dmitry Isaikin.
1852 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1854 JH/02 Support "G" suffix to numbers in ${if comparisons.
1856 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1858 NM/01 Bugzilla 1197 - Spec typo
1859 Bugzilla 1196 - Spec examples corrections
1861 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1863 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1864 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1867 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1868 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1870 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1871 "acl {{name}{arg}...}", and optional args on acl condition
1874 JH/05 Permit multiple router/transport headers_add/remove lines.
1876 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1878 JH/07 Avoid using a waiting database for a single-message-only transport.
1879 Performance patch from Paul Fisher. Bugzilla 1262.
1881 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1884 JH/09 Add $headers_added variable, with content from use of ACL modifier
1885 add_header (but not yet added to the message). Bugzilla 199.
1887 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1888 Pulled from Bugzilla 817 by Wolfgang Breyha.
1890 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1892 (nb: this is the same fix as in Exim 4.80.1)
1894 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1897 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1900 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1902 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1903 advertise SMTP AUTH mechanism to us, instead of a generic
1904 protocol violation error. Also, make Exim more robust to bad
1905 data from the Dovecot auth socket.
1907 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1909 When a queue runner is handling a message, Exim first routes the
1910 recipient addresses, during which it prunes them based on the retry
1911 hints database. After that it attempts to deliver the message to
1912 any remaining recipients. It then updates the hints database using
1915 So if a recipient address works intermittently, it can get repeatedly
1916 deferred at routing time. The retry hints record remains fresh so the
1917 address never reaches the final cutoff time.
1919 This is a fairly common occurrence when a user is bumping up against
1920 their storage quota. Exim had some logic in its local delivery code
1921 to deal with this. However it did not apply to per-recipient defers
1922 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1924 This change adds a proper retry rule check during routing so that the
1925 final cutoff time is checked against the message's age. We only do
1926 this check if there is an address retry record and there is not a
1927 domain retry record; this implies that previous attempts to handle
1928 the address had the retry_use_local_parts option turned on. We use
1929 this as an approximation for the destination being like a local
1930 delivery, as in LMTP.
1932 I suspect this new check makes the old local delivery cutoff check
1933 redundant, but I have not verified this so I left the code in place.
1935 TF/02 Correct gecos expansion when From: is a prefix of the username.
1937 Test 0254 submits a message to Exim with the header
1941 When I ran the test suite under the user fanf2, Exim expanded
1942 the header to contain my full name, whereas it should have added
1943 a Resent-Sender: header. It erroneously treats any prefix of the
1944 username as equal to the username.
1946 This change corrects that bug.
1948 GF/01 DCC debug and logging tidyup
1949 Error conditions log to paniclog rather than rejectlog.
1950 Debug lines prefixed by "DCC: " to remove any ambiguity.
1952 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1954 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1955 Bug spotted by Jeremy Harris; was flawed since initial commit.
1956 Would have resulted in OCSP responses post-SNI triggering an Exim
1957 NULL dereference and crash.
1959 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1961 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1962 Bug detection, analysis and fix by Samuel Thibault.
1963 Bugzilla 1331, Debian bug #698092.
1965 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1967 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1968 Server implementation by Todd Lyons, client by JH.
1969 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1970 config variable "prdr_enable" controls whether the server
1971 advertises the facility. If the client requests PRDR a new
1972 acl_data_smtp_prdr ACL is called once for each recipient, after
1973 the body content is received and before the acl_smtp_data ACL.
1974 The client is controlled by both of: a hosts_try_prdr option
1975 on the smtp transport, and the server advertisement.
1976 Default client logging of deliveries and rejections involving
1977 PRDR are flagged with the string "PRDR".
1979 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1980 fclose(). Diagnosis by Todd Lyons.
1982 PP/17 Update configure.default to handle IPv6 localhost better.
1983 Patch by Alain Williams (plus minor tweaks).
1986 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1987 This is now consistent with GnuTLS, and is now documented: the
1988 previous undocumented portable approach to treating the option as
1989 unset was to force an expansion failure. That still works, and
1990 an empty string is now equivalent.
1992 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1993 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1994 not performing validation itself.
1996 PP/20 Added force_command boolean option to pipe transport.
1997 Patch from Nick Koston, of cPanel Inc.
1999 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2002 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2004 PP/21 Fix eximon continuous updating with timestamped log-files.
2005 Broken in a format-string cleanup in 4.80, missed when I repaired the
2006 other false fix of the same issue.
2007 Report and fix from Heiko Schlichting.
2010 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2011 Report from Prashanth Katuri.
2013 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2014 It's SecureTransport, so affects any MacOS clients which use the
2015 system-integrated TLS libraries, including email clients.
2017 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2018 using a MIME ACL for non-SMTP local injection.
2019 Report and assistance in diagnosis by Warren Baker.
2021 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2023 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2025 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2026 last id that failed so it may be referenced in subsequent ACL's.
2028 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2031 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2032 ldap library initialization, allowing self-signed CA's to be
2033 used. Also properly sets require_cert option later in code by
2034 using NULL (global ldap config) instead of ldap handle (per
2035 session). Bug diagnosis and testing by alxgomz.
2037 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2038 the src/util/ subdirectory.
2040 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2041 renamed to Transport Post Delivery Action by Jeremy Harris, as
2044 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2045 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2046 redis_servers = needs to be configured which will be used by the redis
2047 lookup. Patch from Warren Baker, of The Packet Hub.
2049 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2051 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2052 hostname or reverse DNS when processing a host list. Used suggestions
2053 from multiple comments on this bug.
2055 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2057 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2058 Missed a few lines, added it to make the runtest require no keyboard
2061 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2062 contains upper case chars. Make router use caseful_local_part.
2064 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2065 support when GnuTLS has been built with p11-kit.
2071 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2073 This, or similar/improved, will also be change PP/11 of 4.82.
2079 PP/01 Handle short writes when writing local log-files.
2080 In practice, only affects FreeBSD (8 onwards).
2081 Bugzilla 1053, with thanks to Dmitry Isaikin.
2083 NM/01 Bugzilla 949 - Documentation tweak
2085 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2088 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2090 PP/02 Implemented gsasl authenticator.
2092 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2094 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2095 `pkg-config foo` for cflags/libs.
2097 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2098 with rest of GSASL and with heimdal_gssapi.
2100 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2101 `pkg-config foo` for cflags/libs for the TLS implementation.
2103 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2104 properties get this fed in as external SSF. A number of robustness
2105 and debugging improvements to the cyrus_sasl authenticator.
2107 PP/08 cyrus_sasl server now expands the server_realm option.
2109 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2110 Patch by Jeremy Harris.
2112 PP/10 Added dbmjz lookup type.
2114 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2116 PP/12 MAIL args handles TAB as well as SP, for better interop with
2117 non-compliant senders.
2118 Analysis and variant patch by Todd Lyons.
2120 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2121 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2122 Patch from Dirk Mueller <dmueller@suse.com>
2124 PP/13 tls_peerdn now print-escaped for spool files.
2125 Observed some $tls_peerdn in wild which contained \n, which resulted
2126 in spool file corruption.
2128 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2129 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2130 or write after TLS renegotiation, which otherwise led to messages
2133 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2134 as a tracking header (ie: a signed header comes before the signature).
2135 Patch from Wolfgang Breyha.
2137 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2138 comma-sep list; embedded commas doubled.
2140 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2142 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2144 Report and patch from Dmitry Banschikov.
2146 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2147 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2148 failures. We appear to now support TLS1.1+ with Exim.
2150 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2151 lets Exim select keys and certificates based upon TLS SNI from client.
2152 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2153 before an outbound SMTP session. New log_selector, +tls_sni.
2155 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2156 NULL dereference. Report and patch from Alun Jones.
2158 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2159 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2160 Not seeing resolver debug output on NetBSD, but suspect this is a
2161 resolver implementation change.
2163 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2164 Left warnings. Added "eximon gdb" invocation mode.
2166 PP/21 Defaulting "accept_8bitmime" to true, not false.
2168 PP/22 Added -bw for inetd wait mode support.
2170 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2171 locate the relevant includes and libraries. Made this the default.
2173 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2174 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2176 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2177 This may cause build issues on older platforms.
2179 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2180 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2181 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2182 Added SNI support via GnuTLS too.
2183 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2185 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2187 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2188 Applied second patch from Janne, implementing suggestion to default
2189 multiple-strings-in-record handling to match SPF spec.
2191 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2193 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2194 read-only, out of scope).
2195 Patch from Wolfgang Breyha, report from Stuart Northfield.
2197 PP/29 Fix three issues highlighted by clang analyser static analysis.
2198 Only crash-plausible issue would require the Cambridge-specific
2199 iplookup router and a misconfiguration.
2200 Report from Marcin Mirosław.
2202 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2204 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2205 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2206 As part of this, removing so much warning spew let me fix some minor
2207 real issues in debug logging.
2209 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2210 assignment on my part. Fixed.
2212 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2213 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2214 Janne Snabb (who went above and beyond: thank you).
2216 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2217 string otherwise requires a connection and a bunch more work and it's
2218 relatively easy to get wrong. Should also expose TLS library linkage
2221 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2222 64-bit ${eval} (JH/03).
2224 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2225 GNU libc to support some of the 64-bit stuff, should not lead to
2226 conflicts. Defined before os.h is pulled in, so if a given platform
2227 needs to override this, it can.
2229 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2230 protection layer was required, which is not implemented.
2231 Bugzilla 1254, patch from Wolfgang Breyha.
2233 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2234 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2235 tls_dhparam take prime identifiers. Also unbreak combination of
2236 OpenSSL+DH_params+TLSSNI.
2238 PP/39 Disable SSLv2 by default in OpenSSL support.
2244 PP/01 Solaris build fix for Oracle's LDAP libraries.
2245 Bugzilla 1109, patch from Stephen Usher.
2247 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2249 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2252 TF/02 Fix a couple more cases where we did not log the error message
2253 when unlink() failed. See also change 4.74-TF/03.
2255 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2256 lock up or crash if it happened to be inside a call to libc when it
2257 got a SIGUSR1 from exiwhat.
2259 The SIGUSR1 handler appends the current process status to the process
2260 log which is later printed by exiwhat. It used to use the general
2261 purpose logging code to do this, but several functions it calls are
2262 not safe for signals.
2264 The new output code in the SIGUSR1 handler is specific to the process
2265 log, and simple enough that it's easy to inspect for signal safety.
2266 Removing some special cases also simplifies the general logging code.
2267 Removing the spurious timestamps from the process log simplifies
2270 TF/04 Improved ratelimit ACL condition.
2272 The /noupdate option has been deprecated in favour of /readonly which
2273 has clearer semantics. The /leaky, /strict, and /readonly update modes
2274 are mutually exclusive. The update mode is no longer included in the
2275 database key; it just determines when the database is updated. (This
2276 means that when you upgrade Exim will forget old rate measurements.)
2278 Exim now checks that the per_* options are used with an update mode that
2279 makes sense for the current ACL. For example, when Exim is processing a
2280 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2281 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2282 must specify per_mail/readonly. If you omit the update mode it defaults to
2283 /leaky where that makes sense (as before) or /readonly where required.
2285 The /noupdate option is now undocumented but still supported for
2286 backwards compatibility. It is equivalent to /readonly except that in
2287 ACLs where /readonly is required you may specify /leaky/noupdate or
2288 /strict/noupdate which are treated the same as /readonly.
2290 A useful new feature is the /count= option. This is a generalization
2291 of the per_byte option, so that you can measure the throughput of other
2292 aggregate values. For example, the per_byte option is now equivalent
2293 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2295 The per_rcpt option has been generalized using the /count= mechanism
2296 (though it's more complicated than the per_byte equivalence). When it is
2297 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2298 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2299 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2300 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2301 Note that using per_rcpt with a non-readonly update mode in more than
2302 one ACL will cause the recipients to be double-counted. (The per_mail
2303 and per_byte options don't have this problem.)
2305 The handling of very low rates has changed slightly. If the computed rate
2306 is less than the event's count (usually one) then this event is the first
2307 after a long gap. In this case the rate is set to the same as this event's
2308 count, so that the first message of a spam run is counted properly.
2310 The major new feature is a mechanism for counting the rate of unique
2311 events. The new per_addr option counts the number of different
2312 recipients that someone has sent messages to in the last time period. It
2313 behaves like per_rcpt if all the recipient addresses are different, but
2314 duplicate recipient addresses do not increase the measured rate. Like
2315 the /count= option this is a general mechanism, so the per_addr option
2316 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2317 example, measure the rate that a client uses different sender addresses
2318 with the options per_mail/unique=$sender_address. There are further
2319 details in the main documentation.
2321 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2323 TF/06 Removed a few PCRE remnants.
2325 TF/07 Automatically extract Exim's version number from tags in the git
2326 repository when doing development or release builds.
2328 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2329 Bugzilla 879. Patch from Paul Fisher.
2331 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2332 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2335 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2337 PP/05 Variable $av_failed, true if the AV scanner deferred.
2338 Bugzilla 1078. Patch from John Horne.
2340 PP/06 Stop make process more reliably on build failure.
2341 Bugzilla 1087. Patch from Heiko Schlittermann.
2343 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2344 Bugzilla 1089. Patch from Heiko Schlittermann.
2346 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2347 Bugzilla 1131. Patch from Holger Weiß.
2349 PP/09 Handle IPv6 addresses with SPF.
2350 Bugzilla 860. Patch from Wolfgang Breyha.
2352 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2354 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2357 PP/11 match_* no longer expand right-hand-side by default.
2358 New compile-time build option, EXPAND_LISTMATCH_RHS.
2359 New expansion conditions, "inlist", "inlisti".
2361 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2363 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2365 PP/14 fix log_write() format string regression from TF/03.
2366 Bugzilla 1152. Patch from Dmitry Isaikin.
2372 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2374 PP/02 Harmonised TLS library version reporting; only show if debugging.
2375 Layout now matches that introduced for other libraries in 4.74 PP/03.
2377 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2379 PP/04 New "dns_use_edns0" global option.
2381 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2384 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2385 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2387 TK/01 Updated PolarSSL code to 0.14.2.
2388 Bugzilla 1097. Patch from Andreas Metzler.
2390 PP/07 Catch divide-by-zero in ${eval:...}.
2391 Fixes bugzilla 1102.
2393 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2396 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2397 format-string attack -- SECURITY: remote arbitrary code execution.
2399 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2400 time unintentionally subject to list matching rules, letting the header
2401 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2402 arbitrary string expansion). This allowed for information disclosure.
2404 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2405 INT_MIN/-1 -- value coerced to INT_MAX.
2411 NM/01 Workaround for PCRE version dependency in version reporting
2414 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2415 This fixes portability to compilers other than gcc, notably
2416 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2418 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2419 makefiles for portability to HP-UX and POSIX correctness.
2421 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2422 Also via indented variable definition in the Makefile.
2423 (Debugging by Oliver Heesakkers).
2425 PP/02 Restore caching of spamd results with expanded spamd_address.
2426 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2428 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2429 Improves build reliability. Fix from: Frank Elsner
2431 NM/02 Fix wide character breakage in the rfc2047 coding
2432 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2434 NM/03 Allow underscore in dnslist lookups
2435 Fixes bug 1026. Patch from Graeme Fowler
2437 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2438 Code patches from Adam Ciarcinski of NetBSD.
2440 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2443 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2444 is logged, to avoid truncation. Patch from John Horne.
2446 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2447 Patch from Jakob Hirsch.
2449 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2450 SQL string expansion failure details.
2451 Patch from Andrey Oktyabrski.
2453 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2454 Patch from Simon Arlott.
2456 PP/09 New lookups functionality failed to compile on old gcc which rejects
2457 extern declarations in function scope.
2458 Patch from Oliver Fleischmann
2460 PP/10 Use sig_atomic_t for flags set from signal handlers.
2461 Check getgroups() return and improve debugging.
2462 Fixed developed for diagnosis in bug 927 (which turned out to be
2465 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2466 Patch from Mark Zealey.
2468 PP/12 Bugzilla 1056: Improved spamd server selection.
2469 Patch from Mark Zealey.
2471 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2472 Based on patch from Heiko Schlittermann.
2474 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2475 Patch from Uwe Doering, sign-off by Michael Haardt.
2477 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2478 variable declaration deep within a block. Bug and patch from
2481 PP/15 lookups-Makefile IRIX compatibility coercion.
2483 PP/16 Make DISABLE_DKIM build knob functional.
2485 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2486 Patch by Simon Arlott
2488 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2489 variable argument macros. Our copy now differs from upstream.
2495 TF/01 Failure to get a lock on a hints database can have serious
2496 consequences so log it to the panic log.
2498 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2499 controlled using the smtp_confirmation log selector.
2501 TF/03 Include the error message when we fail to unlink a spool file.
2503 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2504 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2505 for maintaining out-of-tree patches for some time.
2507 PP/01 Bugzilla 139: Documentation and portability issues.
2508 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2509 Handle per-OS dynamic-module compilation flags.
2511 PP/02 Let /dev/null have normal permissions.
2512 The 4.73 fixes were a little too stringent and complained about the
2513 permissions on /dev/null. Exempt it from some checks.
2514 Reported by Andreas M. Kirchwitz.
2516 PP/03 Report version information for many libraries, including
2517 Exim version information for dynamically loaded libraries. Created
2518 version.h, now support a version extension string for distributors
2519 who patch heavily. Dynamic module ABI change.
2521 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2522 privilege escalation vulnerability whereby the Exim run-time user
2523 can cause root to append content of the attacker's choosing to
2526 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2529 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2530 If dropping privileges for untrusted macros, we disabled normal logging
2531 on the basis that it would fail; for the Exim run-time user, this is not
2532 the case, and it resulted in successful deliveries going unlogged.
2533 Fixed. Reported by Andreas Metzler.
2539 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2540 only prepend for the Resent-* case. Fixes regression introduced in
2541 Exim 4.70 by NM/22 for Bugzilla 607.
2543 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2544 increasing numbers of administrators be bitten by this.
2546 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2548 PP/03 Bugzilla 994: added openssl_options main configuration option.
2550 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2552 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2554 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2556 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2557 without a peer certificate, leading to a segfault because of an
2558 assumption that peers always have certificates. Be a little more
2559 paranoid. Problem reported by Martin Tscholak.
2561 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2562 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2563 NB: ClamAV planning to remove STREAM in "middle of 2010".
2564 CL also introduces -bmalware, various -d+acl logging additions and
2565 more caution in buffer sizes.
2567 PP/09 Implemented reverse_ip expansion operator.
2569 PP/10 Bugzilla 937: provide a "debug" ACL control.
2571 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2573 PP/12 Bugzilla 973: Implement --version.
2575 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2577 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2579 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2581 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2582 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2583 ignore trailing whitespace.
2585 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2587 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2590 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2591 Notification from Dr Andrew Aitchison.
2593 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2594 ExtendedDetectionInfo response format.
2595 Notification from John Horne.
2597 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2600 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2601 XSL and documented dependency on system catalogs, with examples of how
2604 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2607 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2608 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2609 configuration file which is writeable by the Exim user or group.
2611 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2612 of configuration files to cover files specified with the -C option if
2613 they are going to be used with root privileges, not just the default
2616 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2617 option (effectively making it always true).
2619 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2620 files to be used while preserving root privileges.
2622 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2623 that rogue child processes cannot use them.
2625 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2626 run-time user, instead of root.
2628 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2629 Exim run-time user without dropping privileges.
2631 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2632 result string, instead of calling string_vformat() twice with the same
2635 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2636 for other users. Others should always drop root privileges if they use
2637 -C on the command line, even for a whitelisted configure file.
2639 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2641 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2647 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2648 $data_path, and $header_path variables; fixed documentation bugs and
2651 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2652 exipick to access non-standard spools, including the "frozen" queue
2655 NM/01 Bugzilla 965: Support mysql stored procedures.
2656 Patch from Alain Williams
2658 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2660 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2661 Patch from Andreas Metzler
2663 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2664 Patch from Kirill Miazine
2666 NM/05 Bugzilla 671: Added umask to procmail example.
2668 JJ/03 installed exipick 20100323.0, fixing doc bug
2670 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2671 directory. Notification and patch from Dan Rosenberg.
2673 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2675 TK/02 Improve log output when DKIM signing operation fails.
2677 MH/01 Treat the transport option dkim_domain as a colon separated
2678 list, not as a single string, and sign the message with each element,
2679 omitting multiple occurences of the same signer.
2681 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2682 Bugzilla 985, 986. Patch by Simon Arlott
2684 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2685 Patch by Simon Arlott
2687 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2688 MBX locking. Notification from Dan Rosenberg.
2694 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2696 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2698 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2700 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2702 NM/04 Bugzilla 915: Flag broken perl installation during build.
2708 TK/01 Added patch by Johannes Berg that expands the main option
2709 "spamd_address" if it starts with a dollar sign.
2711 TK/02 Write list of recipients to X-Envelope-Sender header when building
2712 the mbox-format spool file for content scanning (suggested by Jakob
2715 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2716 (http://www.dcc-servers.net/) support via dccifd. Activated by
2717 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2719 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2720 by Mark Daniel Reidel <mr@df.eu>.
2722 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2723 When building exim an external PCRE library is now needed -
2724 PCRE is a system library on the majority of modern systems.
2725 See entry on PCRE_LIBS in EDITME file.
2727 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2728 conversation. Added nologin parameter to request.
2729 Patch contributed by Kirill Miazine.
2731 TF/01 Do not log submission mode rewrites if they do not change the address.
2733 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2735 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2736 log files in place. Contributed by Roberto Lima.
2738 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2740 TF/03 Bugzilla 615: When checking the local_parts router precondition
2741 after a local_part_suffix or local_part_prefix option, Exim now
2742 does not use the address's named list lookup cache, since this
2743 contains cached lookups for the whole local part.
2745 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2746 Robert Millan. Documentation is in experimental-spec.txt.
2748 TF/04 Bugzilla 668: Fix parallel build (make -j).
2750 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2752 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2753 Patch provided by Jan Srzednicki.
2755 TF/05 Leading white space used to be stripped from $spam_report which
2756 wrecked the formatting. Now it is preserved.
2758 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2759 that they are available at delivery time.
2761 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2763 TF/08 TLS error reporting now respects the incoming_interface and
2764 incoming_port log selectors.
2766 TF/09 Produce a more useful error message if an SMTP transport's hosts
2767 setting expands to an empty string.
2769 NM/06 Bugzilla 744: EXPN did not work under TLS.
2770 Patch provided by Phil Pennock.
2772 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2773 Patch provided by Richard Godbee.
2775 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2776 acl_smtp_notquit, added index entry.
2778 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2779 Patch provided by Eugene Bujak.
2781 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2782 accept(). Patch provided by Maxim Dounin.
2784 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2785 Patch provided by Phil Pennock.
2787 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2789 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2790 Patch provided by Brad "anomie" Jorsch.
2792 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2793 Patch provided by Dean Brooks.
2795 TK/05 Add native DKIM support (does not depend on external libraries).
2797 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2798 Patch provided by Graeme Fowler.
2800 NM/16 Bugzilla 851: Documentation example syntax fix.
2802 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2804 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2807 NM/19 Bugzilla 745: TLS version reporting.
2808 Patch provided by Phil Pennock.
2810 NM/20 Bugzilla 167: bool: condition support.
2811 Patch provided by Phil Pennock.
2813 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2814 clients. Patch provided by Phil Pennock.
2816 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2817 Patch provided by Brad "anomie" Jorsch.
2819 NM/23 Bugzilla 687: Fix misparses in eximstats.
2820 Patch provided by Heiko Schlittermann.
2822 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2823 Patch provided by Heiko Schlittermann.
2825 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2826 plus update to original patch.
2828 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2830 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2831 Patch provided by David Brownlee.
2833 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2835 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2837 NM/30 Bugzilla 888: TLS documentation bugfixes.
2839 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2841 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2842 Unlike the original bugzilla I have changed all shell scripts in src tree.
2844 NM/33 Bugzilla 898: Transport filter timeout fix.
2845 Patch by Todd Rinaldo.
2847 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2848 Patch by Serge Demonchaux.
2850 NM/35 Bugzilla 39: Base64 decode bug fixes.
2851 Patch by Jakob Hirsch.
2853 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2855 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2857 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2859 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2865 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2866 ALT-N's libdkim that I have put here:
2867 http://duncanthrax.net/exim-experimental/
2869 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2870 were called 'true' and it seems that C99 defines that as a reserved
2871 keyword to be used with 'bool' variable types. That means you could
2872 not include C99-style headers which use bools without triggering
2873 build errors in sieve.c.
2875 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2876 as mailq or other aliases. Changed the --help handling significantly
2877 to do whats expected. exim_usage() emits usage/help information.
2879 SC/01 Added the -bylocaldomain option to eximstats.
2881 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2883 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2885 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2891 PH/01 Another patch from the Sieve maintainer.
2893 PH/02 When an IPv6 address is converted to a string for single-key lookup
2894 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2895 dots are used instead of colons so that keys in lsearch files need not
2896 contain colons. This was done some time before quoting was made available
2897 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2898 (notated using the quote facility) so as to distinguish them from IPv4
2899 keys. This meant that lookups for IP addresses in host lists did not work
2900 for iplsearch lookups.
2902 This has been fixed by arranging for IPv6 addresses to be expressed with
2903 colons if the lookup type is iplsearch. This is not incompatible, because
2904 previously such lookups could never work.
2906 The situation is now rather anomalous, since one *can* have colons in
2907 ordinary lsearch keys. However, making the change in all cases is
2908 incompatible and would probably break a number of configurations.
2910 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2913 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2914 conversion specification without a maximum field width, thereby enabling
2915 a rogue spamd server to cause a buffer overflow. While nobody in their
2916 right mind would setup Exim to query an untrusted spamd server, an
2917 attacker that gains access to a server running spamd could potentially
2918 exploit this vulnerability to run arbitrary code as the Exim user.
2920 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2921 $primary_hostname instead of what libspf2 thinks the hosts name is.
2923 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2924 a directory entry by the name of the lookup key. Previously, if a
2925 symlink pointed to a non-existing file or a file in a directory that
2926 Exim lacked permissions to read, a lookup for a key matching that
2927 symlink would fail. Now it is enough that a matching directory entry
2928 exists, symlink or not. (Bugzilla 503.)
2930 PH/03 The body_linecount and body_zerocount variables are now exported in the
2933 PH/04 Added the $dnslist_matched variable.
2935 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2936 This means they are set thereafter only if the connection becomes
2939 PH/06 Added the client_condition to authenticators so that some can be skipped
2940 by clients under certain conditions.
2942 PH/07 The error message for a badly-placed control=no_multiline_responses left
2943 "_responses" off the end of the name.
2945 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2947 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2948 (without spaces) instead of just copying the configuration text.
2950 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2952 PH/11 Added $max_received_linelength.
2954 PH/12 Added +ignore_defer and +include_defer to host lists.
2956 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2957 way in which PCRE > 7.0 is built.
2959 PH/14 Implemented queue_only_load_latch.
2961 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2962 MAIL command. The effect was to mangle the value on 64-bit systems.
2964 PH/16 Another patch from the Sieve maintainer.
2966 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2968 PH/18 If a system quota error occurred while trying to create the file for
2969 a maildir delivery, the message "Mailbox is full" was not appended to the
2970 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2971 only to a quota excession during the actual writing of the file.
2973 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2974 characters?) which causes problems in log lines. The DN values are now
2975 passed through string_printing() before being added to log lines.
2977 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2978 and InterBase are left for another time.)
2980 PH/21 Added message_body_newlines option.
2982 PH/22 Guard against possible overflow in moan_check_errorcopy().
2984 PH/23 POSIX allows open() to be a macro; guard against that.
2986 PH/24 If the recipient of an error message contained an @ in the local part
2987 (suitably quoted, of course), incorrect values were put in $domain and
2988 $local_part during the evaluation of errors_copy.
2994 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2995 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2998 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2999 issue a MAIL command.
3001 PH/02 In an ACL statement such as
3003 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3005 if a client was not listed at all, or was listed with a value other than
3006 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3007 the condition was not true (as it should be), so access was not denied.
3008 The bug was that the ! inversion was incorrectly passed on to the second
3009 item. This has been fixed.
3011 PH/03 Added additional dnslists conditions == and =& which are different from
3012 = and & when the dns lookup returns more than one IP address.
3014 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3015 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3017 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3018 FSYNC, which compiles an option called disable_fsync that allows for
3019 bypassing fsync(). The documentation is heavily laced with warnings.
3021 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3023 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3024 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3025 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3026 including adding "make clean"; (3) Added -fPIC when compiling the test
3027 dynamically loaded module, to get rid of a warning.
3029 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3030 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3031 = 0s. The bug is otherwise harmless.
3033 PH/07 There was a bug in the dovecot authenticator such that the value of
3034 $auth1 could be overwritten, and so not correctly preserved, after a
3035 successful authentication. This usually meant that the value preserved by
3036 the server_setid option was incorrect.
3038 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3040 PH/09 Installed PCRE release 7.0.
3042 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3043 run for batched SMTP input. It is now run at the start of every message
3044 in the batch. While fixing this I discovered that the process information
3045 (output by running exiwhat) was not always getting set for -bs and -bS
3046 input. This is fixed, and it now also says "batched" for BSMTP.
3048 PH/11 Added control=no_pipelining.
3050 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3051 patch, slightly modified), and move the expansion of helo_data till after
3052 the connection is made in the smtp transport (so it can use these
3055 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3057 PH/14 Added log_selector = +pid.
3059 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3061 PH/16 Add ${if forany and ${if forall.
3063 PH/17 Added dsn_from option to vary the From: line in DSNs.
3065 PH/18 Flush SMTP output before performing a callout, unless control =
3066 no_callout_flush is set.
3068 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3069 was true (the default) a successful delivery failed to delete the retry
3070 item, thus causing premature timeout of the address. The bug is now
3073 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3075 PH/21 Long custom messages for fakedefer and fakereject are now split up
3076 into multiline responses in the same way that messages for "deny" and
3077 other ACL rejections are.
3079 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3080 with slight modification.
3082 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3083 draft, changing the syntax and factoring some duplicate code".
3085 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3086 for deliveries of the second and subsequent messages over the same SMTP
3089 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3090 ${reduce, with only minor "tidies".
3092 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3094 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3095 expansion side effects.
3097 PH/27 When a message times out after an over-quota error from an Exim-imposed
3098 quota, the bounce message says "mailbox is full". This message was not
3099 being given when it was a system quota that was exceeded. It now should
3102 MH/03 Made $recipients available in local_scan(). local_scan() already has
3103 better access to the recipient list through recipients_list[], but
3104 $recipients can be useful in postmaster-provided expansion strings.
3106 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3107 in the case of a MAIL command with additional options following the
3108 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3109 were accidentally chopped off.
3111 PH/29 SMTP synchronization checks are implemented when a command is read -
3112 there is a check that no more input is waiting when there shouldn't be
3113 any. However, for some commands, a delay in an ACL can mean that it is
3114 some time before the response is written. In this time, more input might
3115 arrive, invalidly. So now there are extra checks after an ACL has run for
3116 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3117 pipelining has not been advertised.
3119 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3121 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3122 returned in the SMTP response when smtp_return_error_details was false.
3123 This has been fixed.
3125 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3126 instead of the C I/O that was originally supplied, because problems were
3127 reported on Solaris.
3129 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3130 Exim which did not show up earlier: it was assuming that a call to
3131 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3132 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3133 was a macro that became an assignment, so it seemed to work. This has
3134 changed to a proper function call with a void return, hence the compile
3135 error. Exim's code has been fixed.
3137 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3140 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3142 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3144 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3145 because it turns out to be too big or there is a timeout) there is no
3146 "Completed" line in the log. When some messages of this type were
3147 selected by exigrep, they were listed as "not completed". Others were
3148 picked up by some special patterns. I have improved the selection
3149 criteria to be more general.
3151 PH/38 The host_find_failed option in the manualroute router can now be set
3152 to "ignore", to completely ignore a host whose IP address cannot be
3153 found. If all hosts are ignored, the behaviour is controlled by the new
3154 host_all_ignored option.
3156 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3157 homing or because of multiple MX records with /mx) generated more than
3158 one IP address, and the following item turned out to be the local host,
3159 all the secondary addresses of the first item were incorrectly removed
3160 from the list, along with the local host and any following hosts (which
3161 is what is supposed to happen).
3163 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3164 whoever called Exim into the -H file. In the case of the daemon it was
3165 behaving confusingly. When first started, it used values for whoever
3166 started the daemon, but after a SIGHUP it used the Exim user (because it
3167 calls itself on a restart). I have changed the code so that it now always
3170 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3171 message are rejected with the same error (e.g. no authentication or bad
3172 sender address), and a DATA command is nevertheless sent (as can happen
3173 with PIPELINING or a stupid MUA), the error message that was given to the
3174 RCPT commands is included in the rejection of the DATA command. This is
3175 intended to be helpful for MUAs that show only the final error to their
3178 PH/42 Another patch from the Sieve maintainer.
3180 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3181 Eximstats - Fixed some broken HTML links and added missing column headers
3183 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3184 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3186 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3188 PH/43 Yet another patch from the Sieve maintainer.
3190 PH/44 I found a way to check for a TCP/IP connection going away before sending
3191 the response to the final '.' that terminates a message, but only in the
3192 case where the client has not sent further data following the '.'
3193 (unfortunately, this is allowed). However, in many cases there won't be
3194 any further data because there won't be any more messages to send. A call
3195 to select() can be used: if it shows that the input is "ready", there is
3196 either input waiting, or the socket has been closed. An attempt to read
3197 the next input character can distinguish the two cases. Previously, Exim
3198 would have sent an OK response which the client would never have see.
3199 This could lead to message repetition. This fix should cure that, at
3200 least in a lot of common cases.
3202 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3203 advertised in response to EHLO.
3209 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3210 fixed by 4.65/MH/01 (is this a record?) are fixed:
3212 (i) An empty string was always treated as zero by the numeric comparison
3213 operators. This behaviour has been restored.
3215 (ii) It is documented that the numeric comparison operators always treat
3216 their arguments as decimal numbers. This was broken in that numbers
3217 starting with 0 were being interpreted as octal.
3219 While fixing these problems I realized that there was another issue that
3220 hadn't been noticed. Values of message_size_limit (both the global option
3221 and the transport option) were treated as octal if they started with 0.
3222 The documentation was vague. These values are now always treated as
3223 decimal, and I will make that clear in the documentation.
3229 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3230 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3233 MH/01 Don't check that the operands of numeric comparison operators are
3234 integers when their expansion is in "skipping" mode (fixes bug
3235 introduced by 4.64-PH/07).
3237 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3238 child addresses, Exim now panics and dies. Previously, because the count
3239 is held in a short int, deliveries were likely to be lost. As such a
3240 large number of recipients for a single message is ridiculous
3241 (performance will be very, very poor), I have chosen to impose a limit
3242 rather than extend the field.
3248 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3249 leftover -K file (the existence of which was triggered by #402).
3250 While we were at it, introduced process PID as part of the -K
3251 filename. This should rule out race conditions when creating
3254 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3255 processing considerably. Previous code took too long for large mails,
3256 triggering a timeout which in turn triggers #401.
3258 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3259 in the DK code in transports.c. sendfile() is not really portable,
3260 hence the _LINUX specificness.
3262 TF/01 In the add_headers option to the mail command in an Exim filter,
3263 there was a bug that Exim would claim a syntax error in any
3264 header after the first one which had an odd number of characters
3267 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3268 callout verification, Exim cached a "reject" for the entire domain. This
3269 is correct for most verifications, but it is not correct for a recipient
3270 verification with use_sender or use_postmaster set, because in that case
3271 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3272 case of MAIL FROM:<> rejection from other early rejections (e.g.
3273 rejection of HELO). When verifying a recipient using a non-null MAIL
3274 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3275 Whatever the result of the callout, the value of the domain cache is
3276 left unchanged (for any other kind of callout, getting as far as trying
3277 RCPT means that the domain itself is ok).
3279 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3282 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3283 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3284 session, the daemon ignores ECONNECT errors and logs others; it now
3285 ignores EPIPE as well.
3287 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3288 (quoted-printable decoding).
3290 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3291 later the small subsequent patch to fix an introduced bug.
3293 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3295 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3297 PH/08 An error is now given if message_size_limit is specified negative.
3299 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3300 to be given (somewhat) arbitrary names.
3302 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3305 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3306 miscellaneous code fixes
3308 PH/10 Added the log_reject_target ACL modifier to specify where to log
3311 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3312 hostname. This is wrong, because it relates to the incoming message (and
3313 probably the interface on which it is arriving) and not to the outgoing
3314 callout (which could be using a different interface). This has been
3315 changed to use the value of the helo_data option from the smtp transport
3316 instead - this is what is used when a message is actually being sent. If
3317 there is no remote transport (possible with a router that sets up host
3318 addresses), $smtp_active_hostname is used.
3320 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3321 tweaks were necessary in order to get it to work (see also 21 below):
3322 (a) The code assumed that strncpy() returns a negative number on buffer
3323 overflow, which isn't the case. Replaced with Exim's string_format()
3325 (b) There were several signed/unsigned issues. I just did the minimum
3326 hacking in of casts. There is scope for a larger refactoring.
3327 (c) The code used strcasecmp() which is not a standard C function.
3328 Replaced with Exim's strcmpic() function.
3329 (d) The code set only $1; it now sets $auth1 as well.
3330 (e) A simple test gave the error "authentication client didn't specify
3331 service in request". It would seem that Dovecot has changed its
3332 interface. Fortunately there's a specification; I followed it and
3333 changed what the client sends and it appears to be working now.
3335 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3338 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3339 address A is aliased to B and C, where B exists and C does not. Without
3340 -v the output is "A verified" because verification stops after a
3341 successful redirection if more than one address is generated. However,
3342 with -v the child addresses are also verified. Exim was outputting "A
3343 failed to verify" and then showing the successful verification for C,
3344 with its parentage. It now outputs "B failed to verify", showing B's
3345 parentage before showing the successful verification of C.
3347 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3348 look up a TXT record in a specific list after matching in a combined
3351 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3352 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3353 they consult the DNS. I had assumed they would set it the way they
3354 wanted; and indeed my experiments on Linux seem to show that in some
3355 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3356 To be on the safe side, however, I have now made the interface to
3357 host_find_byname() similar to host_find_bydns(), with an argument
3358 containing the DNS resolver options. The host_find_byname() function now
3359 sets these options at its start, just as host_find_bydns() does. The smtp
3360 transport options dns_qualify_single and dns_search_parents are passed to
3361 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3362 of host_find_byname() use the default settings of RES_DEFNAMES
3363 (qualify_single) but not RES_DNSRCH (search_parents).
3365 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3366 spool_read_header() do less string testing, by means of a preliminary
3367 switch on the second character of optional "-foo" lines. (This is
3368 overdue, caused by the large number of possibilities that now exist.
3369 Originally there were few.) While I was there, I also converted the
3370 str(n)cmp tests so they don't re-test the leading "-" and the first
3371 character, in the hope this might squeeze out yet more improvement.
3373 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3374 flag allowing group syntax was set by the header_syntax check but not
3375 turned off, possible causing trouble later; (2) The flag was not being
3376 set at all for the header_verify test, causing "group"-style headers to
3377 be rejected. I have now set it in this case, and also caused header_
3378 verify to ignore an empty address taken from a group. While doing this, I
3379 came across some other cases where the code for allowing group syntax
3380 while scanning a header line wasn't quite right (mostly, not resetting
3381 the flag correctly in the right place). These bugs could have caused
3382 trouble for malformed header lines. I hope it is now all correct.
3384 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3385 with the "reply" argument non-NULL. The code, however (which originally
3386 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3387 but it didn't always do it. This confused somebody who was copying the
3388 code for some other use. I have removed all the tests.
3390 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3391 feature that was used to support insecure browsers during the U.S. crypto
3392 embargo. It requires special client support, and Exim is probably the
3393 only MTA that supported it -- and would never use it because real RSA is
3394 always available. This code has been removed, because it had the bad
3395 effect of slowing Exim down by computing (never used) parameters for the
3396 RSA_EXPORT functionality.
3398 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3399 authenticator to fail if there's a tab character in the incoming data
3400 (there should never be unless someone is messing about, as it's supposed
3401 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3402 if the connection is using TLS or if the remote IP is the same as the
3403 local IP, and the "valid-client-cert option" if a client certificate has
3406 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3407 authenticators. This can be used for authorization after authentication
3408 succeeds. (In the case of plaintext, it servers for both authentication
3411 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3412 if any retry times were supplied.
3414 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3415 connection, where there is no client IP address to check. In this
3416 situation, the verify now always succeeds.
3418 PH/25 Applied John Jetmore's -Mset patch.
3420 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3422 PH/27 In a string expansion for a processed (not raw) header when multiple
3423 headers of the same name were present, leading whitespace was being
3424 removed from all of them, but trailing whitespace was being removed only
3425 from the last one. Now trailing whitespace is removed from each header
3426 before concatenation. Completely empty headers in a concatenation (as
3427 before) are ignored.
3429 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3430 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3432 PH/29 [Removed. This was a change that I later backed out, and forgot to
3433 correct the ChangeLog entry (that I had efficiently created) before
3434 committing the later change.]
3436 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3437 address errors (4xx response to RCPT) over the same connection as other
3438 messages routed to the same hosts. Such deliveries are always "forced",
3439 so retry times are not inspected. This resulted in far too many retries
3440 for the affected addresses. The effect occurred only when there were more
3441 hosts than the hosts_max_try setting in the smtp transport when it had
3442 the 4xx errors. Those hosts that it had tried were not added to the list
3443 of hosts for which the message was waiting, so if all were tried, there
3444 was no problem. Two fixes have been applied:
3446 (i) If there are any address or message errors in an SMTP delivery, none
3447 of the hosts (tried or untried) are now added to the list of hosts
3448 for which the message is waiting, so the message should not be a
3449 candidate for sending over the same connection that was used for a
3450 successful delivery of some other message. This seems entirely
3451 reasonable: after all the message is NOT "waiting for some host".
3452 This is so "obvious" that I'm not sure why it wasn't done
3453 previously. Hope I haven't missed anything, but it can't do any
3454 harm, as the worst effect is to miss an optimization.
3456 (ii) If, despite (i), such a delivery is accidentally attempted, the
3457 routing retry time is respected, so at least it doesn't keep
3458 hammering the server.
3460 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3461 in ${readsocket because some servers need this prod.
3463 PH/32 Added some extra debug output when updating a wait-xxx database.
3465 PH/33 The hint "could be header name not terminated by colon", which has been
3466 given for certain expansion errors for a long time, was not being given
3467 for the ${if def:h_colon_omitted{... case.
3469 PH/34 The spec says: "With one important exception, whenever a domain list is
3470 being scanned, $domain contains the subject domain." There was at least
3471 one case where this was not true.
3473 PH/35 The error "getsockname() failed: connection reset by peer" was being
3474 written to the panic log as well as the main log, but it isn't really
3475 panic-worthy as it just means the connection died rather early on. I have
3476 removed the panic log writing for the ECONNRESET error when getsockname()
3479 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3480 runs only) independently of the message's sender address. This meant
3481 that, if the 4xx error was in fact related to the sender, a different
3482 message to the same recipient with a different sender could confuse
3483 things. In particular, this can happen when sending to a greylisting
3484 server, but other circumstances could also provoke similar problems.
3485 I have changed the default so that the retry time for these errors is now
3486 based a combination of the sender and recipient addresses. This change
3487 can be overridden by setting address_retry_include_sender=false in the
3490 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3491 remote server are returned as part of bounce messages. This was not
3492 happening for LMTP over a pipe (the lmtp transport), but now it is the
3493 same for both kinds of LMTP.
3495 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3496 in header lines that were in fact CNAMEs. This is no longer the case.
3498 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3499 and queue runs started by the daemon processed all messages. This has
3500 been fixed so that -R and -S can now usefully be given with -q<time>.
3502 PH/40 Import PCRE release 6.7 (fixes some bugs).
3504 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3506 PH/42 Give an error if -q is specified more than once.
3508 PH/43 Renamed the variables $interface_address and $interface_port as
3509 $received_ip_address and $received_port, to make it clear that these
3510 values apply to message reception, and not to the outgoing interface when
3511 a message is delivered. (The old names remain recognized, of course.)
3513 PH/44 There was no timeout on the connect() call when using a Unix domain
3514 socket in the ${readsocket expansion. There now is.
3516 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3517 be meaningful with "accept".
3519 SC/01 Eximstats V1.43
3520 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3522 SC/02 Eximstats V1.44
3523 Use a glob alias rather than an array ref in the generated
3524 parser. This improves both readability and performance.
3526 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3527 Collect SpamAssassin and rejection statistics.
3528 Don't display local sender or destination tables unless
3529 there is data to show.
3530 Added average volumes into the top table text output.
3532 SC/04 Eximstats V1.46
3533 Collect data on the number of addresses (recipients)
3534 as well as the number of messages.
3536 SC/05 Eximstats V1.47
3537 Added 'Message too big' to the list of mail rejection
3538 reasons (thanks to Marco Gaiarin).
3540 SC/06 Eximstats V1.48
3541 Mainlog lines which have GMT offsets and are too short to
3542 have a flag are now skipped.
3544 SC/07 Eximstats V1.49 (Alain Williams)
3545 Added the -emptyok flag.
3547 SC/08 Eximstats V1.50
3548 Fixes for obtaining the IP address from reject messages.
3550 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3551 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3552 whitespace changes from 4.64-PH/27
3554 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3557 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3558 are found, allow negative numbers in numeric criteria)
3560 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3562 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3563 to match changes made in 4.64-PH/43
3565 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3567 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3568 "domain" is actually an IP address, and force a failure. This locks out
3569 those revolvers/nameservers that support "A-for-A" lookups, in
3570 contravention of the specifications.
3572 PH/48 When a host name was looked up from an IP address, and the subsequent
3573 forward lookup of the name timed out, the host name was left in
3574 $sender_host_name, contrary to the specification.
3576 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3577 restricted to single-key lookups, Exim was not diagnosing an error if
3578 * or *@ was used with a query-style lookup.
3580 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3582 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3583 long ago, but noone interested enough thought of it. Let's just say that
3584 the "1.1" means that there are some new functions that weren't there at
3585 some point in the past.
3587 PH/51 Error processing for expansion failure of helo_data from an smtp
3588 transport during callout processing was broken.
3590 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3591 tested/used via the -bh/-bhc/-bs options.
3593 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3594 bug, fixed in subsequent PCRE releases).
3596 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3597 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3599 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3605 SC/01 Use a glob alias rather than an array ref in eximstats generated
3606 parser. This improves both readability and performance.
3608 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3609 Don't display local sender or destination tables in eximstats unless
3610 there is data to show.
3611 Added average volumes into the eximstats top table text output.
3613 SC/03 Collect data on the number of addresses (recipients) as well
3614 as the number of messages in eximstats.
3616 TF/01 Correct an error in the documentation for the redirect router. Exim
3617 does (usually) call initgroups() when daemonizing.
3619 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3620 with consistent privilege compared to when running as a daemon.
3622 TF/03 Note in the spec that $authenticated_id is not set for local
3623 submissions from trusted users.
3625 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3626 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3628 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3629 by adding some example configuration directives to the default
3630 configuration file. A little bit of work is required to uncomment the
3631 directives and define how usernames and passwords are checked, but
3632 there is now a framework to start from.
3634 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3635 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3636 without this. I don't know how relevant this is to other LDAP libraries.
3638 PH/02 Add the verb name to the "unknown ACL verb" error.
3640 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3642 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3644 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3645 directory not expanded when it should be if an expanded home directory
3646 was set for the address (which is overridden by the transport).
3648 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3651 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3652 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3653 because it is too late at that time, and has no effect.
3655 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3656 security issue with \' (bugzilla #107). I could not use the
3657 PQescapeStringConn() function, because it needs a PGconn value as one of
3660 PH/08 When testing addresses using -bt, indicate those final addresses that
3661 are duplicates that would not cause an additional delivery. At least one
3662 person was confused, thinking that -bt output corresponded to deliveries.
3663 (Suppressing duplicates isn't a good idea as you lose the information
3664 about possibly different redirections that led to the duplicates.)
3666 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3667 systems where poll() doesn't work, in particular OS X.
3669 PH/10 Added more information to debugging output for retry time not reached.
3671 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3672 operations in malware.c.
3674 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3677 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3678 syslog_duplication set false, log lines that would normally be written
3679 both the the main log and to the reject log were not written to syslog at
3682 PH/14 In the default configuration, change the use of "message" in ACL warn
3683 statements to "add_header".
3685 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3686 not followed by a command (e.g. "seen endif").
3688 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3689 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3692 PH/17 Added extra conditions to the default value of delay_warning_condition
3696 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3697 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3698 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3701 The Auto-Submitted: and various List- headers are standardised, whereas I
3702 don't think Precedence: ever was.
3704 PH/18 Refactored debugging code in route_finduser() to show more information,
3705 in particular, the error code if getpwnam() issues one.
3707 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3708 This is apparently needed in addition to the PH/07 change above to avoid
3709 any possible encoding problems.
3711 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3712 but not after initializing Perl.
3714 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3715 output them only if debugging. By default they are written stderr,
3716 apparently, which is not desirable.
3718 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3721 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3724 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3726 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3727 authenticated or an ident call has been made. Suppress the default
3728 values for $authenticated_id and $authenticated_sender (but permit -oMai
3729 and -oMas) when testing with -bh.
3731 PH/24 Re-jigged the order of the tests in the default configuration so that the
3732 tests for valid domains and recipients precede the DNS black list and CSA
3733 tests, on the grounds that those ones are more expensive.
3735 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3736 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3737 This bug exists in every version of Exim that I still have, right back to
3740 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3741 However, an attempt to turn on case-sensitivity in a regex key by
3742 including (?-i) didn't work because the subject string was already
3743 lowercased, and the effects were non-intuitive. It turns out that a
3744 one-line patch can be used to allow (?-i) to work as expected.
3750 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3751 other effects) broke the use of negated acl sub-conditions.
3753 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3756 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3757 "Deny" causes Exim to reject the incoming connection with a 554 error.
3758 Unfortunately, if there is a major crisis, such as a disk failure,
3759 tcp-wrappers gives "deny", whereas what one would like would be some
3760 kind of temporary error. A kludge has been added to help with this.
3761 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3762 554 error is used if errno is still zero or contains ENOENT (which occurs
3763 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3766 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3768 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3769 errors. Otherwise a message that provokes a temporary error (when other
3770 messages do not) can cause a whole host to time out.
3772 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3773 addresses were routed directly to files or pipes from a redirect router.
3774 File deliveries just didn't batch; pipe deliveries might have suffered
3777 PH/06 A failure to get a lock for a hints database would erroneously always say
3778 "Failed to get write lock", even when it was really a read lock.
3780 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3781 of 0600. This has been changed to use the value of the lockfile_mode
3782 option (which defaults to 0600).
3784 PH/08 Applied small patch from the Sieve maintainer.
3786 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3787 folder from quota calculations, a direct delivery into this folder messed
3788 up the contents of the maildirsize file. This was because the regex was
3789 used only to exclude .Trash (or whatever) when the size of the mailbox
3790 was calculated. There was no check that a delivery was happening into an
3791 excluded directory. This bug has been fixed by ignoring all quota
3792 processing for deliveries into excluded directories.
3794 PH/10 Added the maildirfolder_create_regex option to appendfile.
3800 PH/01 The code for finding all the local interface addresses on a FreeBSD
3801 system running IPv6 was broken. This may well have applied to all BSD
3802 systems, as well as to others that have similar system calls. The broken
3803 code found IPv4 interfaces correctly, but gave incorrect values for the
3804 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3805 that it would not match correctly against @[] and not recognize the IPv6
3808 PH/02 The ipliteral router was not recognizing addresses of the form user@
3809 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3811 PH/03 Added disable_ipv6.
3813 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3814 decoded form, because it is most often used to construct To: headers
3815 lines in autoreplies, and the decoded form may well be syntactically
3816 invalid. However, $reply_address has leading white space removed, and all
3817 newlines turned into spaces so that the autoreply transport does not
3820 PH/05 If group was specified without a user on a router, and no group or user
3821 was specified on a transport, the group from the router was ignored.
3823 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3824 for visible compile-time settings that can be used to change these
3825 numbers, for those that want even more. Backwards compatibility with old
3826 spool files has been maintained. However, going back to a previous Exim
3827 release will lost any variables that are in spool files.
3829 PH/07 Two small changes when running in the test harness: increase delay when
3830 passing a TCP/IP connection to a new process, in case the original
3831 process has to generate a bounce, and remove special handling of
3832 127.0.0.2 (sic), which is no longer necessary.
3834 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3835 be the same on different OS.
3837 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3840 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3841 whether --show-vars was specified or not
3843 JJ/02 exipick: Added support for new ACL variable spool format introduced
3846 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3847 syntactically invalid From: or Reply-to: line, and a filter used this to
3848 generate an autoreply, and therefore failed to obtain an address for the
3849 autoreply, Exim could try to deliver to a non-existent relative file
3850 name, causing unrelated and misleading errors. What now happens is that
3851 it logs this as a hard delivery error, but does not attempt to create a
3854 PH/11 The exinext utility has a -C option for testing purposes, but although
3855 the given file was scanned by exinext itself; it wasn't being passed on
3856 when Exim was called.
3858 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3859 an end-of-file indication when reading a command response.
3861 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3862 compiled. In many other places in Exim, IPv6 addresses are always
3863 recognized, so I have changed this. It also means that IPv4 domain
3864 literals of the form [IPV4:n.n.n.n] are now always recognized.
3866 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3867 used if the router is not running as root, for example, when verifying at
3868 ACL time, or when using -bh. The debugging output from this situation was
3869 non-existent - all you got was a failure to exec. I have made two
3872 (a) Failures to set uid/gid, the current directory, or a process leader
3873 in a subprocess such as that created by queryprogram now generate
3874 suitable debugging output when -d is set.
3876 (b) The queryprogram router detects when it is not running as root,
3877 outputs suitable debugging information if -d is set, and then runs
3878 the subprocess without attempting to change uid/gid.
3880 PH/15 Minor change to Makefile for building test_host (undocumented testing
3883 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3884 additional section of a DNS packet that returns MX or SRV records.
3885 Instead, it always explicitly searches for A/AAAA records. This avoids
3886 major problems that occur when a DNS server includes only records of one
3887 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3888 fixed another bug: if SRV records were looked up and the corresponding
3889 address records were *not* found in the additional section, the port
3890 values from the SRV records were lost.
3892 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3893 using the correct key (the original address) when searching the retry
3894 rules in order to find which one to use for generating the retry hint.
3896 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3897 adding the default one. Similarly, if it contains a Reply-To: header, the
3898 errors_reply_to option, if set, is not used.
3900 PH/19 When calculating a retry time, Exim used to measure the "time since
3901 failure" by looking at the "first failed" field in the retry record. Now
3902 it does not use this if it is later than than the arrival time of the
3903 message. Instead it uses the arrival time. This makes for better
3904 behaviour in cases where some deliveries succeed, thus re-setting the
3905 "first failed" field. An example is a quota failure for a huge message
3906 when small messages continue to be delivered. Without this change, the
3907 "time since failure" will always be short, possible causing more frequent
3908 delivery attempts for the huge message than are intended.
3909 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3911 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3912 $1, $2, $3) because the numerical variables can be reset during some
3913 expansion items (e.g. "match"), thereby losing the authentication data.
3915 PH/21 Make -bV show the size of off_t variables so that the test suite can
3916 decide whether to run tests for quotas > 2G.
3918 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3919 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3920 filecount value is greater than 2G or if a quota value is greater than 2G
3921 on a system where the size of off_t is not greater than 4, a panic error
3924 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3925 never match. The debug and -bh output now contains an explicit error
3926 message indicating a malformed IPv4 address or mask.
3928 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3929 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3930 PH/23 above applies.
3932 PH/25 Do not write to syslog when running in the test harness. The only
3933 occasion when this arises is a failure to open the main or panic logs
3934 (for which there is an explicit test).
3936 PH/26 Added the /no_tell option to "control=freeze".
3938 PH/27 If a host name lookup failed very early in a connection, for example, if
3939 the IP address matched host_lookup and the reverse lookup yielded a name
3940 that did not have a forward lookup, an error message of the form "no IP
3941 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3942 could be logged. Now it outputs the IP address instead of "NULL".
3944 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3945 allows the sender and the authenticated sender to be set when
3946 submitting a message from within Exim. Since child_open_exim() is
3947 documented for local_scan(), the new function should be too.
3949 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3950 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3951 results in an empty string is now treated as unset.
3953 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3955 PH/31 Added sender_verify_fail logging option.
3957 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3958 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3959 tidied the source and removed it altogether.
3961 PH/33 When a queue run was abandoned because the load average was too high, a
3962 log line was always written; now it is written only if the queue_run log
3963 selector is set. In addition, the log line for abandonment now contains
3964 information about the queue run such as the pid. This is always present
3965 in "start" and "stop" lines but was omitted from the "abandon" line.
3967 PH/34 Omit spaces between a header name and the colon in the error message that
3968 is given when verify = headers_syntax fails (if there are lots of them,
3969 the message gets confusing).
3971 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3972 names, as there are now some PTR records that contain slashes. This check
3973 is only to protect against broken name servers that fall over on strange
3974 characters, so the fact that it applies to all lookups doesn't matter.
3976 PH/36 Now that the new test suite is complete, we can remove some of the
3977 special code in Exim that was needed for the old test suite. For example,
3978 sorting DNS records because real resolvers return them in an arbitrary
3979 order. The new test suite's fake resolver always returns records in the
3982 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3983 bounces) except when queue_only is set, to avoid logging races between
3984 the different processes.
3986 PH/38 Panic-die if .include specifies a non-absolute path.
3988 PH/39 A tweak to the "H" retry rule from its user.
3990 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3991 a label. They prevented compilation on older perls.
3993 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3994 a warning to be raised on newish perls.
3996 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3997 on queue. Changes to match documented behaviour of showing count of
3998 messages matching specified criteria.
4000 PH/40 Changed the default ident timeout from 30s to 5s.
4002 PH/41 Added support for the use of login_cap features, on those BSD systems
4003 that have them, for controlling the resources used by pipe deliveries.
4005 PH/42 The content-scanning code uses fopen() to create files in which to put
4006 message data. Previously it was not paying any attention to the mode of
4007 the files. Exim runs with umask(0) because the rest of the code creates
4008 files with open(), and sets the required mode explicitly. Thus, these
4009 files were ending up world-writeable. This was not a big issue, because,
4010 being within the spool directory, they were not world-accessible. I have
4011 created a function called modefopen, which takes an additional mode
4012 argument. It sets umask(777), creates the file, chmods it to the required
4013 mode, then resets the umask. All the relevant calls to fopen() in the
4014 content scanning code have been changed to use this function.
4016 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4017 to 24 hours. This avoids potential overflow problems when processing G
4018 and H retry rules. I suspect nobody ever tinkers with this value.
4020 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4022 PH/45 When the plaintext authenticator is running as a client, the server's
4023 challenges are checked to ensure they are valid base64 strings. By
4024 default, the authentication attempt is cancelled if an invalid string is
4025 received. Setting client_ignore_invalid_base64 true ignores these errors.
4026 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4027 they are received. Thus, the responses can be made to depend on the
4028 challenges. If an invalid string is ignored, an empty string is placed in
4031 PH/46 Messages that are created by the autoreply transport now contains a
4032 References: header, in accordance with RFCs 2822 and 3834.
4034 PH/47 Added authenticated_sender_force to the smtp transport.
4036 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4038 PH/49 Installed latest patch from the Sieve maintainer.
4040 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4041 was also set, the appendfile transport was unnecessarily scanning a
4042 directory of message files (e.g. for maildir delivery) to find the count
4043 of files (along with the size), even though it did not need this
4044 information. It now does the scan only if it needs to find either the
4045 size of the count of files.
4047 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4049 PH/52 Two bugs concerned with error handling when the smtp transport is
4052 (i) Exim was not creating retry information for temporary errors given
4053 for individual recipients after the DATA command when the smtp transport
4054 was used in LMTP mode. This meant that they could be retried too
4055 frequently, and not timed out correctly.
4057 (ii) Exim was setting the flag that allows error details to be returned
4058 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4059 recipients that were returned after the DATA command.
4061 PH/53 This is related to PH/52, but is more general: for any failing address,
4062 when detailed error information was permitted to be returned to the
4063 sender, but the error was temporary, then after the final timeout, only
4064 "retry timeout exceeded" was returned. Now it returns the full error as
4065 well as "retry timeout exceeded".
4067 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4068 do this, and (what is worse) MTAs that accept it.
4070 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4071 will now be deprecated.
4073 PH/56 New os.c-cygwin from the Cygwin maintainer.
4075 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4076 formats (previously only available in exim formats via -bpr, -bpru,
4077 and -bpra. Now also available in native and exiqgrep formats)
4079 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4080 with very large, slow to parse queues
4082 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4084 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4086 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4087 responses to authentication challenges, though it was showing the
4088 challenges; (ii) I've removed the CR characters from the debug output for
4091 PH/58 Allow for the insertion of a newline as well as a space when a string
4092 is turned into more than one encoded-word during RFC 2047 encoding. The
4093 Sieve code now uses this.
4095 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4096 data_4xx, lost_connection, tls_required.
4098 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4099 message was being sent as an SMTP response.
4101 PH/61 Add -l and -k options to exicyclog.
4103 PH/62 When verifying, if an address was redirected to one new address, so that
4104 verification continued, and the new address failed or deferred after
4105 having set something in $address_data, the value of $address_data was not
4106 passed back to the ACL. This was different to the case when no
4107 redirection occurred. The value is now passed back in both cases.
4109 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4110 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4111 use login_cap.h, so on its own it isn't the distinguishing feature. The
4112 new name refers directly to the setclassresources() function.
4114 PH/65 Added configuration files for NetBSD3.
4116 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4118 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4119 is preferred over IPv4.
4121 PH/68 The bounce_return_message and bounce_return_body options were not being
4122 honoured for bounces generated during the reception of non-SMTP messages.
4123 In particular, this applied to messages rejected by the ACL. This bug has
4124 been fixed. However, if bounce_return_message is true and bounce_return_
4125 body is false, the headers that are returned for a non-SMTP message
4126 include only those that have been read before the error was detected.
4127 (In the case of an ACL rejection, they have all been read.)
4129 PH/69 The HTML version of the specification is now built in a directory called
4130 spec_html instead of spec.html, because the latter looks like a path with
4131 a MIME-type, and this confuses some software.
4133 PH/70 Catch two compiler warnings in sieve.c.
4135 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4136 function verify_get_ident() calls ip_connect() to connect a socket, but
4137 if the "connect()" function timed out, ip_connect() used to close the
4138 socket. However, verify_get_ident() also closes the socket later, and in
4139 between Exim writes to the log, which may get opened at this point. When
4140 the socket was closed in ip_connect(), the log could get the same file
4141 descriptor number as the socket. This naturally causes chaos. The fix is
4142 not to close the socket in ip_connect(); the socket should be closed by
4143 the function that creates it. There was only one place in the code where
4144 this was missing, in the iplookup router, which I don't think anybody now
4145 uses, but I've fixed it anyway.
4147 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4148 well as to direct DNS lookups. Otherwise the handling of names in host
4149 lists is inconsistent and therefore confusing.
4155 PH/01 Two changes to the default runtime configuration:
4157 (1) Move the checks for relay_from_hosts and authenticated clients from
4158 after to before the (commented out) DNS black list checks.
4160 (2) Add control=submission to the relay_from_hosts and authenticated
4161 clients checks, on the grounds that messages accepted by these
4162 statements are most likely to be submissions.
4164 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4166 (1) Generate an error if the third argument for the ${prvs expansion is
4169 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4172 (3) Reset the variables that are obtained from the first argument of
4173 ${prvscheck and used in the second argument before leaving the code,
4174 because their memory is reclaimed, so using them afterwards may do
4177 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4178 one (it's much easier than Tom thought :-).
4180 (5) Because of (4), we can now allow for the use of $prvscheck_result
4181 inside the third argument.
4183 PH/03 For some reason, the default setting of PATH when running a command from
4184 a pipe transport was just "/usr/bin". I have changed it to
4187 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4188 anything to be listed in the output from -bV.
4190 PH/05 When a filter generated an autoreply, the entire To: header line was
4191 quoted in the delivery log line, like this:
4193 => >A.N.Other <ano@some.domain> <original@ddress> ...
4195 This has been changed so that it extracts the operative address. There
4196 may be more than one such address. If so, they are comma-separated, like
4199 => >ano@some.domain,ona@other.domain <original@ddress> ...
4201 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4202 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4203 not being looked up in the rDNS to get a host name, Exim was showing the
4204 IP address twice in Received: lines, even though the IP addresses were
4205 identical. For example:
4207 Received: from [1.2.3.4] (helo=[1.2.3.4])
4209 However, if the real host name was known, it was omitting the HELO data
4210 if it matched the actual IP address. This has been tidied up so that it
4211 doesn't show the same IP address twice.
4213 PH/07 When both +timestamp and +memory debugging was on, the value given by
4214 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4215 called by the debug printing, thereby overwriting the timestamp buffer.
4216 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4218 PH/08 When the original message was included in an autoreply transport, it
4219 always said "this is a copy of the message, including all the headers",
4220 even if body_only or headers_only was set. It now gives an appropriate
4223 PH/09 Applied a patch from the Sieve maintainer which:
4225 o fixes some comments
4226 o adds the (disabled) notify extension core
4227 o adds some debug output for the result of if/elsif tests
4228 o points to the current vacation draft in the documentation
4229 and documents the missing references header update
4233 o fixes a bug in processing the envelope test (when testing
4234 multiple envelope elements, the last element determined the
4237 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4238 Electronic Mail") by including:
4240 Auto-submitted: auto-generated
4242 in the messages that it generates (bounce messages and others, such as
4243 warnings). In the case of bounce messages for non-SMTP messages, there was
4244 also a typo: it was using "Auto_submitted" (underscore instead of
4245 hyphen). Since every message generated by Exim is necessarily in response
4246 to another message, thes have all been changed to:
4248 Auto-Submitted: auto-replied
4250 in accordance with these statements in the RFC:
4252 The auto-replied keyword:
4254 - SHOULD be used on messages sent in direct response to another
4255 message by an automatic process,
4257 - MUST NOT be used on manually-generated messages,
4259 - MAY be used on Delivery Status Notifications (DSNs) and Message
4260 Disposition Notifications (MDNs),
4262 - MUST NOT be used on messages generated by automatic or periodic
4263 processes, except for messages which are automatic responses to
4266 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4267 to the default Received: header definition.
4269 PH/12 Added log selector acl_warn_skipped (default on).
4271 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4272 variables because (a) they are in the wrong storage pool and (b) even if
4273 they were copied, it wouldn't work properly because of the caching.
4275 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4276 checking when decoding. Apparently there are clients that generate
4277 overlong encoded strings. Why am I not surprised?
4279 PH/15 If the first argument of "${if match_address" was not empty, but did not
4280 contain an "@" character, Exim crashed. Now it writes a panic log message
4281 and treats the condition as false.
4283 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4285 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4286 extension "envelope-auth". The code is finished and in agreement with
4287 other implementations, but there is no documentation so far and in fact,
4288 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4289 not changing the active code.
4291 Print executed "if" and "elsif" statements when debugging is used. This
4292 helps a great deal to understand what a filter does.
4294 Document more things not specified clearly in RFC3028. I had all this
4295 sorted out, when out of a sudden new issues came to my mind. Oops."
4297 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4300 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4301 canonical form (as documented). However, after a host name lookup from
4302 the IP address, check_host() was doing a simple string comparison with
4303 addresses acquired from the DNS when checking that the found name did
4304 have the original IP as one of its addresses. Since any found IPv6
4305 addresses are likely to be in abbreviated form, the comparison could
4306 fail. Luckily, there already exists a function for doing the comparison
4307 by converting both addresses to binary, so now that is used instead of
4308 the text comparison.
4310 PH/20 There was another similar case to PH/19, when a complete host name was
4311 given in a host list; looking up its IP address could give an abbreviated
4312 form, whereas the current host's name might or might not be abbreviated.
4313 The same fix has been applied.
4319 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4320 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4323 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4324 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4326 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4328 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4329 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4330 Also, for "Auto-Submitted", treat anything other than "no" as
4331 non-personal, in accordance with RFC 3834. (Previously it treated
4332 anything starting "auto-" as non-personal.)
4334 TF/01 The control=submission/name=... option had a problem with syntax
4335 errors if the name included a slash character. The /name= option
4336 now slurps the rest of the string, so it can include any characters
4337 but it must come last in the list of options (after /sender_retain
4340 PH/05 Some modifications to the interface to the fake nameserver for the new
4348 TK/01 Added the "success_on_redirect" address verification option. See
4349 NewStuff for rationale and an example.
4351 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4353 PH/02 Patch to exigrep to allow it to work on syslog lines.
4355 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4356 fread() to skip over the body file's header line, because in Cygwin the
4357 header line is locked and is inaccessible.
4359 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4360 co-exist for some time) to make it clear that it is the Exim ID that is
4361 referenced, not the Message-ID: header line.
4363 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4364 string_format() function, because snprintf() does not exist on all
4367 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4368 ${stat: expansion item.
4370 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4371 protocol synchronization error", to keep the pedants happy.
4373 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4374 well as for IRIX systems, when gcc is being used. See the host.c source
4377 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4379 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4382 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4383 transport and to the smtp transport in LMTP mode.
4385 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4387 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4388 run a filter in a subprocess. This could lead to confusion in subsequent
4389 lookups in the parent process. There should also be a search_tidyup() at
4390 the end of the subprocess.
4392 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4393 only if the host matched helo_try_verify_hosts, which caused the
4394 verification to occur when the EHLO/HELO command was issued. The ACL just
4395 tested the remembered result. Now, if a previous verification attempt has
4396 not happened, "verify = helo" does it there and then.
4398 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4400 TK/03 Fix log output including CR from clamd.
4402 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4403 exist provoked a memory error which could cause a segfault.
4405 PH/15 Installed PCRE 6.2
4407 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4409 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4410 of the problem. Specifically, suggested +O2 rather than +O1 for the
4413 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4415 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4416 record, the port was not correctly logged when the outgoing_port log
4417 selector was set (it logged the transort's default port).
4419 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4420 fallback_hosts, and "hosts" in the smtp transport.
4422 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4423 host errors such as "Connection refused".
4425 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4426 authentication with radiusclient 0.4.9:
4428 - Error returned from rc_read_config was caught wrongly
4429 - Username/password not passed on to radius server due to wrong length.
4431 The presumption is that some radiusclient API changes for 4.51/PH/17
4432 were not taken care of correctly. The code is still untested by me (my
4433 Linux distribution still has 0.3.2 of radiusclient), but it was
4434 contributed by a Radius user.
4436 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4437 expanding the "port" option of the smtp transport.
4439 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4440 while reading a MIME header. Thanks to Tom Hughes for a patch.
4442 PH/24 Include config.h inside local_scan.h so that configuration settings are
4445 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4446 that in an ACL for RCPT (for example), you can examine exactly what was
4449 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4450 commands, but it was not correctly comparing the address with the actual
4451 client host address. Thus, it would show the EHLO address in Received:
4452 header lines when this was not necessary.
4454 PH/27 Added the % operator to ${eval:}.
4456 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4457 it should be ignoring failures (because with -C, for example, it has lost
4458 privilege). It wasn't ignoring creation failures other than "already
4461 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4464 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4465 because an input line was too long, either on its own, or by virtue of
4466 too many continuations, the temporary file was not being removed, and the
4467 return code was incorrect.
4469 PH/31 Missing "BOOL" in function definition in filtertest.c.
4471 PH/32 Applied Sieve patches from the maintainer.
4473 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4475 PH/33 Added "verify = not_blind".
4477 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4478 Local/Makefile (with some defaults set). These are used in built scripts
4479 such as exicyclog, but they have never been used in the exim_install
4480 script (though there are many overriding facilities there). I have
4481 arranged that the exim_install script now takes note of these two
4484 PH/35 Installed configuration files for Dragonfly.
4486 PH/36 When a locally submitted message by a trusted user did not contain a
4487 From: header, and the sender address was obtained from -f or from an SMTP
4488 MAIL command, and the trusted user did not use -F to supply a sender
4489 name, $originator_name was incorrectly used when constructing a From:
4490 header. Furthermore, $originator_name was used for submission mode
4491 messages from external hosts without From: headers in a similar way,
4492 which is clearly wrong.
4494 PH/37 Added control=suppress_local_fixups.
4496 PH/38 When log_selector = +received_sender was set, and the addition of the
4497 sender made the log line's construction buffer exactly full, or one byte
4498 less than full, an overflow happened when the terminating "\n" was
4501 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4502 when the result of a list match is failure because a DNS lookup failed.
4504 PH/40 RM_COMMAND is now used in the building process.
4506 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4507 the "build-* directories that it finds.
4509 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4510 address in a domain literal was a prefix of an interface address.
4512 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4513 when verifying a sender address, unless rewrite_headers is false.
4515 PH/44 Wrote a long comment about why errors_to addresses are verified as
4516 recipients, not senders.
4518 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4519 the ratelimit ACL was added.
4521 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4523 PH/46 Added extra information about PostgreSQL errors to the error string.
4525 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4526 avoiding the need to install special zones in a real server. This is
4527 backwards compatible; if it can't find the fake resolver, it drops back.
4528 Thus, both old and new test suites can be run.
4530 TF/02 Added util/ratelimit.pl
4532 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4533 clock is set back in time.
4535 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4536 Candler <B.Candler@pobox.com>.
4538 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4539 OK for addresses that are the result of redirections.
4541 PH/48 A number of further additions for the benefit of the new test suite,
4542 including a fake gethostbyname() that interfaces to the fake DNS resolver
4545 TF/06 The fix for widen_domains has also been applied to qualify_single and
4546 search_parents which are the other dnslookup options that can cause
4549 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4552 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4554 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4555 service" if it encountered an MX record with an empty target hostname.
4556 The message is now "an MX or SRV record indicated no SMTP service".
4558 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4559 if the DNS of the sending site is misconfigured. This is quite a
4560 common situation. This change restores the behaviour of treating a
4561 helo verification defer as a failure.
4563 PH/51 If self=fail was set on a router, the bounce message did not include the
4564 actual error message.
4570 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4572 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4573 command itself ended in error, the underlying message about the transport
4574 filter timeout was being overwritten with the pipe command error. Now the
4575 underlying error message should be appended to the second error message.
4577 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4579 PH/02 Reduce the amount of output that "make" produces by default. Full output
4580 can still be requested.
4582 PH/03 The warning log line about a condition test deferring for a "warn" verb
4583 was being output only once per connection, rather than after each
4584 occurrence (because it was using the same function as for successful
4585 "warn" verbs). This seems wrong, so I have changed it.
4587 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4588 it should not have, which might have caused a crash in the right
4589 circumstances, but probably never did.
4591 PH/04 Installed a modified version of Tony Finch's patch to make submission
4592 mode fix the return path as well as the Sender: header line, and to
4593 add a /name= option so that you can make the user's friendly name appear
4596 TF/03 Added the control = fakedefer ACL modifier.
4598 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4599 Mark Lowes for thorough testing.
4601 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4603 TK/03 Merged latest SRS patch from Miles Wilton.
4605 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4606 with the definition in sysexits.h (which is #included earlier).
4607 Fortunately, Exim does not actually use EX_OK. The code used to try to
4608 preserve the sysexits.h value, by assuming that macro definitions were
4609 scanned for macro replacements. I have been disabused of this notion,
4610 so now the code just undefines EX_OK before #including unistd.h.
4612 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4613 in the smtp transport. When a block could not be written in a single
4614 write() function, the timeout was being re-applied to each part-write.
4615 This seems wrong - if the receiver was accepting one byte at a time it
4616 would take for ever. The timeout is now adjusted when this happens. It
4617 doesn't have to be particularly precise.
4619 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4620 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4622 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4623 without a domain if the check to <postmaster@domain> fails.
4625 SC/01 Eximstats: added -xls and the ability to specify output files
4626 (patch written by Frank Heydlauf).
4628 SC/02 Eximstats: use FileHandles for outputting results.
4630 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4632 SC/04 Eximstats: fixed display of large numbers with -nvr option
4634 SC/05 Eximstats: fixed merging of reports with empty tables.
4636 SC/06 Eximstats: added the -include_original_destination flag
4638 SC/07 Eximstats: removed tabs and trailing whitespace.
4640 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4642 TK/06 MBOX spool code: Add real "From " MBOX separator line
4643 so the .eml file is really in mbox format (even though
4644 most programs do not really care). Patch from Alex Miller.
4646 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4647 The latter is generated from $received_to and is only set if the
4648 message has one envelope recipient. SA can use these headers,
4649 obviously out-of-the-box. Patch from Alex Miller.
4651 PH/08 The ${def test on a variable was returning false if the variable's
4652 value was "0", contrary to what the specification has always said!
4653 The result should be true unless the variable is empty.
4655 PH/09 The syntax error of a character other than { following "${if
4656 def:variable_name" (after optional whitespace) was not being diagnosed.
4657 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4658 accidental colon was present, for example, could give incorrect results.
4660 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4661 result is used (not including appendfile, where other changes are about
4664 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4665 This involved changing a lot of size variables from int to off_t. It
4666 should work with maildirs and everything.
4668 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4669 spamd dying while we are connected to it.
4671 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4674 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4675 function for BDB 4.3.
4677 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4679 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4680 the sources for PCRE have been re-arranged and more files are now
4683 PH/15 The code I had for printing potentially long long variables in PH/11
4684 above was not the best (it lost precision). The length of off_t variables
4685 is now inspected at build time, and an appropriate printing format (%ld
4686 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4687 to be "long long int" or "long int". This is needed for the internal
4688 formatting function string_vformat().
4690 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4691 the configuration file to be ":syslog", then the script "guesses" where
4692 the logs files are, rather than using the compiled in default. In our
4693 case the guess is not the same as the compiled default, so the script
4694 suddenly stopped working when I started to use syslog. The patch checks
4695 to see if log_file_path is "". If so, it attempts to read it from exim
4696 with no configuration file to get the compiled in version, before it
4697 falls back to the previous guessing code."
4699 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4700 implementing BATV in an Exim configuration. See NewStuff for the gory
4703 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4704 Makefile that are specific to HP-UX.
4706 PH/18 If the "use_postmaster" option was set for a recipient callout together
4707 with the "random" option, the postmaster address was used as the MAIL
4708 FROM address for the random test, but not for the subsequent recipient
4709 test. It is now used for both.
4711 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4712 patch removes a few documentation additions to RFC 3028, because the
4713 latest draft now contains them. It adds the new en;ascii-case comparator
4714 and a new error check for 8bit text in MIME parts. Comparator and
4715 require names are now matched exactly. I enabled the subaddress
4716 extension, but it is not well tested yet (read: it works for me)."
4718 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4719 rework some of the code of TK/09 above to avoid the hardwired use of
4720 "%lld" and "long long". Replaced the call to snprintf() with a call to
4723 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4724 records point to non-existent hosts", "retry timeout exceeded", and
4725 "retry time not reached for any host after a long failure period".
4727 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4728 experimental DomainKeys support:
4730 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4731 (2) On an error such as an illegally used "control", the wrong name for
4732 the control was given.
4734 These problems did NOT occur unless DomainKeys support was compiled.
4736 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4738 PH/24 Added ${if match_ip condition.
4740 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4741 databases so that it will be absolutely obvious if a crash occurs in the
4742 DB library. This is a regular occurrence (often caused by mis-matched
4745 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4746 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4747 picked up on a user's system that detects such things. There doesn't seem
4748 to be a gcc warning option for this - only an attribute that has to be
4749 put on the function's prototype. It seems that in Fedora Core 4 they have
4750 set this on a number of new functions. No doubt there will be more in due
4753 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4754 specify a transport. However, if an address that was verified by such a
4755 router was the subject of a callout, Exim crashed because it tried to
4756 read the rcpt_include_affixes from the non-existent transport. Now it
4757 just assumes that the setting of that option is false. This bug was
4758 introduced by 4.51/PH/31.
4760 PH/28 Changed -d+all to exclude +memory, because that information is very
4761 rarely of interest, but it makes the output a lot bigger. People tend to
4762 do -d+all out of habit.
4764 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4765 code in os-type was giving problems when libc.so lives in lib64, like on
4768 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4769 aren't the modern standard, and it seems that some systems' include files
4770 don't always have them. Exim was already checking for some of the newer
4771 ones like T_AAAA, and defining it itself. I've added checks for all the
4772 record types that Exim uses.
4774 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4775 not automatically generating a new one, as it is supposed to. This
4776 prevented TLS from working. If the file did exist, but contained invalid
4777 data, a new version was generated, as expected. It was only the case of a
4778 non-existent file that was broken.
4780 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4781 with a change in libdomainkeys > 0.64.
4783 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4784 from DNS. If the selector record carries the flag, it now has
4785 precedence over the domain-wide flag.
4787 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4789 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4790 the use of a port name that isn't defined in /etc/services) occurred, the
4791 message was deferred as in a normal delivery, and thus remained on the
4792 spool, instead of being failed because of the mua_wrapper setting. This
4793 is now fixed, and I tidied up some of the mua_wrapper messages at the
4796 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4797 the messages in a hash of arrays rather than using individual hashes.
4798 This is a bit cleaner and results in dramatic memory savings, albeit
4799 at a slight CPU cost.
4801 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4802 as requested by Marc Sherman.
4804 SC/10 Eximstats: added histograms for user specified patterns as requested
4807 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4809 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4810 fopen() in the content-scanning modules that did not already have it.
4816 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4817 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4819 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4821 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4823 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4824 to test Sieve filters that use "vacation".
4826 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4827 that changes the way the GnuTLS parameters are stored in the cache file.
4828 The new format can be generated externally. For backward compatibility,
4829 if the data in the cache doesn't make sense, Exim assumes it has read an
4830 old-format file, and it generates new data and writes a new file. This
4831 means that you can't go back to an older release without removing the
4834 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4835 work if there are any delivery delays because "one_time" forces the
4836 parent to be marked "delivered", so its unseen clone is never tried
4837 again. For this reason, Exim now forbids the simultaneous setting of
4840 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4841 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4842 entry if you want to know the details. The fix, however, neglected to
4843 consider the case when local delivery batching is involved. The test for
4844 "previously delivered" was not happening when checking to see if an
4845 address could be batched with a previous (undelivered) one; under
4846 certain circumstances this could lead to multiple deliveries to the same
4849 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4850 in its include files, and this causes problems building Exim.
4852 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4853 header_syntax) but Exim was just ignoring anything given after a slash.
4854 In particular, this caused confusion with an attempt to use "verify =
4855 reverse_host_lookup/defer_ok". An error is now given when options are
4856 supplied for verify items that do not have them. (Maybe reverse_host_
4857 lookup should have a defer_ok option, but that's a different point.)
4859 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4860 defined by RFC 821) to 2048, because there were problems with some AUTH
4861 commands, and RFC 1869 says the size should be increased for extended
4862 SMTP commands that take arguments.
4864 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4867 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4868 "unknown" error; now it says that the functionality isn't in the binary.
4870 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4871 an address' error message when a string expansion fails (syntax or
4872 whatever). Otherwise the password may appear in the log. Following change
4873 PH/42 below, there is no longer a chance of it appearing in a bounce
4876 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4878 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4879 of its addresses was ever tried. (Bugzilla bug #2.)
4881 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4882 the result incorrectly in the debug output. (It correctly added a newline
4883 to what was transported.)
4885 TF/01 Added $received_time.
4887 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4888 commented out examples of how to interface to a virus scanner and to
4889 SpamAssassin. Also added commented examples of av_scanner and
4890 spamd_address settings.
4892 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4893 and controls are allowed in which ACLs. There were a couple of minor
4894 errors. Some of the entries in the conditions table (which is a table of
4895 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4896 negation of where the condition IS allowed.
4898 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4900 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4901 header file does not have a version number, so I've had to invent a new
4902 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4903 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4904 radiusclient), but it was contributed by a Radius user.
4906 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4907 files or directories, for OpenSSL.
4909 PH/19 When an Exim process that is running as root has to create an Exim log
4910 file, it does so in a subprocess that runs as exim:exim so as to get the
4911 ownership right at creation (otherwise, other Exim processes might see
4912 the file with the wrong ownership). There was no test for failure of this
4913 fork() call, which would lead to the process getting stuck as it waited
4914 for a non-existent subprocess. Forks do occasionally fail when resources
4915 run out. I reviewed all the other calls to fork(); they all seem to check
4918 PH/20 When checking for unexpected SMTP input at connect time (before writing
4919 the banner), Exim was not dealing correctly with a non-positive return
4920 from the read() function. If the client had disconnected by this time,
4921 the result was a log entry for a synchronization error with an empty
4922 string after "input=" when read() returned zero. If read() returned -1
4923 (an event I could not check), uninitialized data bytes were printed.
4924 There were reports of junk text (parts of files, etc) appearing after
4927 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4929 PH/22 Added support for macro redefinition, and (re)definition in between
4930 driver and ACL definitions.
4932 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4933 forgetting to use the resulting value; it was using the unexpanded value.
4935 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4936 hadn't been configured. The fix is from Juergen Kreileder, who
4937 understands it better than I do:
4939 "Here's what I see happening with three configured cyrus_sasl
4940 authenticators configured (plain, login, cram-md5):
4942 On startup auth_cyrus_sasl_init() gets called for each of these.
4943 This means three calls to sasl_listmech() without a specified mech_list.
4944 => SASL tests which mechs of all available mechs actually work
4945 => three warnings about OTP not working
4946 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4948 With the patch, sasl_listmech() also gets called three times. But now
4949 SASL's mech_list option is set to the server_mech specified in the the
4950 authenticator. Or in other words, the answer from sasl_listmech()
4951 gets limited to just the mech you're testing for (which is different
4953 => the return list contains just 'plain' or 'login', 'cram-md5' or
4954 nothing depending on the value of ob->server_mech.
4956 I've just tested the patch: Authentication still works fine,
4957 unavailable mechs specified in the exim configuration are still
4958 caught, and the auth.log warnings about OTP are gone."
4960 PH/25 When debugging is enabled, the contents of the command line are added
4961 to the debugging output, even when log_selector=+arguments is not
4964 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4965 answer is "GNU", and only if the return is "GNU/something" is the answer
4968 PH/27 $acl_verify_message is now set immediately after the failure of a
4969 verification in an ACL, and so is available in subsequent modifiers. In
4970 particular, the message can be preserved by coding like this:
4972 warn !verify = sender
4973 set acl_m0 = $acl_verify_message
4975 Previously, $acl_verify_message was set only while expanding "message"
4976 and "log_message" when a very denied access.
4978 PH/28 Modified OS/os.c-Linux with
4980 -#ifndef OS_LOAD_AVERAGE
4981 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4983 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4984 nomenclature these days.)
4986 PH/29 Installed patch from the Sieve maintainer that adds the options
4987 sieve_useraddress and sieve_subaddress to the redirect router.
4989 PH/30 In these circumstances:
4990 . Two addresses routed to the same list of hosts;
4991 . First host does not offer TLS;
4992 . First host accepts first address;
4993 . First host gives temporary error to second address;
4994 . Second host offers TLS and a TLS session is established;
4995 . Second host accepts second address.
4996 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4997 and peerdn, if requested) that were in fact used only for the second
5000 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5001 attention to any local part prefix or suffix that was matched by the
5002 router that accepted the address. It now behaves in the same way as it
5003 does for delivery: the affixes are removed from the local part unless
5004 rcpt_include_affixes is set on the transport.
5006 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5007 timeout during the DATA phase of an incoming message.
5009 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5010 applied a patch sanctioned by the Sieve maintainer.
5012 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5013 the uid or gid is negative. A case of a negative gid caused this to be
5014 noticed. The fix allows for either to be negative.
5016 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5017 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5020 PH/36 The change PH/12 above was broken. Fixed it.
5022 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5023 the grounds that routing the same address twice would always produce the
5024 same answer. This might have been true once, but it is certainly no
5025 longer true now. Routing a child address may depend on the previous
5026 routing that produced that child. Some complicated redirection strategies
5027 went wrong when messages had multiple recipients, and made Exim's
5028 behaviour dependent on the order in which the addresses were given.
5030 I have moved the duplicate checking until after the routing is complete.
5031 Exim scans the addresses that are assigned to local and remote
5032 transports, and removes any duplicates. This means that more work will be
5033 done, as duplicates will always all be routed, but duplicates are
5034 presumably rare, so I don't expect this is of any significance.
5036 For deliveries to pipes, files, and autoreplies, the duplicate checking
5037 still happens during the routing process, since they are not going to be
5040 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5041 It corrects a timeout issue with spamd. This is Ian's comment: "The
5042 background is that sometimes spamd either never reads data from a
5043 connection it has accepted, or it never writes response data. The exiscan
5044 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5045 blindly assumes that writes won't block so it may never time out."
5047 PH/39 Allow G after quota size as well as K and M.
5049 PH/40 The value set for $authenticated_id in an authenticator may not contain
5050 binary zeroes or newlines because the value is written to log lines and
5051 to spool files. There was no check on this. Now the value is run through
5052 the string_printing() function so that such characters are converted to
5053 printable escape sequences.
5055 PH/41 $message_linecount is a new variable that contains the total number of
5056 lines in the message. Compare $body_linecount, which is the count for the
5059 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5060 bounce and delay warning messages, except in certain special cases, which
5063 (a) An SMTP error message from a remote host;
5064 (b) A message specified in a :fail: redirection;
5065 (c) A message specified in a "fail" command in a system filter;
5066 (d) A message specified in a FAIL return from the queryprogram router;
5067 (e) A message specified by the cannot_route_message router option.
5069 In these cases only, Exim does include the error details in bounce and
5070 warning messages. There are also a few cases where bland messages such
5071 as "unrouteable address" or "local delivery error" are given.
5073 PH/43 $value is now also set for the "else" part of a ${run expansion.
5075 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5076 being worked on, but at least Exim now implements the latest version to
5079 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5080 process to complete was treated as a delivery failure, a timeout while
5081 writing the message to the pipe was logged, but erroneously treated as a
5082 successful delivery. Such timeouts include transport filter timeouts. For
5083 consistency with the overall process timeout, these timeouts are now
5084 treated as errors, giving rise to delivery failures by default. However,
5085 there is now a new Boolean option for the pipe transport called
5086 timeout_defer, which, if set TRUE, converts the failures into defers for
5087 both kinds of timeout. A transport filter timeout is now identified in
5090 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5091 systems where "make" and "gmake" are different, calling "gmake" at top
5092 level broke things. I've arranged for the value of $(MAKE) to be passed
5093 from the Makefile to this script so that it can call the same version of
5097 A note about Exim versions 4.44 and 4.50
5098 ----------------------------------------
5100 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5101 changes of various kinds. As a consequence, a big documentation update was
5102 needed. This delayed the release for rather longer than seemed good, especially
5103 in the light of a couple of (minor) security issues. Therefore, the changes
5104 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5105 release. So 4.44 and 4.50 are in effect two different branches that both start
5108 I have left the 4.50 change log unchanged; it contains all the changes since
5109 4.43. The change log for 4.44 is below; many of its items are identical to
5110 those for 4.50. This seems to be the most sensible way to preserve the
5111 historical information.
5117 1. Minor wording change to the doc/README.SIEVE file.
5119 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5120 computation of the current number of files was incorrect.
5122 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5123 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5126 4. Give more explanation in the error message when the command for a transport
5127 filter fails to execute.
5129 5. There are several places where Exim runs a non-Exim command in a
5130 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5131 was being done only for the command run by the queryprogram router. It is
5132 now done for all such subprocesses. The other cases are: ${run, transport
5133 filters, and the commands run by the lmtp and pipe transports.
5135 6. Added CONFIGURE_GROUP build-time option.
5137 7. Some older OS have a limit of 256 on the maximum number of file
5138 descriptors. Exim was using setrlimit() to set 1000 as a large value
5139 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5140 systems. I've change it so that if it can't get 1000, it tries for 256.
5142 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5143 was an oversight, and furthermore, ever since the addition of extra
5144 controls (e.g. 4.43/32), the checks on when to allow different forms of
5145 "control" were broken. There should now be diagnostics for all cases when a
5146 control that does not make sense is encountered.
5148 9. Added the /retain_sender option to "control=submission".
5150 10. $recipients is now available in the predata ACL (oversight).
5152 11. Tidy the search cache before the fork to do a delivery from a message
5153 received from the command line. Otherwise the child will trigger a lookup
5154 failure and thereby defer the delivery if it tries to use (for example) a
5155 cached ldap connection that the parent has called unbind on.
5157 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5158 of $address_data from the recipient verification was clobbered by the
5159 sender verification.
5161 13. The value of address_data from a sender verification is now available in
5162 $sender_address_data in subsequent conditions in the ACL statement.
5164 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5166 15. Added a new option "connect=<time>" to callout options, to set a different
5169 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5170 was its contents. (It was OK if the option was not defined at all.)
5172 17. A "Completed" log line is now written for messages that are removed from
5173 the spool by the -Mrm option.
5175 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5176 information about exactly what failed.
5178 19. Added -dd to debug only the daemon process.
5180 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5181 handles timeouts, both on the server side and network timeouts. Renamed the
5182 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5184 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5185 It is now set to "smtps".
5187 22. $host_address is now set to the target address during the checking of
5188 ignore_target_hosts.
5190 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5191 being passed; this would have caused $sender_host_name to have been used if
5192 matching the list had actually called for a host name (not very likely,
5193 since this list is usually IP addresses). A host name is now passed as
5196 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5197 code that specifies a non-restarting handler (typically sigaction() in
5198 modern systems) in an attempt to fix a rare and obscure crash bug.
5200 25. Narrowed the window for a race in the daemon that could cause it to ignore
5201 SIGCHLD signals. This is not a major problem, because they are used only to
5202 wake it up if nothing else does.
5204 26. A malformed maildirsize file could cause Exim to calculate negative values
5205 for the mailbox size or file count. Odd effects could occur as a result.
5206 The maildirsize information is now recalculated if the size or filecount
5209 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5210 support for a long time. Removed HAVE_SYS_VFS_H.
5212 28. Installed the latest version of exipick from John Jetmore.
5214 29. In an address list, if the pattern was not a regular expression, an empty
5215 subject address (from a bounce message) matched only if the pattern was an
5216 empty string. Non-empty patterns were not even tested. This was the wrong
5217 because it is perfectly reasonable to use an empty address as part of a
5218 database query. An empty address is now tested by patterns that are
5219 lookups. However, all the other forms of pattern expect the subject to
5220 contain a local part and a domain, and therefore, for them, an empty
5221 address still always fails if the pattern is not itself empty.
5223 30. Exim went into a mad DNS loop when attempting to do a callout where the
5224 host was specified on an smtp transport, and looking it up yielded more
5225 than one IP address.
5227 31. Re-factored the code for checking spool and log partition space into a
5228 function that finds that data and another that does the check. The former
5229 is then used to implement four new variables: $spool_space, $log_space,
5230 $spool_inodes, and $log_inodes.
5232 32. The RFC2047 encoding function was originally intended for short strings
5233 such as real names; it was not keeping to the 75-character limit for
5234 encoded words that the RFC imposes. It now respects the limit, and
5235 generates multiple encoded words if necessary. To be on the safe side, I
5236 have increased the buffer size for the ${rfc2047: expansion operator from
5239 33. It is now permitted to omit both strings after an "if" condition; if the
5240 condition is true, the result is "true". As before, when the second string
5241 is omitted, a false condition yields an empty string. This makes it less
5242 cumbersome to write custom ACL and router conditions.
5244 34. Failure to deliver a bounce message always caused it to be frozen, even if
5245 there was an errors_to setting on the router. The errors_to setting is now
5248 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5249 canonical form (fully expanded) before being placed in
5250 $sender_host_address.
5252 36. The table in the code that translates DNS record types into text (T_A to
5253 "A" for instance) was missing entries for NS and CNAME. It is just possible
5254 that this could have caused confusion if both these types were looked up
5255 for the same domain, because the text type is used as part of Exim's
5256 per-process caching. But the chance of anyone hitting this buglet seems
5259 37. The dnsdb lookup has been extended in a number of ways.
5261 (1) There is a new type, "zns", which walks up the domain tree until it
5262 finds some nameserver records. It should be used with care.
5264 (2) There is a new type, "mxh", which is like "mx" except that it returns
5265 just the host names, not the priorities.
5267 (3) It is now possible to give a list of domains (or IP addresses) to be
5268 looked up. The behaviour when one of the lookups defers can be
5269 controlled by a keyword.
5271 (4) It is now possible to specify the separator character for use when
5272 multiple records are returned.
5274 38. The dnslists ACL condition has been extended: it is now possible to supply
5275 a list of IP addresses and/or domains to be looked up in a particular DNS
5278 39. Added log_selector=+queue_time_overall.
5280 40. When running the queue in the test harness, wait just a tad after forking a
5281 delivery process, to get repeatability of debugging output.
5283 41. Include certificate and key file names in error message when GnuTLS fails
5284 to set them up, because the GnuTLS error message doesn't include the name
5285 of the failing file when there is a problem reading it.
5287 42. Allow both -bf and -bF in the same test run.
5289 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5291 44. The "Exiscan patch" is now merged into the mainline Exim source.
5293 45. Sometimes the final signoff response after QUIT could fail to get
5294 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5295 < 0 before doing a fflush(). This bug looks as though it goes back to the
5296 introduction of TLS in release 3.20, but "sometimes" must have been rare
5297 because the tests only now provoked it.
5299 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5300 (this can affect the format of dates).
5302 47. exim_tidydb, when checking for the continued existence of a message for
5303 which it has found a message-specific retry record, was not finding
5304 messages that were in split spool directories. Consequently, it was
5305 deleting retry records that should have stayed in existence.
5307 48. Steve fixed some bugs in eximstats.
5309 49. The SPA authentication driver was not abandoning authentication and moving
5310 on to the next authenticator when an expansion was forced to fail,
5311 contradicting the general specification for all authenticators. Instead it
5312 was generating a temporary error. It now behaves as specified.
5314 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5315 (the order specifies the preference for clients). The order is now AES256,
5316 AES128, 3DES, ARCFOUR128.
5318 51. Small patch to Sieve code - explicitly set From: when generating an
5321 52. Exim crashed if a remote delivery caused a very long error message to be
5322 recorded - for instance if somebody sent an entire SpamAssassin report back
5323 as a large number of 550 error lines. This bug was coincidentally fixed by
5324 increasing the size of one of Exim's internal buffers (big_buffer) that
5325 happened as part of the Exiscan merge. However, to be on the safe side, I
5326 have made the code more robust (and fixed the comments that describe what
5329 53. Now that there can be additional text after "Completed" in log lines (if
5330 the queue_time_overall log selector is set), a one-byte patch to exigrep
5331 was needed to allow it to recognize "Completed" as not the last thing in
5334 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5335 patch that reportedly fixes this has been added. I am not expert enough to
5336 create a test for it. This is what the patch creator wrote:
5338 "I found a little strange behaviour of ldap code when working with
5339 Windows 2003 AD Domain, where users was placed in more than one
5340 Organization Units. When I tried to give exim partial DN, the exit code
5341 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5342 But simultaneously result of request was absolutely normal ldap result,
5343 so I produce this patch..."
5345 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5346 REFERENCE, so I have modified the code to exclude the patch when that macro
5349 55. Some experimental protocols are using DNS PTR records for new purposes. The
5350 keys for these records are domain names, not reversed IP addresses. The
5351 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5352 leaves it alone. Component reversal etc. now happens only for IP addresses.
5355 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5357 57. Double the size of the debug message buffer (to 2048) so that more of very
5358 long debug lines gets shown.
5360 58. The exicyclog utility now does better if the number of log files to keep
5361 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5363 59. Two changes related to the smtp_active_hostname option:
5365 (1) $smtp_active_hostname is now available as a variable.
5366 (2) The default for smtp_banner uses $smtp_active_hostname instead
5367 of $primary_hostname.
5369 60. The host_aton() function is supposed to be passed a string that is known
5370 to be a valid IP address. However, in the case of IPv6 addresses, it was
5371 not checking this. This is a hostage to fortune. Exim now panics and dies
5372 if the condition is not met. A case was found where this could be provoked
5373 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5374 components; fortuitously, this particular loophole had already been fixed
5375 by change 4.50/55 above.
5377 If there are any other similar loopholes, the new check in host_aton()
5378 itself should stop them being exploited. The report I received stated that
5379 data on the command line could provoke the exploit when Exim was running as
5380 exim, but did not say which command line option was involved. All I could
5381 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5382 running as the user.
5385 61. There was a buffer overflow vulnerability in the SPA authentication code
5386 (which came originally from the Samba project). I have added a test to the
5387 spa_base64_to_bits() function which I hope fixes it.
5390 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5391 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5393 63. The daemon start-up calls getloadavg() while still root for those OS that
5394 need the first call to be done as root, but it missed one case: when
5395 deliver_queue_load_max is set with deliver_drop_privilege. This is
5396 necessary for the benefit of the queue runner, because there is no re-exec
5397 when deliver_drop_privilege is set.
5399 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5400 This has been fixed.
5402 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5403 was in use, was not putting the data itself into the right store pool;
5404 consequently, it could be overwritten for a subsequent message in the same
5405 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5408 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5410 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5411 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5412 the function were treating the return as a boolean value, which happened to
5413 work because 0=false and not-0=true, but is not correct code.
5415 68. The host_aton() function was not handling scoped IPv6 addresses (those
5416 with, for example, "%eth0" on the end) correctly.
5418 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5419 negated items (that is, ~something) in unsigned ints. Some compilers
5420 apparently mutter when there is no cast.
5422 70. If an address verification called from an ACL failed, and did not produce a
5423 user-specific message (i.e. there was only a "system" message), nothing was
5424 put in $acl_verify_message. In this situation, it now puts the system
5427 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5428 session; change 4.31/43 added the unwanted input to the log line - except
5429 that it did not do this in the start of session case. It now does.
5431 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5432 This is wrong and can cause the other end to generate a synchronization
5433 error if it is another Exim or anything else that does the synchronization
5434 check. A QUIT command is no longer sent after a timeout.
5436 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5437 during host lookups.
5439 74. The defer_ok option of callout verification was not working if it was used
5440 when verifying addresses in header lines, that is, for this case:
5442 verify = header_sender/callout=defer_ok
5444 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5445 those file descriptors could be used for SMTP connections. If anything
5446 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5447 could be sent to the SMTP client, causing chaos. The daemon now opens
5448 stdin, stdout, and stderr to /dev/null when it puts itself into the
5451 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5452 log by default. The user can override this with suitable Perl magic.
5454 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5455 the log message when discard triggers, was not working for the DATA ACL or
5456 for the non-SMTP ACL.
5458 78. Error message wording change in sieve.c.
5460 79. If smtp_accept_max_per_host was set, the number of connections could be
5461 restricted to fewer than expected, because the daemon was trying to set up
5462 a new connection before checking whether the processes handling previous
5463 connections had finished. The check for completed processes is now done
5464 earlier. On busy systems, this bug wouldn't be noticed because something
5465 else would have woken the daemon, and it would have reaped the completed
5468 80. If a message was submitted locally by a user whose login name contained one
5469 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5470 It caused a spool format error. I have fixed the spool reading code. A
5471 related problem was that the "from" clause in the Received: line became
5472 illegal because of the space(s). It is now covered by ${quote_local_part.
5474 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5477 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5479 83. Patch from Sieve maintainer to fix unterminated string problem in
5480 "vacation" handling.
5482 84. Some minor changes to the Linux configuration files to help with other
5483 OS variants using glibc.
5485 85. One more patch for Sieve to update vacation handling to latest spec.
5488 ----------------------------------------------------
5489 See the note above about the 4.44 and 4.50 releases.
5490 ----------------------------------------------------
5496 1. Change 4.43/35 introduced a bug that caused file counts to be
5497 incorrectly computed when quota_filecount was set in an appendfile
5500 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5501 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5504 3. Give more explanation in the error message when the command for a transport
5505 filter fails to execute.
5507 4. There are several places where Exim runs a non-Exim command in a
5508 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5509 was being done only for the command run by the queryprogram router. It is
5510 now done for all such subprocesses. The other cases are: ${run, transport
5511 filters, and the commands run by the lmtp and pipe transports.
5513 5. Some older OS have a limit of 256 on the maximum number of file
5514 descriptors. Exim was using setrlimit() to set 1000 as a large value
5515 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5516 systems. I've change it so that if it can't get 1000, it tries for 256.
5518 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5519 was an oversight, and furthermore, ever since the addition of extra
5520 controls (e.g. 4.43/32), the checks on when to allow different forms of
5521 "control" were broken. There should now be diagnostics for all cases when a
5522 control that does not make sense is encountered.
5524 7. $recipients is now available in the predata ACL (oversight).
5526 8. Tidy the search cache before the fork to do a delivery from a message
5527 received from the command line. Otherwise the child will trigger a lookup
5528 failure and thereby defer the delivery if it tries to use (for example) a
5529 cached ldap connection that the parent has called unbind on.
5531 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5532 of $address_data from the recipient verification was clobbered by the
5533 sender verification.
5535 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5536 was its contents. (It was OK if the option was not defined at all.)
5538 11. A "Completed" log line is now written for messages that are removed from
5539 the spool by the -Mrm option.
5541 12. $host_address is now set to the target address during the checking of
5542 ignore_target_hosts.
5544 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5545 being passed; this would have caused $sender_host_name to have been used if
5546 matching the list had actually called for a host name (not very likely,
5547 since this list is usually IP addresses). A host name is now passed as
5550 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5551 code that specifies a non-restarting handler (typically sigaction() in
5552 modern systems) in an attempt to fix a rare and obscure crash bug.
5554 15. Narrowed the window for a race in the daemon that could cause it to ignore
5555 SIGCHLD signals. This is not a major problem, because they are used only to
5556 wake it up if nothing else does.
5558 16. A malformed maildirsize file could cause Exim to calculate negative values
5559 for the mailbox size or file count. Odd effects could occur as a result.
5560 The maildirsize information is now recalculated if the size or filecount
5563 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5564 support for a long time. Removed HAVE_SYS_VFS_H.
5566 18. Updated exipick to current release from John Jetmore.
5568 19. Allow an empty sender to be matched against a lookup in an address list.
5569 Previously the only cases considered were a regular expression, or an
5572 20. Exim went into a mad DNS lookup loop when doing a callout where the
5573 host was specified on the transport, if the DNS lookup yielded more than
5576 21. The RFC2047 encoding function was originally intended for short strings
5577 such as real names; it was not keeping to the 75-character limit for
5578 encoded words that the RFC imposes. It now respects the limit, and
5579 generates multiple encoded words if necessary. To be on the safe side, I
5580 have increased the buffer size for the ${rfc2047: expansion operator from
5583 22. Failure to deliver a bounce message always caused it to be frozen, even if
5584 there was an errors_to setting on the router. The errors_to setting is now
5587 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5588 canonical form (fully expanded) before being placed in
5589 $sender_host_address.
5591 24. Updated eximstats to version 1.33
5593 25. Include certificate and key file names in error message when GnuTLS fails
5594 to set them up, because the GnuTLS error message doesn't include the name
5595 of the failing file when there is a problem reading it.
5597 26. Expand error message when OpenSSL has problems setting up cert/key files.
5600 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5601 (this can affect the format of dates).
5603 28. exim_tidydb, when checking for the continued existence of a message for
5604 which it has found a message-specific retry record, was not finding
5605 messages that were in split spool directories. Consequently, it was
5606 deleting retry records that should have stayed in existence.
5608 29. eximstats updated to version 1.35
5609 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5610 1.35 - bugfix such that pie charts by volume are generated correctly
5612 30. The SPA authentication driver was not abandoning authentication and moving
5613 on to the next authenticator when an expansion was forced to fail,
5614 contradicting the general specification for all authenticators. Instead it
5615 was generating a temporary error. It now behaves as specified.
5617 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5618 (the order specifies the preference for clients). The order is now AES256,
5619 AES128, 3DES, ARCFOUR128.
5621 31. Small patch to Sieve code - explicitly set From: when generating an
5624 32. Exim crashed if a remote delivery caused a very long error message to be
5625 recorded - for instance if somebody sent an entire SpamAssassin report back
5626 as a large number of 550 error lines. This bug was coincidentally fixed by
5627 increasing the size of one of Exim's internal buffers (big_buffer) that
5628 happened as part of the Exiscan merge. However, to be on the safe side, I
5629 have made the code more robust (and fixed the comments that describe what
5632 33. Some experimental protocols are using DNS PTR records for new purposes. The
5633 keys for these records are domain names, not reversed IP addresses. The
5634 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5635 leaves it alone. Component reversal etc. now happens only for IP addresses.
5638 34. The host_aton() function is supposed to be passed a string that is known
5639 to be a valid IP address. However, in the case of IPv6 addresses, it was
5640 not checking this. This is a hostage to fortune. Exim now panics and dies
5641 if the condition is not met. A case was found where this could be provoked
5642 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5643 components; fortuitously, this particular loophole had already been fixed
5644 by change 4.50/55 or 4.44/33 above.
5646 If there are any other similar loopholes, the new check in host_aton()
5647 itself should stop them being exploited. The report I received stated that
5648 data on the command line could provoke the exploit when Exim was running as
5649 exim, but did not say which command line option was involved. All I could
5650 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5651 running as the user.
5654 35. There was a buffer overflow vulnerability in the SPA authentication code
5655 (which came originally from the Samba project). I have added a test to the
5656 spa_base64_to_bits() function which I hope fixes it.
5659 36. The daemon start-up calls getloadavg() while still root for those OS that
5660 need the first call to be done as root, but it missed one case: when
5661 deliver_queue_load_max is set with deliver_drop_privilege. This is
5662 necessary for the benefit of the queue runner, because there is no re-exec
5663 when deliver_drop_privilege is set.
5665 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5666 was in use, was not putting the data itself into the right store pool;
5667 consequently, it could be overwritten for a subsequent message in the same
5668 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5671 38. Sometimes the final signoff response after QUIT could fail to get
5672 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5673 < 0 before doing a fflush(). This bug looks as though it goes back to the
5674 introduction of TLS in release 3.20, but "sometimes" must have been rare
5675 because the tests only now provoked it.
5681 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5682 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5683 Now it returns an integer. A number of calls were still expecting a T/F
5684 return. Fortuitously, in all cases, the tests worked in OK situations,
5685 which is the norm. However, things would have gone wrong on any write
5686 failures on the smtp file descriptor. This function is used when sending
5687 messages over SMTP and also when doing verify callouts.
5689 2. When Exim is called to do synchronous delivery of a locally submitted
5690 message (the -odf or -odi options), it no longer closes stderr before doing
5693 3. Implemented the mua_wrapper option.
5695 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5697 5. Implemented the functions header_remove(), header_testname(),
5698 header_add_at_position(), and receive_remove_recipient(), and exported them
5701 6. If an ACL "warn" statement specified the addition of headers, Exim already
5702 inserted X-ACL-Warn: at the start if there was no header name. However, it
5703 was not making this test for the second and subsequent header lines if
5704 there were newlines in the string. This meant that an invalid header could
5705 be inserted if Exim was badly configured.
5707 7. Allow an ACL "warn" statement to add header lines at the start or after all
5708 the Received: headers, as well as at the end.
5710 8. Added the rcpt_4xx retry error code.
5712 9. Added postmaster_mailfrom=xxx to callout verification option.
5714 10. Added mailfrom=xxxx to the callout verification option, for verify=
5717 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5718 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5720 12. Inserted some casts to stop certain compilers warning when using pointer
5721 differences as field lengths or precisions in printf-type calls (mostly
5722 affecting debugging statements).
5724 13. Added optional readline() support for -be (dynamically loaded).
5726 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5727 same clock tick as a message's arrival, so that its received time was the
5728 same as the "first fail" time on the retry record, and that message
5729 remained on the queue past the ultimate address timeout, every queue runner
5730 would try a delivery (because it was past the ultimate address timeout) but
5731 after another failure, the ultimate address timeout, which should have then
5732 bounced the address, did not kick in. This was a "< instead of <=" error;
5733 in most cases the first failure would have been in the next clock tick
5734 after the received time, and all would be well.
5736 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5737 being recognized when the domain list was tested by the match_domain
5738 condition in an expansion string.
5740 16. Added the ${str2b64: operator.
5742 17. Exim was always calling setrlimit() to set a large limit for the number of
5743 processes, without checking whether the existing limit was already
5744 adequate. (It did check for the limit on file descriptors.) Furthermore,
5745 errors from getrlimit() and setrlimit() were being ignored. Now they are
5746 logged to the main and panic logs, but Exim does carry on, to try to do its
5747 job under whatever limits there are.
5749 18. Imported PCRE 5.0.
5751 19. Trivial typo in log message " temporarily refused connection" (the leading
5754 20. If the log selector return_path_on_delivery was set and an address was
5755 redirected to /dev/null, the delivery process crashed because it assumed
5756 that a return path would always be set for a "successful" delivery. In this
5757 case, the whole delivery is bypassed as an optimization, and therefore no
5760 21. Internal re-arrangement: the function for sending a challenge and reading
5761 a response while authentication was assuming a zero-terminated challenge
5762 string. It's now changed to take a pointer and a length, to allow for
5763 binary data in such strings.
5765 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5767 23. Exim was not respecting finduser_retries when seeking the login of the
5768 uid under which it was called; it was always trying 10 times. (The default
5769 setting of finduser_retries is zero.) Also, it was sleeping after the final
5770 failure, which is pointless.
5772 24. Implemented tls_on_connect_ports.
5774 25. Implemented acl_smtp_predata.
5776 26. If the domain in control=submission is set empty, Exim assumes that the
5777 authenticated id is a complete email address when it generates From: or
5778 Sender: header lines.
5780 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5781 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5782 chown and chgrp in /bin and hostname in /usr/bin.
5784 28. Exim was keeping the "process log" file open after each use, just as it
5785 does for the main log. This opens the possibility of it remaining open for
5786 long periods when the USR1 signal hits a daemon. Occasional processlog
5787 errors were reported, that could have been caused by this. Anyway, it seems
5788 much more sensible not to leave this file open at all, so that is what now
5791 29. The long-running daemon process does not normally write to the log once it
5792 has entered its main loop, and it closes the log before doing so. This is
5793 so that log files can straightforwardly be renamed and moved. However,
5794 there are a couple of unusual error situations where the daemon does write
5795 log entries, and I had neglected to close the log afterwards.
5797 30. The text of an SMTP error response that was received during a remote
5798 delivery was being truncated at 512 bytes. This is too short for some of
5799 the long messages that one sometimes sees. I've increased the limit to
5802 31. It is now possible to make retry rules that apply only when a message has a
5803 specific sender, in particular, an empty sender.
5805 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5806 it possible to be selective about when SMTP synchronization is enforced.
5808 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5810 32. Implemented hosts_connection_nolog.
5812 33. Added an ACL for QUIT.
5814 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5817 35. Added mailbox_size and mailbox_filecount to appendfile.
5819 36. Added control = no_multiline_responses to ACLs.
5821 37. There was a bug in the logic of the code that waits for the clock to tick
5822 in the case where the clock went backwards by a substantial amount such
5823 that the microsecond fraction of "now" was more than the microsecond
5824 fraction of "then" (but the whole seconds number was less).
5826 38. Added support for the libradius Radius client library this is found on
5827 FreeBSD (previously only the radiusclient library was supported).
5833 1. When certain lookups returned multiple values in the form name=value, the
5834 quoting of the values was not always being done properly. Specifically:
5835 (a) If the value started with a double quote, but contained no whitespace,
5837 (b) If the value contained whitespace other than a space character (i.e.
5838 tabs or newlines or carriage returns) it was not quoted.
5839 This fix has been applied to the mysql and pgsql lookups by writing a
5840 separate quoting function and calling it from the lookup code. The fix
5841 should probably also be applied to nisplus, ibase and oracle lookups, but
5842 since I cannot test any of those, I have not disturbed their existing code.
5844 2. A hit in the callout cache for a specific address caused a log line with no
5845 reason for rejecting RCPT. Now it says "Previous (cached) callout
5846 verification failure".
5848 3. There was an off-by-one bug in the queryprogram router. An over-long
5849 return line was truncated at 256 instead of 255 characters, thereby
5850 overflowing its buffer with the terminating zero. As well as fixing this, I
5851 have increased the buffer size to 1024 (and made a note to document this).
5853 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5854 when Exim is waiting for an SMTP response from a remote server, Exim
5855 restarts its select() call on the socket, thereby resetting its timeout.
5856 This is not a problem when such interrupts are rare. Somebody set up a cron
5857 job to run exiwhat every 2 minutes, which is less than the normal select()
5858 timeout (5 or 10 minutes). This meant that the select() timeout never
5859 kicked in because it was always reset. I have fixed this by comparing the
5860 time when an interrupt arrives with the time at the start of the first call
5861 to select(). If more time than the timeout has elapsed, the interrupt is
5862 treated as a timeout.
5864 5. Some internal re-factoring in preparation for the addition of Sieve
5865 extensions (by MH). In particular, the "personal" test is moved to a
5866 separate function, and given an option for scanning Cc: and Bcc: (which is
5867 not set for Exim filters).
5869 6. When Exim created an email address using the login of the caller as the
5870 local part (e.g. when creating a From: or Sender: header line), it was not
5871 quoting the local part when it contained special characters such as @.
5873 7. Installed new OpenBSD configuration files.
5875 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5876 try to make them clearer.
5878 9. Callout options, other than the timeout value, were being ignored when
5879 verifying sender addresses in header lines. For example, when using
5881 verify = header_sender/callout=no_cache
5883 the cache was (incorrectly) being used.
5885 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5886 only the Cygwin environment.
5888 11. When return_path_on_delivery was set as a log selector, if different remote
5889 addresses in the same message used different return paths and parallel
5890 remote delivery occurred, the wrong values would sometimes be logged.
5891 (Whenever a remote delivery process finished, the return path value from
5892 the most recently started remote delivery process was logged.)
5894 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5895 lines when AUTH and/or TLS are in use. This is the "received protocol"
5896 field. Exim used to use "asmtp" for authenticated SMTP, without any
5897 indication (in the protocol name) for TLS use. Now it follows the RFC and
5898 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5899 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5900 names appear in log lines as well as in Received: header lines.
5902 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5903 extensions, and comparison tests, and to fix some bugs.
5905 14. Changes to the "personal" filter test:
5907 (1) The test was buggy in that it was just doing the equivalent of
5908 "contains" tests on header lines. For example, if a user's address was
5909 anne@some.where, the "personal" test would incorrectly be true for
5911 To: susanne@some.where
5913 This test is now done by extracting each address from the header in turn,
5914 and checking the entire address. Other tests that are part of "personal"
5915 are now done using regular expressions (for example, to check local parts
5916 of addresses in From: header lines).
5918 (2) The list of non-personal local parts in From: addresses has been
5919 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5920 taken from the Sieve specification recommendations.
5922 (3) If the message contains any header line starting with "List-" it is
5923 treated as non-personal.
5925 (4) The test for "circular" in the Subject: header line has been removed
5926 because it now seems ill-conceived.
5928 15. Minor typos in src/EDITME comments corrected.
5930 16. Installed latest exipick from John Jetmore.
5932 17. If headers_add on a router specified a text string that was too long for
5933 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5934 of string_sprintf() is now avoided.
5936 18. $message_body_size was not set (it was always zero) when running the DATA
5937 ACL and the local_scan() function.
5939 19. For the "mail" command in an Exim filter, no default was being set for
5940 the once_repeat time, causing a random time value to be used if "once" was
5941 specified. (If the value happened to be <= 0, no repeat happened.) The
5942 default is now 0s, meaning "never repeat". The "vacation" command was OK
5943 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5944 (I found it when inspecting the code).
5946 20. There is now an overall timeout for performing a callout verification. It
5947 defaults to 4 times the callout timeout, which applies to individual SMTP
5948 commands during the callout. The overall timeout applies when there is more
5949 than one host that can be tried. The timeout is checked before trying the
5950 next host. This prevents very long delays if there are a large number of
5951 hosts and all are timing out (e.g. when the network connections are timing
5952 out). The value of the overall timeout can be changed by specifying an
5953 additional sub-option for "callout", called "maxwait". For example:
5955 verify = sender/callout=5s,maxwait=20s
5957 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5958 to the end before writing, but this should make it even safer).
5960 22. Exim was forgetting that it had advertised PIPELINING for the second and
5961 subsequent messages on an SMTP connection. It was also not resetting its
5962 memory on STARTTLS and an internal HELO.
5964 23. When Exim logs an SMTP synchronization error within a session, it now
5965 records whether PIPELINING has been advertised or not.
5967 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5968 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5969 rather than long int.
5971 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5973 26. Added the never_mail option to autoreply.
5979 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5980 crash if the getsockname() call failed; this can happen if a connection is
5981 closed very soon after it is established. The problem was simply in the
5982 order in which certain operations were done, causing Exim to try to write
5983 to the SMTP stream before it had set up the file descriptor. The bug has
5984 been fixed by making things happen in the correct order.
5990 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5991 before the connection was closed, thus losing the rejection response.
5993 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5994 some early Solaris releases, but causes trouble in current releases where
5995 socklen_t is defined.
5997 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6000 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6003 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6004 that releasing the top of it at the end releases what was used for sub-
6005 expansions (unless the block got too big). However, discard this block if
6006 the first thing is a variable or header, so that we can use its block when
6007 it is dynamic (useful for very large $message_headers, for example).
6009 6. Lookups now cache *every* query, not just the most recent. A new, separate
6010 store pool is used for this. It can be recovered when all lookup caches are
6011 flushed. Lookups now release memory at the end of their result strings.
6012 This has involved some general refactoring of the lookup sources.
6014 7. Some code has been added to the store_xxx() functions to reduce the amount
6015 of flapping under certain conditions.
6017 8. log_incoming_interface used to affect only the <= reception log lines. Now
6018 it causes the local interface and port to be added to several more SMTP log
6019 lines, for example "SMTP connection from", and rejection lines.
6021 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6023 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6025 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6026 could be overwritten at the end of the current message (or the start of a
6027 new message if it was set in a HELO ACL). The value is now preserved for
6028 the duration of the SMTP connection.
6030 12. If a transport had a headers_rewrite setting, and a matching header line
6031 contained an unqualified address, that address was qualified, even if it
6032 did not match any rewriting rules. The underlying bug was that the values
6033 of the flags that permit the existence of unqualified sender and recipient
6034 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6035 non-local messages, and by -bnq for local messages) were not being
6036 preserved with the message after it was received.
6038 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6039 "next input=" as part of the text comprising the host identity instead of
6040 the correct text. The code was using the same buffer for two different
6041 strings. However, depending on which order the printing function evaluated
6042 its arguments, the bug did not always show up. Under Linux, for example, my
6043 test suite worked just fine.
6045 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6046 doesn't work with some older versions of Perl. It has been changed to "my",
6047 which in any case is probably the better facility to use.
6049 15. A really picky compiler found some instances of statements for creating
6050 error messages that either had too many or two few arguments for the format
6053 16. The size of the buffer for calls to the DNS resolver has been increased
6054 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6055 for addresses that have a lot of PTR records. This alleviates a problem; it
6056 does not fully solve it.
6058 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6059 buffer now truncates the list and logs the incident, which is the same
6060 action as happens when Exim is looking up a host name and its aliases.
6061 Previously in this situation something unpredictable would happen;
6062 sometimes it was "internal error: store_reset failed".
6064 18. If a server dropped the connection unexpectedly when an Exim client was
6065 using GnuTLS and trying to read a response, the client delivery process
6066 crashed while trying to generate an error log message.
6068 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6069 string, for example:
6071 warn message = H1: something\nH2: something
6073 the text was added as a single header line from Exim's point of view
6074 though it ended up OK in the delivered message. However, searching for the
6075 second and subsequent header lines using $h_h2: did not work. This has been
6076 fixed. Similarly, if a system filter added multiple headers in this way,
6077 the routers could not see them.
6079 20. Expanded the error message when iplsearch is called with an invalid key to
6080 suggest using net-iplsearch in a host list.
6082 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6083 ACLs are no longer actually imposed (and a message to that effect is
6086 22. If a "gecos" field in a passwd entry contained escaped characters, in
6087 particular, if it contained a \" sequence, Exim got it wrong when building
6088 a From: or a Sender: header from that name. A second bug also caused
6089 incorrect handling when an unquoted " was present following a character
6090 that needed quoting.
6092 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6093 was not being matched caselessly.
6095 24. Arranged for all hyphens in the exim.8 source to be escaped with
6098 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6099 properly. Recipient callout cache records were still being keyed to include
6100 the sender, even when use_sender was set false. This led to far more
6101 callouts that were necessary. The sender is no longer included in the key
6102 when use_sender is false.
6104 26. Added "control = submission" modifier to ACLs.
6106 27. Added the ${base62d: operator to decode base 62 numbers.
6108 28. dnsdb lookups can now access SRV records.
6110 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6111 the configuration file.
6113 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6114 (-v) mode. This makes the output for a verbose queue run more intelligible.
6116 31. Added a use_postmaster feature to recipient callouts.
6118 32. Added the $body_zerocount variable, containing the number of binary zero
6119 bytes in the message body.
6121 33. The time of last modification of the "new" subdirectory is now used as the
6122 "mailbox time last read" when there is a quota error for a maildir
6125 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6127 35. Added +ignore_unknown as a special item in host lists.
6129 36. Code for decoding IPv6 addresses in host lists is now included, even if
6130 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6131 address was recognized as an IP address, but was then not correctly decoded
6132 into binary, causing unexpected and incorrect effects when compared with
6139 1. Very minor rewording of debugging text in manualroute to say "list of
6140 hosts" instead of "hostlist".
6142 2. If verify=header_syntax was set, and a header line with an unqualified
6143 address (no domain) and a large number of spaces between the end of the
6144 name and the colon was received, the reception process suffered a buffer
6145 overflow, and (when I tested it) crashed. This was caused by some obsolete
6146 code that should have been removed. The fix is to remove it!
6148 3. When running in the test harness, delay a bit after writing a bounce
6149 message to get a bit more predictability in the log output.
6151 4. Added a call to search_tidyup() just before forking a reception process. In
6152 theory, someone could use a lookup in the expansion of smtp_accept_max_
6153 per_host which, without the tidyup, could leave open a database connection.
6155 5. Added the variables $recipient_data and $sender_data which get set from a
6156 lookup success in an ACL "recipients" or "senders" condition, or a router
6157 "senders" option, similar to $domain_data and $local_part_data.
6159 6. Moved the writing of debug_print from before to after the "senders" test
6162 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6163 problems for message scanning, either using a data ACL, or using
6164 local_scan() because the Received: header was not generated till after they
6165 were called (in order to set the time as the time of reception completion).
6166 I have revised the way this works. The header is now generated after the
6167 body is received, but before the ACL or local_scan() are called. After they
6168 are run, the timestamp in the header is updated.
6174 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6175 before starting a queue runner without re-exec. This happened only when
6176 deliver_drop_privilege was set or when the Exim user was set to root. The
6177 effect of the bug was that timeouts during subsequent deliveries caused
6178 crashes instead of being properly handled. The handler is now left at its
6179 default (and expected) setting.
6181 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6182 message, again when deliver_drop_privilege is set or Exim is run as root.
6183 The bug described in (1) was not present in this case, but the tidying up
6184 of the other signals was missing. I have made the two cases consistent.
6186 3. The ignore_target_hosts setting on a manualroute router was being ignored
6187 for hosts that were looked up using the /MX notation.
6189 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6192 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6193 operated on the sender address. After changing the $sender_address to <>
6194 for the sender address verify, Exim was re-instated it as the original
6195 (before rewriting) address, but remembering that it had rewritten it, so it
6196 wasn't rewriting it again. This bug also had the effect of breaking the
6197 sender address verification caching when the sender address was rewritten.
6199 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6200 This has been changed so that if the ip literal address matches
6201 ignore_target_hosts, the router declines.
6203 7. Added expansion conditions match_domain, match_address, and match_local_
6204 part (NOT match_host).
6206 8. The placeholder for the Received: header didn't have a length field set.
6208 9. Added code to Exim itself and to exim_lock to test for a specific race
6209 condition that could lead to file corruption when using MBX delivery. The
6210 issue is with the lockfile that is created in /tmp. If this file is removed
6211 after a process has opened it but before that process has acquired a lock,
6212 there is the potential for a second process to recreate the file and also
6213 acquire a lock. This could lead to two Exim processes writing to the file
6214 at the same time. The added code performs the same test as UW imapd; it
6215 checks after acquiring the lock that its file descriptor still refers to
6216 the same named file.
6218 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6219 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6222 11. Added the smtp_active_hostname option. If used, this will typically be made
6223 to depend on the incoming interface address. Because $interface_address is
6224 not set up until the daemon has forked a reception process, error responses
6225 that can happen earlier (such as "too many connections") no longer contain
6228 12. If an expansion in a condition on a "warn" statement fails because a lookup
6229 defers, the "warn" statement is abandoned, and the next ACL statement is
6230 processed. Previously this caused the whole ACL to be aborted.
6232 13. Added the iplsearch lookup type.
6234 14. Added ident_timeout as a log selector.
6236 15. Added tls_certificate_verified as a log selector.
6238 16. Added a global option tls_require_ciphers (compare the smtp transport
6239 option of the same name). This controls incoming TLS connections.
6241 17. I finally figured out how to make tls_require_ciphers do a similar thing
6242 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6243 before starting the TLS session.
6245 18. Tabs are now shown as \t in -bP output.
6247 19. If the log selector return_path_on_delivery was set, Exim crashed when
6248 bouncing a message because it had too many Received: header lines.
6250 20. If two routers both had headers_remove settings, and the first one included
6251 a superfluous trailing colon, the final name in the first list and the
6252 first name in the second list were incorrectly joined into one item (with a
6253 colon in the middle).
6259 1. Added -C and -D options to the exinext utility, mainly to make it easier
6260 to include in the automated testing, but these could be helpful when
6261 multiple configurations are in use.
6263 2. The exinext utility was not formatting the output nicely when there was
6264 an alternate port involved in the retry record key, nor when there was a
6265 message id as well (for retries that were specific to a specific message
6266 and a specific host). It was also confused by IPv6 addresses, because of
6267 the additional colons they contain. I have fixed the IPv4 problem, and
6268 patched it up to do a reasonable job for IPv6.
6270 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6271 delivery, the log line now contains "pipelined" if PIPELINING was used.
6273 4. An SMTP transport process used to panic and die if the bind() call to set
6274 an explicit outgoing interface failed. This has been changed; it is now
6275 treated in the same way as a connect() failure.
6277 5. A reference to $sender_host_name in the part of a conditional expansion
6278 that was being skipped was still causing a DNS lookup. This no longer
6281 6. The def: expansion condition was not recognizing references to header lines
6282 that used bh_ and bheader_.
6284 7. Added the _cache feature to named lists.
6286 8. The code for checking quota_filecount in the appendfile transport was
6287 allowing one more file than it should have been.
6289 9. For compatibility with Sendmail, the command line option
6297 and sets the incoming protocol and host name (for trusted callers). The
6298 host name and its colon can be omitted when only the protocol is to be set.
6299 Note the Exim already has two private options, -pd and -ps, that refer to
6300 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6301 "s", but I don't think that's a major issue.
6303 10. A number of refactoring changes to the code, none of which should affect
6306 (a) The number of logging options was getting close to filling up the
6307 32-bit word that was used as a bit map. I have split them into two classes:
6308 those that are passed in the argument to log_write(), and those that are
6309 only ever tested independently outside of that function. These are now in
6310 separate 32-bit words, so there is plenty of room for expansion again.
6311 There is no change in the user interface or the logging behaviour.
6313 (b) When building, for example, log lines, the code previously used a
6314 macro that called string_cat() twice, in order to add two strings. This is
6315 not really sufficiently general. Furthermore, there was one instance where
6316 it was actually wrong because one of the argument was used twice, and in
6317 one call a function was used. (As it happened, calling the function twice
6318 did not affect the overall behaviour.) The macro has been replaced by a
6319 function that can join an arbitrary number of extra strings onto a growing
6322 (c) The code for expansion conditions now uses a table and a binary chop
6323 instead of a serial search (which was left over from when there were very
6324 few conditions). Also, it now recognizes conditions like "pam" even when
6325 the relevant support is not compiled in: a suitably worded error message is
6326 given if an attempt is made to use such a condition.
6328 11. Added ${time_interval:xxxxx}.
6330 12. A bug was causing one of the ddress fields not to be passed back correctly
6331 from remote delivery subprocesses. The field in question was not being
6332 subsequently used, so this caused to problems in practice.
6334 13. Added new log selectors queue_time and deliver_time.
6336 14. Might have fixed a bug in maildirsizefile handling that threw up
6337 "unexpected character" debug warnings, and recalculated the data
6338 unnecessarily. In any case, I expanded the warning message to give more
6341 15. Added the message "Restricted characters in address" to the statements in
6342 the default ACL that block characters like @ and % in local parts.
6344 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6345 Three changes have been made:
6347 (a) There was a serious bug; a negative response to MAIL caused the whole
6348 recipient domain to be cached as invalid, thereby blocking all messages
6349 to all local parts at the same domain, from all senders. This bug has
6350 been fixed. The domain is no longer cached after a negative response to
6351 MAIL if the sender used is not empty.
6353 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6356 (c) A new callout option, "use_sender" has been added for people who want
6357 the modified behaviour.
6363 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6366 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6367 indeed breaks things for older releases.
6369 3. Added additional logging to the case where there is a problem reading data
6370 from a filter that is running in a subprocess using a pipe, in order to
6371 try to track down a specific problem.
6373 4. Testing facility fudge: when running in the test harness and attempting
6374 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6375 getting "No route to host". Convert this to a timeout.
6377 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6380 6. Some OS don't have socklen_t but use size_t instead. This affects the
6381 fifth argument of getsockopt() amongst other things. This is now
6382 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6383 can be set for individual OS. I have set it for SunOS5, OSF1, and
6384 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6385 some earlier ones do not.
6387 7. Change 4.30/15 was not doing the test caselessly.
6389 8. The standard form for an IPv6 address literal was being rejected by address
6390 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6391 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6392 this, as well as the form without the "IPv6" on the front (but only when
6393 address literals are enabled, of course).
6395 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6397 10. Exim crashed if a message with an empty sender address specified by -f
6398 encountered a router with an errors_to setting. This could be provoked only
6399 by a command such as
6403 where an empty string was supplied; "<>" did not hit this bug.
6405 11. Installed PCRE release 4.5.
6407 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6408 remained set. It is now erased.
6410 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6411 times from message ids (which are base 36 rather than the normal 62).
6413 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6414 were being counted as actual protocol errors, and logged if the log
6415 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6416 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6417 and DATA following a set of rejected RCPTs do not count as protocol errors.
6418 In other words, Exim assumes they were pipelined, though this may not
6419 actually be the case. Of course, in all cases the client gets an
6420 appropriate error code.
6422 15. If a lookup fails in an ACL condition, a message about the failure may
6423 be available; it is used if testing the ACL cannot continue, because most
6424 such messages specify what the cause of the deferral is. However, some
6425 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6426 that caused an old message to be retained and used if a later statement
6427 caused a defer, replacing the real cause of the deferral.
6429 16. If an IP address had so many PTR records that the DNS lookup buffer
6430 was not large enough to hold them, Exim could crash while trying to process
6431 the truncated data. It now detects and logs this case.
6433 17. Further to 4.21/58, another change has been made: if (and only if) the
6434 first line of a message (the first header line) ends with CRLF, a bare LF
6435 in a subsequent header line has a space inserted after it, so as not to
6436 terminate the header.
6438 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6439 unnecessarily, used atoi() instead of strtol(), and didn't check the
6440 termination when getting file sizes from file names by regex.
6442 19. Completely re-implemented the support for maildirsize files, in the light
6443 of a number of problems with the previous contributed implementation
6444 (4.30/29). In particular:
6446 . If the quota is zero, the maildirsize file is maintained, but no quota is
6449 . If the maildir directory does not exist, it is created before any attempt
6450 to write a maildirsize file.
6452 . The quota value in the file is just a cache; if the quota is changed in
6453 the transport, the new value overrides.
6455 . A regular expression is available for excluding directories from the
6458 20. The autoreply transport checks the characters in options that define the
6459 message's headers; it allows continued headers, but it was checking with
6460 isspace() after an embedded newline instead of explicitly looking for a
6463 21. If all the "regular" hosts to which an address was routed had passed their
6464 expiry times, and had not reached their retry times, the address was
6465 bounced, even if fallback hosts were defined. Now Exim should go on to try
6468 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6469 equivalent code in the SMTP transport. Some hosts send humungous responses
6470 to HELO/EHLO, more than 1024 it seems.
6472 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6473 gives compiler warnings in some environments. I've now done it "properly",
6476 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6477 (because of problems with the built-in one) was declared to return uschar *
6478 instead of char *, causing compiler failure.
6480 25. Fixed a file descriptor leak when processing alias/forward files.
6482 26. Fixed a minor format string issue in dbfn.c.
6484 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6486 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6487 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6490 29. When a sender address is verified, it is cached, to save repeating the test
6491 when there is more than one recipient in a message. However, when the
6492 verification involves a callout, it is possible for different callout
6493 options to be set for different recipients. It is too complicated to keep
6494 track of this in the cache, so now Exim always runs a verification when a
6495 callout is required, relying on the callout cache for the optimization.
6496 The overhead is duplication of the address routing, but this should not be
6499 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6500 to be verified with callout=postmaster, and the main callout worked but the
6501 postmaster check failed, the verification correctly failed. However, if a
6502 subsequent RCPT command asked for sender verification *without* the
6503 postmaster check, incorrect caching caused this verification also to fail,
6506 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6507 it was not caching the DNS options (qualify_single, search_parents) that
6508 were used when the lookup failed. A subsequent lookup with different
6509 options therefore always gave the same answer, though there were cases
6510 where it should not have. (Example: a "domains = !$mx_any" option on a
6511 dnslookup router: the "domains" option is always processed without any
6512 widening, but the router might have qualify_single set.) Now Exim uses the
6513 cached value only when the same options are set.
6515 32. Added John Jetmore's "exipick" utility to the distribution.
6517 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6518 than a timeout (e.g. a certificate is required, and is not provided), an
6519 Exim server now closes the connection immediately. Previously it waited for
6520 the client to close - but if the client is SSL, it seems that they each
6521 wait for each other, leading to a delay before one of them times out.
6523 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6524 maintained 0.8.x compatibility because I don't think many are using it, and
6525 it is clearly obsolete.
6527 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6530 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6531 client certificate was expired. A simple patch fixes this, though I don't
6532 understand the full logic of why the verify callback is called multiple
6535 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6536 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6537 which causes problems with some clients (such as the Certicom SSL Plus
6538 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6539 disables the coutermeasure allowing Eudora to connect."
6541 38. Exim was not checking that a write() to a log file succeeded. This could
6542 lead to Bad Things if a log got too big, in particular if it hit a file
6543 size limit. Exim now panics and dies if it cannot write to a log file, just
6544 as it does if it cannot open a log file.
6546 39. Modified OS/Makefile-Linux so that it now contains
6548 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6550 The two -D definitions ensure that Exim is compiled with large file
6551 support, which makes it possible to handle log files that are bigger than
6554 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6555 instance) a domain was checked against a named list that involved a lookup,
6556 causing $domain_data to be set, then another domain was checked against the
6557 same list, then the first domain was re-checked, the value of $domain_data
6558 after the final check could be wrong. In particular, if the second check
6559 failed, it could be set empty. This bug probably also applied to
6562 41. The strip_trailing_dot option was not being applied to the address given
6563 with the -f command-line option.
6565 42. The code for reading a message's header from the spool was incrementing
6566 $received_count, but never initializing it. This meant that the value was
6567 incorrect (doubled) while delivering a message in the same process in which
6568 it was received. In the most common configuration of Exim, this never
6569 happens - a fresh exec is done - but it can happen when
6570 deliver_drop_privilege is set.
6572 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6573 it now includes up to 150 characters of the unexpected data in the log
6576 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6577 and building data strings. The size of both of these buffers was 10 000
6578 bytes - far larger than anybody would *ever* want, thought I. Needless to
6579 say, somebody hit the limit. I have increased the maximum line length to
6580 20 000 and the maximum data length of concatenated lines to 100 000. I have
6581 also fixed two bugs, because there was no checking on these buffers. Tsk,
6582 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6583 buffer is too small.
6585 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6586 lsearch lookups. Now it does.
6588 46. When parsing a route_list item in a manualroute router, a fixed-length
6589 buffer was used for the list of hosts. I made this 1024 bytes long,
6590 thinking that nobody would ever have a list of hosts that long. Wrong.
6591 Somebody had a whole pile of complicated expansion conditions, and the
6592 string was silently truncated, leading to an expansion error. It turns out
6593 that it is easier to change to an unlimited length (owing to other changes
6594 that have happened since this code was originally written) than to build
6595 structure for giving a limitation error. The length of the item that
6596 expands into the list of hosts is now unlimited.
6598 47. The lsearch lookup could not handle data where the length of text line was
6599 more than 4095 characters. Such lines were truncated, leading to shortened
6600 data being returned. It should now handle lines of any length.
6602 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6603 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6606 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6607 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6608 no longer happens in comment lines. A list of replacements is now placed
6609 at the head of all of the source files, except those whose only change is
6610 to replace PERL_COMMAND in the very first #! line.
6612 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6613 messages on the queue, with a bottom-up merge sort, using code contributed
6614 by Michael Haardt. This should make operations like -bp somewhat faster on
6615 large queues. It won't affect queue runners, except when queue_run_in_order
6618 51. Installed eximstats 1.31 in the distribution.
6620 52. Added support for SRV lookups to the dnslookup router.
6622 53. If an ACL referred to $message_body or $message_body_end, the value was not
6623 reset for any messages that followed in the same SMTP session.
6625 54. The store-handling optimization for building very long strings was not
6626 differentiating between the different store pools. I don't think this
6627 actually made any difference in practice, but I've tidied it.
6629 55. While running the routers to verify a sender address, $sender_address
6630 was still set to the sender address. This is wrong, because when routing to
6631 send a bounce to the sender, it would be empty. Therefore, I have changed
6632 it so that, while verifying a sender address, $sender_address is set to <>.
6633 (There is no change to what happens when verifying a recipient address.)
6635 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6636 target A or AAAA records (if not already returned) without resetting the
6637 qualify_single or search_parents options of the DNS resolver. These are
6638 inappropriate in this case because the targets of MX and SRV records must
6639 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6640 target that, when qualified, matched something in the local domain. These
6641 two options are now turned off when doing these lookups.
6643 57. It seems that at least some releases of Reiserfs (which does not have the
6644 concept of a fixed number of inodes) returns zero and not -1 for the
6645 number of available inodes. This interacted badly with check_spool_inodes,
6646 which assumed that -1 was the "no such thing" setting. What I have done is
6647 to check that the total number of inodes is greater than zero before doing
6648 the test of how many are available.
6650 58. When a "warn" ACL statement has a log_message modifier, the message is
6651 remembered, and not repeated. This is to avoid a lot of repetition when a
6652 message has many recipients that cause the same warning to be written.
6653 However, Exim was preserving the list of already written lines for an
6654 entire SMTP session, which doesn't seem right. The memory is now reset if a
6655 new message is started.
6657 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6658 debugging output unless log_rewrite was also set.
6660 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6661 of (int)(handle) when we know that handle contains (void *)(-1).
6663 61. The Exim daemon panic-logs an error return when it closes the incoming
6664 connection. However "connection reset by peer" seems to be common, and
6665 isn't really an error worthy of noting specially, so that particular error
6668 62. When Exim is trying to find all the local interfaces, it used to panic and
6669 die if the ioctl to get the interface flags failed. However, it seems that
6670 on at least one OS (Solaris 9) it is possible to have an interface that is
6671 included in the list of interfaces, but for which you get a failure error
6672 for this call. This happens when the interface is not "plumbed" into a
6673 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6674 failure of the "get flags" call assumes that the interface is down.
6676 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6677 makes life easier for people who are doing arithmetic on fields extracted
6678 from dates, where you often get leading zeros that should not be
6679 interpreted as octal.
6681 64. Added qualify_domain to the redirect router, to override the global
6684 65. If a pathologically long header line contained very many addresses (the
6685 report of this problem mentioned 10 000) and each of them was rewritten,
6686 Exim could use up a very large amount of memory. (It kept on making new
6687 copies of the header line as it rewrote, and never released the old ones.)
6688 At the expense of a bit more processing, the header rewriting function has
6689 been changed so that it no longer eats memory in this way.
6691 66. The generation of the Received: header has been moved from the time that a
6692 message starts to be received, to the time that it finishes. The timestamp
6693 in the Received: header should now be very close to that of the <= log
6694 line. There are two side-effects of this change:
6696 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6697 logged header lines no longer include the local Received: line, because
6698 it has not yet been created. The same applies to a copy of the message
6699 that is returned to a non-SMTP sender when a message is rejected.
6701 (b) When a filter file is tested using -bf, no additional Received: header
6702 is added to the test message. After some thought, I decided that this
6705 This change does not affect the value of $received_for. It is still set
6706 after address rewriting, but before local_scan() is called.
6708 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6710 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6711 gave an unhelpful panic error message, and a defer error. I have managed to
6712 change this behaviour so that it now rejects any supplied certificate,
6713 which seems right, as the list of acceptable certificates is empty.
6715 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6716 gave an unhelpful defer error. I have not managed to make this reject any
6717 supplied certificates, but the error message it gives is "no certificate
6718 supplied", which is not helpful.
6720 70. exigrep's output now also includes lines that are not associated with any
6721 message, but which match the given pattern. Implemented by a patch from
6722 Martin Sluka, which also tidied up the Perl a bit.
6724 71. Recipient callout verification, like sender verification, was using <> in
6725 the MAIL FROM command. This isn't really the right thing, since the actual
6726 sender may affect whether the remote host accepts the recipient or not. I
6727 have changed it to use the actual sender in the callout; this means that
6728 the cache record is now keyed on a recipient/sender pair, not just the
6729 recipient address. There doesn't seem to be a real danger of callout loops,
6730 since a callout by the remote host to check the sender would use <>.
6731 [SEE ABOVE: changed after hitting problems.]
6733 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6734 temporary errors. However, in the case of such a code being given after
6735 the end of a data transmission (i.e. after ".") Exim was failing to write
6736 a retry record for the message. (Yes, there was some broken host that was
6737 actually sending 8xx at this point.)
6739 73. An unknown lookup type in a host list could cause Exim to panic-die when
6740 the list was checked. (An example that provoked this was putting <; in the
6741 middle of a list instead of at the start.) If this happened during a DATA
6742 ACL check, a -D file could be left lying around. This kind of configuration
6743 error no longer causes Exim to die; instead it causes a defer error. The
6744 incident is still logged to the main and panic logs.
6746 74. Buglet left over from Exim 3 conversion. The message "too many messages
6747 in one connection" was written to the rejectlog but not the mainlog, except
6748 when address rewriting (yes!) was being logged.
6750 75. Added write_rejectlog option.
6752 76. When a system filter was run not as root (that is, when system_filter_user
6753 was set), the values of the $n variables were not being returned to the
6754 main process; thus, they were not subsequently available in the $sn
6757 77. Added +return_path_on_delivery log selector.
6759 78. A connection timeout was being treated differently from recipients deferred
6760 when testing hosts_max_try with a message that was older than the host's
6761 retry timeout. (The host should not be counted, thus allowing all hosts to
6762 be tried at least once before bouncing.) This may have been the cause of an
6763 occasionally reported bug whereby a message would remain on the queue
6764 longer than the retry timeout, but would be bounced if a delivery was
6765 forced. I say "may" because I never totally pinned down the problem;
6766 setting up timeout/retry tests is difficult. See also the next item.
6768 79. The ultimate address timeout was not being applied to errors that involved
6769 a combination of host plus message (for example, a timeout on a MAIL
6770 command). When an address resolved to a number of possible hosts, and they
6771 were not all tried for each delivery (e.g. because of hosts_max_try), a
6772 message could remain on the queue longer than the retry timeout.
6774 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6777 81. Fixed an obscure SMTP outgoing bug which required at least the following
6778 conditions: (a) there was another message waiting for the same server;
6779 (b) the server returned 5xx to all RCPT commands in the first message so
6780 that the message was not completed; (c) the server dropped the connection
6781 or gave a negative response to the RSET that Exim sends to abort the
6782 transaction. The observed case was a dropped connection after DATA that had
6783 been sent in pipelining mode. That is, the server had advertised PIPELINING
6784 but was not implementing it correctly. The effect of the bug was incorrect
6785 behaviour, such as trying another host, and this could lead to a crash.
6791 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6792 and daemon.c were passed as pointers to ints; they should have been
6793 pointers to socklen_t variables (which are typically unsigned ints).
6795 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6798 3. Fixed a really odd bug that affected only the testing scheme; patching a
6799 certain fixed string in the binary changed the value of another string that
6800 happened to be identical to the end of the original first string.
6802 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6803 name", it returns that address as the IP address. On some operating
6804 systems (e.g. Solaris), it also passes back the IP address string as the
6805 "host name". However, on others (e.g. Linux), it passes back an empty
6806 string. Exim wasn't checking for this, and was changing the host name to an
6807 empty string, assuming it had been canonicalized.
6809 5. Although rare, it is permitted to have more than one PTR record for a given
6810 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6811 all the names associated with an address, because they do in Solaris.
6812 However, it seems that they do not in Linux for data that comes from the
6813 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6814 I found this out when I moved to a new Linux workstation and tried to run
6815 the Exim test suite.
6817 To get round this problem I have changed the code so that it now does its
6818 own call to the DNS to look up PTR records when searching for a host name.
6819 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6820 addresses that are only in /etc/hosts are still found.
6822 This behaviour is, however, controlled by an option called host_lookup_
6823 order, which defaults to "bydns:byaddr". If people want to use the other
6824 order, or indeed, just use one or the other means of lookup, they can
6825 specify it in this variable.
6827 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6828 some operating systems, this comes back from gethostbyaddr() as an empty
6829 string, and this is what Exim used to test for. However, it seems that in
6830 other systems, "." is yielded. Exim now tests for this case too.
6832 7. The values of check_spool_space and check_log_space are now held internally
6833 as a number of kilobytes instead of an absolute number of bytes. If a
6834 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6835 kilobyte. This means that much larger values can be stored.
6837 8. Exim monitor: an attempt to get the action menu when not actually pointing
6838 at a message produces an empty menu entitled "No message selected". This
6839 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6840 no entries in it ("Shell widget menu has zero width and/or height"). So I
6841 have added a single, blank menu entry in this case.
6843 9. Added ${quote_local_part.
6845 10. MIME decoding is now applied to the contents of Subject: header lines when
6848 11. Now that a reference to $sender_host_address automatically causes a reverse
6849 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6850 host lookup before query-style lookups in lists that might use this
6851 variable. This has therefore been abolished, and the "net-" prefix is no
6852 longer necessary for query-style lookups.
6854 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6855 have been a typo for LFLAGS, so it has been changed.
6857 13. The install script calls Exim with "-C /dev/null" in order to find the
6858 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6859 to be output. However, since Exim outputs its version number before the
6860 error, it didn't break the script. It just looked ugly. I fixed this by
6861 always allowing "-C /dev/null" if the caller is root.
6863 14. Ignore overlarge ACL variable number when reading spool file - insurance
6864 against a later release with more variables having written the file.
6866 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6867 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6868 this, as well as the form without the "IPv6" on the front.
6870 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6871 OS/Makefile-Darwin file.
6873 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6874 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6876 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6877 variable. However, some OS use ints for inodes. I've added cast to long int
6878 to get rid of the compiler warning.
6880 19. I had forgotten to lock out "/../" in configuration file names when
6881 ALT_CONFIG_PREFIX was set.
6883 20. Routers used for verification do not need to specify transports. However,
6884 if such a router generated a host list, and callout was configured, Exim
6885 crashed, because it could not find a port number from the (non-existent)
6886 transport. It now assumes port 25 in this circumstance.
6888 21. Added the -t option to exigrep.
6890 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6891 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6892 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6893 them is set, LOOKUP_LSEARCH is forced.
6895 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6897 24. Added sender and host information to the "rejected by local_scan()" log
6898 line; previously there was no indication of these.
6900 25. Added .include_if_exists.
6902 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6903 Linux. It turns out that not all file systems support this. Apparently some
6904 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6905 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6906 not supported on the file descriptor, is now ignored when Exim is trying to
6907 sync a directory. This applies only to Linux.
6909 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6911 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6912 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6915 29. Added support for maildirsize files from supplied patch (modified a bit).
6917 30. The use of :fail: followed by an empty string could lead Exim to respond to
6918 sender verification failures with (e.g.):
6920 550 Verification failed for <xxx>
6921 550 Sender verify failed
6923 where the first response line was missing the '-' that indicates it is not
6924 the final line of the response.
6926 31. The loop for finding the name of the user that called Exim had a hardwired
6927 limit of 10; it now uses the value of finduser_retries, which is used for
6928 all other user lookups.
6930 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6933 33. Exim was neglecting to zero errno before one call of strtol() when
6934 expanding a string and expecting an integer value. On some systems this
6935 resulted in spurious "integer overflow" errors. Also, it was casting the
6936 result into an int without checking.
6938 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6939 did not work. The code looks as if it has *never* worked, though it appears
6940 to have been documented since at least release 1.62. I have made it work.
6942 35. The "timeout_DNS" error in retry rules, also documented since at least
6943 1.62, also never worked. As it isn't clear exactly what this means, and
6944 clearly it isn't a major issue, I have abolished the feature by treating it
6945 as "timeout", and writing a warning to the main and panic logs.
6947 36. The display of retry rules for -brt wasn't always showing the error code
6950 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6951 timeout_connect_A, timeout_connect_MX.
6953 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6954 to the empty sender.
6956 39. The daemon was not analysing the content of -oX till after it had closed
6957 stderr and disconnected from the controlling terminal. This meant that any
6958 syntax errors were only noted on the panic log, and the return code from
6959 the command was 0. By re-arranging the code a little, I've made the
6960 decoding happen first, so such errors now appear on stderr, and the return
6961 code is 1. However, the actual setting up of the sockets still happens in
6962 the disconnected process, so errors there are still only recorded on the
6965 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6966 connections (as happens on some IP stacks) was logged at start up time as
6967 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6968 it from "IPv6 and IPv4", which means that two separate sockets are being
6971 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6972 says whether AF_INET or AF_INET6 was passed as an argument.
6974 42. Exiwhat output was messed up when time zones were included in log
6977 43. Exiwhat now gives more information about the daemon's listening ports,
6978 and whether -tls-on-connect was used.
6980 44. The "port" option of the smtp transport is now expanded.
6982 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6983 silently ignored. Now an error message is written to the main and panic
6986 46. There's a new ACL modifier called "logwrite" which writes to a log file
6987 as soon as it is encountered.
6989 47. Added $local_user_uid and $local_user_gid at routing time.
6991 48. Exim crashed when trying to verify a sender address that was being
6994 49. Exim was recognizing only a space character after ".include". It now also
6995 recognizes a tab character.
6997 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6998 extracting the relevant information from the specification. The man page no
6999 longer contains scrambled data for the -d option, and I've added a section
7000 at the front about calling Exim under different names.
7002 51. Added "extra_headers" argument to the "mail" command in filter files.
7004 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7007 53. Installed eximstats 1.29.
7009 54. Added transport_filter_timeout as a generic transport option.
7011 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7012 Cc: header lines. This was required by RFC 822, but it not required by RFC
7015 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7016 incoming messages that did not have them. Now it does so only if the
7017 message originates locally, that is, if there is no associated remote host
7018 address. When Resent- header lines are present, this applies to the Resent-
7019 lines rather than the non-Resent- lines.
7021 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7022 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7024 58. Messages for configuration errors now include the name of the main
7025 configuration file - useful now that there may be more than one file in a
7026 list (.included file names were always shown).
7028 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7029 for those rare installations that do not start the daemon as root or run it
7030 setuid root. I've cut out the call to initgroups() if the daemon is not
7033 60. The Exim user and group can now be bound into the binary as text strings
7034 that are looked up at the start of Exim's processing.
7036 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7038 62. Added $mailstore_basename variable.
7040 63. Installed patch to sieve.c from Michael Haardt.
7042 64. When Exim failed to open the panic log after failing to open the main log,
7043 the original message it was trying to log was written to stderr and debug
7044 output, but if they were not available (the usual case in production), it
7045 was lost. Now it is written to syslog before the two lines that record the
7046 failures to open the logs.
7048 65. Users' Exim filters run in subprocesses under the user's uid. It is
7049 possible for a "deliver" command or an alias in a "personal" command to
7050 provoke an address rewrite. If logging of address rewriting is configured,
7051 this fails because the process is not running as root or exim. There may be
7052 a better way of dealing with this, but for the moment (because 4.30 needs
7053 to be released), I have disabled address rewrite logging when running a
7054 filter in a non-root, non-exim process.
7060 1. The buildconfig auxiliary program wasn't quoting the value set for
7061 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7062 not defined. This bug was masked in 4.22 by the effect that was fixed in
7065 2. Some messages that were rejected after a message id was allocated were
7066 shown as "incomplete" by exigrep. It no longer does this for messages that
7067 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7069 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7070 have allow_domain_literals set, the ID did not get logged in the <= line.
7071 Domain literals are now always recognized in Message-ID: header lines.
7073 4. The first argument for a ${extract expansion item is the key name or field
7074 number. Leading and trailing spaces in this item were not being ignored,
7075 causing some misleading effects.
7077 5. When deliver_drop_privilege was set, single queue runner processes started
7078 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7079 same command in the process it spins off) were not dropping privilege.
7081 6. When the daemon running as "exim" started a queue runner, it always
7082 re-executed Exim in the spun-off process. This is a waste of effort when
7083 deliver_drop_privilege is set. The new process now just calls the
7084 queue-runner function directly.
7090 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7093 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7094 ignored. Though the use of -oP was forcing the writing of a pid file, it
7095 was always written to the default place.
7097 3. If the message "no IP address found for host xxxx" is generated during
7098 incoming verification, it is now followed by identification of the incoming
7099 connection (so you can more easily find what provoked it).
7101 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7103 5. Added some features to "harden" Exim a bit more against certain attacks:
7105 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7106 be put in Local/Makefile. This is like the never_users runtime option,
7107 but it cannot be overridden. The default setting is "root".
7109 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7110 prefix string with which any file named in a -C command line option
7113 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7114 is retained for -C and -D only if the caller of Exim is root. Without
7115 it, the exim user may also use -C and -D and retain privilege.
7117 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7118 command line option is disabled.
7120 6. Macro names set by the -D option must start with an upper case letter, just
7121 like macro names defined in the configuration file.
7123 7. Added "dereference=" facility to LDAP.
7125 8. Two instances of the typo "uknown" in the source files are fixed.
7127 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7128 the Configure-Makefile script screwed up while processing it.
7130 10. Incorporated PCRE 4.4.
7132 11. The SMTP synchronization check was not operating right at the start of an
7133 SMTP session. For example, it could not catch a HELO sent before the client
7134 waited for the greeting. There is now a check for outstanding input at the
7135 point when the greeting is written. Because of the duplex, asynchronous
7136 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7137 way, but not yet received, when the check is performed.
7139 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7140 on TCP/IP sockets, because this apparently causes some broken clients to
7143 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7144 unchanged) from the Cygwin maintainer.
7146 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7147 was supported instead of testing for mbx. Effectively a typo.
7149 15. The spa authenticator server code was not checking that the input it
7150 received was valid base64.
7152 16. The debug output line for the "set" modifier in ACLs was not showing the
7153 name of the variable that was being set.
7155 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7157 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7158 lookup on its own. Something else was needed to trigger the lookup. For
7159 example, a match in host_lookup or the need for a host name in a host list.
7160 Now, if $sender_host_name is referenced and the host name has not yet been
7161 looked up, a lookup is performed. If the lookup fails, the variable remains
7162 empty, and $host_lookup_failed is set to "1".
7164 19. Added "eqi" as a case-independent comparison operator.
7166 20. The saslauthd authentication condition could segfault if neither service
7167 nor realm was specified.
7169 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7170 error message that was logged was misleading, and incoming SMTP
7171 connections were dropped. The message is now more accurate, and temporary
7172 errors are given to SMTP connections.
7174 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7175 (except RSET) with a 421 error, until QUIT is received. However, it was
7176 failing to send a response to QUIT.
7178 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7179 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7180 ACL is now run after the helo_try_verify_hosts code.
7182 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7183 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7184 case-independent, so other case variants are also recognized). Apparently
7185 some systems use these upper case variants.
7187 25. If more than two messages were waiting for the same host, and a transport
7188 filter was specified for the transport, Exim sent two messages over the
7189 same TCP/IP connection, and then failed with "socket operation on non-
7190 socket" when it tried to send the third.
7192 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7194 27. The extern definition of crypt16() in expand.c was not being excluded when
7195 the OS had its own crypt16() function.
7197 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7198 as a preferred synonym for return_size_limit, both as an option and as an
7201 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7203 30. Changed the default configuration ACL to relax the local part checking rule
7204 for addresses that are not in any local domains. For these addresses,
7205 slashes and pipe symbols are allowed within local parts, but the sequence
7206 /../ is explicitly forbidden.
7208 31. SPA server authentication was not clearing the challenge buffer before
7211 32. log_message in a "warn" ACL statement was writing to the reject log as
7212 well as to the main log, which contradicts the documentation and doesn't
7213 seem right (because no rejection is happening). So I have stopped it.
7215 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7216 However, I am unable to do any testing of this.
7218 34. Fixed an infelicity in the appendfile transport. When checking directories
7219 for a mailbox, to see if any needed to be created, it was accidentally
7220 using path names with one or more superfluous leading slashes; tracing
7221 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7223 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7224 contents are added to the log line that is written for every discarded
7225 recipient. (Previously a log_message setting was ignored.)
7227 36. The ${quote: operator now quotes the string if it is empty.
7229 37. The install script runs exim in order to find its version number. If for
7230 some reason other than non-existence or emptiness, which it checks, it
7231 could not run './exim', it was installing it with an empty version number,
7232 i.e. as "exim-". This error state is now caught, and the installation is
7235 38. An argument was missing from the function that creates an error message
7236 when Exim fails to connect to the socket for saslauthd authentication.
7237 This could cause Exim to crash, or give a corrupted message.
7239 39. Added isip, isip4, and isip6 to ${if conditions.
7241 40. The ACL variables $acl_xx are now saved with the message, and can be
7242 accessed later in routers, transports, and filters.
7244 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7245 strings in the file are not string-expanded.
7247 42. If a MAIL command specified a SIZE value that was too large to fit into an
7248 int variable, the check against message_size_limit failed. Such values are
7249 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7250 day this will have to be increased, but I don't think I want to be around
7251 when emails are that large.
7258 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7259 iconv() is not standard in FreeBSD.
7261 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7262 IPv6 enabled. The observed symptom was a segmentation fault on return from
7263 the function os_common_find_running_interfaces() in src/os.c.
7265 3. In the check_special_case() function in daemon.c I had used "errno" as an
7266 argument name, which causes warnings on some systems. This was basically a
7267 typo, since it was named "eno" in the comments!
7269 4. The code that waits for the clock to tick (at a resolution of some fraction
7270 of a second) so as to ensure message-id uniqueness was always waiting for
7271 at least one whole tick, when it could have waited for less. [This is
7272 almost certainly not relevant at current processor speeds, where it is
7273 unlikely to ever wait at all. But we try to future-proof.]
7275 5. The function that sleeps for a time interval that includes fractions of a
7276 second contained a race. It did not block SIGALRM between setting the
7277 timer, and suspending (a couple of lines later). If the interval was short
7278 and the sigsuspend() was delayed until after it had expired, the suspension
7279 never ended. On busy systems this could lead to processes getting stuck for
7282 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7283 process, before it forks any delivery processes. The open lookup caching
7284 mechanism meant that the open file or database connection was passed into
7285 the delivery process. The problem was that delivery processes always tidy
7286 up cached lookup data. This could cause a problem for the next delivery
7287 process started by the queue runner, because the external queue runner
7288 process does not know about the closure. So the next delivery process
7289 still has data in the lookup cache. In the case of a file lookup, there was
7290 no problem because closing a file descriptor in a subprocess doesn't affect
7291 the parent. However, if the lookup was caching a connection to a database,
7292 the connection was closed, and the second delivery process was likely to
7293 see errors such as "PGSQL: query failed: server closed the connection
7294 unexpectedly". The problem has been fixed by closing all cached lookups
7295 in a queue runner before running a delivery process.
7297 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7298 seem to have the "const" qualifier which it has on other OS. I've
7301 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7302 *of the same type* (client or server) with the same public name that an
7303 error should be diagnosed.
7305 9. When Exim looked up a host name for an IP address, but failed to find the
7306 original IP address when looking up the host name (a safety check), it
7307 output the message "<ip address> does not match any IP for NULL", which was
7308 confusing, to say the least. The bug was that the host name should have
7309 appeared instead of "NULL".
7311 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7312 user that is built into the binary, and the -C or -D options is used, root
7313 privilege is dropped before the configuration file is read. In addition,
7314 logging is switched to stderr instead of the normal log files. If the
7315 configuration then re-defines the Exim user, the unprivileged environment
7316 is probably not what is expected, so Exim logs a panic warning message (but
7319 However, if deliver_drop_privilege is set, the unprivileged state may well
7320 be exactly what is intended, so the warning has been cut out in that case,
7321 and Exim is allowed to try to write to its normal log files.
7327 1. smtp_return_error_details was not giving details for temporary sender
7328 or receiver verification errors.
7330 2. Diagnose a configuration error if two authenticators have the same public
7333 3. Exim used not to create the message log file for a message until the first
7334 delivery attempt. This could be confusing when incoming messages were held
7335 for policy or load reasons. The message log file is now created at the time
7336 the message is received, and an initial "Received" line is written to it.
7338 4. The automatically generated man page for command line options had a minor
7339 bug that caused no ill effects; however, a more serious problem was that
7340 the procedure for building the man page automatically didn't always
7341 operate. Consequently, release 4.20 contains an out-of-date version. This
7342 shouldn't happen again.
7344 5. When building Exim with embedded Perl support, the script that builds the
7345 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7346 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7348 6. The freeze_tell option was not being used for messages that were frozen on
7349 arrival, either by an ACL or by local_scan().
7351 7. Added the smtp_incomplete_transaction log selector.
7353 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7354 was accepting AUTH without a new EHLO.
7356 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7357 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7360 10. Logging of TCP/IP connections (when configured) now happens in the main
7361 daemon process instead of the child process, so that the TCP/IP connection
7362 count is more accurate (but it can never be perfect).
7364 11. The use of "drop" in a nested ACL was not being handled correctly in the
7365 outer ACL. Now, if condition failure induced by the nested "drop" causes
7366 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7367 or "require"), the connection is dropped.
7369 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7370 that yield "discard" can now be used with an "accept" or a "discard" verb,
7371 but an error is generated for any others (because I can't see a useful way
7372 to define what should happen).
7374 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7375 are now processed in the same way as lines in the Exim configuration file.
7376 In particular, continuation lines are supported.
7378 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7380 15. Added -ti meaning -t -i.
7382 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7383 domains, and warn by logging if others are found.
7385 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7386 returned from GIFCONF when Exim is trying to find the list of interfaces on
7387 a host. The code in os.c has been modified to copy each ifreq to an aligned
7388 structure in all cases.
7390 Also, in some cases, the returned ifreq's were being copied to a 'struct
7391 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7392 means the last couple of bytes of an IPv6 address could be chopped if the
7393 ifreq contained only a normal sockaddr (14 bytes storage).
7395 18. Named domain lists were not supported in the hosts_treat_as_local option.
7396 An entry such as +xxxx was not recognized, and was treated as a literal
7399 19. Ensure that header lines added by a DATA ACL are included in the reject log
7400 if the ACL subsequently rejects the message.
7402 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7403 MD5 (which is deprecated).
7405 21. When testing a filter file using -bf, Exim was writing a message when it
7406 took the sender from a "From " line in the message, but it was not doing so
7407 when it took $return_path from a Return-Path: header line. It now does.
7409 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7410 with a valid header line field name (a series of printing characters
7411 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7413 23. Changed "disc" in the source to "disk" to conform to the documentation and
7414 the book and for uniformity.
7416 24. Ignore Sendmail's -Ooption=value command line item.
7418 25. When execve() failed while trying to run a command in a pipe transport,
7419 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7420 could be confused with a return value of 69 from the command itself. This
7421 has been changed to 127, the value the shell returns if it is asked to run
7422 a non-existent command. The wording for the related log line suggests a
7423 non-existent command as the problem.
7425 26. If received_header_text expands to an empty string, do not add a Received:
7426 header line to the message. (Well, it adds a token one on the spool, but
7427 marks it "old" so that it doesn't get used or transmitted.)
7429 27. Installed eximstats 1.28 (addition of -nt option).
7431 28. There was no check for failure on the call to getsockname() in the daemon
7432 code. This can fail if there is a shortage of resources on the system, with
7433 ENOMEM, for example. A temporary error is now given on failure.
7435 29. Contrary to the C standard, it seems that in some environments, the
7436 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7437 program. Exim now does this explicitly; it affects the formatting of
7438 timestamps using strftime().
7440 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7441 complaints. I've now initialized all the variables, to avoid this.
7443 32. Header lines added by a system filter were not being "seen" during
7444 transport-time rewrites.
7446 33. The info_callback() function passed to OpenSSL is set up with type void
7447 (*)(SSL *, int, int), as described somewhere. However, when calling the
7448 function (actually a macro) that sets it up, the type void(*)() is
7449 expected. I've put in a cast to prevent warnings from picky compilers.
7451 34. If a DNS black list lookup found a CNAME record, but there were no A
7452 records associated with the domain it pointed at, Exim crashed.
7454 35. If a DNS black list lookup returned more than one A record, Exim ignored
7455 all but the first. It now scans all returned addresses if a particular IP
7456 value is being sought. In this situation, the contents of the
7457 $dnslist_value variable are a list of all the addresses, separated by a
7460 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7461 to accept a host name and all its aliases if the forward lookup for any of
7462 them yielded the IP address of the incoming connection. Now it accepts only
7463 those names whose forward lookup yields the correct IP address. Any other
7464 names are discarded. This closes a loophole whereby a rogue DNS
7465 administrator could create reverse DNS records to break through a
7466 wildcarded host restriction in an ACL.
7468 37. If a user filter or a system filter that ran in a subprocess used any of
7469 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7470 the wrong values were passed to the pipe command ($thisaddress had the
7471 value of $0, $0 had the value of $1, etc). This bug was introduced by
7472 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7474 38. Improved the line breaking for long SMTP error messages from ACLs.
7475 Previously, if there was no break point between 40 and 75 characters, Exim
7476 left the rest of the message alone. Two changes have been made: (a) I've
7477 reduced the minimum length to 35 characters; (b) if it can't find a break
7478 point between 35 and 75 characters, it looks ahead and uses the first one
7479 that it finds. This may give the occasional overlong line, but at least the
7480 remaining text gets split now.
7482 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7483 file descriptors might be low, and that setting 1000 would always raise it.
7484 It turns out that in some environments, the limit is already over 1000 and
7485 that lowering it causes trouble. So now Exim takes care not to decrease it.
7487 40. When delivering a message, the value of $return_path is set to $sender_
7488 address at the start of routing (routers may change the value). By an
7489 oversight, this default was not being set up when an address was tested by
7490 -bt or -bv, which affected the outcome if any router or filter referred to
7493 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7494 log only when "message" or "log_message" are set. However, if one of the
7495 conditions was an address verification, or a call to a nested ACL, the
7496 messages generated by the underlying test were being passed through. This
7497 no longer happens. The underlying message is available in $acl_verify_
7498 message for both "message" and "log_message" expansions, so it can be
7499 passed through if needed.
7501 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7502 new expansion $bh_ to give the encoded byte string without charset
7503 translation. Translation happens only if iconv() is available; HAVE_ICONV
7504 indicates this at build time. HEADERS_CHARSET gives the charset to
7505 translate to; headers_charset can change it in the configuration, and
7506 "headers charset" can change it in an individual filter file.
7508 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7509 that creates RFC 2047 encoded "words" labels them as that charset instead
7510 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7511 expansion operator; (ii) when Exim creates a From: line for a local
7512 message; (iii) when a header line is rewritten to include a "phrase" part.
7514 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7515 buggy, causing it to skip the first lines of messages whose message ID
7516 ended in 'D'. This would not have bitten before Exim release 4.14, because
7517 message IDs were unlikely to end in 'D' before then. The effect was to have
7518 incorrect size information for certain domains.
7520 45. #include "config.h" was missing at the start of the crypt16.c module. This
7521 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7524 46. If there was a timeout during a "random" callout check, Exim treated it as
7525 a failure of the random address, and carried on sending RSET and the real
7526 address. If the delay was just some slowness somewhere, the response to the
7527 original RCPT would be taken as a response to RSET and so on, causing
7528 mayhem of various kinds.
7530 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7531 when I implemented it. It didn't allow for the fact that some option values
7532 may legitimately be negative (e.g. size_addition), and it didn't even do
7533 the right test for positive values.
7535 48. Domain names in DNS records are case-independent. Exim always looks them up
7536 in lower case. Some resolvers return domain names in exactly the case they
7537 appear in the zone file, that is, they may contain uppercase letters. Not
7538 all resolvers do this - some return always lower case. Exim was treating a
7539 change of case by a resolver as a change of domain, similar to a widening
7540 of a domain abbreviation. This triggered its re-routing code and so it was
7541 trying to route what was effectively the same domain again. This normally
7542 caused routing to fail (because the router wouldn't handle the domain
7543 twice). Now Exim checks for this case specially, and just changes the
7544 casing of the domain that it ultimately uses when it transmits the message
7547 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7550 50. If a filter generated a file delivery with a non-absolute name (possible if
7551 no home directory exists for the router), the forbid_file option was not
7554 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7555 the existing equality matching.
7557 52. Exim was using ints instead of ino_t variables in some places where it was
7558 dealing with inode numbers.
7560 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7561 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7562 called TMPDIR, and if it finds it is different, it changes its value.
7564 54. The smtp_printf() function is now made available to local_scan() so
7565 additional output lines can be written before returning. There is also an
7566 smtp_fflush() function to enable the detection of a dropped connection.
7567 The variables smtp_input and smtp_batched_input are exported to
7570 55. Changed the default runtime configuration: the message "Unknown user"
7571 has been removed from the ACL, and instead placed on the localuser router,
7572 using the cannot_route_message feature. This means that any verification
7573 failures that generate their own messages won't get overridden. Similarly,
7574 the "Unrouteable address" message that was in the ACL for unverifiable
7575 relay addresses has also been removed.
7577 56. Added hosts_avoid_esmtp to the smtp transport.
7579 57. The exicyclog script was not checking for the esoteric option
7580 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7581 will work only if exicyclog is run under the appropriate euid.
7583 58. Following a discussion on the list, the rules by which Exim recognises line
7584 endings on incoming messages have been changed. The -dropcr and drop_cr
7585 options are now no-ops, retained only for backwards compatibility. The
7586 following line terminators are recognized: LF CRLF CR. However, special
7587 processing applies to CR:
7589 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7590 nor a local message in the state where . is a terminator.
7592 (ii) If a bare CR is encountered in a header line, an extra space is added
7593 after the line terminator so as not to end the header. The reasoning
7594 behind this is that bare CRs in header lines are most likely either
7595 to be mistakes, or people trying to play silly games.
7597 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7598 was being incorrectly given as 18 bytes larger than it should have been.
7599 This is a VOB (very old bug).
7601 60. This may never have affected anything current, but just in case it has:
7602 When the local host is found other than at the start of a list of hosts,
7603 the local host, those with the same MX, and any that follow, are discarded.
7604 When the list in question was part of a longer list of hosts, the following
7605 hosts (not currently being processed) were also being discarded. This no
7606 longer happens. I'm not sure if this situation could ever has previously
7609 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7612 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7613 header. This is something that is recommended in a new Internet Draft, and
7614 is something that is documented as being done by Sendmail. There are two
7615 possible values. For messages generated by the autoreply transport, Exim
7618 Auto-Submitted: auto-replied
7620 whereas for all other generated messages (e.g. bounces) it adds
7622 Auto-Submitted: auto-generated
7624 63. The "personal" condition in filters now includes a test for the
7625 Auto-Submitted: header. If it contains the string "auto-" the message it
7626 not considered personal.
7628 64. Added rcpt_include_affixes as a generic transport option.
7630 65. Added queue_only_override (default true).
7632 66. Added the syslog_duplication option.
7634 67. If what should have been the first header line of a message consisted of
7635 a space followed by a colon, Exim was mis-interpreting it as a header line.
7636 It isn't of course - it is syntactically invalid and should therefore be
7637 treated as the start of the message body. The misbehaviour could have
7638 caused a number of strange effects, including loss of data in subsequent
7639 header lines, and spool format errors.
7641 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7642 client host had authenticated. This control can now be exercised by an ACL
7643 for more flexibility.
7645 69. By default, callouts do not happen when testing with -bh. There is now a
7646 variant, -bhc, which does actually run the callout code, including
7647 consulting and updating the callout cache.
7649 70. Added support for saslauthd authentication, courtesy of Alexander
7652 71. If statvfs() failed on the spool or log directories while checking their
7653 size for availability, Exim confusingly gave the error "space shortage".
7654 Furthermore, in debugging mode it crashed with a floating point exception.
7655 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7656 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7657 serious problem, Exim now writes to the main and panic logs when this
7658 happens, with details of the failure. It then refuses to accept the
7659 incoming message, giving the message "spool directory problem" or "log
7660 directory problem" with a 421 code for SMTP messages.
7662 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7663 0, 1, and 2 exist, because some OS complain for execs without them (see
7664 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7665 descriptors. However, the code omitted to check that the open succeeded,
7666 causing mysterious errors if for some reason the permissions on /dev/null
7667 got screwed. Now Exim writes a message to the main and panic logs, and
7668 bombs out if it can't open /dev/null.
7670 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7671 interact so that it is all more flexible. It is supposed to remain
7672 backwards compatible. Also added extra_local_interfaces.
7674 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7675 to bomb out with an assertion failure - to the client this appears as a
7676 connection drop. This problem occurs in the part of the code that was taken
7677 from the Samba project. Fortunately, the assertion is in a very simple
7678 function, so I have fixed this by reproducing the function inline in the
7679 one place where it is called, and arranging for authentication to fail
7680 instead of killing the process with assert().
7682 75. The SPA client code was not working when the server requested OEM rather
7683 than Unicode encoding.
7685 76. Added code to make require_files with a specific uid setting more usable in
7686 the case where statting the file as root fails - usually a non-root-mounted
7687 NFS file system. When this happens and the failure is EACCES, Exim now
7688 forks a subprocess and does the per-uid checking as the relevant uid.
7690 77. Added process_log_path.
7692 78. If log_file_path was not explicitly set, a setting of check_log_space or
7693 check_log_inodes was ignored.
7695 79. If a space check for the spool or log partitions fails, the incident is now
7696 logged. Of course, in the latter case the data may get lost...
7698 80. Added the %p formatting code to string_format() so that it can be used to
7699 print addresses in debug_print(). Adjusted all the address printing in the
7700 debugging in store.c to use %p rather than %d.
7702 81. There was a concern that a line of code in smtp_in.c could overflow a
7703 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7704 initially expressed, the concern was not well-founded, because trailing
7705 spaces are removed early. However, if the trailing spaces were followed by
7706 a NULL, they did not get removed, so the overflow was possible. Two fixes
7709 (a) I re-wrote the offending code in a cleaner fashion.
7710 (b) If an incoming SMTP command contains a NULL character, it is rejected
7713 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7714 runs initgroups(), so that if the Exim user is in any additional groups,
7715 they will be used during message reception.
7721 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob