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.
143 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
144 message following a mime one, the variable was not reset.
146 JH/30 When an pipelined-connect fails at the first response, assume incorrect
147 cached capability (perhaps the peer reneged?) and immediately retry in
154 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
155 side, to decode a TLS 1.3 packet capture.
157 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
158 Previously the default library behaviour applied, sending two, each in
161 JH/03 Debug output for ACL now gives the config file name and line number for
164 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
166 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
168 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
169 buffer overrun for (non-chunking) other transports.
171 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
172 TLS1.3, means that a server rejecting a client certificate is not visible
173 to the client until the first read of encrypted data (typically the
174 response to EHLO). Add detection for that case and treat it as a failed
175 TLS connection attempt, so that the normal retry-in-clear can work (if
176 suitably configured).
178 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
179 and/or domain. Found and fixed by Jason Betts.
181 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
182 configuration). If a CNAME target was not a wellformed name pattern, a
185 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
186 the OS reports them interleaved with other addresses.
188 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
189 used both for input and for a verify callout, both encrypted, SMTP
190 responses being sent by the server could be lost. This resulted in
191 dropped connections and sometimes bounces generated by a peer sending
194 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
195 string. Previously it was possible to cause undefined behaviour in a
196 library routine (usually a crash). Found by "zerons".
198 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
201 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
202 API was removed, so update to use the newer ones.
204 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
205 any timeout set, is taking a long time. Previously we would hang on to a
206 rotated logfile "forever" if the input was arriving with long gaps
207 (a previous attempt to fix addressed lack, for a long time, of initial
210 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
211 shared (NFS) environment. The length of the tempfile name is now
212 4 + 16 ("hdr.$message_exim_id") which might break on file
213 systems which restrict the file name length to lower values.
216 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
217 shared (NFS) environment.
219 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
220 did for all versions <4.90). Notably -M, -m, --invert, -I may be
223 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
224 on some platforms for bit 31.
226 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
227 to changes apparently associated with TLS1.3 handling some of the APIs
228 previously used were either nonfunctional or inappropriate. Strings
229 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
230 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
231 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
232 This affects log line X= elements, the $tls_{in,out}_cipher variables,
233 and the use of specific cipher names in the encrypted= ACL condition.
235 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
237 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
238 verification result was not updated unless hosts_require_ocsp applied.
240 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
241 queue_list_requires_admin set to false, non-admin users were denied the
244 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
245 directory-of-certs mode. Previously they were advertised despite the
248 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
249 A single TCP connection by a client will now hold a TLS connection open
250 for multiple message deliveries, by default. Previoud the default was to
253 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
254 default. If built with the facility, DANE will be used. The facility
255 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
257 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
258 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
259 must be defined and you must still, unless you define DISABLE_TLS, manage
260 the the include-dir and library-file requirements that go with that
261 choice. Non-TLS builds are still supported.
263 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
266 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
267 default. If the platform supports and has the facility enabled, it will
268 be requested on all coneections.
270 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
271 controlled by the build-time option SUPPORT_PIPE_CONNECT.
273 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
275 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
276 success-DSN messages. Previously the From: header was always the default
277 one for these; the option was ignored.
279 JH/28 Fix the timeout on smtp response to apply to the whole response.
280 Previously it was reset for every read, so a teergrubing peer sending
281 single bytes within the time limit could extend the connection for a
282 long time. Credit to Qualsys Security Advisory Team for the discovery.
284 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
285 delivery address, which leaked information of the results of local
286 forwarding. Change to the original envelope recipient address, per
289 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
290 requested. Previously not bounce was generated and a log entry of
291 error ignored was made.
293 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
295 JH/32 Introduce a general tainting mechanism for values read from the input
296 channel, and values derived from them. Refuse to expand any tainted
297 values, to catch one form of exploit.
299 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
300 was unused and the unexpanded text used for the test. Found and
301 fixed by Ruben Jenster.
303 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
304 an attempt to use a TLS library read routine dereffed a nul pointer,
307 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
308 them in our smtp response.
310 JH/36 Have the general router option retry_use_local_part default to true when
311 any of the restrictive preconditions are set (to anything). Previously it
312 was only for check_local user. The change removes one item of manual
313 configuration which is required for proper retries when a remote router
314 handles a subset of addresses for a domain.
316 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
317 link count into consideration.
319 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
320 caused the extension of big_buffer, the following lines were ignored.
322 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
323 accordance with RFC 2308. Previously there was no expiry, so a longlived
324 receive process (eg. due to ACL delays) versus a short SOA value could
327 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
329 JH/39 Promote DMARC support to mainline.
331 JH/40 Bug 2452: Add a References: header to DSNs.
333 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
334 parameters. The relevant library call is documented as "Deprecated: This
335 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
336 3.6.0, DH parameters are negotiated following RFC7919."
338 HS/06 Change the default of dnssec_request_domains to "*"
340 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
341 carried on and emitted a BDAT command, even when PIPELINING was not
344 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
345 buffer was used for the filename, resulting in a trap when tainted
346 arguments (eg. $domain) were used.
348 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
349 recommended to avoid a possible server-load attack. The feature can be
350 re-enabled via the openssl_options main cofiguration option.
352 JH/45 local_scan API: documented the current smtp_printf() call. This changed
353 for version 4.90 - adding a "more data" boolean to the arguments.
354 Bumped the ABI version number also, this having been missed previously;
355 release versions 4.90 to 4.92.3 inclusive were effectively broken in
356 respect of usage of smtp_printf() by either local_scan code or libraries
357 accessed via the ${dlfunc } expansion item. Both will need coding
358 adjustment for any calls to smtp_printf() to match the new function
359 signature; a FALSE value for the new argument is always safe.
361 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
362 the file-offset (which the Linux syscall does, and exim expects); this
363 resulted in an indefinite loop.
365 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
366 to do ARC verification. The Authentication-Results: header line added
367 by the configuration then had no ARC item.
373 JH/01 Remove code calling the customisable local_scan function, unless a new
374 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
376 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
377 non-signal-safe functions being used.
379 JH/03 Bug 2269: When presented with a received message having a stupidly large
380 number of DKIM-Signature headers, disable DKIM verification to avoid
381 a resource-consumption attack. The limit is set at twenty.
383 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
384 report of oldest_pass in ${authres } in consequence, and separate out
385 some descriptions of reasons for verification fail.
387 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
388 files in the spool were present and unlocked. A queue-runner could spot
389 them, resulting in a duplicate delivery. Fix that by doing the unlock
390 after the unlink. Investigation by Tim Stewart. Take the opportunity to
391 add more error-checking on spoolfile handling while that code is being
394 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
395 No known attacks, no CVE, this is defensive hardening.
397 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
398 a queue-runner could start a delivery while other operations were ongoing.
399 Cutthrough delivery was a common victim, resulting in duplicate delivery.
400 Found and investigated by Tim Stewart. Fix by using the open message data
401 file handle rather than opening another, and not locally closing it (which
402 releases a lock) for that case, while creating the temporary .eml format
403 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
405 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
406 $sender_verify_failure/$recipient_verify_failure to "random".
408 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
411 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
412 Previously this would segfault.
414 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
417 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
418 like zero, since the resolver should be doing this for us, But we need one
419 as a CNAME but no MX presence gets the CNAME returned; we need to check
420 that doesn't point to an MX to declare it "no MX returned" rather than
421 "error, loop". A new main option is added so the older capability of
422 following some limited number of chain links is maintained.
424 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
426 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
427 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
428 now-deprecated earlier definitions used only specified the range up to TLS
429 1.2 (in the older-version library docs).
431 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
433 JH/15 Rework TLS client-side context management. Stop using a global, and
434 explicitly pass a context around. This enables future use of TLS for
435 connections to service-daemons (eg. malware scanning) while a client smtp
436 connection is using TLS; with cutthrough connections this is quite likely.
438 JH/16 Fix ARC verification to do AS checks in reverse order.
440 JH/17 Support a "tls" option on the ${readsocket } expansion item.
442 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
443 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
444 Previously the "utf8" would be re-prepended for every additional message.
446 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
447 Previously thery were accepted, resulting in issues when attempting to
448 forward messages to a non-supporting MTA.
450 PP/02 Let -n work with printing macros too, not just options.
452 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
453 one parent address was copied, and bogus data was used at delivery-logging
454 time. Either a crash (after delivery) or bogus log data could result.
455 Discovery and analysis by Tim Stewart.
457 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
458 Previously if the string ended mid-character, we did not insert the
459 promised '?' replacement.
461 PP/04 Documentation: current string operators work on bytes, not codepoints.
463 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
464 should pack well giving a smaller memory footprint so better caching and
465 therefore performance. Group the declarations where this can't be done so
466 that the byte-sized flag variables are not interspersed among pointer
467 variables, giving a better chance of good packing by the compiler.
469 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
470 non-null, to avoid issues with sites running BATV. Previously reports were
471 sent with an empty envelope sender so looked like bounces.
473 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
474 The ignore_error flag wasn't being returned from the filter subprocess so
475 was not set for later routers. Investigation and fix by Matthias Kurz.
477 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
478 and a msg:complete for the whole, when a message is manually removed using
479 -Mrm. Developement by Matthias Kurz, hacked on by JH.
481 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
482 a "Gnu special" function, asprintf() in the DB utility binary builds; I
483 hope that is portable enough.
485 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
486 requiring a known-CA anchor certificate; make it now rely entirely on the
487 TLSA as an anchor. Checking the name on the leaf cert against the name
488 on the A-record for the host is still done for TA (but not for EE mode).
490 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
491 included in delivery lines for non-proxied connections, when compiled with
492 SUPPORT_SOCKS and running with proxy logging enabled.
494 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
495 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
496 move the existing event to fire before the normal logging of message
497 failure so that custom logging is bracketed by normal logging.
499 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
500 msg:fail:internal event. Developement by Matthias Kurz.
502 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
503 far too small for todays use of crypto signatures stored there. Go all
504 the way to the max DNS message size of 64kB, even though this might be
505 overmuch for IOT constrained device use.
507 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
508 copy a string over itself. The library routine is documented as not
509 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
511 JH/32 For main options check_spool_space and check_inode_space, where the
512 platform supports 64b integers, support more than the previous 2^31 kB
513 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
514 the previous G, M, k.
516 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
517 $authenticated_fail_id variable on authentication failure. Previously
520 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
521 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
522 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
523 GNUTLS_SEC_PARAM_MEDIUM.
525 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
526 side. Previously we would continue as if no SNI had been received.
528 JH/36 Harden the handling of string-lists. When a list consisted of a sole
529 "<" character, which should be a list-separator specification, we walked
530 off past the nul-terimation.
532 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
533 causes) even when the retry time is not yet met. Previously they were
534 not, meaning that when (say) an account was over-quota and temp-rejecting,
535 and multiple senders' messages were queued, only one sender would get
536 notified on each configured delay_warning cycle.
538 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
540 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
541 cipher-suites, an error can be left on the stack even for a succeeding
542 accept; this results in impossible error messages when a later operation
545 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
546 return error codes indicating retry. Under TLS1.3 this becomes required.
548 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
549 it only wrote the new authenticators, resulting in a lack of tracking of
550 peer changes of ESMTP extensions until the next cache flush.
552 JH/41 Fix the loop reading a message header line to check for integer overflow,
553 and more-often against header_maxsize. Previously a crafted message could
554 induce a crash of the recive process; now the message is cleanly rejected.
556 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
557 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
563 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
564 When redis_servers is set to a list of > 1 element, and the Redis servers
565 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
566 case of MOVED into a DEFER case instead, thus moving the query onto the
567 next server in the list. For a cluster of N elements, all N servers must
568 be defined in redis_servers.
570 GF/02 Catch and remove uninitialized value warning in exiqsumm
571 Check for existence of @ARGV before looking at $ARGV[0]
573 JH/01 Replace the store_release() internal interface with store_newblock(),
574 which internalises the check required to safely use the old one, plus
575 the allocate and data copy operations duplicated in both (!) of the
576 extant use locations.
578 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
579 modifier. This matches the restriction on the commandline.
581 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
582 Previously only the last row was returned.
584 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
585 we assumed that tags in the header were well-formed, and parsed the
586 element content after inspecting only the first char of the tag.
587 Assumptions at that stage could crash the receive process on malformed
590 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
591 While running the DKIM ACL we operate on the Permanent memory pool so that
592 variables created with "set" persist to the DATA ACL. Also (at any time)
593 DNS lookups that fail create cache records using the Permanent pool. But
594 expansions release any allocations made on the current pool - so a dnsdb
595 lookup expansion done in the DKIM ACL releases the memory used for the
596 DNS negative-cache, and bad things result. Solution is to switch to the
597 Main pool for expansions.
598 While we're in that code, add checks on the DNS cache during store_reset,
599 active in the testsuite.
600 Problem spotted, and debugging aided, by Wolfgang Breyha.
602 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
603 When none of the hosts presented to a transport match an already-open
604 connection, close it and proceed with the list. Previously we would
605 queue the message. Spotted by Lena with Yahoo, probably involving
608 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
609 Previously a spurious "250 OK id=" response was appended to the proper
612 JH/08 The "support for" informational output now, which built with Content
613 Scanning support, has a line for the malware scanner interfaces compiled
614 in. Interface can be individually included or not at build time.
616 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
617 by the template makefile "src/EDITME". The "STREAM" support for an older
618 ClamAV interface method is removed.
620 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
621 rows affected is given instead).
623 JH/11 The runtime Berkeley DB library version is now additionally output by
624 "exim -d -bV". Previously only the compile-time version was shown.
626 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
627 SMTP connection. Previously, when one had more receipients than the
628 first, an abortive onward connection was made. Move to full support for
629 multiple onward connections in sequence, handling cutthrough connection
630 for all multi-message initiating connections.
632 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
633 routers. Previously, a multi-recipient message would fail to match the
634 onward-connection opened for the first recipient, and cause its closure.
636 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
637 a timeout on read on a GnuTLS initiating connection, resulting in the
638 initiating connection being dropped. This mattered most when the callout
639 was marked defer_ok. Fix to keep the two timeout-detection methods
642 JH/15 Relax results from ACL control request to enable cutthrough, in
643 unsupported situations, from error to silently (except under debug)
644 ignoring. This covers use with PRDR, frozen messages, queue-only and
647 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
649 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
650 metadata, resulting in a crash in free().
652 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
653 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
654 Broken also in d185889f4, with init system revamp.
656 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
657 Previously we abruptly closed the connection after reading a malware-
658 found indication; now we go on to read the "scan ok" response line,
661 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
662 ACL. Previously, a crash would result.
664 JH/19 Speed up macro lookups during configuration file read, by skipping non-
665 macro text after a replacement (previously it was only once per line) and
666 by skipping builtin macros when searching for an uppercase lead character.
668 JH/20 DANE support moved from Experimental to mainline. The Makefile control
669 for the build is renamed.
671 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
672 was allocated for every new TLS startup, meaning one per message. Fix
673 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
675 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
676 reported the original. Fix to report (as far as possible) the ACL
677 result replacing the original.
679 JH/23 Fix memory leak during multi-message connections using STARTTLS under
680 OpenSSL. Certificate information is loaded for every new TLS startup,
681 and the resources needed to be freed.
683 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
685 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
688 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
689 DATA response info to the (existing) per-recipient response info for
690 the "C=" log element. It can have useful tracking info from the
691 destination system. Patch from Simon Arlott.
693 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
694 length value. Previously this would segfault.
696 HS/02 Support Avast multiline protoocol, this allows passing flags to
697 newer versions of the scanner.
699 JH/28 Ensure that variables possibly set during message acceptance are marked
700 dead before release of memory in the daemon loop. This stops complaints
701 about them when the debug_store option is enabled. Discovered specifically
702 for sender_rate_period, but applies to a whole set of variables.
703 Do the same for the queue-runner and queue-list loops, for variables set
704 from spool message files. Do the same for the SMTP per-message loop, for
705 certain variables indirectly set in ACL operations.
707 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
708 as a multi-recipient message from a mailinglist manager). The coding had
709 an arbitrary cutoff number of characters while checking for more input;
710 enforced by writing a NUL into the buffer. This corrupted long / fast
711 input. The problem was exposed more widely when more pipelineing of SMTP
712 responses was introduced, and one Exim system was feeding another.
713 The symptom is log complaints of SMTP syntax error (NUL chars) on the
714 receiving system, and refused recipients seen by the sending system
715 (propating to people being dropped from mailing lists).
716 Discovered and pinpointed by David Carter.
718 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
719 replaced by the ${authresults } expansion.
721 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
723 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
724 allows proper process termination in container environments.
726 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
727 Previously the "final dot" had a newline after it; ensure it is CR,LF.
729 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
730 and "err_perm", deprecated since 4.83 when the RFC-defined words
731 "temperror" and "permerror" were introduced.
733 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
734 transport-filters or DKIM-signing. The restriction was lost in the
735 consolidation of verify-callout and delivery SMTP handling.
736 Extend the restriction to also cover ARC-signing.
738 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
739 in defer=pass mode supply a 450 to the initiator. Previously the message
742 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
743 tls_require_ciphers is used as before.
745 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
746 "pass_unscanned". Only tmpfails from the scanner are written to
747 the paniclog, as they may require admin intervention (permission
748 denied, license issues). Other scanner errors (like decompression
749 bombs) do not cause a paniclog entry.
751 JH/36 Fix reinitialisation of DKIM logging variable between messages.
752 Previously it was possible to log spurious information in receive log
755 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
756 triggered odd behaviour from Outlook Express clients.
758 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
761 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
762 since the IETF WG has not yet settled on that versus the original
763 "bare" representation.
765 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
766 Previously the millisecond value corrupted the output.
767 Fix also for syslog_pid=no and log_selector +pid, for which the pid
768 corrupted the output.
774 JH/01 Rework error string handling in TLS interface so that the caller in
775 more cases is responsible for logging. This permits library-sourced
776 string to be attached to addresses during delivery, and collapses
777 pairs of long lines into single ones.
779 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
780 during configuration. Wildcards are allowed and expanded.
782 JH/02 Rework error string handling in DKIM to pass more info back to callers.
783 This permits better logging.
785 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
786 do not close it down and have the child transport start it up again on
787 the passed-on TCP connection. Instead, proxy the child (and any
788 subsequent ones) for TLS via a unix-domain socket channel. Logging is
789 affected: the continued delivery log lines do not have any DNSSEC, TLS
790 Certificate or OCSP information. TLS cipher information is still logged.
792 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
793 identical IP addresses on different listening ports. Will also affect
796 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
797 add noisy ifdef guards to special-case this sillyness.
798 Patch from Bernd Kuhls.
800 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
801 than 255 are no longer allowed.
803 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
804 Disable the session-cache too, which might reduce our load. Since we
805 currrectly use a new context for every connection, both as server and
806 client, there is no benefit for these.
807 GnuTLS appears to not support tickets server-side by default (we don't
808 call gnutls_session_ticket_enable_server()) but client side is enabled
809 by default on recent versions (3.1.3 +) unless the PFS priority string
812 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
813 <https://reproducible-builds.org/specs/source-date-epoch/>.
815 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
816 the check for any unsuccessful recipients did not notice the limit, and
817 erroneously found still-pending ones.
819 JH/08 Pipeline CHUNKING command and data together, on kernels that support
820 MSG_MORE. Only in-clear (not on TLS connections).
822 JH/09 Avoid using a temporary file during transport using dkim. Unless a
823 transport-filter is involved we can buffer the headers in memory for
824 creating the signature, and read the spool data file once for the
825 signature and again for transmission.
827 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
828 4.77 as the kernel support then wasn't solid, having issues in 64bit
829 mode. Now, it's been long enough. Add support for FreeBSD also.
831 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
832 case where the routing stage had gathered several addresses to send to
833 a host before calling the transport for the first, we previously failed
834 to close down TLS in the old transport process before passing the TCP
835 connection to the new process. The new one sent a STARTTLS command
836 which naturally failed, giving a failed delivery and bloating the retry
837 database. Investigation and fix prototype from Wolfgang Breyha.
839 JH/12 Fix check on SMTP command input synchronisation. Previously there were
840 false-negatives in the check that the sender had not preempted a response
841 or prompt from Exim (running as a server), due to that code's lack of
842 awareness of the SMTP input buffering.
844 PP/04 Add commandline_checks_require_admin option.
845 Exim drops privileges sanely, various checks such as -be aren't a
846 security problem, as long as you trust local users with access to their
847 own account. When invoked by services which pass untrusted data to
848 Exim, this might be an issue. Set this option in main configuration
849 AND make fixes to the calling application, such as using `--` to stop
852 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
853 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
854 responses to those, into a single TLS record each way (this usually means
855 a single packet). As a side issue, smtp_enforce_sync now works on TLS
858 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
859 affects you only if you're dancing at the edge of the param size limits.
860 If you are, and this message makes sense to you, then: raise the
861 configured limit or use OpenSSL 1.1. Nothing we can do for older
864 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
865 cmdline element to get the documented default one. Previously it was
868 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
869 are unsuable for non-message-specific reasons - eg. connection timeout,
870 banner-time rejection.
872 JH/16 Fix logging of delivery remote port, when specified by router, under
875 PP/06 Repair manualroute's ability to take options in any order, even if one
876 is the name of a transport.
879 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
881 JH/17 Change the list-building routines interface to use the expanding-string
882 triplet model, for better allocation and copying behaviour.
884 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
885 Previously it was constructed the first time a possibly-matching string
886 was met in the configuration file input during startup; now it is done
889 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
890 compatible one, to avoid the (poorly documented) possibility of a config
891 file in the working directory redirecting the DB files, possibly correpting
892 some existing file. CVE-2017-10140 assigned for BDB.
894 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
895 cache-hot. Previously, although the result was properly cached, the
896 initial verify call returned a defer.
898 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
899 the main verify for receipient in uncached-mode.
901 JH/22 Retire historical build files to an "unsupported" subdir. These are
902 defined as "ones for which we have no current evidence of testing".
904 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
905 if present. Previously it was ignored.
907 JH/24 Start using specified-initialisers in C structure init coding. This is
908 a C99 feature (it's 2017, so now considered safe).
910 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
911 if was a fixed-sized field and bitmask ops via macros; it is now more
914 PP/07 GitHub PR 56: Apply MariaDB build fix.
915 Patch provided by Jaroslav Škarvada.
917 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
918 during Coverity cleanups [4.87 JH/47]
919 Diagnosis and fix provided by Michael Fischer v. Mollard.
921 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
922 the right size to place the terminating semicolon on its own folded
923 line, the header hash was calculated to an incorrect value thanks to
924 the (relaxed) space the fold became.
926 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
927 and confused the parent.
929 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
930 which could crash as a result. This could lead to undeliverable messages.
932 JH/28 Logging: "next input sent too soon" now shows where input was truncated
935 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
936 matters on fast-turnover and PID-randomising systems, which were getting
937 out-of-order delivery.
939 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
940 a possibly-overlapping copy. The symptom was that "Remote host closed
941 connection in response to HELO" was logged instead of the actual 4xx
944 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
945 Previously only that bufferd was discarded, resulting in SYMTP command
948 JH/32 DKIM: when a message has multiple signatures matching an identity given
949 in dkim_verify_signers, run the dkim acl once for each. Previously only
950 one run was done. Bug 2189.
952 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
953 "panic the current process" to "deliberately defer". The panic log is
954 still written with the problem list name; the mail and reject logs now
955 get a temp-reject line for the message that was being handled, saying
956 something like "domains check lookup or other defer". The SMTP 451
957 message is still "Temporary local problem".
959 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
960 A crafted sequence of BDAT commands could result in in-use memory beeing
961 freed. CVE-2017-16943.
963 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
964 from SMTP input. Previously it was always done; now only done for DATA
965 and not BDAT commands. CVE-2017-16944.
967 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
968 to the message (such as an overlong header line). Previously this was
969 not done and we did not exit BDAT mode. Followon from the previous item
970 though a different problem.
976 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
977 than -2003 did; needs libidn2 in addition to libidn.
979 JH/02 The path option on a pipe transport is now expanded before use.
981 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
982 Patch provided by "Björn", documentation fix added too.
984 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
985 missing a wire-to-host endian conversion.
987 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
988 close after a BDAT command line could be taken as a following command,
989 giving a synch failure. Fix by only checking for synch immediately
990 before acknowledging the chunk.
992 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
993 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
994 macro. Patches provided by Josh Soref.
996 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
997 Previously we did not; the RFC seems ambiguous and VRFY is not listed
998 by IANA as a service extension. However, John Klensin suggests that we
1001 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1002 the dkim code may be unix-mode line endings rather than smtp wire-format
1003 CRLF, so prepend a CR to any bare LF.
1005 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1006 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1008 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1009 the first were themselves being wrongly included in the feed into dkim
1010 processing; with most chunk sizes in use this resulted in an incorrect
1011 body hash calculated value.
1013 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1014 DKIM signature block, for verification. Although advised against by
1015 standards it is specifically not ruled illegal.
1017 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1019 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1020 missing a body hash (the bh= tag).
1022 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1023 It seems that HAProxy sends the Proxy Protocol information in clear and
1024 only then does a TLS startup, so do the same.
1026 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1027 TCP connections (such as for Spamd) unless the daemon successfully set
1028 Fast Open mode on its listening sockets. This fixes breakage seen on
1029 too-old kernels or those not configured for Fast Open, at the cost of
1030 requiring both directions being enabled for TFO, and TFO never being used
1031 by non-daemon-related Exim processes.
1033 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1034 endings, at least on the first header line. Try to canonify any that get
1035 past that check, despite the cost.
1037 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1038 now limited to an arbitrary five deep, while parsing addresses with the
1039 strip_excess_angle_brackets option enabled.
1041 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1042 instead leave the unprompted TLS handshake in socket buffer for the
1043 TLS library to consume.
1045 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1047 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1049 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1050 region in one operation, for speed, and this leaves assigned pointers
1051 dangling. Add checks run only under the testsuite which checks all
1052 variables at a store-reset and panics on a dangling pointer; add code
1053 explicitly nulling out all the variables discovered. Fixes one known
1054 bug: a transport crash, where a dangling pointer for $sending_ip_address
1055 originally assigned in a verify callout, is re-used.
1057 PP/06 Drop '.' from @INC in various Perl scripts.
1059 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1061 PP/08 Reduce a number of compilation warnings under clang; building with
1062 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1063 should be warning-free.
1065 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1067 HS/01 Fix portability problems introduced by PP/08 for platforms where
1068 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1070 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1071 chunk. This allows us to accept broken chunked messages. We need a more
1072 general solution here.
1074 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1075 already-broken messages in the queue.
1077 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1079 JH/19 Fix reference counting bug in routing-generated-address tracking.
1085 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1086 supports it and a size is available (ie. the sending peer gave us one).
1088 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1089 years of being deprecated). The replacements are the ACLs
1090 acl_smtp_mime and acl_not_smtp_mime.
1092 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1093 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1094 or even an in-clear connection were permitted. Now, if the host lookup
1095 was dnssec and dane was requested then the host is only used if the
1096 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1097 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1098 if one fails this test.
1099 This means that a poorly-configured remote DNS will make it incommunicado;
1100 but it protects against a DNS-interception attack on it.
1102 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1103 non-noisy when a race steals the message being considered.
1105 JH/05 If main configuration option tls_certificate is unset, generate a
1106 self-signed certificate for inbound TLS connections.
1108 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1109 in rewrites and routers.
1111 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1112 and logged a warning sing 4.83; now they are a configuration file error.
1114 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1115 (lacking @domain). Apply the same qualification processing as RCPT.
1117 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1119 JH/10 Support ${sha256:} applied to a string (as well as the previous
1122 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1123 a cutthrough deliver is pending, as we always want to make a connection.
1124 This also avoids re-routing the message when later placing the cutthrough
1125 connection after a verify cache hit.
1126 Do not update it with the verify result either.
1128 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1129 when routing results in more than one destination address.
1131 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1132 signing (which inhibits the cutthrough capability). Previously only
1133 the presence of an option was tested; now an expansion evaluating as
1134 empty is permissible (obviously it should depend only on data available
1135 when the cutthrough connection is made).
1137 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1138 the relevant preceding SMTP command did not note the pipelining mode.
1140 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1141 Previously they were not counted.
1143 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1144 as one having no matching records. Previously we deferred the message
1145 that needed the lookup.
1147 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1148 distinguished as "(=".
1150 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1151 for missing MX records. Previously it only worked for missing A records.
1153 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1155 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1156 after the data-go-ahead and data-ack. Patch from Jason Betts.
1158 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1159 even for a "none" policy. Patch from Tony Meyer.
1161 JH/22 Fix continued use of a connection for further deliveries. If a port was
1162 specified by a router, it must also match for the delivery to be
1165 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1166 When one of the recipients of a message was unsuitable for the connection
1167 (has no matching addresses), we lost track of needing to mark it
1168 deferred. As a result mail would be lost.
1170 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1172 JH/25 Decoding ACL controls is now done using a binary search; the source code
1173 takes up less space and should be simpler to maintain. Merge the ACL
1174 condition decode tables also, with similar effect.
1176 JH/26 Fix problem with one_time used on a redirect router which returned the
1177 parent address unchanged. A retry would see the parent address marked as
1178 delivered, so not attempt the (identical) child. As a result mail would
1181 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1182 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1183 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1186 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1187 Main config options check_{log,spool}_{inodes,space} are now
1188 100 inodes, 10MB unless set otherwise in the configuration.
1190 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1191 Previously it only applied to the main-section connection policy
1194 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1196 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1197 by me. Added RFC7919 DH primes as an alternative.
1199 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1200 are not in the system include path.
1202 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1203 GnuTLS, when a session startup failed (eg because the client disconnected)
1204 Exim did stdio operations after fclose. This was exposed by a recent
1205 change which nulled out the file handle after the fclose.
1207 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1208 signed directly by the cert-signing cert, rather than an intermediate
1209 OCSP-signing cert. This is the model used by LetsEncrypt.
1211 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1213 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1214 an incoming connection.
1216 HS/02 Bug 1802: Do not half-close the connection after sending a request
1219 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1220 fallback to "prime256v1".
1222 JH/34 SECURITY: Use proper copy of DATA command in error message.
1223 Could leak key material. Remotely exploitable. CVE-2016-9963.
1229 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1230 and 3.4.4 - once the server is enabled to respond to an OCSP request
1231 it does even when not requested, resulting in a stapling non-aware
1232 client dropping the TLS connection.
1234 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1235 support variable-length bit vectors. No functional change.
1237 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1238 The I= interface field on outgoing lines is now after the H= remote
1239 host field, same as incoming lines. There is a separate
1240 outgoing_interface log selector which allows you to disable the
1243 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1244 If not running log_selector +smtp_connection the mainlog would be held
1245 open indefinitely after a "too many connections" event, including to a
1246 deleted file after a log rotate. Leave the per net connection logging
1247 leaving it open for efficiency as that will be quickly detected by the
1248 check on the next write.
1250 HS/01 Bug 1671: Fix post transport crash.
1251 Processing the wait-<transport> messages could crash the delivery
1252 process if the message IDs didn't exist for some reason. When
1253 using 'split_spool_directory=yes' the construction of the spool
1254 file name failed already, exposing the same netto behaviour.
1256 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1257 mime_regex ACL conditions.
1259 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1260 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1261 to HELO, local diagnostic string.
1263 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1264 log line to debug. Even when configured with a tls authenticator many
1265 client connections are expected to not authenticate in this way, so
1266 an authenticate fail is not an error.
1268 HS/02 Add the Exim version string to the process info. This way exiwhat
1269 gives some more detail about the running daemon.
1271 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1272 matter for fast-change records such as DNSBLs.
1274 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1275 retry record, even if constant. There may be multiple transports with
1276 different interface settings and the retry behaviour needs to be kept
1279 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1281 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1283 JH/10 Bug 840: fix log_defer_output option of pipe transport
1285 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1286 hosts_randomize. This matters a lot when combined with mua_wrapper.
1288 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1289 ${quote_pgsql:<string>} operator.
1291 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1293 JH/14 Change header file name for freeradius-client. Relevant if compiling
1294 with Radius support; from the Gentoo tree and checked under Fedora.
1296 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1298 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1299 option transport_filter, meaning no filtering.
1301 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1303 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1304 defaults to "*" (all hosts). The variable is now available when not built
1305 with TLS, default unset, mainly to enable keeping the testsuite sane.
1306 If a server certificate is not supplied (via tls_certificate) an error is
1307 logged, and clients will find TLS connections fail on startup. Presumably
1308 they will retry in-clear.
1309 Packagers of Exim are strongly encouraged to create a server certificate
1310 at installation time.
1312 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1313 with the $config_file variable.
1315 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1316 in transport context, after the attempt, and per-recipient. The latter type
1317 is per host attempted. The event data is the error message, and the errno
1318 information encodes the lookup type (A vs. MX) used for the (first) host,
1319 and the trailing two digits of the smtp 4xx response.
1321 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1322 to write to mainlog (or rejectlog, paniclog) in the window between file
1323 creation and permissions/ownership being changed. Particularly affects
1324 installations where exicyclog is run as root, rather than exim user;
1325 result is that the running daemon panics and dies.
1327 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1329 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1330 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1331 "pri" and "weight". Note that the previous implicit priority given by the
1332 list order is no longer honoured.
1334 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1335 for DKIM processing.
1337 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1338 by defining SUPPORT_SOCKS.
1340 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1341 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1342 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1343 variables are renamed to proxy_{local,external}_{address,port}.
1345 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1346 for a build by defining SUPPORT_I18N
1348 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1349 of the query string, and make ${quote_redis:} do that quoting.
1351 JH/29 Move Events support from Experimental to mainline, enabled by default
1352 and removable for a build by defining DISABLE_EVENT.
1354 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1356 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1357 cached by the daemon.
1359 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1360 by defining LOOKUP_REDIS. The libhiredis library is required.
1362 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1363 keys are given for lookup.
1365 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1366 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1367 only supported when built with TLS support. The PolarSSL SHA routines
1368 are still used when the TLS library is too old for convenient support.
1370 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1371 openssl_options), for security. OpenSSL forces this from version 1.1.0
1372 server-side so match that on older versions.
1374 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1375 allocation for $value could be released as the expansion processing
1376 concluded, but leaving the global pointer active for it.
1378 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1379 and to use the domains and local_parts ACL conditions.
1381 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1382 incorrectly not doubled on cutthrough transmission, hence seen as a
1383 body-termination at the receiving system - resulting in truncated mails.
1384 Commonly the sender saw a TCP-level error, and retransmitted the message
1385 via the normal store-and-forward channel. This could result in duplicates
1386 received - but deduplicating mailstores were liable to retain only the
1387 initial truncated version.
1389 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1391 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1393 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1394 we're in there, support oversigning also; bug 1309.
1396 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1398 HS/04 Add support for keep_environment and add_environment options.
1400 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1401 either intentional arithmetic overflow during PRNG, or testing config-
1404 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1405 delivery resulted in actual delivery. Cancel cutthrough before DATA
1408 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1409 rejects a recipient: pass the reject to the originator.
1411 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1412 Many were false-positives and ignorable, but it's worth fixing the
1415 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1416 for the new environment-manipulation done at startup. Move the routines
1417 from being local to tls.c to being global via the os.c file.
1419 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1420 an extract embedded as result-arg for a map, the first arg for extract
1421 is unavailable so we cannot tell if this is a numbered or keyed
1422 extraction. Accept either.
1428 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1431 JH/02 The smtp transport option "multi_domain" is now expanded.
1433 JH/03 The smtp transport now requests PRDR by default, if the server offers
1436 JH/04 Certificate name checking on server certificates, when exim is a client,
1437 is now done by default. The transport option tls_verify_cert_hostnames
1438 can be used to disable this per-host. The build option
1439 EXPERIMENTAL_CERTNAMES is withdrawn.
1441 JH/05 The value of the tls_verify_certificates smtp transport and main options
1442 default to the word "system" to access the system default CA bundle.
1443 For GnuTLS, only version 3.0.20 or later.
1445 JH/06 Verification of the server certificate for a TLS connection is now tried
1446 (but not required) by default. The verification status is now logged by
1447 default, for both outbound TLS and client-certificate supplying inbound
1450 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1453 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1454 Status Notification (bounce) messages are now MIME format per RFC 3464.
1455 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1456 under the control of the dsn_advertise_hosts option, and routers may
1457 have a dsn_lasthop option.
1459 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1460 default, modifiable by a malware= option. The list separator for
1461 the options can now be changed in the usual way. Bug 68.
1463 JH/10 The smtp_receive_timeout main option is now expanded before use.
1465 JH/11 The incoming_interface log option now also enables logging of the
1466 local interface on delivery outgoing connections.
1468 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1469 if the interface and destination host and port all match.
1471 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1474 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1475 Patch from Andrew Lewis.
1477 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1478 now supports optional time-restrictions, weighting, and priority
1479 modifiers per server. Patch originally by <rommer@active.by>.
1481 JH/16 The spamd_address main option now supports a mixed list of local
1482 and remote servers. Remote servers can be IPv6 addresses, and
1483 specify a port-range.
1485 JH/17 Bug 68: The spamd_address main option now supports an optional
1486 timeout value per server.
1488 JH/18 Bug 1581: Router and transport options headers_add/remove can
1489 now have the list separator specified.
1491 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1494 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1497 JH/21 Support for the A6 type of dns record is withdrawn.
1499 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1500 rather than the verbs used.
1502 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1503 from 255 to 1024 chars.
1505 JH/24 Verification callouts now attempt to use TLS by default.
1507 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1508 are generic router options now. The defaults didn't change.
1510 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1511 Original patch from Alexander Shikoff, worked over by JH.
1513 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1514 files if ZCAT_COMMAND is not executable.
1516 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1518 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1520 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1521 Normally benign, it bites when the pair was led to by a CNAME;
1522 modern usage is to not canonicalize the domain to a CNAME target
1523 (and we were inconsistent anyway for A-only vs AAAA+A).
1525 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1527 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1528 when evaluating $sender_host_dnssec.
1530 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1531 $sender_helo_dnssec variable.
1533 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1535 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1537 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1539 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1540 documented as working, but never had. Support all but $spam_report.
1542 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1543 added for tls authenticator.
1545 HS/03 Add perl_taintmode main config option
1551 TL/01 When running the test suite, the README says that variables such as
1552 no_msglog_check are global and can be placed anywhere in a specific
1553 test's script, however it was observed that placement needed to be near
1554 the beginning for it to behave that way. Changed the runtest perl
1555 script to read through the entire script once to detect and set these
1556 variables, reset to the beginning of the script, and then run through
1557 the script parsing/test process like normal.
1559 TL/02 The BSD's have an arc4random API. One of the functions to induce
1560 adding randomness was arc4random_stir(), but it has been removed in
1561 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1562 function when detected.
1564 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1565 cause callback expansion.
1567 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1568 syntax errors in an expansion can be treated as a string instead of
1569 logging or causing an error, due to the internal use of bool_lax
1570 instead of bool when processing it.
1572 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1573 server certificates when making smtp deliveries.
1575 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1577 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1579 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1581 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1582 Merged patch from Sebastian Wiedenroth.
1584 JH/05 Fix results-pipe from transport process. Several recipients, combined
1585 with certificate use, exposed issues where response data items split
1586 over buffer boundaries were not parsed properly. This eventually
1587 resulted in duplicates being sent. This issue only became common enough
1588 to notice due to the introduction of connection certificate information,
1589 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1591 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1592 size buffer was used, resulting in syntax errors when an expansion
1595 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1596 version 3.3.6 or later.
1598 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1599 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1600 both become "event_action", the variables become $event_name, $event_data
1601 and $event_defer_errno. There is a new variable $verify_mode, usable in
1602 routers, transports and related events. The tls:cert event is now also
1603 raised for inbound connections, if the main configuration event_action
1606 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1607 early OCSP support, but no stapling (appears to be less than 1.0.0).
1609 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1610 server certificate names available under the smtp transport option
1611 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1614 JH/10 Time-related extraction expansions from certificates now use the main
1615 option "timezone" setting for output formatting, and are consistent
1616 between OpenSSL and GnuTLS compilations. Bug 1541.
1618 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1619 encoded parameter in the incoming message. Bug 1558.
1621 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1622 include certificate info, eximon was claiming there were spoolfile
1625 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1627 JH/14 Log delivery-related information more consistently, using the sequence
1628 "H=<name> [<ip>]" wherever possible.
1630 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1631 are problematic for Debian distribution, omit them from the release
1634 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1636 JH/16 Fix string representation of time values on 64bit time_t architectures.
1639 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1640 output list separator was used.
1645 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1646 checkers that were complaining about end of non-void function with no
1649 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1650 This was a regression introduced in 4.83 by another bugfix.
1652 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1654 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1655 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1661 TF/01 Correctly close the server side of TLS when forking for delivery.
1663 When a message was received over SMTP with TLS, Exim failed to clear up
1664 the incoming connection properly after forking off the child process to
1665 deliver the message. In some situations the subsequent outgoing
1666 delivery connection happened to have the same fd number as the incoming
1667 connection previously had. Exim would try to use TLS and fail, logging
1668 a "Bad file descriptor" error.
1670 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1671 utilities have not been installed.
1673 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1674 temporary space as the ACL may create new global variables.
1676 TL/01 LDAP support uses per connection or global context settings, depending
1677 upon the detected version of the libraries at build time.
1679 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1680 to extract and use the src ip:port in logging and expansions as if it
1681 were a direct connection from the outside internet. PPv2 support was
1682 updated based on HAProxy spec change in May 2014.
1684 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1686 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1687 Properly escape header and check for NULL return.
1689 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1692 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1694 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1695 characters in header names, implemented as a verify condition.
1696 Contributed by Michael Fischer v. Mollard.
1698 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1699 results permerror and temperror. Previous values are deprecated but
1700 still accepted. In a future release, err_perm and err_temp will be
1701 completely removed, which will be a backward incompatibility if the
1702 ACL tests for either of these two old results. Patch contributed by
1703 user bes-internal on the mailing list.
1705 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1707 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1708 selectors, in both main and reject logs.
1710 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1713 JH/07 Add malware type "sock" for talking to simple daemon.
1715 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1717 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1718 routers/transports under cutthrough routing.
1720 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1721 numbers. Touch up "bool" conditional to keep the same definition.
1723 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1725 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1727 JH/12 Expand items in router/transport headers_add or headers_remove lists
1728 individually rather than the list as a whole. Bug 1452.
1730 Required for reasonable handling of multiple headers_ options when
1731 they may be empty; requires that headers_remove items with embedded
1732 colons must have them doubled (or the list-separator changed).
1734 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1735 view the policy declared in the DMARC record. Currently, $dmarc_status
1736 is a combined value of both the record presence and the result of the
1739 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1741 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1742 dnslookup router and the smtp transport (applying to the forward
1745 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1746 of ldap servers used for a specific lookup. Patch provided by Heiko
1749 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1750 New variable $lookup_dnssec_authenticated for observability.
1752 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1753 Patch submitted by Lars Timman.
1755 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1757 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1758 Requires trusted mode and valid format message id, aborts otherwise.
1759 Patch contributed by Heiko Schlichting.
1761 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1762 certextract with support for various fields. Bug 1358.
1764 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1765 is requested by default, modifiable by smtp transport option
1768 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1769 operate on certificate variables to give certificate fingerprints
1770 Also new ${sha256:cert_variable}.
1772 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1774 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1775 Christian Aistleitner.
1777 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1779 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1780 file. Patch from Wolfgang Breyha.
1782 JH/25 Expand the coverage of the delivery $host and $host_address to
1783 client authenticators run in verify callout. Bug 1476.
1785 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1786 align with daemon_smtp_ports. Bug 72.
1788 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1789 support and error reporting did not work properly.
1791 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1792 and is readable. Patch from Andrew Colin Kissa.
1794 TL/14 Enhance documentation of ${run expansion and how it parses the
1795 commandline after expansion, particularly in the case when an
1796 unquoted variable expansion results in an empty value.
1798 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1800 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1801 by renaming away from it. Addresses GH issue 15, reported by
1804 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1806 TL/15 SECURITY: prevent double expansion in math comparison functions
1807 (can expand unsanitized data). Not remotely exploitable.
1814 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1816 PP/02 Make -n do something, by making it not do something.
1817 When combined with -bP, the name of an option is not output.
1819 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1822 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1823 $sender_host_name and config options to manage this, and basic check
1826 PP/05 DSCP support for outbound connections and control modifier for inbound.
1828 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1829 (Only plugin which currently uses this is kerberos4, which nobody should
1830 be using, but we should make it available and other future plugins might
1831 conceivably use it, even though it would break NAT; stuff *should* be
1832 using channel bindings instead).
1834 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1835 name; added for Sendmail compatibility; requires admin caller.
1836 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1837 just ignore it); requires trusted caller.
1838 Also parse but ignore: -Ac -Am -X<logfile>
1841 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1843 TL/02 Add +smtp_confirmation as a default logging option.
1845 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1846 Patch by Magnus Holmgren from 2007-02-20.
1848 TL/04 Bugzilla 1281 - Spec typo.
1849 Bugzilla 1283 - Spec typo.
1850 Bugzilla 1290 - Spec grammar fixes.
1852 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1854 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1856 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1857 mailing list by by Dmitry Isaikin.
1859 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1861 JH/02 Support "G" suffix to numbers in ${if comparisons.
1863 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1865 NM/01 Bugzilla 1197 - Spec typo
1866 Bugzilla 1196 - Spec examples corrections
1868 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1870 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1871 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1874 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1875 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1877 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1878 "acl {{name}{arg}...}", and optional args on acl condition
1881 JH/05 Permit multiple router/transport headers_add/remove lines.
1883 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1885 JH/07 Avoid using a waiting database for a single-message-only transport.
1886 Performance patch from Paul Fisher. Bugzilla 1262.
1888 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1891 JH/09 Add $headers_added variable, with content from use of ACL modifier
1892 add_header (but not yet added to the message). Bugzilla 199.
1894 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1895 Pulled from Bugzilla 817 by Wolfgang Breyha.
1897 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1899 (nb: this is the same fix as in Exim 4.80.1)
1901 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1904 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1907 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1909 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1910 advertise SMTP AUTH mechanism to us, instead of a generic
1911 protocol violation error. Also, make Exim more robust to bad
1912 data from the Dovecot auth socket.
1914 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1916 When a queue runner is handling a message, Exim first routes the
1917 recipient addresses, during which it prunes them based on the retry
1918 hints database. After that it attempts to deliver the message to
1919 any remaining recipients. It then updates the hints database using
1922 So if a recipient address works intermittently, it can get repeatedly
1923 deferred at routing time. The retry hints record remains fresh so the
1924 address never reaches the final cutoff time.
1926 This is a fairly common occurrence when a user is bumping up against
1927 their storage quota. Exim had some logic in its local delivery code
1928 to deal with this. However it did not apply to per-recipient defers
1929 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1931 This change adds a proper retry rule check during routing so that the
1932 final cutoff time is checked against the message's age. We only do
1933 this check if there is an address retry record and there is not a
1934 domain retry record; this implies that previous attempts to handle
1935 the address had the retry_use_local_parts option turned on. We use
1936 this as an approximation for the destination being like a local
1937 delivery, as in LMTP.
1939 I suspect this new check makes the old local delivery cutoff check
1940 redundant, but I have not verified this so I left the code in place.
1942 TF/02 Correct gecos expansion when From: is a prefix of the username.
1944 Test 0254 submits a message to Exim with the header
1948 When I ran the test suite under the user fanf2, Exim expanded
1949 the header to contain my full name, whereas it should have added
1950 a Resent-Sender: header. It erroneously treats any prefix of the
1951 username as equal to the username.
1953 This change corrects that bug.
1955 GF/01 DCC debug and logging tidyup
1956 Error conditions log to paniclog rather than rejectlog.
1957 Debug lines prefixed by "DCC: " to remove any ambiguity.
1959 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1961 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1962 Bug spotted by Jeremy Harris; was flawed since initial commit.
1963 Would have resulted in OCSP responses post-SNI triggering an Exim
1964 NULL dereference and crash.
1966 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1968 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1969 Bug detection, analysis and fix by Samuel Thibault.
1970 Bugzilla 1331, Debian bug #698092.
1972 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1974 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1975 Server implementation by Todd Lyons, client by JH.
1976 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1977 config variable "prdr_enable" controls whether the server
1978 advertises the facility. If the client requests PRDR a new
1979 acl_data_smtp_prdr ACL is called once for each recipient, after
1980 the body content is received and before the acl_smtp_data ACL.
1981 The client is controlled by both of: a hosts_try_prdr option
1982 on the smtp transport, and the server advertisement.
1983 Default client logging of deliveries and rejections involving
1984 PRDR are flagged with the string "PRDR".
1986 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1987 fclose(). Diagnosis by Todd Lyons.
1989 PP/17 Update configure.default to handle IPv6 localhost better.
1990 Patch by Alain Williams (plus minor tweaks).
1993 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1994 This is now consistent with GnuTLS, and is now documented: the
1995 previous undocumented portable approach to treating the option as
1996 unset was to force an expansion failure. That still works, and
1997 an empty string is now equivalent.
1999 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2000 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2001 not performing validation itself.
2003 PP/20 Added force_command boolean option to pipe transport.
2004 Patch from Nick Koston, of cPanel Inc.
2006 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2009 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2011 PP/21 Fix eximon continuous updating with timestamped log-files.
2012 Broken in a format-string cleanup in 4.80, missed when I repaired the
2013 other false fix of the same issue.
2014 Report and fix from Heiko Schlichting.
2017 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2018 Report from Prashanth Katuri.
2020 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2021 It's SecureTransport, so affects any MacOS clients which use the
2022 system-integrated TLS libraries, including email clients.
2024 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2025 using a MIME ACL for non-SMTP local injection.
2026 Report and assistance in diagnosis by Warren Baker.
2028 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2030 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2032 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2033 last id that failed so it may be referenced in subsequent ACL's.
2035 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2038 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2039 ldap library initialization, allowing self-signed CA's to be
2040 used. Also properly sets require_cert option later in code by
2041 using NULL (global ldap config) instead of ldap handle (per
2042 session). Bug diagnosis and testing by alxgomz.
2044 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2045 the src/util/ subdirectory.
2047 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2048 renamed to Transport Post Delivery Action by Jeremy Harris, as
2051 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2052 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2053 redis_servers = needs to be configured which will be used by the redis
2054 lookup. Patch from Warren Baker, of The Packet Hub.
2056 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2058 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2059 hostname or reverse DNS when processing a host list. Used suggestions
2060 from multiple comments on this bug.
2062 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2064 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2065 Missed a few lines, added it to make the runtest require no keyboard
2068 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2069 contains upper case chars. Make router use caseful_local_part.
2071 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2072 support when GnuTLS has been built with p11-kit.
2078 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2080 This, or similar/improved, will also be change PP/11 of 4.82.
2086 PP/01 Handle short writes when writing local log-files.
2087 In practice, only affects FreeBSD (8 onwards).
2088 Bugzilla 1053, with thanks to Dmitry Isaikin.
2090 NM/01 Bugzilla 949 - Documentation tweak
2092 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2095 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2097 PP/02 Implemented gsasl authenticator.
2099 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2101 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2102 `pkg-config foo` for cflags/libs.
2104 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2105 with rest of GSASL and with heimdal_gssapi.
2107 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2108 `pkg-config foo` for cflags/libs for the TLS implementation.
2110 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2111 properties get this fed in as external SSF. A number of robustness
2112 and debugging improvements to the cyrus_sasl authenticator.
2114 PP/08 cyrus_sasl server now expands the server_realm option.
2116 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2117 Patch by Jeremy Harris.
2119 PP/10 Added dbmjz lookup type.
2121 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2123 PP/12 MAIL args handles TAB as well as SP, for better interop with
2124 non-compliant senders.
2125 Analysis and variant patch by Todd Lyons.
2127 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2128 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2129 Patch from Dirk Mueller <dmueller@suse.com>
2131 PP/13 tls_peerdn now print-escaped for spool files.
2132 Observed some $tls_peerdn in wild which contained \n, which resulted
2133 in spool file corruption.
2135 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2136 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2137 or write after TLS renegotiation, which otherwise led to messages
2140 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2141 as a tracking header (ie: a signed header comes before the signature).
2142 Patch from Wolfgang Breyha.
2144 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2145 comma-sep list; embedded commas doubled.
2147 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2149 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2151 Report and patch from Dmitry Banschikov.
2153 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2154 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2155 failures. We appear to now support TLS1.1+ with Exim.
2157 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2158 lets Exim select keys and certificates based upon TLS SNI from client.
2159 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2160 before an outbound SMTP session. New log_selector, +tls_sni.
2162 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2163 NULL dereference. Report and patch from Alun Jones.
2165 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2166 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2167 Not seeing resolver debug output on NetBSD, but suspect this is a
2168 resolver implementation change.
2170 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2171 Left warnings. Added "eximon gdb" invocation mode.
2173 PP/21 Defaulting "accept_8bitmime" to true, not false.
2175 PP/22 Added -bw for inetd wait mode support.
2177 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2178 locate the relevant includes and libraries. Made this the default.
2180 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2181 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2183 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2184 This may cause build issues on older platforms.
2186 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2187 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2188 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2189 Added SNI support via GnuTLS too.
2190 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2192 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2194 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2195 Applied second patch from Janne, implementing suggestion to default
2196 multiple-strings-in-record handling to match SPF spec.
2198 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2200 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2201 read-only, out of scope).
2202 Patch from Wolfgang Breyha, report from Stuart Northfield.
2204 PP/29 Fix three issues highlighted by clang analyser static analysis.
2205 Only crash-plausible issue would require the Cambridge-specific
2206 iplookup router and a misconfiguration.
2207 Report from Marcin Mirosław.
2209 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2211 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2212 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2213 As part of this, removing so much warning spew let me fix some minor
2214 real issues in debug logging.
2216 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2217 assignment on my part. Fixed.
2219 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2220 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2221 Janne Snabb (who went above and beyond: thank you).
2223 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2224 string otherwise requires a connection and a bunch more work and it's
2225 relatively easy to get wrong. Should also expose TLS library linkage
2228 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2229 64-bit ${eval} (JH/03).
2231 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2232 GNU libc to support some of the 64-bit stuff, should not lead to
2233 conflicts. Defined before os.h is pulled in, so if a given platform
2234 needs to override this, it can.
2236 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2237 protection layer was required, which is not implemented.
2238 Bugzilla 1254, patch from Wolfgang Breyha.
2240 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2241 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2242 tls_dhparam take prime identifiers. Also unbreak combination of
2243 OpenSSL+DH_params+TLSSNI.
2245 PP/39 Disable SSLv2 by default in OpenSSL support.
2251 PP/01 Solaris build fix for Oracle's LDAP libraries.
2252 Bugzilla 1109, patch from Stephen Usher.
2254 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2256 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2259 TF/02 Fix a couple more cases where we did not log the error message
2260 when unlink() failed. See also change 4.74-TF/03.
2262 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2263 lock up or crash if it happened to be inside a call to libc when it
2264 got a SIGUSR1 from exiwhat.
2266 The SIGUSR1 handler appends the current process status to the process
2267 log which is later printed by exiwhat. It used to use the general
2268 purpose logging code to do this, but several functions it calls are
2269 not safe for signals.
2271 The new output code in the SIGUSR1 handler is specific to the process
2272 log, and simple enough that it's easy to inspect for signal safety.
2273 Removing some special cases also simplifies the general logging code.
2274 Removing the spurious timestamps from the process log simplifies
2277 TF/04 Improved ratelimit ACL condition.
2279 The /noupdate option has been deprecated in favour of /readonly which
2280 has clearer semantics. The /leaky, /strict, and /readonly update modes
2281 are mutually exclusive. The update mode is no longer included in the
2282 database key; it just determines when the database is updated. (This
2283 means that when you upgrade Exim will forget old rate measurements.)
2285 Exim now checks that the per_* options are used with an update mode that
2286 makes sense for the current ACL. For example, when Exim is processing a
2287 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2288 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2289 must specify per_mail/readonly. If you omit the update mode it defaults to
2290 /leaky where that makes sense (as before) or /readonly where required.
2292 The /noupdate option is now undocumented but still supported for
2293 backwards compatibility. It is equivalent to /readonly except that in
2294 ACLs where /readonly is required you may specify /leaky/noupdate or
2295 /strict/noupdate which are treated the same as /readonly.
2297 A useful new feature is the /count= option. This is a generalization
2298 of the per_byte option, so that you can measure the throughput of other
2299 aggregate values. For example, the per_byte option is now equivalent
2300 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2302 The per_rcpt option has been generalized using the /count= mechanism
2303 (though it's more complicated than the per_byte equivalence). When it is
2304 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2305 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2306 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2307 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2308 Note that using per_rcpt with a non-readonly update mode in more than
2309 one ACL will cause the recipients to be double-counted. (The per_mail
2310 and per_byte options don't have this problem.)
2312 The handling of very low rates has changed slightly. If the computed rate
2313 is less than the event's count (usually one) then this event is the first
2314 after a long gap. In this case the rate is set to the same as this event's
2315 count, so that the first message of a spam run is counted properly.
2317 The major new feature is a mechanism for counting the rate of unique
2318 events. The new per_addr option counts the number of different
2319 recipients that someone has sent messages to in the last time period. It
2320 behaves like per_rcpt if all the recipient addresses are different, but
2321 duplicate recipient addresses do not increase the measured rate. Like
2322 the /count= option this is a general mechanism, so the per_addr option
2323 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2324 example, measure the rate that a client uses different sender addresses
2325 with the options per_mail/unique=$sender_address. There are further
2326 details in the main documentation.
2328 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2330 TF/06 Removed a few PCRE remnants.
2332 TF/07 Automatically extract Exim's version number from tags in the git
2333 repository when doing development or release builds.
2335 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2336 Bugzilla 879. Patch from Paul Fisher.
2338 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2339 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2342 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2344 PP/05 Variable $av_failed, true if the AV scanner deferred.
2345 Bugzilla 1078. Patch from John Horne.
2347 PP/06 Stop make process more reliably on build failure.
2348 Bugzilla 1087. Patch from Heiko Schlittermann.
2350 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2351 Bugzilla 1089. Patch from Heiko Schlittermann.
2353 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2354 Bugzilla 1131. Patch from Holger Weiß.
2356 PP/09 Handle IPv6 addresses with SPF.
2357 Bugzilla 860. Patch from Wolfgang Breyha.
2359 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2361 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2364 PP/11 match_* no longer expand right-hand-side by default.
2365 New compile-time build option, EXPAND_LISTMATCH_RHS.
2366 New expansion conditions, "inlist", "inlisti".
2368 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2370 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2372 PP/14 fix log_write() format string regression from TF/03.
2373 Bugzilla 1152. Patch from Dmitry Isaikin.
2379 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2381 PP/02 Harmonised TLS library version reporting; only show if debugging.
2382 Layout now matches that introduced for other libraries in 4.74 PP/03.
2384 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2386 PP/04 New "dns_use_edns0" global option.
2388 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2391 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2392 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2394 TK/01 Updated PolarSSL code to 0.14.2.
2395 Bugzilla 1097. Patch from Andreas Metzler.
2397 PP/07 Catch divide-by-zero in ${eval:...}.
2398 Fixes bugzilla 1102.
2400 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2403 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2404 format-string attack -- SECURITY: remote arbitrary code execution.
2406 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2407 time unintentionally subject to list matching rules, letting the header
2408 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2409 arbitrary string expansion). This allowed for information disclosure.
2411 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2412 INT_MIN/-1 -- value coerced to INT_MAX.
2418 NM/01 Workaround for PCRE version dependency in version reporting
2421 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2422 This fixes portability to compilers other than gcc, notably
2423 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2425 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2426 makefiles for portability to HP-UX and POSIX correctness.
2428 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2429 Also via indented variable definition in the Makefile.
2430 (Debugging by Oliver Heesakkers).
2432 PP/02 Restore caching of spamd results with expanded spamd_address.
2433 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2435 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2436 Improves build reliability. Fix from: Frank Elsner
2438 NM/02 Fix wide character breakage in the rfc2047 coding
2439 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2441 NM/03 Allow underscore in dnslist lookups
2442 Fixes bug 1026. Patch from Graeme Fowler
2444 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2445 Code patches from Adam Ciarcinski of NetBSD.
2447 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2450 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2451 is logged, to avoid truncation. Patch from John Horne.
2453 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2454 Patch from Jakob Hirsch.
2456 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2457 SQL string expansion failure details.
2458 Patch from Andrey Oktyabrski.
2460 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2461 Patch from Simon Arlott.
2463 PP/09 New lookups functionality failed to compile on old gcc which rejects
2464 extern declarations in function scope.
2465 Patch from Oliver Fleischmann
2467 PP/10 Use sig_atomic_t for flags set from signal handlers.
2468 Check getgroups() return and improve debugging.
2469 Fixed developed for diagnosis in bug 927 (which turned out to be
2472 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2473 Patch from Mark Zealey.
2475 PP/12 Bugzilla 1056: Improved spamd server selection.
2476 Patch from Mark Zealey.
2478 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2479 Based on patch from Heiko Schlittermann.
2481 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2482 Patch from Uwe Doering, sign-off by Michael Haardt.
2484 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2485 variable declaration deep within a block. Bug and patch from
2488 PP/15 lookups-Makefile IRIX compatibility coercion.
2490 PP/16 Make DISABLE_DKIM build knob functional.
2492 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2493 Patch by Simon Arlott
2495 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2496 variable argument macros. Our copy now differs from upstream.
2502 TF/01 Failure to get a lock on a hints database can have serious
2503 consequences so log it to the panic log.
2505 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2506 controlled using the smtp_confirmation log selector.
2508 TF/03 Include the error message when we fail to unlink a spool file.
2510 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2511 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2512 for maintaining out-of-tree patches for some time.
2514 PP/01 Bugzilla 139: Documentation and portability issues.
2515 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2516 Handle per-OS dynamic-module compilation flags.
2518 PP/02 Let /dev/null have normal permissions.
2519 The 4.73 fixes were a little too stringent and complained about the
2520 permissions on /dev/null. Exempt it from some checks.
2521 Reported by Andreas M. Kirchwitz.
2523 PP/03 Report version information for many libraries, including
2524 Exim version information for dynamically loaded libraries. Created
2525 version.h, now support a version extension string for distributors
2526 who patch heavily. Dynamic module ABI change.
2528 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2529 privilege escalation vulnerability whereby the Exim run-time user
2530 can cause root to append content of the attacker's choosing to
2533 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2536 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2537 If dropping privileges for untrusted macros, we disabled normal logging
2538 on the basis that it would fail; for the Exim run-time user, this is not
2539 the case, and it resulted in successful deliveries going unlogged.
2540 Fixed. Reported by Andreas Metzler.
2546 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2547 only prepend for the Resent-* case. Fixes regression introduced in
2548 Exim 4.70 by NM/22 for Bugzilla 607.
2550 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2551 increasing numbers of administrators be bitten by this.
2553 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2555 PP/03 Bugzilla 994: added openssl_options main configuration option.
2557 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2559 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2561 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2563 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2564 without a peer certificate, leading to a segfault because of an
2565 assumption that peers always have certificates. Be a little more
2566 paranoid. Problem reported by Martin Tscholak.
2568 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2569 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2570 NB: ClamAV planning to remove STREAM in "middle of 2010".
2571 CL also introduces -bmalware, various -d+acl logging additions and
2572 more caution in buffer sizes.
2574 PP/09 Implemented reverse_ip expansion operator.
2576 PP/10 Bugzilla 937: provide a "debug" ACL control.
2578 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2580 PP/12 Bugzilla 973: Implement --version.
2582 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2584 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2586 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2588 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2589 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2590 ignore trailing whitespace.
2592 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2594 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2597 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2598 Notification from Dr Andrew Aitchison.
2600 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2601 ExtendedDetectionInfo response format.
2602 Notification from John Horne.
2604 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2607 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2608 XSL and documented dependency on system catalogs, with examples of how
2611 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2614 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2615 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2616 configuration file which is writeable by the Exim user or group.
2618 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2619 of configuration files to cover files specified with the -C option if
2620 they are going to be used with root privileges, not just the default
2623 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2624 option (effectively making it always true).
2626 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2627 files to be used while preserving root privileges.
2629 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2630 that rogue child processes cannot use them.
2632 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2633 run-time user, instead of root.
2635 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2636 Exim run-time user without dropping privileges.
2638 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2639 result string, instead of calling string_vformat() twice with the same
2642 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2643 for other users. Others should always drop root privileges if they use
2644 -C on the command line, even for a whitelisted configure file.
2646 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2648 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2654 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2655 $data_path, and $header_path variables; fixed documentation bugs and
2658 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2659 exipick to access non-standard spools, including the "frozen" queue
2662 NM/01 Bugzilla 965: Support mysql stored procedures.
2663 Patch from Alain Williams
2665 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2667 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2668 Patch from Andreas Metzler
2670 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2671 Patch from Kirill Miazine
2673 NM/05 Bugzilla 671: Added umask to procmail example.
2675 JJ/03 installed exipick 20100323.0, fixing doc bug
2677 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2678 directory. Notification and patch from Dan Rosenberg.
2680 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2682 TK/02 Improve log output when DKIM signing operation fails.
2684 MH/01 Treat the transport option dkim_domain as a colon separated
2685 list, not as a single string, and sign the message with each element,
2686 omitting multiple occurences of the same signer.
2688 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2689 Bugzilla 985, 986. Patch by Simon Arlott
2691 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2692 Patch by Simon Arlott
2694 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2695 MBX locking. Notification from Dan Rosenberg.
2701 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2703 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2705 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2707 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2709 NM/04 Bugzilla 915: Flag broken perl installation during build.
2715 TK/01 Added patch by Johannes Berg that expands the main option
2716 "spamd_address" if it starts with a dollar sign.
2718 TK/02 Write list of recipients to X-Envelope-Sender header when building
2719 the mbox-format spool file for content scanning (suggested by Jakob
2722 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2723 (http://www.dcc-servers.net/) support via dccifd. Activated by
2724 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2726 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2727 by Mark Daniel Reidel <mr@df.eu>.
2729 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2730 When building exim an external PCRE library is now needed -
2731 PCRE is a system library on the majority of modern systems.
2732 See entry on PCRE_LIBS in EDITME file.
2734 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2735 conversation. Added nologin parameter to request.
2736 Patch contributed by Kirill Miazine.
2738 TF/01 Do not log submission mode rewrites if they do not change the address.
2740 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2742 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2743 log files in place. Contributed by Roberto Lima.
2745 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2747 TF/03 Bugzilla 615: When checking the local_parts router precondition
2748 after a local_part_suffix or local_part_prefix option, Exim now
2749 does not use the address's named list lookup cache, since this
2750 contains cached lookups for the whole local part.
2752 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2753 Robert Millan. Documentation is in experimental-spec.txt.
2755 TF/04 Bugzilla 668: Fix parallel build (make -j).
2757 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2759 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2760 Patch provided by Jan Srzednicki.
2762 TF/05 Leading white space used to be stripped from $spam_report which
2763 wrecked the formatting. Now it is preserved.
2765 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2766 that they are available at delivery time.
2768 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2770 TF/08 TLS error reporting now respects the incoming_interface and
2771 incoming_port log selectors.
2773 TF/09 Produce a more useful error message if an SMTP transport's hosts
2774 setting expands to an empty string.
2776 NM/06 Bugzilla 744: EXPN did not work under TLS.
2777 Patch provided by Phil Pennock.
2779 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2780 Patch provided by Richard Godbee.
2782 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2783 acl_smtp_notquit, added index entry.
2785 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2786 Patch provided by Eugene Bujak.
2788 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2789 accept(). Patch provided by Maxim Dounin.
2791 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2792 Patch provided by Phil Pennock.
2794 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2796 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2797 Patch provided by Brad "anomie" Jorsch.
2799 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2800 Patch provided by Dean Brooks.
2802 TK/05 Add native DKIM support (does not depend on external libraries).
2804 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2805 Patch provided by Graeme Fowler.
2807 NM/16 Bugzilla 851: Documentation example syntax fix.
2809 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2811 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2814 NM/19 Bugzilla 745: TLS version reporting.
2815 Patch provided by Phil Pennock.
2817 NM/20 Bugzilla 167: bool: condition support.
2818 Patch provided by Phil Pennock.
2820 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2821 clients. Patch provided by Phil Pennock.
2823 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2824 Patch provided by Brad "anomie" Jorsch.
2826 NM/23 Bugzilla 687: Fix misparses in eximstats.
2827 Patch provided by Heiko Schlittermann.
2829 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2830 Patch provided by Heiko Schlittermann.
2832 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2833 plus update to original patch.
2835 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2837 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2838 Patch provided by David Brownlee.
2840 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2842 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2844 NM/30 Bugzilla 888: TLS documentation bugfixes.
2846 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2848 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2849 Unlike the original bugzilla I have changed all shell scripts in src tree.
2851 NM/33 Bugzilla 898: Transport filter timeout fix.
2852 Patch by Todd Rinaldo.
2854 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2855 Patch by Serge Demonchaux.
2857 NM/35 Bugzilla 39: Base64 decode bug fixes.
2858 Patch by Jakob Hirsch.
2860 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2862 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2864 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2866 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2872 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2873 ALT-N's libdkim that I have put here:
2874 http://duncanthrax.net/exim-experimental/
2876 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2877 were called 'true' and it seems that C99 defines that as a reserved
2878 keyword to be used with 'bool' variable types. That means you could
2879 not include C99-style headers which use bools without triggering
2880 build errors in sieve.c.
2882 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2883 as mailq or other aliases. Changed the --help handling significantly
2884 to do whats expected. exim_usage() emits usage/help information.
2886 SC/01 Added the -bylocaldomain option to eximstats.
2888 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2890 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2892 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2898 PH/01 Another patch from the Sieve maintainer.
2900 PH/02 When an IPv6 address is converted to a string for single-key lookup
2901 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2902 dots are used instead of colons so that keys in lsearch files need not
2903 contain colons. This was done some time before quoting was made available
2904 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2905 (notated using the quote facility) so as to distinguish them from IPv4
2906 keys. This meant that lookups for IP addresses in host lists did not work
2907 for iplsearch lookups.
2909 This has been fixed by arranging for IPv6 addresses to be expressed with
2910 colons if the lookup type is iplsearch. This is not incompatible, because
2911 previously such lookups could never work.
2913 The situation is now rather anomalous, since one *can* have colons in
2914 ordinary lsearch keys. However, making the change in all cases is
2915 incompatible and would probably break a number of configurations.
2917 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2920 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2921 conversion specification without a maximum field width, thereby enabling
2922 a rogue spamd server to cause a buffer overflow. While nobody in their
2923 right mind would setup Exim to query an untrusted spamd server, an
2924 attacker that gains access to a server running spamd could potentially
2925 exploit this vulnerability to run arbitrary code as the Exim user.
2927 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2928 $primary_hostname instead of what libspf2 thinks the hosts name is.
2930 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2931 a directory entry by the name of the lookup key. Previously, if a
2932 symlink pointed to a non-existing file or a file in a directory that
2933 Exim lacked permissions to read, a lookup for a key matching that
2934 symlink would fail. Now it is enough that a matching directory entry
2935 exists, symlink or not. (Bugzilla 503.)
2937 PH/03 The body_linecount and body_zerocount variables are now exported in the
2940 PH/04 Added the $dnslist_matched variable.
2942 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2943 This means they are set thereafter only if the connection becomes
2946 PH/06 Added the client_condition to authenticators so that some can be skipped
2947 by clients under certain conditions.
2949 PH/07 The error message for a badly-placed control=no_multiline_responses left
2950 "_responses" off the end of the name.
2952 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2954 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2955 (without spaces) instead of just copying the configuration text.
2957 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2959 PH/11 Added $max_received_linelength.
2961 PH/12 Added +ignore_defer and +include_defer to host lists.
2963 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2964 way in which PCRE > 7.0 is built.
2966 PH/14 Implemented queue_only_load_latch.
2968 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2969 MAIL command. The effect was to mangle the value on 64-bit systems.
2971 PH/16 Another patch from the Sieve maintainer.
2973 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2975 PH/18 If a system quota error occurred while trying to create the file for
2976 a maildir delivery, the message "Mailbox is full" was not appended to the
2977 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2978 only to a quota excession during the actual writing of the file.
2980 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2981 characters?) which causes problems in log lines. The DN values are now
2982 passed through string_printing() before being added to log lines.
2984 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2985 and InterBase are left for another time.)
2987 PH/21 Added message_body_newlines option.
2989 PH/22 Guard against possible overflow in moan_check_errorcopy().
2991 PH/23 POSIX allows open() to be a macro; guard against that.
2993 PH/24 If the recipient of an error message contained an @ in the local part
2994 (suitably quoted, of course), incorrect values were put in $domain and
2995 $local_part during the evaluation of errors_copy.
3001 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3002 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3005 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3006 issue a MAIL command.
3008 PH/02 In an ACL statement such as
3010 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3012 if a client was not listed at all, or was listed with a value other than
3013 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3014 the condition was not true (as it should be), so access was not denied.
3015 The bug was that the ! inversion was incorrectly passed on to the second
3016 item. This has been fixed.
3018 PH/03 Added additional dnslists conditions == and =& which are different from
3019 = and & when the dns lookup returns more than one IP address.
3021 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3022 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3024 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3025 FSYNC, which compiles an option called disable_fsync that allows for
3026 bypassing fsync(). The documentation is heavily laced with warnings.
3028 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3030 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3031 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3032 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3033 including adding "make clean"; (3) Added -fPIC when compiling the test
3034 dynamically loaded module, to get rid of a warning.
3036 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3037 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3038 = 0s. The bug is otherwise harmless.
3040 PH/07 There was a bug in the dovecot authenticator such that the value of
3041 $auth1 could be overwritten, and so not correctly preserved, after a
3042 successful authentication. This usually meant that the value preserved by
3043 the server_setid option was incorrect.
3045 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3047 PH/09 Installed PCRE release 7.0.
3049 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3050 run for batched SMTP input. It is now run at the start of every message
3051 in the batch. While fixing this I discovered that the process information
3052 (output by running exiwhat) was not always getting set for -bs and -bS
3053 input. This is fixed, and it now also says "batched" for BSMTP.
3055 PH/11 Added control=no_pipelining.
3057 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3058 patch, slightly modified), and move the expansion of helo_data till after
3059 the connection is made in the smtp transport (so it can use these
3062 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3064 PH/14 Added log_selector = +pid.
3066 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3068 PH/16 Add ${if forany and ${if forall.
3070 PH/17 Added dsn_from option to vary the From: line in DSNs.
3072 PH/18 Flush SMTP output before performing a callout, unless control =
3073 no_callout_flush is set.
3075 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3076 was true (the default) a successful delivery failed to delete the retry
3077 item, thus causing premature timeout of the address. The bug is now
3080 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3082 PH/21 Long custom messages for fakedefer and fakereject are now split up
3083 into multiline responses in the same way that messages for "deny" and
3084 other ACL rejections are.
3086 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3087 with slight modification.
3089 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3090 draft, changing the syntax and factoring some duplicate code".
3092 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3093 for deliveries of the second and subsequent messages over the same SMTP
3096 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3097 ${reduce, with only minor "tidies".
3099 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3101 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3102 expansion side effects.
3104 PH/27 When a message times out after an over-quota error from an Exim-imposed
3105 quota, the bounce message says "mailbox is full". This message was not
3106 being given when it was a system quota that was exceeded. It now should
3109 MH/03 Made $recipients available in local_scan(). local_scan() already has
3110 better access to the recipient list through recipients_list[], but
3111 $recipients can be useful in postmaster-provided expansion strings.
3113 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3114 in the case of a MAIL command with additional options following the
3115 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3116 were accidentally chopped off.
3118 PH/29 SMTP synchronization checks are implemented when a command is read -
3119 there is a check that no more input is waiting when there shouldn't be
3120 any. However, for some commands, a delay in an ACL can mean that it is
3121 some time before the response is written. In this time, more input might
3122 arrive, invalidly. So now there are extra checks after an ACL has run for
3123 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3124 pipelining has not been advertised.
3126 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3128 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3129 returned in the SMTP response when smtp_return_error_details was false.
3130 This has been fixed.
3132 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3133 instead of the C I/O that was originally supplied, because problems were
3134 reported on Solaris.
3136 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3137 Exim which did not show up earlier: it was assuming that a call to
3138 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3139 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3140 was a macro that became an assignment, so it seemed to work. This has
3141 changed to a proper function call with a void return, hence the compile
3142 error. Exim's code has been fixed.
3144 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3147 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3149 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3151 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3152 because it turns out to be too big or there is a timeout) there is no
3153 "Completed" line in the log. When some messages of this type were
3154 selected by exigrep, they were listed as "not completed". Others were
3155 picked up by some special patterns. I have improved the selection
3156 criteria to be more general.
3158 PH/38 The host_find_failed option in the manualroute router can now be set
3159 to "ignore", to completely ignore a host whose IP address cannot be
3160 found. If all hosts are ignored, the behaviour is controlled by the new
3161 host_all_ignored option.
3163 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3164 homing or because of multiple MX records with /mx) generated more than
3165 one IP address, and the following item turned out to be the local host,
3166 all the secondary addresses of the first item were incorrectly removed
3167 from the list, along with the local host and any following hosts (which
3168 is what is supposed to happen).
3170 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3171 whoever called Exim into the -H file. In the case of the daemon it was
3172 behaving confusingly. When first started, it used values for whoever
3173 started the daemon, but after a SIGHUP it used the Exim user (because it
3174 calls itself on a restart). I have changed the code so that it now always
3177 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3178 message are rejected with the same error (e.g. no authentication or bad
3179 sender address), and a DATA command is nevertheless sent (as can happen
3180 with PIPELINING or a stupid MUA), the error message that was given to the
3181 RCPT commands is included in the rejection of the DATA command. This is
3182 intended to be helpful for MUAs that show only the final error to their
3185 PH/42 Another patch from the Sieve maintainer.
3187 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3188 Eximstats - Fixed some broken HTML links and added missing column headers
3190 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3191 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3193 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3195 PH/43 Yet another patch from the Sieve maintainer.
3197 PH/44 I found a way to check for a TCP/IP connection going away before sending
3198 the response to the final '.' that terminates a message, but only in the
3199 case where the client has not sent further data following the '.'
3200 (unfortunately, this is allowed). However, in many cases there won't be
3201 any further data because there won't be any more messages to send. A call
3202 to select() can be used: if it shows that the input is "ready", there is
3203 either input waiting, or the socket has been closed. An attempt to read
3204 the next input character can distinguish the two cases. Previously, Exim
3205 would have sent an OK response which the client would never have see.
3206 This could lead to message repetition. This fix should cure that, at
3207 least in a lot of common cases.
3209 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3210 advertised in response to EHLO.
3216 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3217 fixed by 4.65/MH/01 (is this a record?) are fixed:
3219 (i) An empty string was always treated as zero by the numeric comparison
3220 operators. This behaviour has been restored.
3222 (ii) It is documented that the numeric comparison operators always treat
3223 their arguments as decimal numbers. This was broken in that numbers
3224 starting with 0 were being interpreted as octal.
3226 While fixing these problems I realized that there was another issue that
3227 hadn't been noticed. Values of message_size_limit (both the global option
3228 and the transport option) were treated as octal if they started with 0.
3229 The documentation was vague. These values are now always treated as
3230 decimal, and I will make that clear in the documentation.
3236 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3237 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3240 MH/01 Don't check that the operands of numeric comparison operators are
3241 integers when their expansion is in "skipping" mode (fixes bug
3242 introduced by 4.64-PH/07).
3244 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3245 child addresses, Exim now panics and dies. Previously, because the count
3246 is held in a short int, deliveries were likely to be lost. As such a
3247 large number of recipients for a single message is ridiculous
3248 (performance will be very, very poor), I have chosen to impose a limit
3249 rather than extend the field.
3255 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3256 leftover -K file (the existence of which was triggered by #402).
3257 While we were at it, introduced process PID as part of the -K
3258 filename. This should rule out race conditions when creating
3261 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3262 processing considerably. Previous code took too long for large mails,
3263 triggering a timeout which in turn triggers #401.
3265 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3266 in the DK code in transports.c. sendfile() is not really portable,
3267 hence the _LINUX specificness.
3269 TF/01 In the add_headers option to the mail command in an Exim filter,
3270 there was a bug that Exim would claim a syntax error in any
3271 header after the first one which had an odd number of characters
3274 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3275 callout verification, Exim cached a "reject" for the entire domain. This
3276 is correct for most verifications, but it is not correct for a recipient
3277 verification with use_sender or use_postmaster set, because in that case
3278 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3279 case of MAIL FROM:<> rejection from other early rejections (e.g.
3280 rejection of HELO). When verifying a recipient using a non-null MAIL
3281 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3282 Whatever the result of the callout, the value of the domain cache is
3283 left unchanged (for any other kind of callout, getting as far as trying
3284 RCPT means that the domain itself is ok).
3286 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3289 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3290 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3291 session, the daemon ignores ECONNECT errors and logs others; it now
3292 ignores EPIPE as well.
3294 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3295 (quoted-printable decoding).
3297 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3298 later the small subsequent patch to fix an introduced bug.
3300 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3302 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3304 PH/08 An error is now given if message_size_limit is specified negative.
3306 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3307 to be given (somewhat) arbitrary names.
3309 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3312 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3313 miscellaneous code fixes
3315 PH/10 Added the log_reject_target ACL modifier to specify where to log
3318 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3319 hostname. This is wrong, because it relates to the incoming message (and
3320 probably the interface on which it is arriving) and not to the outgoing
3321 callout (which could be using a different interface). This has been
3322 changed to use the value of the helo_data option from the smtp transport
3323 instead - this is what is used when a message is actually being sent. If
3324 there is no remote transport (possible with a router that sets up host
3325 addresses), $smtp_active_hostname is used.
3327 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3328 tweaks were necessary in order to get it to work (see also 21 below):
3329 (a) The code assumed that strncpy() returns a negative number on buffer
3330 overflow, which isn't the case. Replaced with Exim's string_format()
3332 (b) There were several signed/unsigned issues. I just did the minimum
3333 hacking in of casts. There is scope for a larger refactoring.
3334 (c) The code used strcasecmp() which is not a standard C function.
3335 Replaced with Exim's strcmpic() function.
3336 (d) The code set only $1; it now sets $auth1 as well.
3337 (e) A simple test gave the error "authentication client didn't specify
3338 service in request". It would seem that Dovecot has changed its
3339 interface. Fortunately there's a specification; I followed it and
3340 changed what the client sends and it appears to be working now.
3342 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3345 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3346 address A is aliased to B and C, where B exists and C does not. Without
3347 -v the output is "A verified" because verification stops after a
3348 successful redirection if more than one address is generated. However,
3349 with -v the child addresses are also verified. Exim was outputting "A
3350 failed to verify" and then showing the successful verification for C,
3351 with its parentage. It now outputs "B failed to verify", showing B's
3352 parentage before showing the successful verification of C.
3354 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3355 look up a TXT record in a specific list after matching in a combined
3358 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3359 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3360 they consult the DNS. I had assumed they would set it the way they
3361 wanted; and indeed my experiments on Linux seem to show that in some
3362 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3363 To be on the safe side, however, I have now made the interface to
3364 host_find_byname() similar to host_find_bydns(), with an argument
3365 containing the DNS resolver options. The host_find_byname() function now
3366 sets these options at its start, just as host_find_bydns() does. The smtp
3367 transport options dns_qualify_single and dns_search_parents are passed to
3368 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3369 of host_find_byname() use the default settings of RES_DEFNAMES
3370 (qualify_single) but not RES_DNSRCH (search_parents).
3372 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3373 spool_read_header() do less string testing, by means of a preliminary
3374 switch on the second character of optional "-foo" lines. (This is
3375 overdue, caused by the large number of possibilities that now exist.
3376 Originally there were few.) While I was there, I also converted the
3377 str(n)cmp tests so they don't re-test the leading "-" and the first
3378 character, in the hope this might squeeze out yet more improvement.
3380 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3381 flag allowing group syntax was set by the header_syntax check but not
3382 turned off, possible causing trouble later; (2) The flag was not being
3383 set at all for the header_verify test, causing "group"-style headers to
3384 be rejected. I have now set it in this case, and also caused header_
3385 verify to ignore an empty address taken from a group. While doing this, I
3386 came across some other cases where the code for allowing group syntax
3387 while scanning a header line wasn't quite right (mostly, not resetting
3388 the flag correctly in the right place). These bugs could have caused
3389 trouble for malformed header lines. I hope it is now all correct.
3391 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3392 with the "reply" argument non-NULL. The code, however (which originally
3393 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3394 but it didn't always do it. This confused somebody who was copying the
3395 code for some other use. I have removed all the tests.
3397 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3398 feature that was used to support insecure browsers during the U.S. crypto
3399 embargo. It requires special client support, and Exim is probably the
3400 only MTA that supported it -- and would never use it because real RSA is
3401 always available. This code has been removed, because it had the bad
3402 effect of slowing Exim down by computing (never used) parameters for the
3403 RSA_EXPORT functionality.
3405 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3406 authenticator to fail if there's a tab character in the incoming data
3407 (there should never be unless someone is messing about, as it's supposed
3408 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3409 if the connection is using TLS or if the remote IP is the same as the
3410 local IP, and the "valid-client-cert option" if a client certificate has
3413 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3414 authenticators. This can be used for authorization after authentication
3415 succeeds. (In the case of plaintext, it servers for both authentication
3418 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3419 if any retry times were supplied.
3421 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3422 connection, where there is no client IP address to check. In this
3423 situation, the verify now always succeeds.
3425 PH/25 Applied John Jetmore's -Mset patch.
3427 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3429 PH/27 In a string expansion for a processed (not raw) header when multiple
3430 headers of the same name were present, leading whitespace was being
3431 removed from all of them, but trailing whitespace was being removed only
3432 from the last one. Now trailing whitespace is removed from each header
3433 before concatenation. Completely empty headers in a concatenation (as
3434 before) are ignored.
3436 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3437 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3439 PH/29 [Removed. This was a change that I later backed out, and forgot to
3440 correct the ChangeLog entry (that I had efficiently created) before
3441 committing the later change.]
3443 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3444 address errors (4xx response to RCPT) over the same connection as other
3445 messages routed to the same hosts. Such deliveries are always "forced",
3446 so retry times are not inspected. This resulted in far too many retries
3447 for the affected addresses. The effect occurred only when there were more
3448 hosts than the hosts_max_try setting in the smtp transport when it had
3449 the 4xx errors. Those hosts that it had tried were not added to the list
3450 of hosts for which the message was waiting, so if all were tried, there
3451 was no problem. Two fixes have been applied:
3453 (i) If there are any address or message errors in an SMTP delivery, none
3454 of the hosts (tried or untried) are now added to the list of hosts
3455 for which the message is waiting, so the message should not be a
3456 candidate for sending over the same connection that was used for a
3457 successful delivery of some other message. This seems entirely
3458 reasonable: after all the message is NOT "waiting for some host".
3459 This is so "obvious" that I'm not sure why it wasn't done
3460 previously. Hope I haven't missed anything, but it can't do any
3461 harm, as the worst effect is to miss an optimization.
3463 (ii) If, despite (i), such a delivery is accidentally attempted, the
3464 routing retry time is respected, so at least it doesn't keep
3465 hammering the server.
3467 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3468 in ${readsocket because some servers need this prod.
3470 PH/32 Added some extra debug output when updating a wait-xxx database.
3472 PH/33 The hint "could be header name not terminated by colon", which has been
3473 given for certain expansion errors for a long time, was not being given
3474 for the ${if def:h_colon_omitted{... case.
3476 PH/34 The spec says: "With one important exception, whenever a domain list is
3477 being scanned, $domain contains the subject domain." There was at least
3478 one case where this was not true.
3480 PH/35 The error "getsockname() failed: connection reset by peer" was being
3481 written to the panic log as well as the main log, but it isn't really
3482 panic-worthy as it just means the connection died rather early on. I have
3483 removed the panic log writing for the ECONNRESET error when getsockname()
3486 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3487 runs only) independently of the message's sender address. This meant
3488 that, if the 4xx error was in fact related to the sender, a different
3489 message to the same recipient with a different sender could confuse
3490 things. In particular, this can happen when sending to a greylisting
3491 server, but other circumstances could also provoke similar problems.
3492 I have changed the default so that the retry time for these errors is now
3493 based a combination of the sender and recipient addresses. This change
3494 can be overridden by setting address_retry_include_sender=false in the
3497 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3498 remote server are returned as part of bounce messages. This was not
3499 happening for LMTP over a pipe (the lmtp transport), but now it is the
3500 same for both kinds of LMTP.
3502 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3503 in header lines that were in fact CNAMEs. This is no longer the case.
3505 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3506 and queue runs started by the daemon processed all messages. This has
3507 been fixed so that -R and -S can now usefully be given with -q<time>.
3509 PH/40 Import PCRE release 6.7 (fixes some bugs).
3511 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3513 PH/42 Give an error if -q is specified more than once.
3515 PH/43 Renamed the variables $interface_address and $interface_port as
3516 $received_ip_address and $received_port, to make it clear that these
3517 values apply to message reception, and not to the outgoing interface when
3518 a message is delivered. (The old names remain recognized, of course.)
3520 PH/44 There was no timeout on the connect() call when using a Unix domain
3521 socket in the ${readsocket expansion. There now is.
3523 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3524 be meaningful with "accept".
3526 SC/01 Eximstats V1.43
3527 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3529 SC/02 Eximstats V1.44
3530 Use a glob alias rather than an array ref in the generated
3531 parser. This improves both readability and performance.
3533 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3534 Collect SpamAssassin and rejection statistics.
3535 Don't display local sender or destination tables unless
3536 there is data to show.
3537 Added average volumes into the top table text output.
3539 SC/04 Eximstats V1.46
3540 Collect data on the number of addresses (recipients)
3541 as well as the number of messages.
3543 SC/05 Eximstats V1.47
3544 Added 'Message too big' to the list of mail rejection
3545 reasons (thanks to Marco Gaiarin).
3547 SC/06 Eximstats V1.48
3548 Mainlog lines which have GMT offsets and are too short to
3549 have a flag are now skipped.
3551 SC/07 Eximstats V1.49 (Alain Williams)
3552 Added the -emptyok flag.
3554 SC/08 Eximstats V1.50
3555 Fixes for obtaining the IP address from reject messages.
3557 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3558 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3559 whitespace changes from 4.64-PH/27
3561 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3564 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3565 are found, allow negative numbers in numeric criteria)
3567 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3569 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3570 to match changes made in 4.64-PH/43
3572 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3574 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3575 "domain" is actually an IP address, and force a failure. This locks out
3576 those revolvers/nameservers that support "A-for-A" lookups, in
3577 contravention of the specifications.
3579 PH/48 When a host name was looked up from an IP address, and the subsequent
3580 forward lookup of the name timed out, the host name was left in
3581 $sender_host_name, contrary to the specification.
3583 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3584 restricted to single-key lookups, Exim was not diagnosing an error if
3585 * or *@ was used with a query-style lookup.
3587 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3589 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3590 long ago, but noone interested enough thought of it. Let's just say that
3591 the "1.1" means that there are some new functions that weren't there at
3592 some point in the past.
3594 PH/51 Error processing for expansion failure of helo_data from an smtp
3595 transport during callout processing was broken.
3597 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3598 tested/used via the -bh/-bhc/-bs options.
3600 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3601 bug, fixed in subsequent PCRE releases).
3603 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3604 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3606 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3612 SC/01 Use a glob alias rather than an array ref in eximstats generated
3613 parser. This improves both readability and performance.
3615 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3616 Don't display local sender or destination tables in eximstats unless
3617 there is data to show.
3618 Added average volumes into the eximstats top table text output.
3620 SC/03 Collect data on the number of addresses (recipients) as well
3621 as the number of messages in eximstats.
3623 TF/01 Correct an error in the documentation for the redirect router. Exim
3624 does (usually) call initgroups() when daemonizing.
3626 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3627 with consistent privilege compared to when running as a daemon.
3629 TF/03 Note in the spec that $authenticated_id is not set for local
3630 submissions from trusted users.
3632 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3633 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3635 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3636 by adding some example configuration directives to the default
3637 configuration file. A little bit of work is required to uncomment the
3638 directives and define how usernames and passwords are checked, but
3639 there is now a framework to start from.
3641 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3642 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3643 without this. I don't know how relevant this is to other LDAP libraries.
3645 PH/02 Add the verb name to the "unknown ACL verb" error.
3647 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3649 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3651 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3652 directory not expanded when it should be if an expanded home directory
3653 was set for the address (which is overridden by the transport).
3655 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3658 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3659 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3660 because it is too late at that time, and has no effect.
3662 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3663 security issue with \' (bugzilla #107). I could not use the
3664 PQescapeStringConn() function, because it needs a PGconn value as one of
3667 PH/08 When testing addresses using -bt, indicate those final addresses that
3668 are duplicates that would not cause an additional delivery. At least one
3669 person was confused, thinking that -bt output corresponded to deliveries.
3670 (Suppressing duplicates isn't a good idea as you lose the information
3671 about possibly different redirections that led to the duplicates.)
3673 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3674 systems where poll() doesn't work, in particular OS X.
3676 PH/10 Added more information to debugging output for retry time not reached.
3678 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3679 operations in malware.c.
3681 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3684 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3685 syslog_duplication set false, log lines that would normally be written
3686 both the the main log and to the reject log were not written to syslog at
3689 PH/14 In the default configuration, change the use of "message" in ACL warn
3690 statements to "add_header".
3692 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3693 not followed by a command (e.g. "seen endif").
3695 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3696 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3699 PH/17 Added extra conditions to the default value of delay_warning_condition
3703 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3704 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3705 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3708 The Auto-Submitted: and various List- headers are standardised, whereas I
3709 don't think Precedence: ever was.
3711 PH/18 Refactored debugging code in route_finduser() to show more information,
3712 in particular, the error code if getpwnam() issues one.
3714 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3715 This is apparently needed in addition to the PH/07 change above to avoid
3716 any possible encoding problems.
3718 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3719 but not after initializing Perl.
3721 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3722 output them only if debugging. By default they are written stderr,
3723 apparently, which is not desirable.
3725 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3728 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3731 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3733 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3734 authenticated or an ident call has been made. Suppress the default
3735 values for $authenticated_id and $authenticated_sender (but permit -oMai
3736 and -oMas) when testing with -bh.
3738 PH/24 Re-jigged the order of the tests in the default configuration so that the
3739 tests for valid domains and recipients precede the DNS black list and CSA
3740 tests, on the grounds that those ones are more expensive.
3742 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3743 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3744 This bug exists in every version of Exim that I still have, right back to
3747 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3748 However, an attempt to turn on case-sensitivity in a regex key by
3749 including (?-i) didn't work because the subject string was already
3750 lowercased, and the effects were non-intuitive. It turns out that a
3751 one-line patch can be used to allow (?-i) to work as expected.
3757 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3758 other effects) broke the use of negated acl sub-conditions.
3760 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3763 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3764 "Deny" causes Exim to reject the incoming connection with a 554 error.
3765 Unfortunately, if there is a major crisis, such as a disk failure,
3766 tcp-wrappers gives "deny", whereas what one would like would be some
3767 kind of temporary error. A kludge has been added to help with this.
3768 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3769 554 error is used if errno is still zero or contains ENOENT (which occurs
3770 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3773 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3775 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3776 errors. Otherwise a message that provokes a temporary error (when other
3777 messages do not) can cause a whole host to time out.
3779 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3780 addresses were routed directly to files or pipes from a redirect router.
3781 File deliveries just didn't batch; pipe deliveries might have suffered
3784 PH/06 A failure to get a lock for a hints database would erroneously always say
3785 "Failed to get write lock", even when it was really a read lock.
3787 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3788 of 0600. This has been changed to use the value of the lockfile_mode
3789 option (which defaults to 0600).
3791 PH/08 Applied small patch from the Sieve maintainer.
3793 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3794 folder from quota calculations, a direct delivery into this folder messed
3795 up the contents of the maildirsize file. This was because the regex was
3796 used only to exclude .Trash (or whatever) when the size of the mailbox
3797 was calculated. There was no check that a delivery was happening into an
3798 excluded directory. This bug has been fixed by ignoring all quota
3799 processing for deliveries into excluded directories.
3801 PH/10 Added the maildirfolder_create_regex option to appendfile.
3807 PH/01 The code for finding all the local interface addresses on a FreeBSD
3808 system running IPv6 was broken. This may well have applied to all BSD
3809 systems, as well as to others that have similar system calls. The broken
3810 code found IPv4 interfaces correctly, but gave incorrect values for the
3811 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3812 that it would not match correctly against @[] and not recognize the IPv6
3815 PH/02 The ipliteral router was not recognizing addresses of the form user@
3816 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3818 PH/03 Added disable_ipv6.
3820 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3821 decoded form, because it is most often used to construct To: headers
3822 lines in autoreplies, and the decoded form may well be syntactically
3823 invalid. However, $reply_address has leading white space removed, and all
3824 newlines turned into spaces so that the autoreply transport does not
3827 PH/05 If group was specified without a user on a router, and no group or user
3828 was specified on a transport, the group from the router was ignored.
3830 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3831 for visible compile-time settings that can be used to change these
3832 numbers, for those that want even more. Backwards compatibility with old
3833 spool files has been maintained. However, going back to a previous Exim
3834 release will lost any variables that are in spool files.
3836 PH/07 Two small changes when running in the test harness: increase delay when
3837 passing a TCP/IP connection to a new process, in case the original
3838 process has to generate a bounce, and remove special handling of
3839 127.0.0.2 (sic), which is no longer necessary.
3841 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3842 be the same on different OS.
3844 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3847 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3848 whether --show-vars was specified or not
3850 JJ/02 exipick: Added support for new ACL variable spool format introduced
3853 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3854 syntactically invalid From: or Reply-to: line, and a filter used this to
3855 generate an autoreply, and therefore failed to obtain an address for the
3856 autoreply, Exim could try to deliver to a non-existent relative file
3857 name, causing unrelated and misleading errors. What now happens is that
3858 it logs this as a hard delivery error, but does not attempt to create a
3861 PH/11 The exinext utility has a -C option for testing purposes, but although
3862 the given file was scanned by exinext itself; it wasn't being passed on
3863 when Exim was called.
3865 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3866 an end-of-file indication when reading a command response.
3868 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3869 compiled. In many other places in Exim, IPv6 addresses are always
3870 recognized, so I have changed this. It also means that IPv4 domain
3871 literals of the form [IPV4:n.n.n.n] are now always recognized.
3873 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3874 used if the router is not running as root, for example, when verifying at
3875 ACL time, or when using -bh. The debugging output from this situation was
3876 non-existent - all you got was a failure to exec. I have made two
3879 (a) Failures to set uid/gid, the current directory, or a process leader
3880 in a subprocess such as that created by queryprogram now generate
3881 suitable debugging output when -d is set.
3883 (b) The queryprogram router detects when it is not running as root,
3884 outputs suitable debugging information if -d is set, and then runs
3885 the subprocess without attempting to change uid/gid.
3887 PH/15 Minor change to Makefile for building test_host (undocumented testing
3890 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3891 additional section of a DNS packet that returns MX or SRV records.
3892 Instead, it always explicitly searches for A/AAAA records. This avoids
3893 major problems that occur when a DNS server includes only records of one
3894 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3895 fixed another bug: if SRV records were looked up and the corresponding
3896 address records were *not* found in the additional section, the port
3897 values from the SRV records were lost.
3899 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3900 using the correct key (the original address) when searching the retry
3901 rules in order to find which one to use for generating the retry hint.
3903 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3904 adding the default one. Similarly, if it contains a Reply-To: header, the
3905 errors_reply_to option, if set, is not used.
3907 PH/19 When calculating a retry time, Exim used to measure the "time since
3908 failure" by looking at the "first failed" field in the retry record. Now
3909 it does not use this if it is later than than the arrival time of the
3910 message. Instead it uses the arrival time. This makes for better
3911 behaviour in cases where some deliveries succeed, thus re-setting the
3912 "first failed" field. An example is a quota failure for a huge message
3913 when small messages continue to be delivered. Without this change, the
3914 "time since failure" will always be short, possible causing more frequent
3915 delivery attempts for the huge message than are intended.
3916 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3918 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3919 $1, $2, $3) because the numerical variables can be reset during some
3920 expansion items (e.g. "match"), thereby losing the authentication data.
3922 PH/21 Make -bV show the size of off_t variables so that the test suite can
3923 decide whether to run tests for quotas > 2G.
3925 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3926 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3927 filecount value is greater than 2G or if a quota value is greater than 2G
3928 on a system where the size of off_t is not greater than 4, a panic error
3931 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3932 never match. The debug and -bh output now contains an explicit error
3933 message indicating a malformed IPv4 address or mask.
3935 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3936 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3937 PH/23 above applies.
3939 PH/25 Do not write to syslog when running in the test harness. The only
3940 occasion when this arises is a failure to open the main or panic logs
3941 (for which there is an explicit test).
3943 PH/26 Added the /no_tell option to "control=freeze".
3945 PH/27 If a host name lookup failed very early in a connection, for example, if
3946 the IP address matched host_lookup and the reverse lookup yielded a name
3947 that did not have a forward lookup, an error message of the form "no IP
3948 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3949 could be logged. Now it outputs the IP address instead of "NULL".
3951 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3952 allows the sender and the authenticated sender to be set when
3953 submitting a message from within Exim. Since child_open_exim() is
3954 documented for local_scan(), the new function should be too.
3956 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3957 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3958 results in an empty string is now treated as unset.
3960 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3962 PH/31 Added sender_verify_fail logging option.
3964 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3965 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3966 tidied the source and removed it altogether.
3968 PH/33 When a queue run was abandoned because the load average was too high, a
3969 log line was always written; now it is written only if the queue_run log
3970 selector is set. In addition, the log line for abandonment now contains
3971 information about the queue run such as the pid. This is always present
3972 in "start" and "stop" lines but was omitted from the "abandon" line.
3974 PH/34 Omit spaces between a header name and the colon in the error message that
3975 is given when verify = headers_syntax fails (if there are lots of them,
3976 the message gets confusing).
3978 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3979 names, as there are now some PTR records that contain slashes. This check
3980 is only to protect against broken name servers that fall over on strange
3981 characters, so the fact that it applies to all lookups doesn't matter.
3983 PH/36 Now that the new test suite is complete, we can remove some of the
3984 special code in Exim that was needed for the old test suite. For example,
3985 sorting DNS records because real resolvers return them in an arbitrary
3986 order. The new test suite's fake resolver always returns records in the
3989 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3990 bounces) except when queue_only is set, to avoid logging races between
3991 the different processes.
3993 PH/38 Panic-die if .include specifies a non-absolute path.
3995 PH/39 A tweak to the "H" retry rule from its user.
3997 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3998 a label. They prevented compilation on older perls.
4000 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4001 a warning to be raised on newish perls.
4003 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4004 on queue. Changes to match documented behaviour of showing count of
4005 messages matching specified criteria.
4007 PH/40 Changed the default ident timeout from 30s to 5s.
4009 PH/41 Added support for the use of login_cap features, on those BSD systems
4010 that have them, for controlling the resources used by pipe deliveries.
4012 PH/42 The content-scanning code uses fopen() to create files in which to put
4013 message data. Previously it was not paying any attention to the mode of
4014 the files. Exim runs with umask(0) because the rest of the code creates
4015 files with open(), and sets the required mode explicitly. Thus, these
4016 files were ending up world-writeable. This was not a big issue, because,
4017 being within the spool directory, they were not world-accessible. I have
4018 created a function called modefopen, which takes an additional mode
4019 argument. It sets umask(777), creates the file, chmods it to the required
4020 mode, then resets the umask. All the relevant calls to fopen() in the
4021 content scanning code have been changed to use this function.
4023 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4024 to 24 hours. This avoids potential overflow problems when processing G
4025 and H retry rules. I suspect nobody ever tinkers with this value.
4027 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4029 PH/45 When the plaintext authenticator is running as a client, the server's
4030 challenges are checked to ensure they are valid base64 strings. By
4031 default, the authentication attempt is cancelled if an invalid string is
4032 received. Setting client_ignore_invalid_base64 true ignores these errors.
4033 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4034 they are received. Thus, the responses can be made to depend on the
4035 challenges. If an invalid string is ignored, an empty string is placed in
4038 PH/46 Messages that are created by the autoreply transport now contains a
4039 References: header, in accordance with RFCs 2822 and 3834.
4041 PH/47 Added authenticated_sender_force to the smtp transport.
4043 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4045 PH/49 Installed latest patch from the Sieve maintainer.
4047 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4048 was also set, the appendfile transport was unnecessarily scanning a
4049 directory of message files (e.g. for maildir delivery) to find the count
4050 of files (along with the size), even though it did not need this
4051 information. It now does the scan only if it needs to find either the
4052 size of the count of files.
4054 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4056 PH/52 Two bugs concerned with error handling when the smtp transport is
4059 (i) Exim was not creating retry information for temporary errors given
4060 for individual recipients after the DATA command when the smtp transport
4061 was used in LMTP mode. This meant that they could be retried too
4062 frequently, and not timed out correctly.
4064 (ii) Exim was setting the flag that allows error details to be returned
4065 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4066 recipients that were returned after the DATA command.
4068 PH/53 This is related to PH/52, but is more general: for any failing address,
4069 when detailed error information was permitted to be returned to the
4070 sender, but the error was temporary, then after the final timeout, only
4071 "retry timeout exceeded" was returned. Now it returns the full error as
4072 well as "retry timeout exceeded".
4074 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4075 do this, and (what is worse) MTAs that accept it.
4077 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4078 will now be deprecated.
4080 PH/56 New os.c-cygwin from the Cygwin maintainer.
4082 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4083 formats (previously only available in exim formats via -bpr, -bpru,
4084 and -bpra. Now also available in native and exiqgrep formats)
4086 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4087 with very large, slow to parse queues
4089 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4091 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4093 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4094 responses to authentication challenges, though it was showing the
4095 challenges; (ii) I've removed the CR characters from the debug output for
4098 PH/58 Allow for the insertion of a newline as well as a space when a string
4099 is turned into more than one encoded-word during RFC 2047 encoding. The
4100 Sieve code now uses this.
4102 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4103 data_4xx, lost_connection, tls_required.
4105 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4106 message was being sent as an SMTP response.
4108 PH/61 Add -l and -k options to exicyclog.
4110 PH/62 When verifying, if an address was redirected to one new address, so that
4111 verification continued, and the new address failed or deferred after
4112 having set something in $address_data, the value of $address_data was not
4113 passed back to the ACL. This was different to the case when no
4114 redirection occurred. The value is now passed back in both cases.
4116 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4117 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4118 use login_cap.h, so on its own it isn't the distinguishing feature. The
4119 new name refers directly to the setclassresources() function.
4121 PH/65 Added configuration files for NetBSD3.
4123 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4125 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4126 is preferred over IPv4.
4128 PH/68 The bounce_return_message and bounce_return_body options were not being
4129 honoured for bounces generated during the reception of non-SMTP messages.
4130 In particular, this applied to messages rejected by the ACL. This bug has
4131 been fixed. However, if bounce_return_message is true and bounce_return_
4132 body is false, the headers that are returned for a non-SMTP message
4133 include only those that have been read before the error was detected.
4134 (In the case of an ACL rejection, they have all been read.)
4136 PH/69 The HTML version of the specification is now built in a directory called
4137 spec_html instead of spec.html, because the latter looks like a path with
4138 a MIME-type, and this confuses some software.
4140 PH/70 Catch two compiler warnings in sieve.c.
4142 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4143 function verify_get_ident() calls ip_connect() to connect a socket, but
4144 if the "connect()" function timed out, ip_connect() used to close the
4145 socket. However, verify_get_ident() also closes the socket later, and in
4146 between Exim writes to the log, which may get opened at this point. When
4147 the socket was closed in ip_connect(), the log could get the same file
4148 descriptor number as the socket. This naturally causes chaos. The fix is
4149 not to close the socket in ip_connect(); the socket should be closed by
4150 the function that creates it. There was only one place in the code where
4151 this was missing, in the iplookup router, which I don't think anybody now
4152 uses, but I've fixed it anyway.
4154 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4155 well as to direct DNS lookups. Otherwise the handling of names in host
4156 lists is inconsistent and therefore confusing.
4162 PH/01 Two changes to the default runtime configuration:
4164 (1) Move the checks for relay_from_hosts and authenticated clients from
4165 after to before the (commented out) DNS black list checks.
4167 (2) Add control=submission to the relay_from_hosts and authenticated
4168 clients checks, on the grounds that messages accepted by these
4169 statements are most likely to be submissions.
4171 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4173 (1) Generate an error if the third argument for the ${prvs expansion is
4176 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4179 (3) Reset the variables that are obtained from the first argument of
4180 ${prvscheck and used in the second argument before leaving the code,
4181 because their memory is reclaimed, so using them afterwards may do
4184 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4185 one (it's much easier than Tom thought :-).
4187 (5) Because of (4), we can now allow for the use of $prvscheck_result
4188 inside the third argument.
4190 PH/03 For some reason, the default setting of PATH when running a command from
4191 a pipe transport was just "/usr/bin". I have changed it to
4194 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4195 anything to be listed in the output from -bV.
4197 PH/05 When a filter generated an autoreply, the entire To: header line was
4198 quoted in the delivery log line, like this:
4200 => >A.N.Other <ano@some.domain> <original@ddress> ...
4202 This has been changed so that it extracts the operative address. There
4203 may be more than one such address. If so, they are comma-separated, like
4206 => >ano@some.domain,ona@other.domain <original@ddress> ...
4208 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4209 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4210 not being looked up in the rDNS to get a host name, Exim was showing the
4211 IP address twice in Received: lines, even though the IP addresses were
4212 identical. For example:
4214 Received: from [1.2.3.4] (helo=[1.2.3.4])
4216 However, if the real host name was known, it was omitting the HELO data
4217 if it matched the actual IP address. This has been tidied up so that it
4218 doesn't show the same IP address twice.
4220 PH/07 When both +timestamp and +memory debugging was on, the value given by
4221 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4222 called by the debug printing, thereby overwriting the timestamp buffer.
4223 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4225 PH/08 When the original message was included in an autoreply transport, it
4226 always said "this is a copy of the message, including all the headers",
4227 even if body_only or headers_only was set. It now gives an appropriate
4230 PH/09 Applied a patch from the Sieve maintainer which:
4232 o fixes some comments
4233 o adds the (disabled) notify extension core
4234 o adds some debug output for the result of if/elsif tests
4235 o points to the current vacation draft in the documentation
4236 and documents the missing references header update
4240 o fixes a bug in processing the envelope test (when testing
4241 multiple envelope elements, the last element determined the
4244 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4245 Electronic Mail") by including:
4247 Auto-submitted: auto-generated
4249 in the messages that it generates (bounce messages and others, such as
4250 warnings). In the case of bounce messages for non-SMTP messages, there was
4251 also a typo: it was using "Auto_submitted" (underscore instead of
4252 hyphen). Since every message generated by Exim is necessarily in response
4253 to another message, thes have all been changed to:
4255 Auto-Submitted: auto-replied
4257 in accordance with these statements in the RFC:
4259 The auto-replied keyword:
4261 - SHOULD be used on messages sent in direct response to another
4262 message by an automatic process,
4264 - MUST NOT be used on manually-generated messages,
4266 - MAY be used on Delivery Status Notifications (DSNs) and Message
4267 Disposition Notifications (MDNs),
4269 - MUST NOT be used on messages generated by automatic or periodic
4270 processes, except for messages which are automatic responses to
4273 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4274 to the default Received: header definition.
4276 PH/12 Added log selector acl_warn_skipped (default on).
4278 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4279 variables because (a) they are in the wrong storage pool and (b) even if
4280 they were copied, it wouldn't work properly because of the caching.
4282 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4283 checking when decoding. Apparently there are clients that generate
4284 overlong encoded strings. Why am I not surprised?
4286 PH/15 If the first argument of "${if match_address" was not empty, but did not
4287 contain an "@" character, Exim crashed. Now it writes a panic log message
4288 and treats the condition as false.
4290 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4292 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4293 extension "envelope-auth". The code is finished and in agreement with
4294 other implementations, but there is no documentation so far and in fact,
4295 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4296 not changing the active code.
4298 Print executed "if" and "elsif" statements when debugging is used. This
4299 helps a great deal to understand what a filter does.
4301 Document more things not specified clearly in RFC3028. I had all this
4302 sorted out, when out of a sudden new issues came to my mind. Oops."
4304 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4307 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4308 canonical form (as documented). However, after a host name lookup from
4309 the IP address, check_host() was doing a simple string comparison with
4310 addresses acquired from the DNS when checking that the found name did
4311 have the original IP as one of its addresses. Since any found IPv6
4312 addresses are likely to be in abbreviated form, the comparison could
4313 fail. Luckily, there already exists a function for doing the comparison
4314 by converting both addresses to binary, so now that is used instead of
4315 the text comparison.
4317 PH/20 There was another similar case to PH/19, when a complete host name was
4318 given in a host list; looking up its IP address could give an abbreviated
4319 form, whereas the current host's name might or might not be abbreviated.
4320 The same fix has been applied.
4326 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4327 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4330 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4331 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4333 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4335 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4336 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4337 Also, for "Auto-Submitted", treat anything other than "no" as
4338 non-personal, in accordance with RFC 3834. (Previously it treated
4339 anything starting "auto-" as non-personal.)
4341 TF/01 The control=submission/name=... option had a problem with syntax
4342 errors if the name included a slash character. The /name= option
4343 now slurps the rest of the string, so it can include any characters
4344 but it must come last in the list of options (after /sender_retain
4347 PH/05 Some modifications to the interface to the fake nameserver for the new
4355 TK/01 Added the "success_on_redirect" address verification option. See
4356 NewStuff for rationale and an example.
4358 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4360 PH/02 Patch to exigrep to allow it to work on syslog lines.
4362 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4363 fread() to skip over the body file's header line, because in Cygwin the
4364 header line is locked and is inaccessible.
4366 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4367 co-exist for some time) to make it clear that it is the Exim ID that is
4368 referenced, not the Message-ID: header line.
4370 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4371 string_format() function, because snprintf() does not exist on all
4374 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4375 ${stat: expansion item.
4377 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4378 protocol synchronization error", to keep the pedants happy.
4380 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4381 well as for IRIX systems, when gcc is being used. See the host.c source
4384 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4386 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4389 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4390 transport and to the smtp transport in LMTP mode.
4392 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4394 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4395 run a filter in a subprocess. This could lead to confusion in subsequent
4396 lookups in the parent process. There should also be a search_tidyup() at
4397 the end of the subprocess.
4399 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4400 only if the host matched helo_try_verify_hosts, which caused the
4401 verification to occur when the EHLO/HELO command was issued. The ACL just
4402 tested the remembered result. Now, if a previous verification attempt has
4403 not happened, "verify = helo" does it there and then.
4405 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4407 TK/03 Fix log output including CR from clamd.
4409 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4410 exist provoked a memory error which could cause a segfault.
4412 PH/15 Installed PCRE 6.2
4414 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4416 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4417 of the problem. Specifically, suggested +O2 rather than +O1 for the
4420 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4422 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4423 record, the port was not correctly logged when the outgoing_port log
4424 selector was set (it logged the transort's default port).
4426 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4427 fallback_hosts, and "hosts" in the smtp transport.
4429 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4430 host errors such as "Connection refused".
4432 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4433 authentication with radiusclient 0.4.9:
4435 - Error returned from rc_read_config was caught wrongly
4436 - Username/password not passed on to radius server due to wrong length.
4438 The presumption is that some radiusclient API changes for 4.51/PH/17
4439 were not taken care of correctly. The code is still untested by me (my
4440 Linux distribution still has 0.3.2 of radiusclient), but it was
4441 contributed by a Radius user.
4443 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4444 expanding the "port" option of the smtp transport.
4446 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4447 while reading a MIME header. Thanks to Tom Hughes for a patch.
4449 PH/24 Include config.h inside local_scan.h so that configuration settings are
4452 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4453 that in an ACL for RCPT (for example), you can examine exactly what was
4456 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4457 commands, but it was not correctly comparing the address with the actual
4458 client host address. Thus, it would show the EHLO address in Received:
4459 header lines when this was not necessary.
4461 PH/27 Added the % operator to ${eval:}.
4463 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4464 it should be ignoring failures (because with -C, for example, it has lost
4465 privilege). It wasn't ignoring creation failures other than "already
4468 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4471 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4472 because an input line was too long, either on its own, or by virtue of
4473 too many continuations, the temporary file was not being removed, and the
4474 return code was incorrect.
4476 PH/31 Missing "BOOL" in function definition in filtertest.c.
4478 PH/32 Applied Sieve patches from the maintainer.
4480 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4482 PH/33 Added "verify = not_blind".
4484 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4485 Local/Makefile (with some defaults set). These are used in built scripts
4486 such as exicyclog, but they have never been used in the exim_install
4487 script (though there are many overriding facilities there). I have
4488 arranged that the exim_install script now takes note of these two
4491 PH/35 Installed configuration files for Dragonfly.
4493 PH/36 When a locally submitted message by a trusted user did not contain a
4494 From: header, and the sender address was obtained from -f or from an SMTP
4495 MAIL command, and the trusted user did not use -F to supply a sender
4496 name, $originator_name was incorrectly used when constructing a From:
4497 header. Furthermore, $originator_name was used for submission mode
4498 messages from external hosts without From: headers in a similar way,
4499 which is clearly wrong.
4501 PH/37 Added control=suppress_local_fixups.
4503 PH/38 When log_selector = +received_sender was set, and the addition of the
4504 sender made the log line's construction buffer exactly full, or one byte
4505 less than full, an overflow happened when the terminating "\n" was
4508 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4509 when the result of a list match is failure because a DNS lookup failed.
4511 PH/40 RM_COMMAND is now used in the building process.
4513 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4514 the "build-* directories that it finds.
4516 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4517 address in a domain literal was a prefix of an interface address.
4519 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4520 when verifying a sender address, unless rewrite_headers is false.
4522 PH/44 Wrote a long comment about why errors_to addresses are verified as
4523 recipients, not senders.
4525 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4526 the ratelimit ACL was added.
4528 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4530 PH/46 Added extra information about PostgreSQL errors to the error string.
4532 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4533 avoiding the need to install special zones in a real server. This is
4534 backwards compatible; if it can't find the fake resolver, it drops back.
4535 Thus, both old and new test suites can be run.
4537 TF/02 Added util/ratelimit.pl
4539 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4540 clock is set back in time.
4542 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4543 Candler <B.Candler@pobox.com>.
4545 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4546 OK for addresses that are the result of redirections.
4548 PH/48 A number of further additions for the benefit of the new test suite,
4549 including a fake gethostbyname() that interfaces to the fake DNS resolver
4552 TF/06 The fix for widen_domains has also been applied to qualify_single and
4553 search_parents which are the other dnslookup options that can cause
4556 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4559 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4561 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4562 service" if it encountered an MX record with an empty target hostname.
4563 The message is now "an MX or SRV record indicated no SMTP service".
4565 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4566 if the DNS of the sending site is misconfigured. This is quite a
4567 common situation. This change restores the behaviour of treating a
4568 helo verification defer as a failure.
4570 PH/51 If self=fail was set on a router, the bounce message did not include the
4571 actual error message.
4577 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4579 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4580 command itself ended in error, the underlying message about the transport
4581 filter timeout was being overwritten with the pipe command error. Now the
4582 underlying error message should be appended to the second error message.
4584 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4586 PH/02 Reduce the amount of output that "make" produces by default. Full output
4587 can still be requested.
4589 PH/03 The warning log line about a condition test deferring for a "warn" verb
4590 was being output only once per connection, rather than after each
4591 occurrence (because it was using the same function as for successful
4592 "warn" verbs). This seems wrong, so I have changed it.
4594 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4595 it should not have, which might have caused a crash in the right
4596 circumstances, but probably never did.
4598 PH/04 Installed a modified version of Tony Finch's patch to make submission
4599 mode fix the return path as well as the Sender: header line, and to
4600 add a /name= option so that you can make the user's friendly name appear
4603 TF/03 Added the control = fakedefer ACL modifier.
4605 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4606 Mark Lowes for thorough testing.
4608 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4610 TK/03 Merged latest SRS patch from Miles Wilton.
4612 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4613 with the definition in sysexits.h (which is #included earlier).
4614 Fortunately, Exim does not actually use EX_OK. The code used to try to
4615 preserve the sysexits.h value, by assuming that macro definitions were
4616 scanned for macro replacements. I have been disabused of this notion,
4617 so now the code just undefines EX_OK before #including unistd.h.
4619 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4620 in the smtp transport. When a block could not be written in a single
4621 write() function, the timeout was being re-applied to each part-write.
4622 This seems wrong - if the receiver was accepting one byte at a time it
4623 would take for ever. The timeout is now adjusted when this happens. It
4624 doesn't have to be particularly precise.
4626 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4627 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4629 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4630 without a domain if the check to <postmaster@domain> fails.
4632 SC/01 Eximstats: added -xls and the ability to specify output files
4633 (patch written by Frank Heydlauf).
4635 SC/02 Eximstats: use FileHandles for outputting results.
4637 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4639 SC/04 Eximstats: fixed display of large numbers with -nvr option
4641 SC/05 Eximstats: fixed merging of reports with empty tables.
4643 SC/06 Eximstats: added the -include_original_destination flag
4645 SC/07 Eximstats: removed tabs and trailing whitespace.
4647 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4649 TK/06 MBOX spool code: Add real "From " MBOX separator line
4650 so the .eml file is really in mbox format (even though
4651 most programs do not really care). Patch from Alex Miller.
4653 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4654 The latter is generated from $received_to and is only set if the
4655 message has one envelope recipient. SA can use these headers,
4656 obviously out-of-the-box. Patch from Alex Miller.
4658 PH/08 The ${def test on a variable was returning false if the variable's
4659 value was "0", contrary to what the specification has always said!
4660 The result should be true unless the variable is empty.
4662 PH/09 The syntax error of a character other than { following "${if
4663 def:variable_name" (after optional whitespace) was not being diagnosed.
4664 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4665 accidental colon was present, for example, could give incorrect results.
4667 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4668 result is used (not including appendfile, where other changes are about
4671 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4672 This involved changing a lot of size variables from int to off_t. It
4673 should work with maildirs and everything.
4675 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4676 spamd dying while we are connected to it.
4678 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4681 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4682 function for BDB 4.3.
4684 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4686 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4687 the sources for PCRE have been re-arranged and more files are now
4690 PH/15 The code I had for printing potentially long long variables in PH/11
4691 above was not the best (it lost precision). The length of off_t variables
4692 is now inspected at build time, and an appropriate printing format (%ld
4693 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4694 to be "long long int" or "long int". This is needed for the internal
4695 formatting function string_vformat().
4697 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4698 the configuration file to be ":syslog", then the script "guesses" where
4699 the logs files are, rather than using the compiled in default. In our
4700 case the guess is not the same as the compiled default, so the script
4701 suddenly stopped working when I started to use syslog. The patch checks
4702 to see if log_file_path is "". If so, it attempts to read it from exim
4703 with no configuration file to get the compiled in version, before it
4704 falls back to the previous guessing code."
4706 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4707 implementing BATV in an Exim configuration. See NewStuff for the gory
4710 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4711 Makefile that are specific to HP-UX.
4713 PH/18 If the "use_postmaster" option was set for a recipient callout together
4714 with the "random" option, the postmaster address was used as the MAIL
4715 FROM address for the random test, but not for the subsequent recipient
4716 test. It is now used for both.
4718 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4719 patch removes a few documentation additions to RFC 3028, because the
4720 latest draft now contains them. It adds the new en;ascii-case comparator
4721 and a new error check for 8bit text in MIME parts. Comparator and
4722 require names are now matched exactly. I enabled the subaddress
4723 extension, but it is not well tested yet (read: it works for me)."
4725 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4726 rework some of the code of TK/09 above to avoid the hardwired use of
4727 "%lld" and "long long". Replaced the call to snprintf() with a call to
4730 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4731 records point to non-existent hosts", "retry timeout exceeded", and
4732 "retry time not reached for any host after a long failure period".
4734 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4735 experimental DomainKeys support:
4737 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4738 (2) On an error such as an illegally used "control", the wrong name for
4739 the control was given.
4741 These problems did NOT occur unless DomainKeys support was compiled.
4743 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4745 PH/24 Added ${if match_ip condition.
4747 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4748 databases so that it will be absolutely obvious if a crash occurs in the
4749 DB library. This is a regular occurrence (often caused by mis-matched
4752 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4753 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4754 picked up on a user's system that detects such things. There doesn't seem
4755 to be a gcc warning option for this - only an attribute that has to be
4756 put on the function's prototype. It seems that in Fedora Core 4 they have
4757 set this on a number of new functions. No doubt there will be more in due
4760 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4761 specify a transport. However, if an address that was verified by such a
4762 router was the subject of a callout, Exim crashed because it tried to
4763 read the rcpt_include_affixes from the non-existent transport. Now it
4764 just assumes that the setting of that option is false. This bug was
4765 introduced by 4.51/PH/31.
4767 PH/28 Changed -d+all to exclude +memory, because that information is very
4768 rarely of interest, but it makes the output a lot bigger. People tend to
4769 do -d+all out of habit.
4771 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4772 code in os-type was giving problems when libc.so lives in lib64, like on
4775 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4776 aren't the modern standard, and it seems that some systems' include files
4777 don't always have them. Exim was already checking for some of the newer
4778 ones like T_AAAA, and defining it itself. I've added checks for all the
4779 record types that Exim uses.
4781 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4782 not automatically generating a new one, as it is supposed to. This
4783 prevented TLS from working. If the file did exist, but contained invalid
4784 data, a new version was generated, as expected. It was only the case of a
4785 non-existent file that was broken.
4787 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4788 with a change in libdomainkeys > 0.64.
4790 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4791 from DNS. If the selector record carries the flag, it now has
4792 precedence over the domain-wide flag.
4794 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4796 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4797 the use of a port name that isn't defined in /etc/services) occurred, the
4798 message was deferred as in a normal delivery, and thus remained on the
4799 spool, instead of being failed because of the mua_wrapper setting. This
4800 is now fixed, and I tidied up some of the mua_wrapper messages at the
4803 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4804 the messages in a hash of arrays rather than using individual hashes.
4805 This is a bit cleaner and results in dramatic memory savings, albeit
4806 at a slight CPU cost.
4808 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4809 as requested by Marc Sherman.
4811 SC/10 Eximstats: added histograms for user specified patterns as requested
4814 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4816 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4817 fopen() in the content-scanning modules that did not already have it.
4823 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4824 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4826 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4828 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4830 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4831 to test Sieve filters that use "vacation".
4833 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4834 that changes the way the GnuTLS parameters are stored in the cache file.
4835 The new format can be generated externally. For backward compatibility,
4836 if the data in the cache doesn't make sense, Exim assumes it has read an
4837 old-format file, and it generates new data and writes a new file. This
4838 means that you can't go back to an older release without removing the
4841 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4842 work if there are any delivery delays because "one_time" forces the
4843 parent to be marked "delivered", so its unseen clone is never tried
4844 again. For this reason, Exim now forbids the simultaneous setting of
4847 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4848 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4849 entry if you want to know the details. The fix, however, neglected to
4850 consider the case when local delivery batching is involved. The test for
4851 "previously delivered" was not happening when checking to see if an
4852 address could be batched with a previous (undelivered) one; under
4853 certain circumstances this could lead to multiple deliveries to the same
4856 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4857 in its include files, and this causes problems building Exim.
4859 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4860 header_syntax) but Exim was just ignoring anything given after a slash.
4861 In particular, this caused confusion with an attempt to use "verify =
4862 reverse_host_lookup/defer_ok". An error is now given when options are
4863 supplied for verify items that do not have them. (Maybe reverse_host_
4864 lookup should have a defer_ok option, but that's a different point.)
4866 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4867 defined by RFC 821) to 2048, because there were problems with some AUTH
4868 commands, and RFC 1869 says the size should be increased for extended
4869 SMTP commands that take arguments.
4871 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4874 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4875 "unknown" error; now it says that the functionality isn't in the binary.
4877 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4878 an address' error message when a string expansion fails (syntax or
4879 whatever). Otherwise the password may appear in the log. Following change
4880 PH/42 below, there is no longer a chance of it appearing in a bounce
4883 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4885 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4886 of its addresses was ever tried. (Bugzilla bug #2.)
4888 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4889 the result incorrectly in the debug output. (It correctly added a newline
4890 to what was transported.)
4892 TF/01 Added $received_time.
4894 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4895 commented out examples of how to interface to a virus scanner and to
4896 SpamAssassin. Also added commented examples of av_scanner and
4897 spamd_address settings.
4899 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4900 and controls are allowed in which ACLs. There were a couple of minor
4901 errors. Some of the entries in the conditions table (which is a table of
4902 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4903 negation of where the condition IS allowed.
4905 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4907 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4908 header file does not have a version number, so I've had to invent a new
4909 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4910 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4911 radiusclient), but it was contributed by a Radius user.
4913 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4914 files or directories, for OpenSSL.
4916 PH/19 When an Exim process that is running as root has to create an Exim log
4917 file, it does so in a subprocess that runs as exim:exim so as to get the
4918 ownership right at creation (otherwise, other Exim processes might see
4919 the file with the wrong ownership). There was no test for failure of this
4920 fork() call, which would lead to the process getting stuck as it waited
4921 for a non-existent subprocess. Forks do occasionally fail when resources
4922 run out. I reviewed all the other calls to fork(); they all seem to check
4925 PH/20 When checking for unexpected SMTP input at connect time (before writing
4926 the banner), Exim was not dealing correctly with a non-positive return
4927 from the read() function. If the client had disconnected by this time,
4928 the result was a log entry for a synchronization error with an empty
4929 string after "input=" when read() returned zero. If read() returned -1
4930 (an event I could not check), uninitialized data bytes were printed.
4931 There were reports of junk text (parts of files, etc) appearing after
4934 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4936 PH/22 Added support for macro redefinition, and (re)definition in between
4937 driver and ACL definitions.
4939 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4940 forgetting to use the resulting value; it was using the unexpanded value.
4942 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4943 hadn't been configured. The fix is from Juergen Kreileder, who
4944 understands it better than I do:
4946 "Here's what I see happening with three configured cyrus_sasl
4947 authenticators configured (plain, login, cram-md5):
4949 On startup auth_cyrus_sasl_init() gets called for each of these.
4950 This means three calls to sasl_listmech() without a specified mech_list.
4951 => SASL tests which mechs of all available mechs actually work
4952 => three warnings about OTP not working
4953 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4955 With the patch, sasl_listmech() also gets called three times. But now
4956 SASL's mech_list option is set to the server_mech specified in the the
4957 authenticator. Or in other words, the answer from sasl_listmech()
4958 gets limited to just the mech you're testing for (which is different
4960 => the return list contains just 'plain' or 'login', 'cram-md5' or
4961 nothing depending on the value of ob->server_mech.
4963 I've just tested the patch: Authentication still works fine,
4964 unavailable mechs specified in the exim configuration are still
4965 caught, and the auth.log warnings about OTP are gone."
4967 PH/25 When debugging is enabled, the contents of the command line are added
4968 to the debugging output, even when log_selector=+arguments is not
4971 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4972 answer is "GNU", and only if the return is "GNU/something" is the answer
4975 PH/27 $acl_verify_message is now set immediately after the failure of a
4976 verification in an ACL, and so is available in subsequent modifiers. In
4977 particular, the message can be preserved by coding like this:
4979 warn !verify = sender
4980 set acl_m0 = $acl_verify_message
4982 Previously, $acl_verify_message was set only while expanding "message"
4983 and "log_message" when a very denied access.
4985 PH/28 Modified OS/os.c-Linux with
4987 -#ifndef OS_LOAD_AVERAGE
4988 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4990 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4991 nomenclature these days.)
4993 PH/29 Installed patch from the Sieve maintainer that adds the options
4994 sieve_useraddress and sieve_subaddress to the redirect router.
4996 PH/30 In these circumstances:
4997 . Two addresses routed to the same list of hosts;
4998 . First host does not offer TLS;
4999 . First host accepts first address;
5000 . First host gives temporary error to second address;
5001 . Second host offers TLS and a TLS session is established;
5002 . Second host accepts second address.
5003 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5004 and peerdn, if requested) that were in fact used only for the second
5007 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5008 attention to any local part prefix or suffix that was matched by the
5009 router that accepted the address. It now behaves in the same way as it
5010 does for delivery: the affixes are removed from the local part unless
5011 rcpt_include_affixes is set on the transport.
5013 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5014 timeout during the DATA phase of an incoming message.
5016 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5017 applied a patch sanctioned by the Sieve maintainer.
5019 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5020 the uid or gid is negative. A case of a negative gid caused this to be
5021 noticed. The fix allows for either to be negative.
5023 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5024 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5027 PH/36 The change PH/12 above was broken. Fixed it.
5029 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5030 the grounds that routing the same address twice would always produce the
5031 same answer. This might have been true once, but it is certainly no
5032 longer true now. Routing a child address may depend on the previous
5033 routing that produced that child. Some complicated redirection strategies
5034 went wrong when messages had multiple recipients, and made Exim's
5035 behaviour dependent on the order in which the addresses were given.
5037 I have moved the duplicate checking until after the routing is complete.
5038 Exim scans the addresses that are assigned to local and remote
5039 transports, and removes any duplicates. This means that more work will be
5040 done, as duplicates will always all be routed, but duplicates are
5041 presumably rare, so I don't expect this is of any significance.
5043 For deliveries to pipes, files, and autoreplies, the duplicate checking
5044 still happens during the routing process, since they are not going to be
5047 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5048 It corrects a timeout issue with spamd. This is Ian's comment: "The
5049 background is that sometimes spamd either never reads data from a
5050 connection it has accepted, or it never writes response data. The exiscan
5051 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5052 blindly assumes that writes won't block so it may never time out."
5054 PH/39 Allow G after quota size as well as K and M.
5056 PH/40 The value set for $authenticated_id in an authenticator may not contain
5057 binary zeroes or newlines because the value is written to log lines and
5058 to spool files. There was no check on this. Now the value is run through
5059 the string_printing() function so that such characters are converted to
5060 printable escape sequences.
5062 PH/41 $message_linecount is a new variable that contains the total number of
5063 lines in the message. Compare $body_linecount, which is the count for the
5066 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5067 bounce and delay warning messages, except in certain special cases, which
5070 (a) An SMTP error message from a remote host;
5071 (b) A message specified in a :fail: redirection;
5072 (c) A message specified in a "fail" command in a system filter;
5073 (d) A message specified in a FAIL return from the queryprogram router;
5074 (e) A message specified by the cannot_route_message router option.
5076 In these cases only, Exim does include the error details in bounce and
5077 warning messages. There are also a few cases where bland messages such
5078 as "unrouteable address" or "local delivery error" are given.
5080 PH/43 $value is now also set for the "else" part of a ${run expansion.
5082 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5083 being worked on, but at least Exim now implements the latest version to
5086 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5087 process to complete was treated as a delivery failure, a timeout while
5088 writing the message to the pipe was logged, but erroneously treated as a
5089 successful delivery. Such timeouts include transport filter timeouts. For
5090 consistency with the overall process timeout, these timeouts are now
5091 treated as errors, giving rise to delivery failures by default. However,
5092 there is now a new Boolean option for the pipe transport called
5093 timeout_defer, which, if set TRUE, converts the failures into defers for
5094 both kinds of timeout. A transport filter timeout is now identified in
5097 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5098 systems where "make" and "gmake" are different, calling "gmake" at top
5099 level broke things. I've arranged for the value of $(MAKE) to be passed
5100 from the Makefile to this script so that it can call the same version of
5104 A note about Exim versions 4.44 and 4.50
5105 ----------------------------------------
5107 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5108 changes of various kinds. As a consequence, a big documentation update was
5109 needed. This delayed the release for rather longer than seemed good, especially
5110 in the light of a couple of (minor) security issues. Therefore, the changes
5111 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5112 release. So 4.44 and 4.50 are in effect two different branches that both start
5115 I have left the 4.50 change log unchanged; it contains all the changes since
5116 4.43. The change log for 4.44 is below; many of its items are identical to
5117 those for 4.50. This seems to be the most sensible way to preserve the
5118 historical information.
5124 1. Minor wording change to the doc/README.SIEVE file.
5126 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5127 computation of the current number of files was incorrect.
5129 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5130 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5133 4. Give more explanation in the error message when the command for a transport
5134 filter fails to execute.
5136 5. There are several places where Exim runs a non-Exim command in a
5137 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5138 was being done only for the command run by the queryprogram router. It is
5139 now done for all such subprocesses. The other cases are: ${run, transport
5140 filters, and the commands run by the lmtp and pipe transports.
5142 6. Added CONFIGURE_GROUP build-time option.
5144 7. Some older OS have a limit of 256 on the maximum number of file
5145 descriptors. Exim was using setrlimit() to set 1000 as a large value
5146 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5147 systems. I've change it so that if it can't get 1000, it tries for 256.
5149 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5150 was an oversight, and furthermore, ever since the addition of extra
5151 controls (e.g. 4.43/32), the checks on when to allow different forms of
5152 "control" were broken. There should now be diagnostics for all cases when a
5153 control that does not make sense is encountered.
5155 9. Added the /retain_sender option to "control=submission".
5157 10. $recipients is now available in the predata ACL (oversight).
5159 11. Tidy the search cache before the fork to do a delivery from a message
5160 received from the command line. Otherwise the child will trigger a lookup
5161 failure and thereby defer the delivery if it tries to use (for example) a
5162 cached ldap connection that the parent has called unbind on.
5164 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5165 of $address_data from the recipient verification was clobbered by the
5166 sender verification.
5168 13. The value of address_data from a sender verification is now available in
5169 $sender_address_data in subsequent conditions in the ACL statement.
5171 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5173 15. Added a new option "connect=<time>" to callout options, to set a different
5176 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5177 was its contents. (It was OK if the option was not defined at all.)
5179 17. A "Completed" log line is now written for messages that are removed from
5180 the spool by the -Mrm option.
5182 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5183 information about exactly what failed.
5185 19. Added -dd to debug only the daemon process.
5187 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5188 handles timeouts, both on the server side and network timeouts. Renamed the
5189 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5191 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5192 It is now set to "smtps".
5194 22. $host_address is now set to the target address during the checking of
5195 ignore_target_hosts.
5197 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5198 being passed; this would have caused $sender_host_name to have been used if
5199 matching the list had actually called for a host name (not very likely,
5200 since this list is usually IP addresses). A host name is now passed as
5203 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5204 code that specifies a non-restarting handler (typically sigaction() in
5205 modern systems) in an attempt to fix a rare and obscure crash bug.
5207 25. Narrowed the window for a race in the daemon that could cause it to ignore
5208 SIGCHLD signals. This is not a major problem, because they are used only to
5209 wake it up if nothing else does.
5211 26. A malformed maildirsize file could cause Exim to calculate negative values
5212 for the mailbox size or file count. Odd effects could occur as a result.
5213 The maildirsize information is now recalculated if the size or filecount
5216 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5217 support for a long time. Removed HAVE_SYS_VFS_H.
5219 28. Installed the latest version of exipick from John Jetmore.
5221 29. In an address list, if the pattern was not a regular expression, an empty
5222 subject address (from a bounce message) matched only if the pattern was an
5223 empty string. Non-empty patterns were not even tested. This was the wrong
5224 because it is perfectly reasonable to use an empty address as part of a
5225 database query. An empty address is now tested by patterns that are
5226 lookups. However, all the other forms of pattern expect the subject to
5227 contain a local part and a domain, and therefore, for them, an empty
5228 address still always fails if the pattern is not itself empty.
5230 30. Exim went into a mad DNS loop when attempting to do a callout where the
5231 host was specified on an smtp transport, and looking it up yielded more
5232 than one IP address.
5234 31. Re-factored the code for checking spool and log partition space into a
5235 function that finds that data and another that does the check. The former
5236 is then used to implement four new variables: $spool_space, $log_space,
5237 $spool_inodes, and $log_inodes.
5239 32. The RFC2047 encoding function was originally intended for short strings
5240 such as real names; it was not keeping to the 75-character limit for
5241 encoded words that the RFC imposes. It now respects the limit, and
5242 generates multiple encoded words if necessary. To be on the safe side, I
5243 have increased the buffer size for the ${rfc2047: expansion operator from
5246 33. It is now permitted to omit both strings after an "if" condition; if the
5247 condition is true, the result is "true". As before, when the second string
5248 is omitted, a false condition yields an empty string. This makes it less
5249 cumbersome to write custom ACL and router conditions.
5251 34. Failure to deliver a bounce message always caused it to be frozen, even if
5252 there was an errors_to setting on the router. The errors_to setting is now
5255 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5256 canonical form (fully expanded) before being placed in
5257 $sender_host_address.
5259 36. The table in the code that translates DNS record types into text (T_A to
5260 "A" for instance) was missing entries for NS and CNAME. It is just possible
5261 that this could have caused confusion if both these types were looked up
5262 for the same domain, because the text type is used as part of Exim's
5263 per-process caching. But the chance of anyone hitting this buglet seems
5266 37. The dnsdb lookup has been extended in a number of ways.
5268 (1) There is a new type, "zns", which walks up the domain tree until it
5269 finds some nameserver records. It should be used with care.
5271 (2) There is a new type, "mxh", which is like "mx" except that it returns
5272 just the host names, not the priorities.
5274 (3) It is now possible to give a list of domains (or IP addresses) to be
5275 looked up. The behaviour when one of the lookups defers can be
5276 controlled by a keyword.
5278 (4) It is now possible to specify the separator character for use when
5279 multiple records are returned.
5281 38. The dnslists ACL condition has been extended: it is now possible to supply
5282 a list of IP addresses and/or domains to be looked up in a particular DNS
5285 39. Added log_selector=+queue_time_overall.
5287 40. When running the queue in the test harness, wait just a tad after forking a
5288 delivery process, to get repeatability of debugging output.
5290 41. Include certificate and key file names in error message when GnuTLS fails
5291 to set them up, because the GnuTLS error message doesn't include the name
5292 of the failing file when there is a problem reading it.
5294 42. Allow both -bf and -bF in the same test run.
5296 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5298 44. The "Exiscan patch" is now merged into the mainline Exim source.
5300 45. Sometimes the final signoff response after QUIT could fail to get
5301 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5302 < 0 before doing a fflush(). This bug looks as though it goes back to the
5303 introduction of TLS in release 3.20, but "sometimes" must have been rare
5304 because the tests only now provoked it.
5306 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5307 (this can affect the format of dates).
5309 47. exim_tidydb, when checking for the continued existence of a message for
5310 which it has found a message-specific retry record, was not finding
5311 messages that were in split spool directories. Consequently, it was
5312 deleting retry records that should have stayed in existence.
5314 48. Steve fixed some bugs in eximstats.
5316 49. The SPA authentication driver was not abandoning authentication and moving
5317 on to the next authenticator when an expansion was forced to fail,
5318 contradicting the general specification for all authenticators. Instead it
5319 was generating a temporary error. It now behaves as specified.
5321 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5322 (the order specifies the preference for clients). The order is now AES256,
5323 AES128, 3DES, ARCFOUR128.
5325 51. Small patch to Sieve code - explicitly set From: when generating an
5328 52. Exim crashed if a remote delivery caused a very long error message to be
5329 recorded - for instance if somebody sent an entire SpamAssassin report back
5330 as a large number of 550 error lines. This bug was coincidentally fixed by
5331 increasing the size of one of Exim's internal buffers (big_buffer) that
5332 happened as part of the Exiscan merge. However, to be on the safe side, I
5333 have made the code more robust (and fixed the comments that describe what
5336 53. Now that there can be additional text after "Completed" in log lines (if
5337 the queue_time_overall log selector is set), a one-byte patch to exigrep
5338 was needed to allow it to recognize "Completed" as not the last thing in
5341 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5342 patch that reportedly fixes this has been added. I am not expert enough to
5343 create a test for it. This is what the patch creator wrote:
5345 "I found a little strange behaviour of ldap code when working with
5346 Windows 2003 AD Domain, where users was placed in more than one
5347 Organization Units. When I tried to give exim partial DN, the exit code
5348 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5349 But simultaneously result of request was absolutely normal ldap result,
5350 so I produce this patch..."
5352 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5353 REFERENCE, so I have modified the code to exclude the patch when that macro
5356 55. Some experimental protocols are using DNS PTR records for new purposes. The
5357 keys for these records are domain names, not reversed IP addresses. The
5358 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5359 leaves it alone. Component reversal etc. now happens only for IP addresses.
5362 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5364 57. Double the size of the debug message buffer (to 2048) so that more of very
5365 long debug lines gets shown.
5367 58. The exicyclog utility now does better if the number of log files to keep
5368 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5370 59. Two changes related to the smtp_active_hostname option:
5372 (1) $smtp_active_hostname is now available as a variable.
5373 (2) The default for smtp_banner uses $smtp_active_hostname instead
5374 of $primary_hostname.
5376 60. The host_aton() function is supposed to be passed a string that is known
5377 to be a valid IP address. However, in the case of IPv6 addresses, it was
5378 not checking this. This is a hostage to fortune. Exim now panics and dies
5379 if the condition is not met. A case was found where this could be provoked
5380 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5381 components; fortuitously, this particular loophole had already been fixed
5382 by change 4.50/55 above.
5384 If there are any other similar loopholes, the new check in host_aton()
5385 itself should stop them being exploited. The report I received stated that
5386 data on the command line could provoke the exploit when Exim was running as
5387 exim, but did not say which command line option was involved. All I could
5388 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5389 running as the user.
5392 61. There was a buffer overflow vulnerability in the SPA authentication code
5393 (which came originally from the Samba project). I have added a test to the
5394 spa_base64_to_bits() function which I hope fixes it.
5397 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5398 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5400 63. The daemon start-up calls getloadavg() while still root for those OS that
5401 need the first call to be done as root, but it missed one case: when
5402 deliver_queue_load_max is set with deliver_drop_privilege. This is
5403 necessary for the benefit of the queue runner, because there is no re-exec
5404 when deliver_drop_privilege is set.
5406 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5407 This has been fixed.
5409 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5410 was in use, was not putting the data itself into the right store pool;
5411 consequently, it could be overwritten for a subsequent message in the same
5412 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5415 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5417 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5418 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5419 the function were treating the return as a boolean value, which happened to
5420 work because 0=false and not-0=true, but is not correct code.
5422 68. The host_aton() function was not handling scoped IPv6 addresses (those
5423 with, for example, "%eth0" on the end) correctly.
5425 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5426 negated items (that is, ~something) in unsigned ints. Some compilers
5427 apparently mutter when there is no cast.
5429 70. If an address verification called from an ACL failed, and did not produce a
5430 user-specific message (i.e. there was only a "system" message), nothing was
5431 put in $acl_verify_message. In this situation, it now puts the system
5434 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5435 session; change 4.31/43 added the unwanted input to the log line - except
5436 that it did not do this in the start of session case. It now does.
5438 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5439 This is wrong and can cause the other end to generate a synchronization
5440 error if it is another Exim or anything else that does the synchronization
5441 check. A QUIT command is no longer sent after a timeout.
5443 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5444 during host lookups.
5446 74. The defer_ok option of callout verification was not working if it was used
5447 when verifying addresses in header lines, that is, for this case:
5449 verify = header_sender/callout=defer_ok
5451 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5452 those file descriptors could be used for SMTP connections. If anything
5453 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5454 could be sent to the SMTP client, causing chaos. The daemon now opens
5455 stdin, stdout, and stderr to /dev/null when it puts itself into the
5458 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5459 log by default. The user can override this with suitable Perl magic.
5461 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5462 the log message when discard triggers, was not working for the DATA ACL or
5463 for the non-SMTP ACL.
5465 78. Error message wording change in sieve.c.
5467 79. If smtp_accept_max_per_host was set, the number of connections could be
5468 restricted to fewer than expected, because the daemon was trying to set up
5469 a new connection before checking whether the processes handling previous
5470 connections had finished. The check for completed processes is now done
5471 earlier. On busy systems, this bug wouldn't be noticed because something
5472 else would have woken the daemon, and it would have reaped the completed
5475 80. If a message was submitted locally by a user whose login name contained one
5476 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5477 It caused a spool format error. I have fixed the spool reading code. A
5478 related problem was that the "from" clause in the Received: line became
5479 illegal because of the space(s). It is now covered by ${quote_local_part.
5481 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5484 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5486 83. Patch from Sieve maintainer to fix unterminated string problem in
5487 "vacation" handling.
5489 84. Some minor changes to the Linux configuration files to help with other
5490 OS variants using glibc.
5492 85. One more patch for Sieve to update vacation handling to latest spec.
5495 ----------------------------------------------------
5496 See the note above about the 4.44 and 4.50 releases.
5497 ----------------------------------------------------
5503 1. Change 4.43/35 introduced a bug that caused file counts to be
5504 incorrectly computed when quota_filecount was set in an appendfile
5507 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5508 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5511 3. Give more explanation in the error message when the command for a transport
5512 filter fails to execute.
5514 4. There are several places where Exim runs a non-Exim command in a
5515 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5516 was being done only for the command run by the queryprogram router. It is
5517 now done for all such subprocesses. The other cases are: ${run, transport
5518 filters, and the commands run by the lmtp and pipe transports.
5520 5. Some older OS have a limit of 256 on the maximum number of file
5521 descriptors. Exim was using setrlimit() to set 1000 as a large value
5522 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5523 systems. I've change it so that if it can't get 1000, it tries for 256.
5525 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5526 was an oversight, and furthermore, ever since the addition of extra
5527 controls (e.g. 4.43/32), the checks on when to allow different forms of
5528 "control" were broken. There should now be diagnostics for all cases when a
5529 control that does not make sense is encountered.
5531 7. $recipients is now available in the predata ACL (oversight).
5533 8. Tidy the search cache before the fork to do a delivery from a message
5534 received from the command line. Otherwise the child will trigger a lookup
5535 failure and thereby defer the delivery if it tries to use (for example) a
5536 cached ldap connection that the parent has called unbind on.
5538 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5539 of $address_data from the recipient verification was clobbered by the
5540 sender verification.
5542 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5543 was its contents. (It was OK if the option was not defined at all.)
5545 11. A "Completed" log line is now written for messages that are removed from
5546 the spool by the -Mrm option.
5548 12. $host_address is now set to the target address during the checking of
5549 ignore_target_hosts.
5551 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5552 being passed; this would have caused $sender_host_name to have been used if
5553 matching the list had actually called for a host name (not very likely,
5554 since this list is usually IP addresses). A host name is now passed as
5557 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5558 code that specifies a non-restarting handler (typically sigaction() in
5559 modern systems) in an attempt to fix a rare and obscure crash bug.
5561 15. Narrowed the window for a race in the daemon that could cause it to ignore
5562 SIGCHLD signals. This is not a major problem, because they are used only to
5563 wake it up if nothing else does.
5565 16. A malformed maildirsize file could cause Exim to calculate negative values
5566 for the mailbox size or file count. Odd effects could occur as a result.
5567 The maildirsize information is now recalculated if the size or filecount
5570 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5571 support for a long time. Removed HAVE_SYS_VFS_H.
5573 18. Updated exipick to current release from John Jetmore.
5575 19. Allow an empty sender to be matched against a lookup in an address list.
5576 Previously the only cases considered were a regular expression, or an
5579 20. Exim went into a mad DNS lookup loop when doing a callout where the
5580 host was specified on the transport, if the DNS lookup yielded more than
5583 21. The RFC2047 encoding function was originally intended for short strings
5584 such as real names; it was not keeping to the 75-character limit for
5585 encoded words that the RFC imposes. It now respects the limit, and
5586 generates multiple encoded words if necessary. To be on the safe side, I
5587 have increased the buffer size for the ${rfc2047: expansion operator from
5590 22. Failure to deliver a bounce message always caused it to be frozen, even if
5591 there was an errors_to setting on the router. The errors_to setting is now
5594 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5595 canonical form (fully expanded) before being placed in
5596 $sender_host_address.
5598 24. Updated eximstats to version 1.33
5600 25. Include certificate and key file names in error message when GnuTLS fails
5601 to set them up, because the GnuTLS error message doesn't include the name
5602 of the failing file when there is a problem reading it.
5604 26. Expand error message when OpenSSL has problems setting up cert/key files.
5607 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5608 (this can affect the format of dates).
5610 28. exim_tidydb, when checking for the continued existence of a message for
5611 which it has found a message-specific retry record, was not finding
5612 messages that were in split spool directories. Consequently, it was
5613 deleting retry records that should have stayed in existence.
5615 29. eximstats updated to version 1.35
5616 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5617 1.35 - bugfix such that pie charts by volume are generated correctly
5619 30. The SPA authentication driver was not abandoning authentication and moving
5620 on to the next authenticator when an expansion was forced to fail,
5621 contradicting the general specification for all authenticators. Instead it
5622 was generating a temporary error. It now behaves as specified.
5624 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5625 (the order specifies the preference for clients). The order is now AES256,
5626 AES128, 3DES, ARCFOUR128.
5628 31. Small patch to Sieve code - explicitly set From: when generating an
5631 32. Exim crashed if a remote delivery caused a very long error message to be
5632 recorded - for instance if somebody sent an entire SpamAssassin report back
5633 as a large number of 550 error lines. This bug was coincidentally fixed by
5634 increasing the size of one of Exim's internal buffers (big_buffer) that
5635 happened as part of the Exiscan merge. However, to be on the safe side, I
5636 have made the code more robust (and fixed the comments that describe what
5639 33. Some experimental protocols are using DNS PTR records for new purposes. The
5640 keys for these records are domain names, not reversed IP addresses. The
5641 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5642 leaves it alone. Component reversal etc. now happens only for IP addresses.
5645 34. The host_aton() function is supposed to be passed a string that is known
5646 to be a valid IP address. However, in the case of IPv6 addresses, it was
5647 not checking this. This is a hostage to fortune. Exim now panics and dies
5648 if the condition is not met. A case was found where this could be provoked
5649 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5650 components; fortuitously, this particular loophole had already been fixed
5651 by change 4.50/55 or 4.44/33 above.
5653 If there are any other similar loopholes, the new check in host_aton()
5654 itself should stop them being exploited. The report I received stated that
5655 data on the command line could provoke the exploit when Exim was running as
5656 exim, but did not say which command line option was involved. All I could
5657 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5658 running as the user.
5661 35. There was a buffer overflow vulnerability in the SPA authentication code
5662 (which came originally from the Samba project). I have added a test to the
5663 spa_base64_to_bits() function which I hope fixes it.
5666 36. The daemon start-up calls getloadavg() while still root for those OS that
5667 need the first call to be done as root, but it missed one case: when
5668 deliver_queue_load_max is set with deliver_drop_privilege. This is
5669 necessary for the benefit of the queue runner, because there is no re-exec
5670 when deliver_drop_privilege is set.
5672 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5673 was in use, was not putting the data itself into the right store pool;
5674 consequently, it could be overwritten for a subsequent message in the same
5675 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5678 38. Sometimes the final signoff response after QUIT could fail to get
5679 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5680 < 0 before doing a fflush(). This bug looks as though it goes back to the
5681 introduction of TLS in release 3.20, but "sometimes" must have been rare
5682 because the tests only now provoked it.
5688 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5689 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5690 Now it returns an integer. A number of calls were still expecting a T/F
5691 return. Fortuitously, in all cases, the tests worked in OK situations,
5692 which is the norm. However, things would have gone wrong on any write
5693 failures on the smtp file descriptor. This function is used when sending
5694 messages over SMTP and also when doing verify callouts.
5696 2. When Exim is called to do synchronous delivery of a locally submitted
5697 message (the -odf or -odi options), it no longer closes stderr before doing
5700 3. Implemented the mua_wrapper option.
5702 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5704 5. Implemented the functions header_remove(), header_testname(),
5705 header_add_at_position(), and receive_remove_recipient(), and exported them
5708 6. If an ACL "warn" statement specified the addition of headers, Exim already
5709 inserted X-ACL-Warn: at the start if there was no header name. However, it
5710 was not making this test for the second and subsequent header lines if
5711 there were newlines in the string. This meant that an invalid header could
5712 be inserted if Exim was badly configured.
5714 7. Allow an ACL "warn" statement to add header lines at the start or after all
5715 the Received: headers, as well as at the end.
5717 8. Added the rcpt_4xx retry error code.
5719 9. Added postmaster_mailfrom=xxx to callout verification option.
5721 10. Added mailfrom=xxxx to the callout verification option, for verify=
5724 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5725 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5727 12. Inserted some casts to stop certain compilers warning when using pointer
5728 differences as field lengths or precisions in printf-type calls (mostly
5729 affecting debugging statements).
5731 13. Added optional readline() support for -be (dynamically loaded).
5733 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5734 same clock tick as a message's arrival, so that its received time was the
5735 same as the "first fail" time on the retry record, and that message
5736 remained on the queue past the ultimate address timeout, every queue runner
5737 would try a delivery (because it was past the ultimate address timeout) but
5738 after another failure, the ultimate address timeout, which should have then
5739 bounced the address, did not kick in. This was a "< instead of <=" error;
5740 in most cases the first failure would have been in the next clock tick
5741 after the received time, and all would be well.
5743 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5744 being recognized when the domain list was tested by the match_domain
5745 condition in an expansion string.
5747 16. Added the ${str2b64: operator.
5749 17. Exim was always calling setrlimit() to set a large limit for the number of
5750 processes, without checking whether the existing limit was already
5751 adequate. (It did check for the limit on file descriptors.) Furthermore,
5752 errors from getrlimit() and setrlimit() were being ignored. Now they are
5753 logged to the main and panic logs, but Exim does carry on, to try to do its
5754 job under whatever limits there are.
5756 18. Imported PCRE 5.0.
5758 19. Trivial typo in log message " temporarily refused connection" (the leading
5761 20. If the log selector return_path_on_delivery was set and an address was
5762 redirected to /dev/null, the delivery process crashed because it assumed
5763 that a return path would always be set for a "successful" delivery. In this
5764 case, the whole delivery is bypassed as an optimization, and therefore no
5767 21. Internal re-arrangement: the function for sending a challenge and reading
5768 a response while authentication was assuming a zero-terminated challenge
5769 string. It's now changed to take a pointer and a length, to allow for
5770 binary data in such strings.
5772 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5774 23. Exim was not respecting finduser_retries when seeking the login of the
5775 uid under which it was called; it was always trying 10 times. (The default
5776 setting of finduser_retries is zero.) Also, it was sleeping after the final
5777 failure, which is pointless.
5779 24. Implemented tls_on_connect_ports.
5781 25. Implemented acl_smtp_predata.
5783 26. If the domain in control=submission is set empty, Exim assumes that the
5784 authenticated id is a complete email address when it generates From: or
5785 Sender: header lines.
5787 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5788 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5789 chown and chgrp in /bin and hostname in /usr/bin.
5791 28. Exim was keeping the "process log" file open after each use, just as it
5792 does for the main log. This opens the possibility of it remaining open for
5793 long periods when the USR1 signal hits a daemon. Occasional processlog
5794 errors were reported, that could have been caused by this. Anyway, it seems
5795 much more sensible not to leave this file open at all, so that is what now
5798 29. The long-running daemon process does not normally write to the log once it
5799 has entered its main loop, and it closes the log before doing so. This is
5800 so that log files can straightforwardly be renamed and moved. However,
5801 there are a couple of unusual error situations where the daemon does write
5802 log entries, and I had neglected to close the log afterwards.
5804 30. The text of an SMTP error response that was received during a remote
5805 delivery was being truncated at 512 bytes. This is too short for some of
5806 the long messages that one sometimes sees. I've increased the limit to
5809 31. It is now possible to make retry rules that apply only when a message has a
5810 specific sender, in particular, an empty sender.
5812 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5813 it possible to be selective about when SMTP synchronization is enforced.
5815 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5817 32. Implemented hosts_connection_nolog.
5819 33. Added an ACL for QUIT.
5821 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5824 35. Added mailbox_size and mailbox_filecount to appendfile.
5826 36. Added control = no_multiline_responses to ACLs.
5828 37. There was a bug in the logic of the code that waits for the clock to tick
5829 in the case where the clock went backwards by a substantial amount such
5830 that the microsecond fraction of "now" was more than the microsecond
5831 fraction of "then" (but the whole seconds number was less).
5833 38. Added support for the libradius Radius client library this is found on
5834 FreeBSD (previously only the radiusclient library was supported).
5840 1. When certain lookups returned multiple values in the form name=value, the
5841 quoting of the values was not always being done properly. Specifically:
5842 (a) If the value started with a double quote, but contained no whitespace,
5844 (b) If the value contained whitespace other than a space character (i.e.
5845 tabs or newlines or carriage returns) it was not quoted.
5846 This fix has been applied to the mysql and pgsql lookups by writing a
5847 separate quoting function and calling it from the lookup code. The fix
5848 should probably also be applied to nisplus, ibase and oracle lookups, but
5849 since I cannot test any of those, I have not disturbed their existing code.
5851 2. A hit in the callout cache for a specific address caused a log line with no
5852 reason for rejecting RCPT. Now it says "Previous (cached) callout
5853 verification failure".
5855 3. There was an off-by-one bug in the queryprogram router. An over-long
5856 return line was truncated at 256 instead of 255 characters, thereby
5857 overflowing its buffer with the terminating zero. As well as fixing this, I
5858 have increased the buffer size to 1024 (and made a note to document this).
5860 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5861 when Exim is waiting for an SMTP response from a remote server, Exim
5862 restarts its select() call on the socket, thereby resetting its timeout.
5863 This is not a problem when such interrupts are rare. Somebody set up a cron
5864 job to run exiwhat every 2 minutes, which is less than the normal select()
5865 timeout (5 or 10 minutes). This meant that the select() timeout never
5866 kicked in because it was always reset. I have fixed this by comparing the
5867 time when an interrupt arrives with the time at the start of the first call
5868 to select(). If more time than the timeout has elapsed, the interrupt is
5869 treated as a timeout.
5871 5. Some internal re-factoring in preparation for the addition of Sieve
5872 extensions (by MH). In particular, the "personal" test is moved to a
5873 separate function, and given an option for scanning Cc: and Bcc: (which is
5874 not set for Exim filters).
5876 6. When Exim created an email address using the login of the caller as the
5877 local part (e.g. when creating a From: or Sender: header line), it was not
5878 quoting the local part when it contained special characters such as @.
5880 7. Installed new OpenBSD configuration files.
5882 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5883 try to make them clearer.
5885 9. Callout options, other than the timeout value, were being ignored when
5886 verifying sender addresses in header lines. For example, when using
5888 verify = header_sender/callout=no_cache
5890 the cache was (incorrectly) being used.
5892 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5893 only the Cygwin environment.
5895 11. When return_path_on_delivery was set as a log selector, if different remote
5896 addresses in the same message used different return paths and parallel
5897 remote delivery occurred, the wrong values would sometimes be logged.
5898 (Whenever a remote delivery process finished, the return path value from
5899 the most recently started remote delivery process was logged.)
5901 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5902 lines when AUTH and/or TLS are in use. This is the "received protocol"
5903 field. Exim used to use "asmtp" for authenticated SMTP, without any
5904 indication (in the protocol name) for TLS use. Now it follows the RFC and
5905 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5906 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5907 names appear in log lines as well as in Received: header lines.
5909 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5910 extensions, and comparison tests, and to fix some bugs.
5912 14. Changes to the "personal" filter test:
5914 (1) The test was buggy in that it was just doing the equivalent of
5915 "contains" tests on header lines. For example, if a user's address was
5916 anne@some.where, the "personal" test would incorrectly be true for
5918 To: susanne@some.where
5920 This test is now done by extracting each address from the header in turn,
5921 and checking the entire address. Other tests that are part of "personal"
5922 are now done using regular expressions (for example, to check local parts
5923 of addresses in From: header lines).
5925 (2) The list of non-personal local parts in From: addresses has been
5926 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5927 taken from the Sieve specification recommendations.
5929 (3) If the message contains any header line starting with "List-" it is
5930 treated as non-personal.
5932 (4) The test for "circular" in the Subject: header line has been removed
5933 because it now seems ill-conceived.
5935 15. Minor typos in src/EDITME comments corrected.
5937 16. Installed latest exipick from John Jetmore.
5939 17. If headers_add on a router specified a text string that was too long for
5940 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5941 of string_sprintf() is now avoided.
5943 18. $message_body_size was not set (it was always zero) when running the DATA
5944 ACL and the local_scan() function.
5946 19. For the "mail" command in an Exim filter, no default was being set for
5947 the once_repeat time, causing a random time value to be used if "once" was
5948 specified. (If the value happened to be <= 0, no repeat happened.) The
5949 default is now 0s, meaning "never repeat". The "vacation" command was OK
5950 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5951 (I found it when inspecting the code).
5953 20. There is now an overall timeout for performing a callout verification. It
5954 defaults to 4 times the callout timeout, which applies to individual SMTP
5955 commands during the callout. The overall timeout applies when there is more
5956 than one host that can be tried. The timeout is checked before trying the
5957 next host. This prevents very long delays if there are a large number of
5958 hosts and all are timing out (e.g. when the network connections are timing
5959 out). The value of the overall timeout can be changed by specifying an
5960 additional sub-option for "callout", called "maxwait". For example:
5962 verify = sender/callout=5s,maxwait=20s
5964 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5965 to the end before writing, but this should make it even safer).
5967 22. Exim was forgetting that it had advertised PIPELINING for the second and
5968 subsequent messages on an SMTP connection. It was also not resetting its
5969 memory on STARTTLS and an internal HELO.
5971 23. When Exim logs an SMTP synchronization error within a session, it now
5972 records whether PIPELINING has been advertised or not.
5974 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5975 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5976 rather than long int.
5978 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5980 26. Added the never_mail option to autoreply.
5986 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5987 crash if the getsockname() call failed; this can happen if a connection is
5988 closed very soon after it is established. The problem was simply in the
5989 order in which certain operations were done, causing Exim to try to write
5990 to the SMTP stream before it had set up the file descriptor. The bug has
5991 been fixed by making things happen in the correct order.
5997 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5998 before the connection was closed, thus losing the rejection response.
6000 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6001 some early Solaris releases, but causes trouble in current releases where
6002 socklen_t is defined.
6004 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6007 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6010 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6011 that releasing the top of it at the end releases what was used for sub-
6012 expansions (unless the block got too big). However, discard this block if
6013 the first thing is a variable or header, so that we can use its block when
6014 it is dynamic (useful for very large $message_headers, for example).
6016 6. Lookups now cache *every* query, not just the most recent. A new, separate
6017 store pool is used for this. It can be recovered when all lookup caches are
6018 flushed. Lookups now release memory at the end of their result strings.
6019 This has involved some general refactoring of the lookup sources.
6021 7. Some code has been added to the store_xxx() functions to reduce the amount
6022 of flapping under certain conditions.
6024 8. log_incoming_interface used to affect only the <= reception log lines. Now
6025 it causes the local interface and port to be added to several more SMTP log
6026 lines, for example "SMTP connection from", and rejection lines.
6028 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6030 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6032 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6033 could be overwritten at the end of the current message (or the start of a
6034 new message if it was set in a HELO ACL). The value is now preserved for
6035 the duration of the SMTP connection.
6037 12. If a transport had a headers_rewrite setting, and a matching header line
6038 contained an unqualified address, that address was qualified, even if it
6039 did not match any rewriting rules. The underlying bug was that the values
6040 of the flags that permit the existence of unqualified sender and recipient
6041 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6042 non-local messages, and by -bnq for local messages) were not being
6043 preserved with the message after it was received.
6045 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6046 "next input=" as part of the text comprising the host identity instead of
6047 the correct text. The code was using the same buffer for two different
6048 strings. However, depending on which order the printing function evaluated
6049 its arguments, the bug did not always show up. Under Linux, for example, my
6050 test suite worked just fine.
6052 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6053 doesn't work with some older versions of Perl. It has been changed to "my",
6054 which in any case is probably the better facility to use.
6056 15. A really picky compiler found some instances of statements for creating
6057 error messages that either had too many or two few arguments for the format
6060 16. The size of the buffer for calls to the DNS resolver has been increased
6061 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6062 for addresses that have a lot of PTR records. This alleviates a problem; it
6063 does not fully solve it.
6065 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6066 buffer now truncates the list and logs the incident, which is the same
6067 action as happens when Exim is looking up a host name and its aliases.
6068 Previously in this situation something unpredictable would happen;
6069 sometimes it was "internal error: store_reset failed".
6071 18. If a server dropped the connection unexpectedly when an Exim client was
6072 using GnuTLS and trying to read a response, the client delivery process
6073 crashed while trying to generate an error log message.
6075 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6076 string, for example:
6078 warn message = H1: something\nH2: something
6080 the text was added as a single header line from Exim's point of view
6081 though it ended up OK in the delivered message. However, searching for the
6082 second and subsequent header lines using $h_h2: did not work. This has been
6083 fixed. Similarly, if a system filter added multiple headers in this way,
6084 the routers could not see them.
6086 20. Expanded the error message when iplsearch is called with an invalid key to
6087 suggest using net-iplsearch in a host list.
6089 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6090 ACLs are no longer actually imposed (and a message to that effect is
6093 22. If a "gecos" field in a passwd entry contained escaped characters, in
6094 particular, if it contained a \" sequence, Exim got it wrong when building
6095 a From: or a Sender: header from that name. A second bug also caused
6096 incorrect handling when an unquoted " was present following a character
6097 that needed quoting.
6099 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6100 was not being matched caselessly.
6102 24. Arranged for all hyphens in the exim.8 source to be escaped with
6105 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6106 properly. Recipient callout cache records were still being keyed to include
6107 the sender, even when use_sender was set false. This led to far more
6108 callouts that were necessary. The sender is no longer included in the key
6109 when use_sender is false.
6111 26. Added "control = submission" modifier to ACLs.
6113 27. Added the ${base62d: operator to decode base 62 numbers.
6115 28. dnsdb lookups can now access SRV records.
6117 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6118 the configuration file.
6120 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6121 (-v) mode. This makes the output for a verbose queue run more intelligible.
6123 31. Added a use_postmaster feature to recipient callouts.
6125 32. Added the $body_zerocount variable, containing the number of binary zero
6126 bytes in the message body.
6128 33. The time of last modification of the "new" subdirectory is now used as the
6129 "mailbox time last read" when there is a quota error for a maildir
6132 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6134 35. Added +ignore_unknown as a special item in host lists.
6136 36. Code for decoding IPv6 addresses in host lists is now included, even if
6137 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6138 address was recognized as an IP address, but was then not correctly decoded
6139 into binary, causing unexpected and incorrect effects when compared with
6146 1. Very minor rewording of debugging text in manualroute to say "list of
6147 hosts" instead of "hostlist".
6149 2. If verify=header_syntax was set, and a header line with an unqualified
6150 address (no domain) and a large number of spaces between the end of the
6151 name and the colon was received, the reception process suffered a buffer
6152 overflow, and (when I tested it) crashed. This was caused by some obsolete
6153 code that should have been removed. The fix is to remove it!
6155 3. When running in the test harness, delay a bit after writing a bounce
6156 message to get a bit more predictability in the log output.
6158 4. Added a call to search_tidyup() just before forking a reception process. In
6159 theory, someone could use a lookup in the expansion of smtp_accept_max_
6160 per_host which, without the tidyup, could leave open a database connection.
6162 5. Added the variables $recipient_data and $sender_data which get set from a
6163 lookup success in an ACL "recipients" or "senders" condition, or a router
6164 "senders" option, similar to $domain_data and $local_part_data.
6166 6. Moved the writing of debug_print from before to after the "senders" test
6169 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6170 problems for message scanning, either using a data ACL, or using
6171 local_scan() because the Received: header was not generated till after they
6172 were called (in order to set the time as the time of reception completion).
6173 I have revised the way this works. The header is now generated after the
6174 body is received, but before the ACL or local_scan() are called. After they
6175 are run, the timestamp in the header is updated.
6181 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6182 before starting a queue runner without re-exec. This happened only when
6183 deliver_drop_privilege was set or when the Exim user was set to root. The
6184 effect of the bug was that timeouts during subsequent deliveries caused
6185 crashes instead of being properly handled. The handler is now left at its
6186 default (and expected) setting.
6188 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6189 message, again when deliver_drop_privilege is set or Exim is run as root.
6190 The bug described in (1) was not present in this case, but the tidying up
6191 of the other signals was missing. I have made the two cases consistent.
6193 3. The ignore_target_hosts setting on a manualroute router was being ignored
6194 for hosts that were looked up using the /MX notation.
6196 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6199 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6200 operated on the sender address. After changing the $sender_address to <>
6201 for the sender address verify, Exim was re-instated it as the original
6202 (before rewriting) address, but remembering that it had rewritten it, so it
6203 wasn't rewriting it again. This bug also had the effect of breaking the
6204 sender address verification caching when the sender address was rewritten.
6206 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6207 This has been changed so that if the ip literal address matches
6208 ignore_target_hosts, the router declines.
6210 7. Added expansion conditions match_domain, match_address, and match_local_
6211 part (NOT match_host).
6213 8. The placeholder for the Received: header didn't have a length field set.
6215 9. Added code to Exim itself and to exim_lock to test for a specific race
6216 condition that could lead to file corruption when using MBX delivery. The
6217 issue is with the lockfile that is created in /tmp. If this file is removed
6218 after a process has opened it but before that process has acquired a lock,
6219 there is the potential for a second process to recreate the file and also
6220 acquire a lock. This could lead to two Exim processes writing to the file
6221 at the same time. The added code performs the same test as UW imapd; it
6222 checks after acquiring the lock that its file descriptor still refers to
6223 the same named file.
6225 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6226 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6229 11. Added the smtp_active_hostname option. If used, this will typically be made
6230 to depend on the incoming interface address. Because $interface_address is
6231 not set up until the daemon has forked a reception process, error responses
6232 that can happen earlier (such as "too many connections") no longer contain
6235 12. If an expansion in a condition on a "warn" statement fails because a lookup
6236 defers, the "warn" statement is abandoned, and the next ACL statement is
6237 processed. Previously this caused the whole ACL to be aborted.
6239 13. Added the iplsearch lookup type.
6241 14. Added ident_timeout as a log selector.
6243 15. Added tls_certificate_verified as a log selector.
6245 16. Added a global option tls_require_ciphers (compare the smtp transport
6246 option of the same name). This controls incoming TLS connections.
6248 17. I finally figured out how to make tls_require_ciphers do a similar thing
6249 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6250 before starting the TLS session.
6252 18. Tabs are now shown as \t in -bP output.
6254 19. If the log selector return_path_on_delivery was set, Exim crashed when
6255 bouncing a message because it had too many Received: header lines.
6257 20. If two routers both had headers_remove settings, and the first one included
6258 a superfluous trailing colon, the final name in the first list and the
6259 first name in the second list were incorrectly joined into one item (with a
6260 colon in the middle).
6266 1. Added -C and -D options to the exinext utility, mainly to make it easier
6267 to include in the automated testing, but these could be helpful when
6268 multiple configurations are in use.
6270 2. The exinext utility was not formatting the output nicely when there was
6271 an alternate port involved in the retry record key, nor when there was a
6272 message id as well (for retries that were specific to a specific message
6273 and a specific host). It was also confused by IPv6 addresses, because of
6274 the additional colons they contain. I have fixed the IPv4 problem, and
6275 patched it up to do a reasonable job for IPv6.
6277 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6278 delivery, the log line now contains "pipelined" if PIPELINING was used.
6280 4. An SMTP transport process used to panic and die if the bind() call to set
6281 an explicit outgoing interface failed. This has been changed; it is now
6282 treated in the same way as a connect() failure.
6284 5. A reference to $sender_host_name in the part of a conditional expansion
6285 that was being skipped was still causing a DNS lookup. This no longer
6288 6. The def: expansion condition was not recognizing references to header lines
6289 that used bh_ and bheader_.
6291 7. Added the _cache feature to named lists.
6293 8. The code for checking quota_filecount in the appendfile transport was
6294 allowing one more file than it should have been.
6296 9. For compatibility with Sendmail, the command line option
6304 and sets the incoming protocol and host name (for trusted callers). The
6305 host name and its colon can be omitted when only the protocol is to be set.
6306 Note the Exim already has two private options, -pd and -ps, that refer to
6307 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6308 "s", but I don't think that's a major issue.
6310 10. A number of refactoring changes to the code, none of which should affect
6313 (a) The number of logging options was getting close to filling up the
6314 32-bit word that was used as a bit map. I have split them into two classes:
6315 those that are passed in the argument to log_write(), and those that are
6316 only ever tested independently outside of that function. These are now in
6317 separate 32-bit words, so there is plenty of room for expansion again.
6318 There is no change in the user interface or the logging behaviour.
6320 (b) When building, for example, log lines, the code previously used a
6321 macro that called string_cat() twice, in order to add two strings. This is
6322 not really sufficiently general. Furthermore, there was one instance where
6323 it was actually wrong because one of the argument was used twice, and in
6324 one call a function was used. (As it happened, calling the function twice
6325 did not affect the overall behaviour.) The macro has been replaced by a
6326 function that can join an arbitrary number of extra strings onto a growing
6329 (c) The code for expansion conditions now uses a table and a binary chop
6330 instead of a serial search (which was left over from when there were very
6331 few conditions). Also, it now recognizes conditions like "pam" even when
6332 the relevant support is not compiled in: a suitably worded error message is
6333 given if an attempt is made to use such a condition.
6335 11. Added ${time_interval:xxxxx}.
6337 12. A bug was causing one of the ddress fields not to be passed back correctly
6338 from remote delivery subprocesses. The field in question was not being
6339 subsequently used, so this caused to problems in practice.
6341 13. Added new log selectors queue_time and deliver_time.
6343 14. Might have fixed a bug in maildirsizefile handling that threw up
6344 "unexpected character" debug warnings, and recalculated the data
6345 unnecessarily. In any case, I expanded the warning message to give more
6348 15. Added the message "Restricted characters in address" to the statements in
6349 the default ACL that block characters like @ and % in local parts.
6351 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6352 Three changes have been made:
6354 (a) There was a serious bug; a negative response to MAIL caused the whole
6355 recipient domain to be cached as invalid, thereby blocking all messages
6356 to all local parts at the same domain, from all senders. This bug has
6357 been fixed. The domain is no longer cached after a negative response to
6358 MAIL if the sender used is not empty.
6360 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6363 (c) A new callout option, "use_sender" has been added for people who want
6364 the modified behaviour.
6370 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6373 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6374 indeed breaks things for older releases.
6376 3. Added additional logging to the case where there is a problem reading data
6377 from a filter that is running in a subprocess using a pipe, in order to
6378 try to track down a specific problem.
6380 4. Testing facility fudge: when running in the test harness and attempting
6381 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6382 getting "No route to host". Convert this to a timeout.
6384 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6387 6. Some OS don't have socklen_t but use size_t instead. This affects the
6388 fifth argument of getsockopt() amongst other things. This is now
6389 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6390 can be set for individual OS. I have set it for SunOS5, OSF1, and
6391 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6392 some earlier ones do not.
6394 7. Change 4.30/15 was not doing the test caselessly.
6396 8. The standard form for an IPv6 address literal was being rejected by address
6397 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6398 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6399 this, as well as the form without the "IPv6" on the front (but only when
6400 address literals are enabled, of course).
6402 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6404 10. Exim crashed if a message with an empty sender address specified by -f
6405 encountered a router with an errors_to setting. This could be provoked only
6406 by a command such as
6410 where an empty string was supplied; "<>" did not hit this bug.
6412 11. Installed PCRE release 4.5.
6414 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6415 remained set. It is now erased.
6417 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6418 times from message ids (which are base 36 rather than the normal 62).
6420 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6421 were being counted as actual protocol errors, and logged if the log
6422 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6423 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6424 and DATA following a set of rejected RCPTs do not count as protocol errors.
6425 In other words, Exim assumes they were pipelined, though this may not
6426 actually be the case. Of course, in all cases the client gets an
6427 appropriate error code.
6429 15. If a lookup fails in an ACL condition, a message about the failure may
6430 be available; it is used if testing the ACL cannot continue, because most
6431 such messages specify what the cause of the deferral is. However, some
6432 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6433 that caused an old message to be retained and used if a later statement
6434 caused a defer, replacing the real cause of the deferral.
6436 16. If an IP address had so many PTR records that the DNS lookup buffer
6437 was not large enough to hold them, Exim could crash while trying to process
6438 the truncated data. It now detects and logs this case.
6440 17. Further to 4.21/58, another change has been made: if (and only if) the
6441 first line of a message (the first header line) ends with CRLF, a bare LF
6442 in a subsequent header line has a space inserted after it, so as not to
6443 terminate the header.
6445 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6446 unnecessarily, used atoi() instead of strtol(), and didn't check the
6447 termination when getting file sizes from file names by regex.
6449 19. Completely re-implemented the support for maildirsize files, in the light
6450 of a number of problems with the previous contributed implementation
6451 (4.30/29). In particular:
6453 . If the quota is zero, the maildirsize file is maintained, but no quota is
6456 . If the maildir directory does not exist, it is created before any attempt
6457 to write a maildirsize file.
6459 . The quota value in the file is just a cache; if the quota is changed in
6460 the transport, the new value overrides.
6462 . A regular expression is available for excluding directories from the
6465 20. The autoreply transport checks the characters in options that define the
6466 message's headers; it allows continued headers, but it was checking with
6467 isspace() after an embedded newline instead of explicitly looking for a
6470 21. If all the "regular" hosts to which an address was routed had passed their
6471 expiry times, and had not reached their retry times, the address was
6472 bounced, even if fallback hosts were defined. Now Exim should go on to try
6475 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6476 equivalent code in the SMTP transport. Some hosts send humungous responses
6477 to HELO/EHLO, more than 1024 it seems.
6479 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6480 gives compiler warnings in some environments. I've now done it "properly",
6483 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6484 (because of problems with the built-in one) was declared to return uschar *
6485 instead of char *, causing compiler failure.
6487 25. Fixed a file descriptor leak when processing alias/forward files.
6489 26. Fixed a minor format string issue in dbfn.c.
6491 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6493 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6494 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6497 29. When a sender address is verified, it is cached, to save repeating the test
6498 when there is more than one recipient in a message. However, when the
6499 verification involves a callout, it is possible for different callout
6500 options to be set for different recipients. It is too complicated to keep
6501 track of this in the cache, so now Exim always runs a verification when a
6502 callout is required, relying on the callout cache for the optimization.
6503 The overhead is duplication of the address routing, but this should not be
6506 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6507 to be verified with callout=postmaster, and the main callout worked but the
6508 postmaster check failed, the verification correctly failed. However, if a
6509 subsequent RCPT command asked for sender verification *without* the
6510 postmaster check, incorrect caching caused this verification also to fail,
6513 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6514 it was not caching the DNS options (qualify_single, search_parents) that
6515 were used when the lookup failed. A subsequent lookup with different
6516 options therefore always gave the same answer, though there were cases
6517 where it should not have. (Example: a "domains = !$mx_any" option on a
6518 dnslookup router: the "domains" option is always processed without any
6519 widening, but the router might have qualify_single set.) Now Exim uses the
6520 cached value only when the same options are set.
6522 32. Added John Jetmore's "exipick" utility to the distribution.
6524 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6525 than a timeout (e.g. a certificate is required, and is not provided), an
6526 Exim server now closes the connection immediately. Previously it waited for
6527 the client to close - but if the client is SSL, it seems that they each
6528 wait for each other, leading to a delay before one of them times out.
6530 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6531 maintained 0.8.x compatibility because I don't think many are using it, and
6532 it is clearly obsolete.
6534 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6537 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6538 client certificate was expired. A simple patch fixes this, though I don't
6539 understand the full logic of why the verify callback is called multiple
6542 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6543 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6544 which causes problems with some clients (such as the Certicom SSL Plus
6545 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6546 disables the coutermeasure allowing Eudora to connect."
6548 38. Exim was not checking that a write() to a log file succeeded. This could
6549 lead to Bad Things if a log got too big, in particular if it hit a file
6550 size limit. Exim now panics and dies if it cannot write to a log file, just
6551 as it does if it cannot open a log file.
6553 39. Modified OS/Makefile-Linux so that it now contains
6555 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6557 The two -D definitions ensure that Exim is compiled with large file
6558 support, which makes it possible to handle log files that are bigger than
6561 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6562 instance) a domain was checked against a named list that involved a lookup,
6563 causing $domain_data to be set, then another domain was checked against the
6564 same list, then the first domain was re-checked, the value of $domain_data
6565 after the final check could be wrong. In particular, if the second check
6566 failed, it could be set empty. This bug probably also applied to
6569 41. The strip_trailing_dot option was not being applied to the address given
6570 with the -f command-line option.
6572 42. The code for reading a message's header from the spool was incrementing
6573 $received_count, but never initializing it. This meant that the value was
6574 incorrect (doubled) while delivering a message in the same process in which
6575 it was received. In the most common configuration of Exim, this never
6576 happens - a fresh exec is done - but it can happen when
6577 deliver_drop_privilege is set.
6579 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6580 it now includes up to 150 characters of the unexpected data in the log
6583 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6584 and building data strings. The size of both of these buffers was 10 000
6585 bytes - far larger than anybody would *ever* want, thought I. Needless to
6586 say, somebody hit the limit. I have increased the maximum line length to
6587 20 000 and the maximum data length of concatenated lines to 100 000. I have
6588 also fixed two bugs, because there was no checking on these buffers. Tsk,
6589 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6590 buffer is too small.
6592 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6593 lsearch lookups. Now it does.
6595 46. When parsing a route_list item in a manualroute router, a fixed-length
6596 buffer was used for the list of hosts. I made this 1024 bytes long,
6597 thinking that nobody would ever have a list of hosts that long. Wrong.
6598 Somebody had a whole pile of complicated expansion conditions, and the
6599 string was silently truncated, leading to an expansion error. It turns out
6600 that it is easier to change to an unlimited length (owing to other changes
6601 that have happened since this code was originally written) than to build
6602 structure for giving a limitation error. The length of the item that
6603 expands into the list of hosts is now unlimited.
6605 47. The lsearch lookup could not handle data where the length of text line was
6606 more than 4095 characters. Such lines were truncated, leading to shortened
6607 data being returned. It should now handle lines of any length.
6609 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6610 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6613 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6614 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6615 no longer happens in comment lines. A list of replacements is now placed
6616 at the head of all of the source files, except those whose only change is
6617 to replace PERL_COMMAND in the very first #! line.
6619 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6620 messages on the queue, with a bottom-up merge sort, using code contributed
6621 by Michael Haardt. This should make operations like -bp somewhat faster on
6622 large queues. It won't affect queue runners, except when queue_run_in_order
6625 51. Installed eximstats 1.31 in the distribution.
6627 52. Added support for SRV lookups to the dnslookup router.
6629 53. If an ACL referred to $message_body or $message_body_end, the value was not
6630 reset for any messages that followed in the same SMTP session.
6632 54. The store-handling optimization for building very long strings was not
6633 differentiating between the different store pools. I don't think this
6634 actually made any difference in practice, but I've tidied it.
6636 55. While running the routers to verify a sender address, $sender_address
6637 was still set to the sender address. This is wrong, because when routing to
6638 send a bounce to the sender, it would be empty. Therefore, I have changed
6639 it so that, while verifying a sender address, $sender_address is set to <>.
6640 (There is no change to what happens when verifying a recipient address.)
6642 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6643 target A or AAAA records (if not already returned) without resetting the
6644 qualify_single or search_parents options of the DNS resolver. These are
6645 inappropriate in this case because the targets of MX and SRV records must
6646 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6647 target that, when qualified, matched something in the local domain. These
6648 two options are now turned off when doing these lookups.
6650 57. It seems that at least some releases of Reiserfs (which does not have the
6651 concept of a fixed number of inodes) returns zero and not -1 for the
6652 number of available inodes. This interacted badly with check_spool_inodes,
6653 which assumed that -1 was the "no such thing" setting. What I have done is
6654 to check that the total number of inodes is greater than zero before doing
6655 the test of how many are available.
6657 58. When a "warn" ACL statement has a log_message modifier, the message is
6658 remembered, and not repeated. This is to avoid a lot of repetition when a
6659 message has many recipients that cause the same warning to be written.
6660 However, Exim was preserving the list of already written lines for an
6661 entire SMTP session, which doesn't seem right. The memory is now reset if a
6662 new message is started.
6664 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6665 debugging output unless log_rewrite was also set.
6667 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6668 of (int)(handle) when we know that handle contains (void *)(-1).
6670 61. The Exim daemon panic-logs an error return when it closes the incoming
6671 connection. However "connection reset by peer" seems to be common, and
6672 isn't really an error worthy of noting specially, so that particular error
6675 62. When Exim is trying to find all the local interfaces, it used to panic and
6676 die if the ioctl to get the interface flags failed. However, it seems that
6677 on at least one OS (Solaris 9) it is possible to have an interface that is
6678 included in the list of interfaces, but for which you get a failure error
6679 for this call. This happens when the interface is not "plumbed" into a
6680 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6681 failure of the "get flags" call assumes that the interface is down.
6683 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6684 makes life easier for people who are doing arithmetic on fields extracted
6685 from dates, where you often get leading zeros that should not be
6686 interpreted as octal.
6688 64. Added qualify_domain to the redirect router, to override the global
6691 65. If a pathologically long header line contained very many addresses (the
6692 report of this problem mentioned 10 000) and each of them was rewritten,
6693 Exim could use up a very large amount of memory. (It kept on making new
6694 copies of the header line as it rewrote, and never released the old ones.)
6695 At the expense of a bit more processing, the header rewriting function has
6696 been changed so that it no longer eats memory in this way.
6698 66. The generation of the Received: header has been moved from the time that a
6699 message starts to be received, to the time that it finishes. The timestamp
6700 in the Received: header should now be very close to that of the <= log
6701 line. There are two side-effects of this change:
6703 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6704 logged header lines no longer include the local Received: line, because
6705 it has not yet been created. The same applies to a copy of the message
6706 that is returned to a non-SMTP sender when a message is rejected.
6708 (b) When a filter file is tested using -bf, no additional Received: header
6709 is added to the test message. After some thought, I decided that this
6712 This change does not affect the value of $received_for. It is still set
6713 after address rewriting, but before local_scan() is called.
6715 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6717 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6718 gave an unhelpful panic error message, and a defer error. I have managed to
6719 change this behaviour so that it now rejects any supplied certificate,
6720 which seems right, as the list of acceptable certificates is empty.
6722 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6723 gave an unhelpful defer error. I have not managed to make this reject any
6724 supplied certificates, but the error message it gives is "no certificate
6725 supplied", which is not helpful.
6727 70. exigrep's output now also includes lines that are not associated with any
6728 message, but which match the given pattern. Implemented by a patch from
6729 Martin Sluka, which also tidied up the Perl a bit.
6731 71. Recipient callout verification, like sender verification, was using <> in
6732 the MAIL FROM command. This isn't really the right thing, since the actual
6733 sender may affect whether the remote host accepts the recipient or not. I
6734 have changed it to use the actual sender in the callout; this means that
6735 the cache record is now keyed on a recipient/sender pair, not just the
6736 recipient address. There doesn't seem to be a real danger of callout loops,
6737 since a callout by the remote host to check the sender would use <>.
6738 [SEE ABOVE: changed after hitting problems.]
6740 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6741 temporary errors. However, in the case of such a code being given after
6742 the end of a data transmission (i.e. after ".") Exim was failing to write
6743 a retry record for the message. (Yes, there was some broken host that was
6744 actually sending 8xx at this point.)
6746 73. An unknown lookup type in a host list could cause Exim to panic-die when
6747 the list was checked. (An example that provoked this was putting <; in the
6748 middle of a list instead of at the start.) If this happened during a DATA
6749 ACL check, a -D file could be left lying around. This kind of configuration
6750 error no longer causes Exim to die; instead it causes a defer error. The
6751 incident is still logged to the main and panic logs.
6753 74. Buglet left over from Exim 3 conversion. The message "too many messages
6754 in one connection" was written to the rejectlog but not the mainlog, except
6755 when address rewriting (yes!) was being logged.
6757 75. Added write_rejectlog option.
6759 76. When a system filter was run not as root (that is, when system_filter_user
6760 was set), the values of the $n variables were not being returned to the
6761 main process; thus, they were not subsequently available in the $sn
6764 77. Added +return_path_on_delivery log selector.
6766 78. A connection timeout was being treated differently from recipients deferred
6767 when testing hosts_max_try with a message that was older than the host's
6768 retry timeout. (The host should not be counted, thus allowing all hosts to
6769 be tried at least once before bouncing.) This may have been the cause of an
6770 occasionally reported bug whereby a message would remain on the queue
6771 longer than the retry timeout, but would be bounced if a delivery was
6772 forced. I say "may" because I never totally pinned down the problem;
6773 setting up timeout/retry tests is difficult. See also the next item.
6775 79. The ultimate address timeout was not being applied to errors that involved
6776 a combination of host plus message (for example, a timeout on a MAIL
6777 command). When an address resolved to a number of possible hosts, and they
6778 were not all tried for each delivery (e.g. because of hosts_max_try), a
6779 message could remain on the queue longer than the retry timeout.
6781 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6784 81. Fixed an obscure SMTP outgoing bug which required at least the following
6785 conditions: (a) there was another message waiting for the same server;
6786 (b) the server returned 5xx to all RCPT commands in the first message so
6787 that the message was not completed; (c) the server dropped the connection
6788 or gave a negative response to the RSET that Exim sends to abort the
6789 transaction. The observed case was a dropped connection after DATA that had
6790 been sent in pipelining mode. That is, the server had advertised PIPELINING
6791 but was not implementing it correctly. The effect of the bug was incorrect
6792 behaviour, such as trying another host, and this could lead to a crash.
6798 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6799 and daemon.c were passed as pointers to ints; they should have been
6800 pointers to socklen_t variables (which are typically unsigned ints).
6802 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6805 3. Fixed a really odd bug that affected only the testing scheme; patching a
6806 certain fixed string in the binary changed the value of another string that
6807 happened to be identical to the end of the original first string.
6809 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6810 name", it returns that address as the IP address. On some operating
6811 systems (e.g. Solaris), it also passes back the IP address string as the
6812 "host name". However, on others (e.g. Linux), it passes back an empty
6813 string. Exim wasn't checking for this, and was changing the host name to an
6814 empty string, assuming it had been canonicalized.
6816 5. Although rare, it is permitted to have more than one PTR record for a given
6817 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6818 all the names associated with an address, because they do in Solaris.
6819 However, it seems that they do not in Linux for data that comes from the
6820 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6821 I found this out when I moved to a new Linux workstation and tried to run
6822 the Exim test suite.
6824 To get round this problem I have changed the code so that it now does its
6825 own call to the DNS to look up PTR records when searching for a host name.
6826 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6827 addresses that are only in /etc/hosts are still found.
6829 This behaviour is, however, controlled by an option called host_lookup_
6830 order, which defaults to "bydns:byaddr". If people want to use the other
6831 order, or indeed, just use one or the other means of lookup, they can
6832 specify it in this variable.
6834 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6835 some operating systems, this comes back from gethostbyaddr() as an empty
6836 string, and this is what Exim used to test for. However, it seems that in
6837 other systems, "." is yielded. Exim now tests for this case too.
6839 7. The values of check_spool_space and check_log_space are now held internally
6840 as a number of kilobytes instead of an absolute number of bytes. If a
6841 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6842 kilobyte. This means that much larger values can be stored.
6844 8. Exim monitor: an attempt to get the action menu when not actually pointing
6845 at a message produces an empty menu entitled "No message selected". This
6846 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6847 no entries in it ("Shell widget menu has zero width and/or height"). So I
6848 have added a single, blank menu entry in this case.
6850 9. Added ${quote_local_part.
6852 10. MIME decoding is now applied to the contents of Subject: header lines when
6855 11. Now that a reference to $sender_host_address automatically causes a reverse
6856 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6857 host lookup before query-style lookups in lists that might use this
6858 variable. This has therefore been abolished, and the "net-" prefix is no
6859 longer necessary for query-style lookups.
6861 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6862 have been a typo for LFLAGS, so it has been changed.
6864 13. The install script calls Exim with "-C /dev/null" in order to find the
6865 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6866 to be output. However, since Exim outputs its version number before the
6867 error, it didn't break the script. It just looked ugly. I fixed this by
6868 always allowing "-C /dev/null" if the caller is root.
6870 14. Ignore overlarge ACL variable number when reading spool file - insurance
6871 against a later release with more variables having written the file.
6873 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6874 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6875 this, as well as the form without the "IPv6" on the front.
6877 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6878 OS/Makefile-Darwin file.
6880 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6881 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6883 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6884 variable. However, some OS use ints for inodes. I've added cast to long int
6885 to get rid of the compiler warning.
6887 19. I had forgotten to lock out "/../" in configuration file names when
6888 ALT_CONFIG_PREFIX was set.
6890 20. Routers used for verification do not need to specify transports. However,
6891 if such a router generated a host list, and callout was configured, Exim
6892 crashed, because it could not find a port number from the (non-existent)
6893 transport. It now assumes port 25 in this circumstance.
6895 21. Added the -t option to exigrep.
6897 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6898 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6899 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6900 them is set, LOOKUP_LSEARCH is forced.
6902 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6904 24. Added sender and host information to the "rejected by local_scan()" log
6905 line; previously there was no indication of these.
6907 25. Added .include_if_exists.
6909 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6910 Linux. It turns out that not all file systems support this. Apparently some
6911 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6912 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6913 not supported on the file descriptor, is now ignored when Exim is trying to
6914 sync a directory. This applies only to Linux.
6916 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6918 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6919 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6922 29. Added support for maildirsize files from supplied patch (modified a bit).
6924 30. The use of :fail: followed by an empty string could lead Exim to respond to
6925 sender verification failures with (e.g.):
6927 550 Verification failed for <xxx>
6928 550 Sender verify failed
6930 where the first response line was missing the '-' that indicates it is not
6931 the final line of the response.
6933 31. The loop for finding the name of the user that called Exim had a hardwired
6934 limit of 10; it now uses the value of finduser_retries, which is used for
6935 all other user lookups.
6937 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6940 33. Exim was neglecting to zero errno before one call of strtol() when
6941 expanding a string and expecting an integer value. On some systems this
6942 resulted in spurious "integer overflow" errors. Also, it was casting the
6943 result into an int without checking.
6945 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6946 did not work. The code looks as if it has *never* worked, though it appears
6947 to have been documented since at least release 1.62. I have made it work.
6949 35. The "timeout_DNS" error in retry rules, also documented since at least
6950 1.62, also never worked. As it isn't clear exactly what this means, and
6951 clearly it isn't a major issue, I have abolished the feature by treating it
6952 as "timeout", and writing a warning to the main and panic logs.
6954 36. The display of retry rules for -brt wasn't always showing the error code
6957 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6958 timeout_connect_A, timeout_connect_MX.
6960 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6961 to the empty sender.
6963 39. The daemon was not analysing the content of -oX till after it had closed
6964 stderr and disconnected from the controlling terminal. This meant that any
6965 syntax errors were only noted on the panic log, and the return code from
6966 the command was 0. By re-arranging the code a little, I've made the
6967 decoding happen first, so such errors now appear on stderr, and the return
6968 code is 1. However, the actual setting up of the sockets still happens in
6969 the disconnected process, so errors there are still only recorded on the
6972 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6973 connections (as happens on some IP stacks) was logged at start up time as
6974 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6975 it from "IPv6 and IPv4", which means that two separate sockets are being
6978 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6979 says whether AF_INET or AF_INET6 was passed as an argument.
6981 42. Exiwhat output was messed up when time zones were included in log
6984 43. Exiwhat now gives more information about the daemon's listening ports,
6985 and whether -tls-on-connect was used.
6987 44. The "port" option of the smtp transport is now expanded.
6989 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6990 silently ignored. Now an error message is written to the main and panic
6993 46. There's a new ACL modifier called "logwrite" which writes to a log file
6994 as soon as it is encountered.
6996 47. Added $local_user_uid and $local_user_gid at routing time.
6998 48. Exim crashed when trying to verify a sender address that was being
7001 49. Exim was recognizing only a space character after ".include". It now also
7002 recognizes a tab character.
7004 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7005 extracting the relevant information from the specification. The man page no
7006 longer contains scrambled data for the -d option, and I've added a section
7007 at the front about calling Exim under different names.
7009 51. Added "extra_headers" argument to the "mail" command in filter files.
7011 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7014 53. Installed eximstats 1.29.
7016 54. Added transport_filter_timeout as a generic transport option.
7018 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7019 Cc: header lines. This was required by RFC 822, but it not required by RFC
7022 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7023 incoming messages that did not have them. Now it does so only if the
7024 message originates locally, that is, if there is no associated remote host
7025 address. When Resent- header lines are present, this applies to the Resent-
7026 lines rather than the non-Resent- lines.
7028 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7029 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7031 58. Messages for configuration errors now include the name of the main
7032 configuration file - useful now that there may be more than one file in a
7033 list (.included file names were always shown).
7035 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7036 for those rare installations that do not start the daemon as root or run it
7037 setuid root. I've cut out the call to initgroups() if the daemon is not
7040 60. The Exim user and group can now be bound into the binary as text strings
7041 that are looked up at the start of Exim's processing.
7043 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7045 62. Added $mailstore_basename variable.
7047 63. Installed patch to sieve.c from Michael Haardt.
7049 64. When Exim failed to open the panic log after failing to open the main log,
7050 the original message it was trying to log was written to stderr and debug
7051 output, but if they were not available (the usual case in production), it
7052 was lost. Now it is written to syslog before the two lines that record the
7053 failures to open the logs.
7055 65. Users' Exim filters run in subprocesses under the user's uid. It is
7056 possible for a "deliver" command or an alias in a "personal" command to
7057 provoke an address rewrite. If logging of address rewriting is configured,
7058 this fails because the process is not running as root or exim. There may be
7059 a better way of dealing with this, but for the moment (because 4.30 needs
7060 to be released), I have disabled address rewrite logging when running a
7061 filter in a non-root, non-exim process.
7067 1. The buildconfig auxiliary program wasn't quoting the value set for
7068 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7069 not defined. This bug was masked in 4.22 by the effect that was fixed in
7072 2. Some messages that were rejected after a message id was allocated were
7073 shown as "incomplete" by exigrep. It no longer does this for messages that
7074 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7076 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7077 have allow_domain_literals set, the ID did not get logged in the <= line.
7078 Domain literals are now always recognized in Message-ID: header lines.
7080 4. The first argument for a ${extract expansion item is the key name or field
7081 number. Leading and trailing spaces in this item were not being ignored,
7082 causing some misleading effects.
7084 5. When deliver_drop_privilege was set, single queue runner processes started
7085 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7086 same command in the process it spins off) were not dropping privilege.
7088 6. When the daemon running as "exim" started a queue runner, it always
7089 re-executed Exim in the spun-off process. This is a waste of effort when
7090 deliver_drop_privilege is set. The new process now just calls the
7091 queue-runner function directly.
7097 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7100 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7101 ignored. Though the use of -oP was forcing the writing of a pid file, it
7102 was always written to the default place.
7104 3. If the message "no IP address found for host xxxx" is generated during
7105 incoming verification, it is now followed by identification of the incoming
7106 connection (so you can more easily find what provoked it).
7108 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7110 5. Added some features to "harden" Exim a bit more against certain attacks:
7112 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7113 be put in Local/Makefile. This is like the never_users runtime option,
7114 but it cannot be overridden. The default setting is "root".
7116 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7117 prefix string with which any file named in a -C command line option
7120 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7121 is retained for -C and -D only if the caller of Exim is root. Without
7122 it, the exim user may also use -C and -D and retain privilege.
7124 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7125 command line option is disabled.
7127 6. Macro names set by the -D option must start with an upper case letter, just
7128 like macro names defined in the configuration file.
7130 7. Added "dereference=" facility to LDAP.
7132 8. Two instances of the typo "uknown" in the source files are fixed.
7134 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7135 the Configure-Makefile script screwed up while processing it.
7137 10. Incorporated PCRE 4.4.
7139 11. The SMTP synchronization check was not operating right at the start of an
7140 SMTP session. For example, it could not catch a HELO sent before the client
7141 waited for the greeting. There is now a check for outstanding input at the
7142 point when the greeting is written. Because of the duplex, asynchronous
7143 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7144 way, but not yet received, when the check is performed.
7146 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7147 on TCP/IP sockets, because this apparently causes some broken clients to
7150 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7151 unchanged) from the Cygwin maintainer.
7153 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7154 was supported instead of testing for mbx. Effectively a typo.
7156 15. The spa authenticator server code was not checking that the input it
7157 received was valid base64.
7159 16. The debug output line for the "set" modifier in ACLs was not showing the
7160 name of the variable that was being set.
7162 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7164 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7165 lookup on its own. Something else was needed to trigger the lookup. For
7166 example, a match in host_lookup or the need for a host name in a host list.
7167 Now, if $sender_host_name is referenced and the host name has not yet been
7168 looked up, a lookup is performed. If the lookup fails, the variable remains
7169 empty, and $host_lookup_failed is set to "1".
7171 19. Added "eqi" as a case-independent comparison operator.
7173 20. The saslauthd authentication condition could segfault if neither service
7174 nor realm was specified.
7176 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7177 error message that was logged was misleading, and incoming SMTP
7178 connections were dropped. The message is now more accurate, and temporary
7179 errors are given to SMTP connections.
7181 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7182 (except RSET) with a 421 error, until QUIT is received. However, it was
7183 failing to send a response to QUIT.
7185 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7186 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7187 ACL is now run after the helo_try_verify_hosts code.
7189 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7190 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7191 case-independent, so other case variants are also recognized). Apparently
7192 some systems use these upper case variants.
7194 25. If more than two messages were waiting for the same host, and a transport
7195 filter was specified for the transport, Exim sent two messages over the
7196 same TCP/IP connection, and then failed with "socket operation on non-
7197 socket" when it tried to send the third.
7199 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7201 27. The extern definition of crypt16() in expand.c was not being excluded when
7202 the OS had its own crypt16() function.
7204 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7205 as a preferred synonym for return_size_limit, both as an option and as an
7208 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7210 30. Changed the default configuration ACL to relax the local part checking rule
7211 for addresses that are not in any local domains. For these addresses,
7212 slashes and pipe symbols are allowed within local parts, but the sequence
7213 /../ is explicitly forbidden.
7215 31. SPA server authentication was not clearing the challenge buffer before
7218 32. log_message in a "warn" ACL statement was writing to the reject log as
7219 well as to the main log, which contradicts the documentation and doesn't
7220 seem right (because no rejection is happening). So I have stopped it.
7222 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7223 However, I am unable to do any testing of this.
7225 34. Fixed an infelicity in the appendfile transport. When checking directories
7226 for a mailbox, to see if any needed to be created, it was accidentally
7227 using path names with one or more superfluous leading slashes; tracing
7228 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7230 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7231 contents are added to the log line that is written for every discarded
7232 recipient. (Previously a log_message setting was ignored.)
7234 36. The ${quote: operator now quotes the string if it is empty.
7236 37. The install script runs exim in order to find its version number. If for
7237 some reason other than non-existence or emptiness, which it checks, it
7238 could not run './exim', it was installing it with an empty version number,
7239 i.e. as "exim-". This error state is now caught, and the installation is
7242 38. An argument was missing from the function that creates an error message
7243 when Exim fails to connect to the socket for saslauthd authentication.
7244 This could cause Exim to crash, or give a corrupted message.
7246 39. Added isip, isip4, and isip6 to ${if conditions.
7248 40. The ACL variables $acl_xx are now saved with the message, and can be
7249 accessed later in routers, transports, and filters.
7251 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7252 strings in the file are not string-expanded.
7254 42. If a MAIL command specified a SIZE value that was too large to fit into an
7255 int variable, the check against message_size_limit failed. Such values are
7256 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7257 day this will have to be increased, but I don't think I want to be around
7258 when emails are that large.
7265 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7266 iconv() is not standard in FreeBSD.
7268 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7269 IPv6 enabled. The observed symptom was a segmentation fault on return from
7270 the function os_common_find_running_interfaces() in src/os.c.
7272 3. In the check_special_case() function in daemon.c I had used "errno" as an
7273 argument name, which causes warnings on some systems. This was basically a
7274 typo, since it was named "eno" in the comments!
7276 4. The code that waits for the clock to tick (at a resolution of some fraction
7277 of a second) so as to ensure message-id uniqueness was always waiting for
7278 at least one whole tick, when it could have waited for less. [This is
7279 almost certainly not relevant at current processor speeds, where it is
7280 unlikely to ever wait at all. But we try to future-proof.]
7282 5. The function that sleeps for a time interval that includes fractions of a
7283 second contained a race. It did not block SIGALRM between setting the
7284 timer, and suspending (a couple of lines later). If the interval was short
7285 and the sigsuspend() was delayed until after it had expired, the suspension
7286 never ended. On busy systems this could lead to processes getting stuck for
7289 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7290 process, before it forks any delivery processes. The open lookup caching
7291 mechanism meant that the open file or database connection was passed into
7292 the delivery process. The problem was that delivery processes always tidy
7293 up cached lookup data. This could cause a problem for the next delivery
7294 process started by the queue runner, because the external queue runner
7295 process does not know about the closure. So the next delivery process
7296 still has data in the lookup cache. In the case of a file lookup, there was
7297 no problem because closing a file descriptor in a subprocess doesn't affect
7298 the parent. However, if the lookup was caching a connection to a database,
7299 the connection was closed, and the second delivery process was likely to
7300 see errors such as "PGSQL: query failed: server closed the connection
7301 unexpectedly". The problem has been fixed by closing all cached lookups
7302 in a queue runner before running a delivery process.
7304 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7305 seem to have the "const" qualifier which it has on other OS. I've
7308 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7309 *of the same type* (client or server) with the same public name that an
7310 error should be diagnosed.
7312 9. When Exim looked up a host name for an IP address, but failed to find the
7313 original IP address when looking up the host name (a safety check), it
7314 output the message "<ip address> does not match any IP for NULL", which was
7315 confusing, to say the least. The bug was that the host name should have
7316 appeared instead of "NULL".
7318 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7319 user that is built into the binary, and the -C or -D options is used, root
7320 privilege is dropped before the configuration file is read. In addition,
7321 logging is switched to stderr instead of the normal log files. If the
7322 configuration then re-defines the Exim user, the unprivileged environment
7323 is probably not what is expected, so Exim logs a panic warning message (but
7326 However, if deliver_drop_privilege is set, the unprivileged state may well
7327 be exactly what is intended, so the warning has been cut out in that case,
7328 and Exim is allowed to try to write to its normal log files.
7334 1. smtp_return_error_details was not giving details for temporary sender
7335 or receiver verification errors.
7337 2. Diagnose a configuration error if two authenticators have the same public
7340 3. Exim used not to create the message log file for a message until the first
7341 delivery attempt. This could be confusing when incoming messages were held
7342 for policy or load reasons. The message log file is now created at the time
7343 the message is received, and an initial "Received" line is written to it.
7345 4. The automatically generated man page for command line options had a minor
7346 bug that caused no ill effects; however, a more serious problem was that
7347 the procedure for building the man page automatically didn't always
7348 operate. Consequently, release 4.20 contains an out-of-date version. This
7349 shouldn't happen again.
7351 5. When building Exim with embedded Perl support, the script that builds the
7352 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7353 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7355 6. The freeze_tell option was not being used for messages that were frozen on
7356 arrival, either by an ACL or by local_scan().
7358 7. Added the smtp_incomplete_transaction log selector.
7360 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7361 was accepting AUTH without a new EHLO.
7363 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7364 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7367 10. Logging of TCP/IP connections (when configured) now happens in the main
7368 daemon process instead of the child process, so that the TCP/IP connection
7369 count is more accurate (but it can never be perfect).
7371 11. The use of "drop" in a nested ACL was not being handled correctly in the
7372 outer ACL. Now, if condition failure induced by the nested "drop" causes
7373 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7374 or "require"), the connection is dropped.
7376 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7377 that yield "discard" can now be used with an "accept" or a "discard" verb,
7378 but an error is generated for any others (because I can't see a useful way
7379 to define what should happen).
7381 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7382 are now processed in the same way as lines in the Exim configuration file.
7383 In particular, continuation lines are supported.
7385 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7387 15. Added -ti meaning -t -i.
7389 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7390 domains, and warn by logging if others are found.
7392 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7393 returned from GIFCONF when Exim is trying to find the list of interfaces on
7394 a host. The code in os.c has been modified to copy each ifreq to an aligned
7395 structure in all cases.
7397 Also, in some cases, the returned ifreq's were being copied to a 'struct
7398 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7399 means the last couple of bytes of an IPv6 address could be chopped if the
7400 ifreq contained only a normal sockaddr (14 bytes storage).
7402 18. Named domain lists were not supported in the hosts_treat_as_local option.
7403 An entry such as +xxxx was not recognized, and was treated as a literal
7406 19. Ensure that header lines added by a DATA ACL are included in the reject log
7407 if the ACL subsequently rejects the message.
7409 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7410 MD5 (which is deprecated).
7412 21. When testing a filter file using -bf, Exim was writing a message when it
7413 took the sender from a "From " line in the message, but it was not doing so
7414 when it took $return_path from a Return-Path: header line. It now does.
7416 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7417 with a valid header line field name (a series of printing characters
7418 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7420 23. Changed "disc" in the source to "disk" to conform to the documentation and
7421 the book and for uniformity.
7423 24. Ignore Sendmail's -Ooption=value command line item.
7425 25. When execve() failed while trying to run a command in a pipe transport,
7426 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7427 could be confused with a return value of 69 from the command itself. This
7428 has been changed to 127, the value the shell returns if it is asked to run
7429 a non-existent command. The wording for the related log line suggests a
7430 non-existent command as the problem.
7432 26. If received_header_text expands to an empty string, do not add a Received:
7433 header line to the message. (Well, it adds a token one on the spool, but
7434 marks it "old" so that it doesn't get used or transmitted.)
7436 27. Installed eximstats 1.28 (addition of -nt option).
7438 28. There was no check for failure on the call to getsockname() in the daemon
7439 code. This can fail if there is a shortage of resources on the system, with
7440 ENOMEM, for example. A temporary error is now given on failure.
7442 29. Contrary to the C standard, it seems that in some environments, the
7443 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7444 program. Exim now does this explicitly; it affects the formatting of
7445 timestamps using strftime().
7447 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7448 complaints. I've now initialized all the variables, to avoid this.
7450 32. Header lines added by a system filter were not being "seen" during
7451 transport-time rewrites.
7453 33. The info_callback() function passed to OpenSSL is set up with type void
7454 (*)(SSL *, int, int), as described somewhere. However, when calling the
7455 function (actually a macro) that sets it up, the type void(*)() is
7456 expected. I've put in a cast to prevent warnings from picky compilers.
7458 34. If a DNS black list lookup found a CNAME record, but there were no A
7459 records associated with the domain it pointed at, Exim crashed.
7461 35. If a DNS black list lookup returned more than one A record, Exim ignored
7462 all but the first. It now scans all returned addresses if a particular IP
7463 value is being sought. In this situation, the contents of the
7464 $dnslist_value variable are a list of all the addresses, separated by a
7467 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7468 to accept a host name and all its aliases if the forward lookup for any of
7469 them yielded the IP address of the incoming connection. Now it accepts only
7470 those names whose forward lookup yields the correct IP address. Any other
7471 names are discarded. This closes a loophole whereby a rogue DNS
7472 administrator could create reverse DNS records to break through a
7473 wildcarded host restriction in an ACL.
7475 37. If a user filter or a system filter that ran in a subprocess used any of
7476 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7477 the wrong values were passed to the pipe command ($thisaddress had the
7478 value of $0, $0 had the value of $1, etc). This bug was introduced by
7479 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7481 38. Improved the line breaking for long SMTP error messages from ACLs.
7482 Previously, if there was no break point between 40 and 75 characters, Exim
7483 left the rest of the message alone. Two changes have been made: (a) I've
7484 reduced the minimum length to 35 characters; (b) if it can't find a break
7485 point between 35 and 75 characters, it looks ahead and uses the first one
7486 that it finds. This may give the occasional overlong line, but at least the
7487 remaining text gets split now.
7489 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7490 file descriptors might be low, and that setting 1000 would always raise it.
7491 It turns out that in some environments, the limit is already over 1000 and
7492 that lowering it causes trouble. So now Exim takes care not to decrease it.
7494 40. When delivering a message, the value of $return_path is set to $sender_
7495 address at the start of routing (routers may change the value). By an
7496 oversight, this default was not being set up when an address was tested by
7497 -bt or -bv, which affected the outcome if any router or filter referred to
7500 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7501 log only when "message" or "log_message" are set. However, if one of the
7502 conditions was an address verification, or a call to a nested ACL, the
7503 messages generated by the underlying test were being passed through. This
7504 no longer happens. The underlying message is available in $acl_verify_
7505 message for both "message" and "log_message" expansions, so it can be
7506 passed through if needed.
7508 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7509 new expansion $bh_ to give the encoded byte string without charset
7510 translation. Translation happens only if iconv() is available; HAVE_ICONV
7511 indicates this at build time. HEADERS_CHARSET gives the charset to
7512 translate to; headers_charset can change it in the configuration, and
7513 "headers charset" can change it in an individual filter file.
7515 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7516 that creates RFC 2047 encoded "words" labels them as that charset instead
7517 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7518 expansion operator; (ii) when Exim creates a From: line for a local
7519 message; (iii) when a header line is rewritten to include a "phrase" part.
7521 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7522 buggy, causing it to skip the first lines of messages whose message ID
7523 ended in 'D'. This would not have bitten before Exim release 4.14, because
7524 message IDs were unlikely to end in 'D' before then. The effect was to have
7525 incorrect size information for certain domains.
7527 45. #include "config.h" was missing at the start of the crypt16.c module. This
7528 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7531 46. If there was a timeout during a "random" callout check, Exim treated it as
7532 a failure of the random address, and carried on sending RSET and the real
7533 address. If the delay was just some slowness somewhere, the response to the
7534 original RCPT would be taken as a response to RSET and so on, causing
7535 mayhem of various kinds.
7537 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7538 when I implemented it. It didn't allow for the fact that some option values
7539 may legitimately be negative (e.g. size_addition), and it didn't even do
7540 the right test for positive values.
7542 48. Domain names in DNS records are case-independent. Exim always looks them up
7543 in lower case. Some resolvers return domain names in exactly the case they
7544 appear in the zone file, that is, they may contain uppercase letters. Not
7545 all resolvers do this - some return always lower case. Exim was treating a
7546 change of case by a resolver as a change of domain, similar to a widening
7547 of a domain abbreviation. This triggered its re-routing code and so it was
7548 trying to route what was effectively the same domain again. This normally
7549 caused routing to fail (because the router wouldn't handle the domain
7550 twice). Now Exim checks for this case specially, and just changes the
7551 casing of the domain that it ultimately uses when it transmits the message
7554 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7557 50. If a filter generated a file delivery with a non-absolute name (possible if
7558 no home directory exists for the router), the forbid_file option was not
7561 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7562 the existing equality matching.
7564 52. Exim was using ints instead of ino_t variables in some places where it was
7565 dealing with inode numbers.
7567 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7568 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7569 called TMPDIR, and if it finds it is different, it changes its value.
7571 54. The smtp_printf() function is now made available to local_scan() so
7572 additional output lines can be written before returning. There is also an
7573 smtp_fflush() function to enable the detection of a dropped connection.
7574 The variables smtp_input and smtp_batched_input are exported to
7577 55. Changed the default runtime configuration: the message "Unknown user"
7578 has been removed from the ACL, and instead placed on the localuser router,
7579 using the cannot_route_message feature. This means that any verification
7580 failures that generate their own messages won't get overridden. Similarly,
7581 the "Unrouteable address" message that was in the ACL for unverifiable
7582 relay addresses has also been removed.
7584 56. Added hosts_avoid_esmtp to the smtp transport.
7586 57. The exicyclog script was not checking for the esoteric option
7587 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7588 will work only if exicyclog is run under the appropriate euid.
7590 58. Following a discussion on the list, the rules by which Exim recognises line
7591 endings on incoming messages have been changed. The -dropcr and drop_cr
7592 options are now no-ops, retained only for backwards compatibility. The
7593 following line terminators are recognized: LF CRLF CR. However, special
7594 processing applies to CR:
7596 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7597 nor a local message in the state where . is a terminator.
7599 (ii) If a bare CR is encountered in a header line, an extra space is added
7600 after the line terminator so as not to end the header. The reasoning
7601 behind this is that bare CRs in header lines are most likely either
7602 to be mistakes, or people trying to play silly games.
7604 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7605 was being incorrectly given as 18 bytes larger than it should have been.
7606 This is a VOB (very old bug).
7608 60. This may never have affected anything current, but just in case it has:
7609 When the local host is found other than at the start of a list of hosts,
7610 the local host, those with the same MX, and any that follow, are discarded.
7611 When the list in question was part of a longer list of hosts, the following
7612 hosts (not currently being processed) were also being discarded. This no
7613 longer happens. I'm not sure if this situation could ever has previously
7616 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7619 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7620 header. This is something that is recommended in a new Internet Draft, and
7621 is something that is documented as being done by Sendmail. There are two
7622 possible values. For messages generated by the autoreply transport, Exim
7625 Auto-Submitted: auto-replied
7627 whereas for all other generated messages (e.g. bounces) it adds
7629 Auto-Submitted: auto-generated
7631 63. The "personal" condition in filters now includes a test for the
7632 Auto-Submitted: header. If it contains the string "auto-" the message it
7633 not considered personal.
7635 64. Added rcpt_include_affixes as a generic transport option.
7637 65. Added queue_only_override (default true).
7639 66. Added the syslog_duplication option.
7641 67. If what should have been the first header line of a message consisted of
7642 a space followed by a colon, Exim was mis-interpreting it as a header line.
7643 It isn't of course - it is syntactically invalid and should therefore be
7644 treated as the start of the message body. The misbehaviour could have
7645 caused a number of strange effects, including loss of data in subsequent
7646 header lines, and spool format errors.
7648 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7649 client host had authenticated. This control can now be exercised by an ACL
7650 for more flexibility.
7652 69. By default, callouts do not happen when testing with -bh. There is now a
7653 variant, -bhc, which does actually run the callout code, including
7654 consulting and updating the callout cache.
7656 70. Added support for saslauthd authentication, courtesy of Alexander
7659 71. If statvfs() failed on the spool or log directories while checking their
7660 size for availability, Exim confusingly gave the error "space shortage".
7661 Furthermore, in debugging mode it crashed with a floating point exception.
7662 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7663 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7664 serious problem, Exim now writes to the main and panic logs when this
7665 happens, with details of the failure. It then refuses to accept the
7666 incoming message, giving the message "spool directory problem" or "log
7667 directory problem" with a 421 code for SMTP messages.
7669 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7670 0, 1, and 2 exist, because some OS complain for execs without them (see
7671 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7672 descriptors. However, the code omitted to check that the open succeeded,
7673 causing mysterious errors if for some reason the permissions on /dev/null
7674 got screwed. Now Exim writes a message to the main and panic logs, and
7675 bombs out if it can't open /dev/null.
7677 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7678 interact so that it is all more flexible. It is supposed to remain
7679 backwards compatible. Also added extra_local_interfaces.
7681 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7682 to bomb out with an assertion failure - to the client this appears as a
7683 connection drop. This problem occurs in the part of the code that was taken
7684 from the Samba project. Fortunately, the assertion is in a very simple
7685 function, so I have fixed this by reproducing the function inline in the
7686 one place where it is called, and arranging for authentication to fail
7687 instead of killing the process with assert().
7689 75. The SPA client code was not working when the server requested OEM rather
7690 than Unicode encoding.
7692 76. Added code to make require_files with a specific uid setting more usable in
7693 the case where statting the file as root fails - usually a non-root-mounted
7694 NFS file system. When this happens and the failure is EACCES, Exim now
7695 forks a subprocess and does the per-uid checking as the relevant uid.
7697 77. Added process_log_path.
7699 78. If log_file_path was not explicitly set, a setting of check_log_space or
7700 check_log_inodes was ignored.
7702 79. If a space check for the spool or log partitions fails, the incident is now
7703 logged. Of course, in the latter case the data may get lost...
7705 80. Added the %p formatting code to string_format() so that it can be used to
7706 print addresses in debug_print(). Adjusted all the address printing in the
7707 debugging in store.c to use %p rather than %d.
7709 81. There was a concern that a line of code in smtp_in.c could overflow a
7710 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7711 initially expressed, the concern was not well-founded, because trailing
7712 spaces are removed early. However, if the trailing spaces were followed by
7713 a NULL, they did not get removed, so the overflow was possible. Two fixes
7716 (a) I re-wrote the offending code in a cleaner fashion.
7717 (b) If an incoming SMTP command contains a NULL character, it is rejected
7720 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7721 runs initgroups(), so that if the Exim user is in any additional groups,
7722 they will be used during message reception.
7728 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob