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.
128 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
129 side, to decode a TLS 1.3 packet capture.
131 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
132 Previously the default library behaviour applied, sending two, each in
135 JH/03 Debug output for ACL now gives the config file name and line number for
138 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
140 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
142 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
143 buffer overrun for (non-chunking) other transports.
145 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
146 TLS1.3, means that a server rejecting a client certificate is not visible
147 to the client until the first read of encrypted data (typically the
148 response to EHLO). Add detection for that case and treat it as a failed
149 TLS connection attempt, so that the normal retry-in-clear can work (if
150 suitably configured).
152 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
153 and/or domain. Found and fixed by Jason Betts.
155 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
156 configuration). If a CNAME target was not a wellformed name pattern, a
159 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
160 the OS reports them interleaved with other addresses.
162 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
163 used both for input and for a verify callout, both encrypted, SMTP
164 responses being sent by the server could be lost. This resulted in
165 dropped connections and sometimes bounces generated by a peer sending
168 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
169 string. Previously it was possible to cause undefined behaviour in a
170 library routine (usually a crash). Found by "zerons".
172 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
175 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
176 API was removed, so update to use the newer ones.
178 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
179 any timeout set, is taking a long time. Previously we would hang on to a
180 rotated logfile "forever" if the input was arriving with long gaps
181 (a previous attempt to fix addressed lack, for a long time, of initial
184 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
185 shared (NFS) environment. The length of the tempfile name is now
186 4 + 16 ("hdr.$message_exim_id") which might break on file
187 systems which restrict the file name length to lower values.
190 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
191 shared (NFS) environment.
193 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
194 did for all versions <4.90). Notably -M, -m, --invert, -I may be
197 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
198 on some platforms for bit 31.
200 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
201 to changes apparently associated with TLS1.3 handling some of the APIs
202 previously used were either nonfunctional or inappropriate. Strings
203 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
204 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
205 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
206 This affects log line X= elements, the $tls_{in,out}_cipher variables,
207 and the use of specific cipher names in the encrypted= ACL condition.
209 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
211 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
212 verification result was not updated unless hosts_require_ocsp applied.
214 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
215 queue_list_requires_admin set to false, non-admin users were denied the
218 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
219 directory-of-certs mode. Previously they were advertised despite the
222 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
223 A single TCP connection by a client will now hold a TLS connection open
224 for multiple message deliveries, by default. Previoud the default was to
227 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
228 default. If built with the facility, DANE will be used. The facility
229 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
231 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
232 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
233 must be defined and you must still, unless you define DISABLE_TLS, manage
234 the the include-dir and library-file requirements that go with that
235 choice. Non-TLS builds are still supported.
237 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
240 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
241 default. If the platform supports and has the facility enabled, it will
242 be requested on all coneections.
244 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
245 controlled by the build-time option SUPPORT_PIPE_CONNECT.
247 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
249 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
250 success-DSN messages. Previously the From: header was always the default
251 one for these; the option was ignored.
253 JH/28 Fix the timeout on smtp response to apply to the whole response.
254 Previously it was reset for every read, so a teergrubing peer sending
255 single bytes within the time limit could extend the connection for a
256 long time. Credit to Qualsys Security Advisory Team for the discovery.
258 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
259 delivery address, which leaked information of the results of local
260 forwarding. Change to the original envelope recipient address, per
263 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
264 requested. Previously not bounce was generated and a log entry of
265 error ignored was made.
267 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
269 JH/32 Introduce a general tainting mechanism for values read from the input
270 channel, and values derived from them. Refuse to expand any tainted
271 values, to catch one form of exploit.
273 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
274 was unused and the unexpanded text used for the test. Found and
275 fixed by Ruben Jenster.
277 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
278 an attempt to use a TLS library read routine dereffed a nul pointer,
281 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
282 them in our smtp response.
284 JH/36 Have the general router option retry_use_local_part default to true when
285 any of the restrictive preconditions are set (to anything). Previously it
286 was only for check_local user. The change removes one item of manual
287 configuration which is required for proper retries when a remote router
288 handles a subset of addresses for a domain.
290 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
291 link count into consideration.
293 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
294 caused the extension of big_buffer, the following lines were ignored.
296 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
297 accordance with RFC 2308. Previously there was no expiry, so a longlived
298 receive process (eg. due to ACL delays) versus a short SOA value could
301 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
303 JH/39 Promote DMARC support to mainline.
305 JH/40 Bug 2452: Add a References: header to DSNs.
307 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
308 parameters. The relevant library call is documented as "Deprecated: This
309 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
310 3.6.0, DH parameters are negotiated following RFC7919."
312 HS/06 Change the default of dnssec_request_domains to "*"
314 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
315 carried on and emitted a BDAT command, even when PIPELINING was not
318 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
319 buffer was used for the filename, resulting in a trap when tainted
320 arguments (eg. $domain) were used.
322 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
323 recommended to avoid a possible server-load attack. The feature can be
324 re-enabled via the openssl_options main cofiguration option.
326 JH/45 local_scan API: documented the current smtp_printf() call. This changed
327 for version 4.90 - adding a "more data" boolean to the arguments.
328 Bumped the ABI version number also, this having been missed previously;
329 release versions 4.90 to 4.92.3 inclusive were effectively broken in
330 respect of usage of smtp_printf() by either local_scan code or libraries
331 accessed via the ${dlfunc } expansion item. Both will need coding
332 adjustment for any calls to smtp_printf() to match the new function
333 signature; a FALSE value for the new argument is always safe.
335 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
336 the file-offset (which the Linux syscall does, and exim expects); this
337 resulted in an indefinite loop.
339 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
340 to do ARC verification. The Authentication-Results: header line added
341 by the configuration then had no ARC item.
347 JH/01 Remove code calling the customisable local_scan function, unless a new
348 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
350 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
351 non-signal-safe functions being used.
353 JH/03 Bug 2269: When presented with a received message having a stupidly large
354 number of DKIM-Signature headers, disable DKIM verification to avoid
355 a resource-consumption attack. The limit is set at twenty.
357 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
358 report of oldest_pass in ${authres } in consequence, and separate out
359 some descriptions of reasons for verification fail.
361 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
362 files in the spool were present and unlocked. A queue-runner could spot
363 them, resulting in a duplicate delivery. Fix that by doing the unlock
364 after the unlink. Investigation by Tim Stewart. Take the opportunity to
365 add more error-checking on spoolfile handling while that code is being
368 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
369 No known attacks, no CVE, this is defensive hardening.
371 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
372 a queue-runner could start a delivery while other operations were ongoing.
373 Cutthrough delivery was a common victim, resulting in duplicate delivery.
374 Found and investigated by Tim Stewart. Fix by using the open message data
375 file handle rather than opening another, and not locally closing it (which
376 releases a lock) for that case, while creating the temporary .eml format
377 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
379 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
380 $sender_verify_failure/$recipient_verify_failure to "random".
382 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
385 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
386 Previously this would segfault.
388 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
391 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
392 like zero, since the resolver should be doing this for us, But we need one
393 as a CNAME but no MX presence gets the CNAME returned; we need to check
394 that doesn't point to an MX to declare it "no MX returned" rather than
395 "error, loop". A new main option is added so the older capability of
396 following some limited number of chain links is maintained.
398 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
400 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
401 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
402 now-deprecated earlier definitions used only specified the range up to TLS
403 1.2 (in the older-version library docs).
405 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
407 JH/15 Rework TLS client-side context management. Stop using a global, and
408 explicitly pass a context around. This enables future use of TLS for
409 connections to service-daemons (eg. malware scanning) while a client smtp
410 connection is using TLS; with cutthrough connections this is quite likely.
412 JH/16 Fix ARC verification to do AS checks in reverse order.
414 JH/17 Support a "tls" option on the ${readsocket } expansion item.
416 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
417 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
418 Previously the "utf8" would be re-prepended for every additional message.
420 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
421 Previously thery were accepted, resulting in issues when attempting to
422 forward messages to a non-supporting MTA.
424 PP/02 Let -n work with printing macros too, not just options.
426 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
427 one parent address was copied, and bogus data was used at delivery-logging
428 time. Either a crash (after delivery) or bogus log data could result.
429 Discovery and analysis by Tim Stewart.
431 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
432 Previously if the string ended mid-character, we did not insert the
433 promised '?' replacement.
435 PP/04 Documentation: current string operators work on bytes, not codepoints.
437 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
438 should pack well giving a smaller memory footprint so better caching and
439 therefore performance. Group the declarations where this can't be done so
440 that the byte-sized flag variables are not interspersed among pointer
441 variables, giving a better chance of good packing by the compiler.
443 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
444 non-null, to avoid issues with sites running BATV. Previously reports were
445 sent with an empty envelope sender so looked like bounces.
447 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
448 The ignore_error flag wasn't being returned from the filter subprocess so
449 was not set for later routers. Investigation and fix by Matthias Kurz.
451 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
452 and a msg:complete for the whole, when a message is manually removed using
453 -Mrm. Developement by Matthias Kurz, hacked on by JH.
455 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
456 a "Gnu special" function, asprintf() in the DB utility binary builds; I
457 hope that is portable enough.
459 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
460 requiring a known-CA anchor certificate; make it now rely entirely on the
461 TLSA as an anchor. Checking the name on the leaf cert against the name
462 on the A-record for the host is still done for TA (but not for EE mode).
464 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
465 included in delivery lines for non-proxied connections, when compiled with
466 SUPPORT_SOCKS and running with proxy logging enabled.
468 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
469 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
470 move the existing event to fire before the normal logging of message
471 failure so that custom logging is bracketed by normal logging.
473 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
474 msg:fail:internal event. Developement by Matthias Kurz.
476 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
477 far too small for todays use of crypto signatures stored there. Go all
478 the way to the max DNS message size of 64kB, even though this might be
479 overmuch for IOT constrained device use.
481 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
482 copy a string over itself. The library routine is documented as not
483 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
485 JH/32 For main options check_spool_space and check_inode_space, where the
486 platform supports 64b integers, support more than the previous 2^31 kB
487 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
488 the previous G, M, k.
490 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
491 $authenticated_fail_id variable on authentication failure. Previously
494 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
495 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
496 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
497 GNUTLS_SEC_PARAM_MEDIUM.
499 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
500 side. Previously we would continue as if no SNI had been received.
502 JH/36 Harden the handling of string-lists. When a list consisted of a sole
503 "<" character, which should be a list-separator specification, we walked
504 off past the nul-terimation.
506 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
507 causes) even when the retry time is not yet met. Previously they were
508 not, meaning that when (say) an account was over-quota and temp-rejecting,
509 and multiple senders' messages were queued, only one sender would get
510 notified on each configured delay_warning cycle.
512 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
514 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
515 cipher-suites, an error can be left on the stack even for a succeeding
516 accept; this results in impossible error messages when a later operation
519 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
520 return error codes indicating retry. Under TLS1.3 this becomes required.
522 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
523 it only wrote the new authenticators, resulting in a lack of tracking of
524 peer changes of ESMTP extensions until the next cache flush.
526 JH/41 Fix the loop reading a message header line to check for integer overflow,
527 and more-often against header_maxsize. Previously a crafted message could
528 induce a crash of the recive process; now the message is cleanly rejected.
530 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
531 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
537 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
538 When redis_servers is set to a list of > 1 element, and the Redis servers
539 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
540 case of MOVED into a DEFER case instead, thus moving the query onto the
541 next server in the list. For a cluster of N elements, all N servers must
542 be defined in redis_servers.
544 GF/02 Catch and remove uninitialized value warning in exiqsumm
545 Check for existence of @ARGV before looking at $ARGV[0]
547 JH/01 Replace the store_release() internal interface with store_newblock(),
548 which internalises the check required to safely use the old one, plus
549 the allocate and data copy operations duplicated in both (!) of the
550 extant use locations.
552 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
553 modifier. This matches the restriction on the commandline.
555 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
556 Previously only the last row was returned.
558 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
559 we assumed that tags in the header were well-formed, and parsed the
560 element content after inspecting only the first char of the tag.
561 Assumptions at that stage could crash the receive process on malformed
564 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
565 While running the DKIM ACL we operate on the Permanent memory pool so that
566 variables created with "set" persist to the DATA ACL. Also (at any time)
567 DNS lookups that fail create cache records using the Permanent pool. But
568 expansions release any allocations made on the current pool - so a dnsdb
569 lookup expansion done in the DKIM ACL releases the memory used for the
570 DNS negative-cache, and bad things result. Solution is to switch to the
571 Main pool for expansions.
572 While we're in that code, add checks on the DNS cache during store_reset,
573 active in the testsuite.
574 Problem spotted, and debugging aided, by Wolfgang Breyha.
576 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
577 When none of the hosts presented to a transport match an already-open
578 connection, close it and proceed with the list. Previously we would
579 queue the message. Spotted by Lena with Yahoo, probably involving
582 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
583 Previously a spurious "250 OK id=" response was appended to the proper
586 JH/08 The "support for" informational output now, which built with Content
587 Scanning support, has a line for the malware scanner interfaces compiled
588 in. Interface can be individually included or not at build time.
590 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
591 by the template makefile "src/EDITME". The "STREAM" support for an older
592 ClamAV interface method is removed.
594 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
595 rows affected is given instead).
597 JH/11 The runtime Berkeley DB library version is now additionally output by
598 "exim -d -bV". Previously only the compile-time version was shown.
600 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
601 SMTP connection. Previously, when one had more receipients than the
602 first, an abortive onward connection was made. Move to full support for
603 multiple onward connections in sequence, handling cutthrough connection
604 for all multi-message initiating connections.
606 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
607 routers. Previously, a multi-recipient message would fail to match the
608 onward-connection opened for the first recipient, and cause its closure.
610 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
611 a timeout on read on a GnuTLS initiating connection, resulting in the
612 initiating connection being dropped. This mattered most when the callout
613 was marked defer_ok. Fix to keep the two timeout-detection methods
616 JH/15 Relax results from ACL control request to enable cutthrough, in
617 unsupported situations, from error to silently (except under debug)
618 ignoring. This covers use with PRDR, frozen messages, queue-only and
621 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
623 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
624 metadata, resulting in a crash in free().
626 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
627 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
628 Broken also in d185889f4, with init system revamp.
630 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
631 Previously we abruptly closed the connection after reading a malware-
632 found indication; now we go on to read the "scan ok" response line,
635 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
636 ACL. Previously, a crash would result.
638 JH/19 Speed up macro lookups during configuration file read, by skipping non-
639 macro text after a replacement (previously it was only once per line) and
640 by skipping builtin macros when searching for an uppercase lead character.
642 JH/20 DANE support moved from Experimental to mainline. The Makefile control
643 for the build is renamed.
645 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
646 was allocated for every new TLS startup, meaning one per message. Fix
647 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
649 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
650 reported the original. Fix to report (as far as possible) the ACL
651 result replacing the original.
653 JH/23 Fix memory leak during multi-message connections using STARTTLS under
654 OpenSSL. Certificate information is loaded for every new TLS startup,
655 and the resources needed to be freed.
657 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
659 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
662 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
663 DATA response info to the (existing) per-recipient response info for
664 the "C=" log element. It can have useful tracking info from the
665 destination system. Patch from Simon Arlott.
667 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
668 length value. Previously this would segfault.
670 HS/02 Support Avast multiline protoocol, this allows passing flags to
671 newer versions of the scanner.
673 JH/28 Ensure that variables possibly set during message acceptance are marked
674 dead before release of memory in the daemon loop. This stops complaints
675 about them when the debug_store option is enabled. Discovered specifically
676 for sender_rate_period, but applies to a whole set of variables.
677 Do the same for the queue-runner and queue-list loops, for variables set
678 from spool message files. Do the same for the SMTP per-message loop, for
679 certain variables indirectly set in ACL operations.
681 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
682 as a multi-recipient message from a mailinglist manager). The coding had
683 an arbitrary cutoff number of characters while checking for more input;
684 enforced by writing a NUL into the buffer. This corrupted long / fast
685 input. The problem was exposed more widely when more pipelineing of SMTP
686 responses was introduced, and one Exim system was feeding another.
687 The symptom is log complaints of SMTP syntax error (NUL chars) on the
688 receiving system, and refused recipients seen by the sending system
689 (propating to people being dropped from mailing lists).
690 Discovered and pinpointed by David Carter.
692 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
693 replaced by the ${authresults } expansion.
695 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
697 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
698 allows proper process termination in container environments.
700 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
701 Previously the "final dot" had a newline after it; ensure it is CR,LF.
703 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
704 and "err_perm", deprecated since 4.83 when the RFC-defined words
705 "temperror" and "permerror" were introduced.
707 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
708 transport-filters or DKIM-signing. The restriction was lost in the
709 consolidation of verify-callout and delivery SMTP handling.
710 Extend the restriction to also cover ARC-signing.
712 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
713 in defer=pass mode supply a 450 to the initiator. Previously the message
716 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
717 tls_require_ciphers is used as before.
719 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
720 "pass_unscanned". Only tmpfails from the scanner are written to
721 the paniclog, as they may require admin intervention (permission
722 denied, license issues). Other scanner errors (like decompression
723 bombs) do not cause a paniclog entry.
725 JH/36 Fix reinitialisation of DKIM logging variable between messages.
726 Previously it was possible to log spurious information in receive log
729 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
730 triggered odd behaviour from Outlook Express clients.
732 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
735 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
736 since the IETF WG has not yet settled on that versus the original
737 "bare" representation.
739 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
740 Previously the millisecond value corrupted the output.
741 Fix also for syslog_pid=no and log_selector +pid, for which the pid
742 corrupted the output.
748 JH/01 Rework error string handling in TLS interface so that the caller in
749 more cases is responsible for logging. This permits library-sourced
750 string to be attached to addresses during delivery, and collapses
751 pairs of long lines into single ones.
753 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
754 during configuration. Wildcards are allowed and expanded.
756 JH/02 Rework error string handling in DKIM to pass more info back to callers.
757 This permits better logging.
759 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
760 do not close it down and have the child transport start it up again on
761 the passed-on TCP connection. Instead, proxy the child (and any
762 subsequent ones) for TLS via a unix-domain socket channel. Logging is
763 affected: the continued delivery log lines do not have any DNSSEC, TLS
764 Certificate or OCSP information. TLS cipher information is still logged.
766 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
767 identical IP addresses on different listening ports. Will also affect
770 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
771 add noisy ifdef guards to special-case this sillyness.
772 Patch from Bernd Kuhls.
774 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
775 than 255 are no longer allowed.
777 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
778 Disable the session-cache too, which might reduce our load. Since we
779 currrectly use a new context for every connection, both as server and
780 client, there is no benefit for these.
781 GnuTLS appears to not support tickets server-side by default (we don't
782 call gnutls_session_ticket_enable_server()) but client side is enabled
783 by default on recent versions (3.1.3 +) unless the PFS priority string
786 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
787 <https://reproducible-builds.org/specs/source-date-epoch/>.
789 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
790 the check for any unsuccessful recipients did not notice the limit, and
791 erroneously found still-pending ones.
793 JH/08 Pipeline CHUNKING command and data together, on kernels that support
794 MSG_MORE. Only in-clear (not on TLS connections).
796 JH/09 Avoid using a temporary file during transport using dkim. Unless a
797 transport-filter is involved we can buffer the headers in memory for
798 creating the signature, and read the spool data file once for the
799 signature and again for transmission.
801 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
802 4.77 as the kernel support then wasn't solid, having issues in 64bit
803 mode. Now, it's been long enough. Add support for FreeBSD also.
805 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
806 case where the routing stage had gathered several addresses to send to
807 a host before calling the transport for the first, we previously failed
808 to close down TLS in the old transport process before passing the TCP
809 connection to the new process. The new one sent a STARTTLS command
810 which naturally failed, giving a failed delivery and bloating the retry
811 database. Investigation and fix prototype from Wolfgang Breyha.
813 JH/12 Fix check on SMTP command input synchronisation. Previously there were
814 false-negatives in the check that the sender had not preempted a response
815 or prompt from Exim (running as a server), due to that code's lack of
816 awareness of the SMTP input buffering.
818 PP/04 Add commandline_checks_require_admin option.
819 Exim drops privileges sanely, various checks such as -be aren't a
820 security problem, as long as you trust local users with access to their
821 own account. When invoked by services which pass untrusted data to
822 Exim, this might be an issue. Set this option in main configuration
823 AND make fixes to the calling application, such as using `--` to stop
826 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
827 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
828 responses to those, into a single TLS record each way (this usually means
829 a single packet). As a side issue, smtp_enforce_sync now works on TLS
832 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
833 affects you only if you're dancing at the edge of the param size limits.
834 If you are, and this message makes sense to you, then: raise the
835 configured limit or use OpenSSL 1.1. Nothing we can do for older
838 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
839 cmdline element to get the documented default one. Previously it was
842 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
843 are unsuable for non-message-specific reasons - eg. connection timeout,
844 banner-time rejection.
846 JH/16 Fix logging of delivery remote port, when specified by router, under
849 PP/06 Repair manualroute's ability to take options in any order, even if one
850 is the name of a transport.
853 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
855 JH/17 Change the list-building routines interface to use the expanding-string
856 triplet model, for better allocation and copying behaviour.
858 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
859 Previously it was constructed the first time a possibly-matching string
860 was met in the configuration file input during startup; now it is done
863 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
864 compatible one, to avoid the (poorly documented) possibility of a config
865 file in the working directory redirecting the DB files, possibly correpting
866 some existing file. CVE-2017-10140 assigned for BDB.
868 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
869 cache-hot. Previously, although the result was properly cached, the
870 initial verify call returned a defer.
872 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
873 the main verify for receipient in uncached-mode.
875 JH/22 Retire historical build files to an "unsupported" subdir. These are
876 defined as "ones for which we have no current evidence of testing".
878 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
879 if present. Previously it was ignored.
881 JH/24 Start using specified-initialisers in C structure init coding. This is
882 a C99 feature (it's 2017, so now considered safe).
884 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
885 if was a fixed-sized field and bitmask ops via macros; it is now more
888 PP/07 GitHub PR 56: Apply MariaDB build fix.
889 Patch provided by Jaroslav Škarvada.
891 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
892 during Coverity cleanups [4.87 JH/47]
893 Diagnosis and fix provided by Michael Fischer v. Mollard.
895 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
896 the right size to place the terminating semicolon on its own folded
897 line, the header hash was calculated to an incorrect value thanks to
898 the (relaxed) space the fold became.
900 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
901 and confused the parent.
903 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
904 which could crash as a result. This could lead to undeliverable messages.
906 JH/28 Logging: "next input sent too soon" now shows where input was truncated
909 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
910 matters on fast-turnover and PID-randomising systems, which were getting
911 out-of-order delivery.
913 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
914 a possibly-overlapping copy. The symptom was that "Remote host closed
915 connection in response to HELO" was logged instead of the actual 4xx
918 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
919 Previously only that bufferd was discarded, resulting in SYMTP command
922 JH/32 DKIM: when a message has multiple signatures matching an identity given
923 in dkim_verify_signers, run the dkim acl once for each. Previously only
924 one run was done. Bug 2189.
926 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
927 "panic the current process" to "deliberately defer". The panic log is
928 still written with the problem list name; the mail and reject logs now
929 get a temp-reject line for the message that was being handled, saying
930 something like "domains check lookup or other defer". The SMTP 451
931 message is still "Temporary local problem".
933 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
934 A crafted sequence of BDAT commands could result in in-use memory beeing
935 freed. CVE-2017-16943.
937 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
938 from SMTP input. Previously it was always done; now only done for DATA
939 and not BDAT commands. CVE-2017-16944.
941 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
942 to the message (such as an overlong header line). Previously this was
943 not done and we did not exit BDAT mode. Followon from the previous item
944 though a different problem.
950 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
951 than -2003 did; needs libidn2 in addition to libidn.
953 JH/02 The path option on a pipe transport is now expanded before use.
955 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
956 Patch provided by "Björn", documentation fix added too.
958 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
959 missing a wire-to-host endian conversion.
961 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
962 close after a BDAT command line could be taken as a following command,
963 giving a synch failure. Fix by only checking for synch immediately
964 before acknowledging the chunk.
966 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
967 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
968 macro. Patches provided by Josh Soref.
970 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
971 Previously we did not; the RFC seems ambiguous and VRFY is not listed
972 by IANA as a service extension. However, John Klensin suggests that we
975 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
976 the dkim code may be unix-mode line endings rather than smtp wire-format
977 CRLF, so prepend a CR to any bare LF.
979 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
980 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
982 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
983 the first were themselves being wrongly included in the feed into dkim
984 processing; with most chunk sizes in use this resulted in an incorrect
985 body hash calculated value.
987 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
988 DKIM signature block, for verification. Although advised against by
989 standards it is specifically not ruled illegal.
991 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
993 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
994 missing a body hash (the bh= tag).
996 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
997 It seems that HAProxy sends the Proxy Protocol information in clear and
998 only then does a TLS startup, so do the same.
1000 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1001 TCP connections (such as for Spamd) unless the daemon successfully set
1002 Fast Open mode on its listening sockets. This fixes breakage seen on
1003 too-old kernels or those not configured for Fast Open, at the cost of
1004 requiring both directions being enabled for TFO, and TFO never being used
1005 by non-daemon-related Exim processes.
1007 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1008 endings, at least on the first header line. Try to canonify any that get
1009 past that check, despite the cost.
1011 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1012 now limited to an arbitrary five deep, while parsing addresses with the
1013 strip_excess_angle_brackets option enabled.
1015 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1016 instead leave the unprompted TLS handshake in socket buffer for the
1017 TLS library to consume.
1019 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1021 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1023 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1024 region in one operation, for speed, and this leaves assigned pointers
1025 dangling. Add checks run only under the testsuite which checks all
1026 variables at a store-reset and panics on a dangling pointer; add code
1027 explicitly nulling out all the variables discovered. Fixes one known
1028 bug: a transport crash, where a dangling pointer for $sending_ip_address
1029 originally assigned in a verify callout, is re-used.
1031 PP/06 Drop '.' from @INC in various Perl scripts.
1033 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1035 PP/08 Reduce a number of compilation warnings under clang; building with
1036 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1037 should be warning-free.
1039 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1041 HS/01 Fix portability problems introduced by PP/08 for platforms where
1042 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1044 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1045 chunk. This allows us to accept broken chunked messages. We need a more
1046 general solution here.
1048 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1049 already-broken messages in the queue.
1051 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1053 JH/19 Fix reference counting bug in routing-generated-address tracking.
1059 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1060 supports it and a size is available (ie. the sending peer gave us one).
1062 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1063 years of being deprecated). The replacements are the ACLs
1064 acl_smtp_mime and acl_not_smtp_mime.
1066 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1067 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1068 or even an in-clear connection were permitted. Now, if the host lookup
1069 was dnssec and dane was requested then the host is only used if the
1070 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1071 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1072 if one fails this test.
1073 This means that a poorly-configured remote DNS will make it incommunicado;
1074 but it protects against a DNS-interception attack on it.
1076 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1077 non-noisy when a race steals the message being considered.
1079 JH/05 If main configuration option tls_certificate is unset, generate a
1080 self-signed certificate for inbound TLS connections.
1082 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1083 in rewrites and routers.
1085 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1086 and logged a warning sing 4.83; now they are a configuration file error.
1088 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1089 (lacking @domain). Apply the same qualification processing as RCPT.
1091 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1093 JH/10 Support ${sha256:} applied to a string (as well as the previous
1096 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1097 a cutthrough deliver is pending, as we always want to make a connection.
1098 This also avoids re-routing the message when later placing the cutthrough
1099 connection after a verify cache hit.
1100 Do not update it with the verify result either.
1102 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1103 when routing results in more than one destination address.
1105 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1106 signing (which inhibits the cutthrough capability). Previously only
1107 the presence of an option was tested; now an expansion evaluating as
1108 empty is permissible (obviously it should depend only on data available
1109 when the cutthrough connection is made).
1111 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1112 the relevant preceding SMTP command did not note the pipelining mode.
1114 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1115 Previously they were not counted.
1117 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1118 as one having no matching records. Previously we deferred the message
1119 that needed the lookup.
1121 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1122 distinguished as "(=".
1124 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1125 for missing MX records. Previously it only worked for missing A records.
1127 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1129 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1130 after the data-go-ahead and data-ack. Patch from Jason Betts.
1132 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1133 even for a "none" policy. Patch from Tony Meyer.
1135 JH/22 Fix continued use of a connection for further deliveries. If a port was
1136 specified by a router, it must also match for the delivery to be
1139 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1140 When one of the recipients of a message was unsuitable for the connection
1141 (has no matching addresses), we lost track of needing to mark it
1142 deferred. As a result mail would be lost.
1144 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1146 JH/25 Decoding ACL controls is now done using a binary search; the source code
1147 takes up less space and should be simpler to maintain. Merge the ACL
1148 condition decode tables also, with similar effect.
1150 JH/26 Fix problem with one_time used on a redirect router which returned the
1151 parent address unchanged. A retry would see the parent address marked as
1152 delivered, so not attempt the (identical) child. As a result mail would
1155 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1156 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1157 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1160 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1161 Main config options check_{log,spool}_{inodes,space} are now
1162 100 inodes, 10MB unless set otherwise in the configuration.
1164 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1165 Previously it only applied to the main-section connection policy
1168 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1170 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1171 by me. Added RFC7919 DH primes as an alternative.
1173 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1174 are not in the system include path.
1176 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1177 GnuTLS, when a session startup failed (eg because the client disconnected)
1178 Exim did stdio operations after fclose. This was exposed by a recent
1179 change which nulled out the file handle after the fclose.
1181 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1182 signed directly by the cert-signing cert, rather than an intermediate
1183 OCSP-signing cert. This is the model used by LetsEncrypt.
1185 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1187 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1188 an incoming connection.
1190 HS/02 Bug 1802: Do not half-close the connection after sending a request
1193 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1194 fallback to "prime256v1".
1196 JH/34 SECURITY: Use proper copy of DATA command in error message.
1197 Could leak key material. Remotely exploitable. CVE-2016-9963.
1203 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1204 and 3.4.4 - once the server is enabled to respond to an OCSP request
1205 it does even when not requested, resulting in a stapling non-aware
1206 client dropping the TLS connection.
1208 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1209 support variable-length bit vectors. No functional change.
1211 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1212 The I= interface field on outgoing lines is now after the H= remote
1213 host field, same as incoming lines. There is a separate
1214 outgoing_interface log selector which allows you to disable the
1217 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1218 If not running log_selector +smtp_connection the mainlog would be held
1219 open indefinitely after a "too many connections" event, including to a
1220 deleted file after a log rotate. Leave the per net connection logging
1221 leaving it open for efficiency as that will be quickly detected by the
1222 check on the next write.
1224 HS/01 Bug 1671: Fix post transport crash.
1225 Processing the wait-<transport> messages could crash the delivery
1226 process if the message IDs didn't exist for some reason. When
1227 using 'split_spool_directory=yes' the construction of the spool
1228 file name failed already, exposing the same netto behaviour.
1230 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1231 mime_regex ACL conditions.
1233 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1234 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1235 to HELO, local diagnostic string.
1237 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1238 log line to debug. Even when configured with a tls authenticator many
1239 client connections are expected to not authenticate in this way, so
1240 an authenticate fail is not an error.
1242 HS/02 Add the Exim version string to the process info. This way exiwhat
1243 gives some more detail about the running daemon.
1245 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1246 matter for fast-change records such as DNSBLs.
1248 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1249 retry record, even if constant. There may be multiple transports with
1250 different interface settings and the retry behaviour needs to be kept
1253 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1255 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1257 JH/10 Bug 840: fix log_defer_output option of pipe transport
1259 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1260 hosts_randomize. This matters a lot when combined with mua_wrapper.
1262 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1263 ${quote_pgsql:<string>} operator.
1265 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1267 JH/14 Change header file name for freeradius-client. Relevant if compiling
1268 with Radius support; from the Gentoo tree and checked under Fedora.
1270 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1272 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1273 option transport_filter, meaning no filtering.
1275 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1277 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1278 defaults to "*" (all hosts). The variable is now available when not built
1279 with TLS, default unset, mainly to enable keeping the testsuite sane.
1280 If a server certificate is not supplied (via tls_certificate) an error is
1281 logged, and clients will find TLS connections fail on startup. Presumably
1282 they will retry in-clear.
1283 Packagers of Exim are strongly encouraged to create a server certificate
1284 at installation time.
1286 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1287 with the $config_file variable.
1289 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1290 in transport context, after the attempt, and per-recipient. The latter type
1291 is per host attempted. The event data is the error message, and the errno
1292 information encodes the lookup type (A vs. MX) used for the (first) host,
1293 and the trailing two digits of the smtp 4xx response.
1295 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1296 to write to mainlog (or rejectlog, paniclog) in the window between file
1297 creation and permissions/ownership being changed. Particularly affects
1298 installations where exicyclog is run as root, rather than exim user;
1299 result is that the running daemon panics and dies.
1301 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1303 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1304 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1305 "pri" and "weight". Note that the previous implicit priority given by the
1306 list order is no longer honoured.
1308 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1309 for DKIM processing.
1311 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1312 by defining SUPPORT_SOCKS.
1314 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1315 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1316 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1317 variables are renamed to proxy_{local,external}_{address,port}.
1319 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1320 for a build by defining SUPPORT_I18N
1322 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1323 of the query string, and make ${quote_redis:} do that quoting.
1325 JH/29 Move Events support from Experimental to mainline, enabled by default
1326 and removable for a build by defining DISABLE_EVENT.
1328 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1330 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1331 cached by the daemon.
1333 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1334 by defining LOOKUP_REDIS. The libhiredis library is required.
1336 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1337 keys are given for lookup.
1339 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1340 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1341 only supported when built with TLS support. The PolarSSL SHA routines
1342 are still used when the TLS library is too old for convenient support.
1344 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1345 openssl_options), for security. OpenSSL forces this from version 1.1.0
1346 server-side so match that on older versions.
1348 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1349 allocation for $value could be released as the expansion processing
1350 concluded, but leaving the global pointer active for it.
1352 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1353 and to use the domains and local_parts ACL conditions.
1355 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1356 incorrectly not doubled on cutthrough transmission, hence seen as a
1357 body-termination at the receiving system - resulting in truncated mails.
1358 Commonly the sender saw a TCP-level error, and retransmitted the message
1359 via the normal store-and-forward channel. This could result in duplicates
1360 received - but deduplicating mailstores were liable to retain only the
1361 initial truncated version.
1363 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1365 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1367 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1368 we're in there, support oversigning also; bug 1309.
1370 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1372 HS/04 Add support for keep_environment and add_environment options.
1374 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1375 either intentional arithmetic overflow during PRNG, or testing config-
1378 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1379 delivery resulted in actual delivery. Cancel cutthrough before DATA
1382 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1383 rejects a recipient: pass the reject to the originator.
1385 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1386 Many were false-positives and ignorable, but it's worth fixing the
1389 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1390 for the new environment-manipulation done at startup. Move the routines
1391 from being local to tls.c to being global via the os.c file.
1393 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1394 an extract embedded as result-arg for a map, the first arg for extract
1395 is unavailable so we cannot tell if this is a numbered or keyed
1396 extraction. Accept either.
1402 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1405 JH/02 The smtp transport option "multi_domain" is now expanded.
1407 JH/03 The smtp transport now requests PRDR by default, if the server offers
1410 JH/04 Certificate name checking on server certificates, when exim is a client,
1411 is now done by default. The transport option tls_verify_cert_hostnames
1412 can be used to disable this per-host. The build option
1413 EXPERIMENTAL_CERTNAMES is withdrawn.
1415 JH/05 The value of the tls_verify_certificates smtp transport and main options
1416 default to the word "system" to access the system default CA bundle.
1417 For GnuTLS, only version 3.0.20 or later.
1419 JH/06 Verification of the server certificate for a TLS connection is now tried
1420 (but not required) by default. The verification status is now logged by
1421 default, for both outbound TLS and client-certificate supplying inbound
1424 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1427 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1428 Status Notification (bounce) messages are now MIME format per RFC 3464.
1429 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1430 under the control of the dsn_advertise_hosts option, and routers may
1431 have a dsn_lasthop option.
1433 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1434 default, modifiable by a malware= option. The list separator for
1435 the options can now be changed in the usual way. Bug 68.
1437 JH/10 The smtp_receive_timeout main option is now expanded before use.
1439 JH/11 The incoming_interface log option now also enables logging of the
1440 local interface on delivery outgoing connections.
1442 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1443 if the interface and destination host and port all match.
1445 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1448 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1449 Patch from Andrew Lewis.
1451 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1452 now supports optional time-restrictions, weighting, and priority
1453 modifiers per server. Patch originally by <rommer@active.by>.
1455 JH/16 The spamd_address main option now supports a mixed list of local
1456 and remote servers. Remote servers can be IPv6 addresses, and
1457 specify a port-range.
1459 JH/17 Bug 68: The spamd_address main option now supports an optional
1460 timeout value per server.
1462 JH/18 Bug 1581: Router and transport options headers_add/remove can
1463 now have the list separator specified.
1465 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1468 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1471 JH/21 Support for the A6 type of dns record is withdrawn.
1473 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1474 rather than the verbs used.
1476 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1477 from 255 to 1024 chars.
1479 JH/24 Verification callouts now attempt to use TLS by default.
1481 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1482 are generic router options now. The defaults didn't change.
1484 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1485 Original patch from Alexander Shikoff, worked over by JH.
1487 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1488 files if ZCAT_COMMAND is not executable.
1490 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1492 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1494 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1495 Normally benign, it bites when the pair was led to by a CNAME;
1496 modern usage is to not canonicalize the domain to a CNAME target
1497 (and we were inconsistent anyway for A-only vs AAAA+A).
1499 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1501 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1502 when evaluating $sender_host_dnssec.
1504 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1505 $sender_helo_dnssec variable.
1507 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1509 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1511 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1513 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1514 documented as working, but never had. Support all but $spam_report.
1516 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1517 added for tls authenticator.
1519 HS/03 Add perl_taintmode main config option
1525 TL/01 When running the test suite, the README says that variables such as
1526 no_msglog_check are global and can be placed anywhere in a specific
1527 test's script, however it was observed that placement needed to be near
1528 the beginning for it to behave that way. Changed the runtest perl
1529 script to read through the entire script once to detect and set these
1530 variables, reset to the beginning of the script, and then run through
1531 the script parsing/test process like normal.
1533 TL/02 The BSD's have an arc4random API. One of the functions to induce
1534 adding randomness was arc4random_stir(), but it has been removed in
1535 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1536 function when detected.
1538 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1539 cause callback expansion.
1541 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1542 syntax errors in an expansion can be treated as a string instead of
1543 logging or causing an error, due to the internal use of bool_lax
1544 instead of bool when processing it.
1546 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1547 server certificates when making smtp deliveries.
1549 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1551 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1553 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1555 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1556 Merged patch from Sebastian Wiedenroth.
1558 JH/05 Fix results-pipe from transport process. Several recipients, combined
1559 with certificate use, exposed issues where response data items split
1560 over buffer boundaries were not parsed properly. This eventually
1561 resulted in duplicates being sent. This issue only became common enough
1562 to notice due to the introduction of connection certificate information,
1563 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1565 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1566 size buffer was used, resulting in syntax errors when an expansion
1569 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1570 version 3.3.6 or later.
1572 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1573 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1574 both become "event_action", the variables become $event_name, $event_data
1575 and $event_defer_errno. There is a new variable $verify_mode, usable in
1576 routers, transports and related events. The tls:cert event is now also
1577 raised for inbound connections, if the main configuration event_action
1580 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1581 early OCSP support, but no stapling (appears to be less than 1.0.0).
1583 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1584 server certificate names available under the smtp transport option
1585 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1588 JH/10 Time-related extraction expansions from certificates now use the main
1589 option "timezone" setting for output formatting, and are consistent
1590 between OpenSSL and GnuTLS compilations. Bug 1541.
1592 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1593 encoded parameter in the incoming message. Bug 1558.
1595 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1596 include certificate info, eximon was claiming there were spoolfile
1599 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1601 JH/14 Log delivery-related information more consistently, using the sequence
1602 "H=<name> [<ip>]" wherever possible.
1604 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1605 are problematic for Debian distribution, omit them from the release
1608 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1610 JH/16 Fix string representation of time values on 64bit time_t architectures.
1613 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1614 output list separator was used.
1619 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1620 checkers that were complaining about end of non-void function with no
1623 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1624 This was a regression introduced in 4.83 by another bugfix.
1626 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1628 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1629 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1635 TF/01 Correctly close the server side of TLS when forking for delivery.
1637 When a message was received over SMTP with TLS, Exim failed to clear up
1638 the incoming connection properly after forking off the child process to
1639 deliver the message. In some situations the subsequent outgoing
1640 delivery connection happened to have the same fd number as the incoming
1641 connection previously had. Exim would try to use TLS and fail, logging
1642 a "Bad file descriptor" error.
1644 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1645 utilities have not been installed.
1647 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1648 temporary space as the ACL may create new global variables.
1650 TL/01 LDAP support uses per connection or global context settings, depending
1651 upon the detected version of the libraries at build time.
1653 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1654 to extract and use the src ip:port in logging and expansions as if it
1655 were a direct connection from the outside internet. PPv2 support was
1656 updated based on HAProxy spec change in May 2014.
1658 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1660 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1661 Properly escape header and check for NULL return.
1663 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1666 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1668 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1669 characters in header names, implemented as a verify condition.
1670 Contributed by Michael Fischer v. Mollard.
1672 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1673 results permerror and temperror. Previous values are deprecated but
1674 still accepted. In a future release, err_perm and err_temp will be
1675 completely removed, which will be a backward incompatibility if the
1676 ACL tests for either of these two old results. Patch contributed by
1677 user bes-internal on the mailing list.
1679 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1681 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1682 selectors, in both main and reject logs.
1684 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1687 JH/07 Add malware type "sock" for talking to simple daemon.
1689 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1691 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1692 routers/transports under cutthrough routing.
1694 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1695 numbers. Touch up "bool" conditional to keep the same definition.
1697 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1699 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1701 JH/12 Expand items in router/transport headers_add or headers_remove lists
1702 individually rather than the list as a whole. Bug 1452.
1704 Required for reasonable handling of multiple headers_ options when
1705 they may be empty; requires that headers_remove items with embedded
1706 colons must have them doubled (or the list-separator changed).
1708 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1709 view the policy declared in the DMARC record. Currently, $dmarc_status
1710 is a combined value of both the record presence and the result of the
1713 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1715 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1716 dnslookup router and the smtp transport (applying to the forward
1719 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1720 of ldap servers used for a specific lookup. Patch provided by Heiko
1723 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1724 New variable $lookup_dnssec_authenticated for observability.
1726 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1727 Patch submitted by Lars Timman.
1729 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1731 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1732 Requires trusted mode and valid format message id, aborts otherwise.
1733 Patch contributed by Heiko Schlichting.
1735 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1736 certextract with support for various fields. Bug 1358.
1738 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1739 is requested by default, modifiable by smtp transport option
1742 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1743 operate on certificate variables to give certificate fingerprints
1744 Also new ${sha256:cert_variable}.
1746 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1748 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1749 Christian Aistleitner.
1751 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1753 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1754 file. Patch from Wolfgang Breyha.
1756 JH/25 Expand the coverage of the delivery $host and $host_address to
1757 client authenticators run in verify callout. Bug 1476.
1759 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1760 align with daemon_smtp_ports. Bug 72.
1762 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1763 support and error reporting did not work properly.
1765 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1766 and is readable. Patch from Andrew Colin Kissa.
1768 TL/14 Enhance documentation of ${run expansion and how it parses the
1769 commandline after expansion, particularly in the case when an
1770 unquoted variable expansion results in an empty value.
1772 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1774 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1775 by renaming away from it. Addresses GH issue 15, reported by
1778 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1780 TL/15 SECURITY: prevent double expansion in math comparison functions
1781 (can expand unsanitized data). Not remotely exploitable.
1788 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1790 PP/02 Make -n do something, by making it not do something.
1791 When combined with -bP, the name of an option is not output.
1793 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1796 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1797 $sender_host_name and config options to manage this, and basic check
1800 PP/05 DSCP support for outbound connections and control modifier for inbound.
1802 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1803 (Only plugin which currently uses this is kerberos4, which nobody should
1804 be using, but we should make it available and other future plugins might
1805 conceivably use it, even though it would break NAT; stuff *should* be
1806 using channel bindings instead).
1808 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1809 name; added for Sendmail compatibility; requires admin caller.
1810 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1811 just ignore it); requires trusted caller.
1812 Also parse but ignore: -Ac -Am -X<logfile>
1815 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1817 TL/02 Add +smtp_confirmation as a default logging option.
1819 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1820 Patch by Magnus Holmgren from 2007-02-20.
1822 TL/04 Bugzilla 1281 - Spec typo.
1823 Bugzilla 1283 - Spec typo.
1824 Bugzilla 1290 - Spec grammar fixes.
1826 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1828 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1830 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1831 mailing list by by Dmitry Isaikin.
1833 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1835 JH/02 Support "G" suffix to numbers in ${if comparisons.
1837 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1839 NM/01 Bugzilla 1197 - Spec typo
1840 Bugzilla 1196 - Spec examples corrections
1842 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1844 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1845 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1848 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1849 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1851 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1852 "acl {{name}{arg}...}", and optional args on acl condition
1855 JH/05 Permit multiple router/transport headers_add/remove lines.
1857 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1859 JH/07 Avoid using a waiting database for a single-message-only transport.
1860 Performance patch from Paul Fisher. Bugzilla 1262.
1862 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1865 JH/09 Add $headers_added variable, with content from use of ACL modifier
1866 add_header (but not yet added to the message). Bugzilla 199.
1868 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1869 Pulled from Bugzilla 817 by Wolfgang Breyha.
1871 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1873 (nb: this is the same fix as in Exim 4.80.1)
1875 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1878 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1881 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1883 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1884 advertise SMTP AUTH mechanism to us, instead of a generic
1885 protocol violation error. Also, make Exim more robust to bad
1886 data from the Dovecot auth socket.
1888 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1890 When a queue runner is handling a message, Exim first routes the
1891 recipient addresses, during which it prunes them based on the retry
1892 hints database. After that it attempts to deliver the message to
1893 any remaining recipients. It then updates the hints database using
1896 So if a recipient address works intermittently, it can get repeatedly
1897 deferred at routing time. The retry hints record remains fresh so the
1898 address never reaches the final cutoff time.
1900 This is a fairly common occurrence when a user is bumping up against
1901 their storage quota. Exim had some logic in its local delivery code
1902 to deal with this. However it did not apply to per-recipient defers
1903 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1905 This change adds a proper retry rule check during routing so that the
1906 final cutoff time is checked against the message's age. We only do
1907 this check if there is an address retry record and there is not a
1908 domain retry record; this implies that previous attempts to handle
1909 the address had the retry_use_local_parts option turned on. We use
1910 this as an approximation for the destination being like a local
1911 delivery, as in LMTP.
1913 I suspect this new check makes the old local delivery cutoff check
1914 redundant, but I have not verified this so I left the code in place.
1916 TF/02 Correct gecos expansion when From: is a prefix of the username.
1918 Test 0254 submits a message to Exim with the header
1922 When I ran the test suite under the user fanf2, Exim expanded
1923 the header to contain my full name, whereas it should have added
1924 a Resent-Sender: header. It erroneously treats any prefix of the
1925 username as equal to the username.
1927 This change corrects that bug.
1929 GF/01 DCC debug and logging tidyup
1930 Error conditions log to paniclog rather than rejectlog.
1931 Debug lines prefixed by "DCC: " to remove any ambiguity.
1933 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1935 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1936 Bug spotted by Jeremy Harris; was flawed since initial commit.
1937 Would have resulted in OCSP responses post-SNI triggering an Exim
1938 NULL dereference and crash.
1940 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1942 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1943 Bug detection, analysis and fix by Samuel Thibault.
1944 Bugzilla 1331, Debian bug #698092.
1946 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1948 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1949 Server implementation by Todd Lyons, client by JH.
1950 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1951 config variable "prdr_enable" controls whether the server
1952 advertises the facility. If the client requests PRDR a new
1953 acl_data_smtp_prdr ACL is called once for each recipient, after
1954 the body content is received and before the acl_smtp_data ACL.
1955 The client is controlled by both of: a hosts_try_prdr option
1956 on the smtp transport, and the server advertisement.
1957 Default client logging of deliveries and rejections involving
1958 PRDR are flagged with the string "PRDR".
1960 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1961 fclose(). Diagnosis by Todd Lyons.
1963 PP/17 Update configure.default to handle IPv6 localhost better.
1964 Patch by Alain Williams (plus minor tweaks).
1967 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1968 This is now consistent with GnuTLS, and is now documented: the
1969 previous undocumented portable approach to treating the option as
1970 unset was to force an expansion failure. That still works, and
1971 an empty string is now equivalent.
1973 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1974 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1975 not performing validation itself.
1977 PP/20 Added force_command boolean option to pipe transport.
1978 Patch from Nick Koston, of cPanel Inc.
1980 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1983 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1985 PP/21 Fix eximon continuous updating with timestamped log-files.
1986 Broken in a format-string cleanup in 4.80, missed when I repaired the
1987 other false fix of the same issue.
1988 Report and fix from Heiko Schlichting.
1991 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1992 Report from Prashanth Katuri.
1994 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1995 It's SecureTransport, so affects any MacOS clients which use the
1996 system-integrated TLS libraries, including email clients.
1998 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1999 using a MIME ACL for non-SMTP local injection.
2000 Report and assistance in diagnosis by Warren Baker.
2002 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2004 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2006 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2007 last id that failed so it may be referenced in subsequent ACL's.
2009 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2012 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2013 ldap library initialization, allowing self-signed CA's to be
2014 used. Also properly sets require_cert option later in code by
2015 using NULL (global ldap config) instead of ldap handle (per
2016 session). Bug diagnosis and testing by alxgomz.
2018 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2019 the src/util/ subdirectory.
2021 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2022 renamed to Transport Post Delivery Action by Jeremy Harris, as
2025 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2026 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2027 redis_servers = needs to be configured which will be used by the redis
2028 lookup. Patch from Warren Baker, of The Packet Hub.
2030 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2032 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2033 hostname or reverse DNS when processing a host list. Used suggestions
2034 from multiple comments on this bug.
2036 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2038 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2039 Missed a few lines, added it to make the runtest require no keyboard
2042 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2043 contains upper case chars. Make router use caseful_local_part.
2045 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2046 support when GnuTLS has been built with p11-kit.
2052 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2054 This, or similar/improved, will also be change PP/11 of 4.82.
2060 PP/01 Handle short writes when writing local log-files.
2061 In practice, only affects FreeBSD (8 onwards).
2062 Bugzilla 1053, with thanks to Dmitry Isaikin.
2064 NM/01 Bugzilla 949 - Documentation tweak
2066 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2069 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2071 PP/02 Implemented gsasl authenticator.
2073 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2075 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2076 `pkg-config foo` for cflags/libs.
2078 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2079 with rest of GSASL and with heimdal_gssapi.
2081 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2082 `pkg-config foo` for cflags/libs for the TLS implementation.
2084 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2085 properties get this fed in as external SSF. A number of robustness
2086 and debugging improvements to the cyrus_sasl authenticator.
2088 PP/08 cyrus_sasl server now expands the server_realm option.
2090 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2091 Patch by Jeremy Harris.
2093 PP/10 Added dbmjz lookup type.
2095 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2097 PP/12 MAIL args handles TAB as well as SP, for better interop with
2098 non-compliant senders.
2099 Analysis and variant patch by Todd Lyons.
2101 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2102 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2103 Patch from Dirk Mueller <dmueller@suse.com>
2105 PP/13 tls_peerdn now print-escaped for spool files.
2106 Observed some $tls_peerdn in wild which contained \n, which resulted
2107 in spool file corruption.
2109 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2110 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2111 or write after TLS renegotiation, which otherwise led to messages
2114 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2115 as a tracking header (ie: a signed header comes before the signature).
2116 Patch from Wolfgang Breyha.
2118 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2119 comma-sep list; embedded commas doubled.
2121 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2123 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2125 Report and patch from Dmitry Banschikov.
2127 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2128 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2129 failures. We appear to now support TLS1.1+ with Exim.
2131 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2132 lets Exim select keys and certificates based upon TLS SNI from client.
2133 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2134 before an outbound SMTP session. New log_selector, +tls_sni.
2136 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2137 NULL dereference. Report and patch from Alun Jones.
2139 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2140 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2141 Not seeing resolver debug output on NetBSD, but suspect this is a
2142 resolver implementation change.
2144 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2145 Left warnings. Added "eximon gdb" invocation mode.
2147 PP/21 Defaulting "accept_8bitmime" to true, not false.
2149 PP/22 Added -bw for inetd wait mode support.
2151 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2152 locate the relevant includes and libraries. Made this the default.
2154 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2155 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2157 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2158 This may cause build issues on older platforms.
2160 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2161 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2162 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2163 Added SNI support via GnuTLS too.
2164 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2166 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2168 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2169 Applied second patch from Janne, implementing suggestion to default
2170 multiple-strings-in-record handling to match SPF spec.
2172 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2174 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2175 read-only, out of scope).
2176 Patch from Wolfgang Breyha, report from Stuart Northfield.
2178 PP/29 Fix three issues highlighted by clang analyser static analysis.
2179 Only crash-plausible issue would require the Cambridge-specific
2180 iplookup router and a misconfiguration.
2181 Report from Marcin Mirosław.
2183 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2185 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2186 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2187 As part of this, removing so much warning spew let me fix some minor
2188 real issues in debug logging.
2190 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2191 assignment on my part. Fixed.
2193 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2194 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2195 Janne Snabb (who went above and beyond: thank you).
2197 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2198 string otherwise requires a connection and a bunch more work and it's
2199 relatively easy to get wrong. Should also expose TLS library linkage
2202 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2203 64-bit ${eval} (JH/03).
2205 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2206 GNU libc to support some of the 64-bit stuff, should not lead to
2207 conflicts. Defined before os.h is pulled in, so if a given platform
2208 needs to override this, it can.
2210 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2211 protection layer was required, which is not implemented.
2212 Bugzilla 1254, patch from Wolfgang Breyha.
2214 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2215 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2216 tls_dhparam take prime identifiers. Also unbreak combination of
2217 OpenSSL+DH_params+TLSSNI.
2219 PP/39 Disable SSLv2 by default in OpenSSL support.
2225 PP/01 Solaris build fix for Oracle's LDAP libraries.
2226 Bugzilla 1109, patch from Stephen Usher.
2228 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2230 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2233 TF/02 Fix a couple more cases where we did not log the error message
2234 when unlink() failed. See also change 4.74-TF/03.
2236 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2237 lock up or crash if it happened to be inside a call to libc when it
2238 got a SIGUSR1 from exiwhat.
2240 The SIGUSR1 handler appends the current process status to the process
2241 log which is later printed by exiwhat. It used to use the general
2242 purpose logging code to do this, but several functions it calls are
2243 not safe for signals.
2245 The new output code in the SIGUSR1 handler is specific to the process
2246 log, and simple enough that it's easy to inspect for signal safety.
2247 Removing some special cases also simplifies the general logging code.
2248 Removing the spurious timestamps from the process log simplifies
2251 TF/04 Improved ratelimit ACL condition.
2253 The /noupdate option has been deprecated in favour of /readonly which
2254 has clearer semantics. The /leaky, /strict, and /readonly update modes
2255 are mutually exclusive. The update mode is no longer included in the
2256 database key; it just determines when the database is updated. (This
2257 means that when you upgrade Exim will forget old rate measurements.)
2259 Exim now checks that the per_* options are used with an update mode that
2260 makes sense for the current ACL. For example, when Exim is processing a
2261 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2262 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2263 must specify per_mail/readonly. If you omit the update mode it defaults to
2264 /leaky where that makes sense (as before) or /readonly where required.
2266 The /noupdate option is now undocumented but still supported for
2267 backwards compatibility. It is equivalent to /readonly except that in
2268 ACLs where /readonly is required you may specify /leaky/noupdate or
2269 /strict/noupdate which are treated the same as /readonly.
2271 A useful new feature is the /count= option. This is a generalization
2272 of the per_byte option, so that you can measure the throughput of other
2273 aggregate values. For example, the per_byte option is now equivalent
2274 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2276 The per_rcpt option has been generalized using the /count= mechanism
2277 (though it's more complicated than the per_byte equivalence). When it is
2278 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2279 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2280 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2281 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2282 Note that using per_rcpt with a non-readonly update mode in more than
2283 one ACL will cause the recipients to be double-counted. (The per_mail
2284 and per_byte options don't have this problem.)
2286 The handling of very low rates has changed slightly. If the computed rate
2287 is less than the event's count (usually one) then this event is the first
2288 after a long gap. In this case the rate is set to the same as this event's
2289 count, so that the first message of a spam run is counted properly.
2291 The major new feature is a mechanism for counting the rate of unique
2292 events. The new per_addr option counts the number of different
2293 recipients that someone has sent messages to in the last time period. It
2294 behaves like per_rcpt if all the recipient addresses are different, but
2295 duplicate recipient addresses do not increase the measured rate. Like
2296 the /count= option this is a general mechanism, so the per_addr option
2297 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2298 example, measure the rate that a client uses different sender addresses
2299 with the options per_mail/unique=$sender_address. There are further
2300 details in the main documentation.
2302 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2304 TF/06 Removed a few PCRE remnants.
2306 TF/07 Automatically extract Exim's version number from tags in the git
2307 repository when doing development or release builds.
2309 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2310 Bugzilla 879. Patch from Paul Fisher.
2312 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2313 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2316 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2318 PP/05 Variable $av_failed, true if the AV scanner deferred.
2319 Bugzilla 1078. Patch from John Horne.
2321 PP/06 Stop make process more reliably on build failure.
2322 Bugzilla 1087. Patch from Heiko Schlittermann.
2324 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2325 Bugzilla 1089. Patch from Heiko Schlittermann.
2327 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2328 Bugzilla 1131. Patch from Holger Weiß.
2330 PP/09 Handle IPv6 addresses with SPF.
2331 Bugzilla 860. Patch from Wolfgang Breyha.
2333 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2335 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2338 PP/11 match_* no longer expand right-hand-side by default.
2339 New compile-time build option, EXPAND_LISTMATCH_RHS.
2340 New expansion conditions, "inlist", "inlisti".
2342 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2344 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2346 PP/14 fix log_write() format string regression from TF/03.
2347 Bugzilla 1152. Patch from Dmitry Isaikin.
2353 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2355 PP/02 Harmonised TLS library version reporting; only show if debugging.
2356 Layout now matches that introduced for other libraries in 4.74 PP/03.
2358 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2360 PP/04 New "dns_use_edns0" global option.
2362 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2365 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2366 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2368 TK/01 Updated PolarSSL code to 0.14.2.
2369 Bugzilla 1097. Patch from Andreas Metzler.
2371 PP/07 Catch divide-by-zero in ${eval:...}.
2372 Fixes bugzilla 1102.
2374 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2377 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2378 format-string attack -- SECURITY: remote arbitrary code execution.
2380 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2381 time unintentionally subject to list matching rules, letting the header
2382 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2383 arbitrary string expansion). This allowed for information disclosure.
2385 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2386 INT_MIN/-1 -- value coerced to INT_MAX.
2392 NM/01 Workaround for PCRE version dependency in version reporting
2395 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2396 This fixes portability to compilers other than gcc, notably
2397 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2399 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2400 makefiles for portability to HP-UX and POSIX correctness.
2402 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2403 Also via indented variable definition in the Makefile.
2404 (Debugging by Oliver Heesakkers).
2406 PP/02 Restore caching of spamd results with expanded spamd_address.
2407 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2409 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2410 Improves build reliability. Fix from: Frank Elsner
2412 NM/02 Fix wide character breakage in the rfc2047 coding
2413 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2415 NM/03 Allow underscore in dnslist lookups
2416 Fixes bug 1026. Patch from Graeme Fowler
2418 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2419 Code patches from Adam Ciarcinski of NetBSD.
2421 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2424 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2425 is logged, to avoid truncation. Patch from John Horne.
2427 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2428 Patch from Jakob Hirsch.
2430 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2431 SQL string expansion failure details.
2432 Patch from Andrey Oktyabrski.
2434 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2435 Patch from Simon Arlott.
2437 PP/09 New lookups functionality failed to compile on old gcc which rejects
2438 extern declarations in function scope.
2439 Patch from Oliver Fleischmann
2441 PP/10 Use sig_atomic_t for flags set from signal handlers.
2442 Check getgroups() return and improve debugging.
2443 Fixed developed for diagnosis in bug 927 (which turned out to be
2446 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2447 Patch from Mark Zealey.
2449 PP/12 Bugzilla 1056: Improved spamd server selection.
2450 Patch from Mark Zealey.
2452 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2453 Based on patch from Heiko Schlittermann.
2455 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2456 Patch from Uwe Doering, sign-off by Michael Haardt.
2458 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2459 variable declaration deep within a block. Bug and patch from
2462 PP/15 lookups-Makefile IRIX compatibility coercion.
2464 PP/16 Make DISABLE_DKIM build knob functional.
2466 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2467 Patch by Simon Arlott
2469 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2470 variable argument macros. Our copy now differs from upstream.
2476 TF/01 Failure to get a lock on a hints database can have serious
2477 consequences so log it to the panic log.
2479 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2480 controlled using the smtp_confirmation log selector.
2482 TF/03 Include the error message when we fail to unlink a spool file.
2484 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2485 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2486 for maintaining out-of-tree patches for some time.
2488 PP/01 Bugzilla 139: Documentation and portability issues.
2489 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2490 Handle per-OS dynamic-module compilation flags.
2492 PP/02 Let /dev/null have normal permissions.
2493 The 4.73 fixes were a little too stringent and complained about the
2494 permissions on /dev/null. Exempt it from some checks.
2495 Reported by Andreas M. Kirchwitz.
2497 PP/03 Report version information for many libraries, including
2498 Exim version information for dynamically loaded libraries. Created
2499 version.h, now support a version extension string for distributors
2500 who patch heavily. Dynamic module ABI change.
2502 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2503 privilege escalation vulnerability whereby the Exim run-time user
2504 can cause root to append content of the attacker's choosing to
2507 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2510 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2511 If dropping privileges for untrusted macros, we disabled normal logging
2512 on the basis that it would fail; for the Exim run-time user, this is not
2513 the case, and it resulted in successful deliveries going unlogged.
2514 Fixed. Reported by Andreas Metzler.
2520 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2521 only prepend for the Resent-* case. Fixes regression introduced in
2522 Exim 4.70 by NM/22 for Bugzilla 607.
2524 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2525 increasing numbers of administrators be bitten by this.
2527 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2529 PP/03 Bugzilla 994: added openssl_options main configuration option.
2531 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2533 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2535 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2537 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2538 without a peer certificate, leading to a segfault because of an
2539 assumption that peers always have certificates. Be a little more
2540 paranoid. Problem reported by Martin Tscholak.
2542 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2543 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2544 NB: ClamAV planning to remove STREAM in "middle of 2010".
2545 CL also introduces -bmalware, various -d+acl logging additions and
2546 more caution in buffer sizes.
2548 PP/09 Implemented reverse_ip expansion operator.
2550 PP/10 Bugzilla 937: provide a "debug" ACL control.
2552 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2554 PP/12 Bugzilla 973: Implement --version.
2556 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2558 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2560 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2562 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2563 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2564 ignore trailing whitespace.
2566 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2568 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2571 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2572 Notification from Dr Andrew Aitchison.
2574 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2575 ExtendedDetectionInfo response format.
2576 Notification from John Horne.
2578 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2581 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2582 XSL and documented dependency on system catalogs, with examples of how
2585 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2588 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2589 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2590 configuration file which is writeable by the Exim user or group.
2592 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2593 of configuration files to cover files specified with the -C option if
2594 they are going to be used with root privileges, not just the default
2597 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2598 option (effectively making it always true).
2600 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2601 files to be used while preserving root privileges.
2603 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2604 that rogue child processes cannot use them.
2606 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2607 run-time user, instead of root.
2609 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2610 Exim run-time user without dropping privileges.
2612 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2613 result string, instead of calling string_vformat() twice with the same
2616 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2617 for other users. Others should always drop root privileges if they use
2618 -C on the command line, even for a whitelisted configure file.
2620 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2622 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2628 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2629 $data_path, and $header_path variables; fixed documentation bugs and
2632 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2633 exipick to access non-standard spools, including the "frozen" queue
2636 NM/01 Bugzilla 965: Support mysql stored procedures.
2637 Patch from Alain Williams
2639 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2641 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2642 Patch from Andreas Metzler
2644 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2645 Patch from Kirill Miazine
2647 NM/05 Bugzilla 671: Added umask to procmail example.
2649 JJ/03 installed exipick 20100323.0, fixing doc bug
2651 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2652 directory. Notification and patch from Dan Rosenberg.
2654 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2656 TK/02 Improve log output when DKIM signing operation fails.
2658 MH/01 Treat the transport option dkim_domain as a colon separated
2659 list, not as a single string, and sign the message with each element,
2660 omitting multiple occurences of the same signer.
2662 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2663 Bugzilla 985, 986. Patch by Simon Arlott
2665 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2666 Patch by Simon Arlott
2668 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2669 MBX locking. Notification from Dan Rosenberg.
2675 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2677 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2679 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2681 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2683 NM/04 Bugzilla 915: Flag broken perl installation during build.
2689 TK/01 Added patch by Johannes Berg that expands the main option
2690 "spamd_address" if it starts with a dollar sign.
2692 TK/02 Write list of recipients to X-Envelope-Sender header when building
2693 the mbox-format spool file for content scanning (suggested by Jakob
2696 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2697 (http://www.dcc-servers.net/) support via dccifd. Activated by
2698 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2700 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2701 by Mark Daniel Reidel <mr@df.eu>.
2703 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2704 When building exim an external PCRE library is now needed -
2705 PCRE is a system library on the majority of modern systems.
2706 See entry on PCRE_LIBS in EDITME file.
2708 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2709 conversation. Added nologin parameter to request.
2710 Patch contributed by Kirill Miazine.
2712 TF/01 Do not log submission mode rewrites if they do not change the address.
2714 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2716 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2717 log files in place. Contributed by Roberto Lima.
2719 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2721 TF/03 Bugzilla 615: When checking the local_parts router precondition
2722 after a local_part_suffix or local_part_prefix option, Exim now
2723 does not use the address's named list lookup cache, since this
2724 contains cached lookups for the whole local part.
2726 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2727 Robert Millan. Documentation is in experimental-spec.txt.
2729 TF/04 Bugzilla 668: Fix parallel build (make -j).
2731 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2733 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2734 Patch provided by Jan Srzednicki.
2736 TF/05 Leading white space used to be stripped from $spam_report which
2737 wrecked the formatting. Now it is preserved.
2739 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2740 that they are available at delivery time.
2742 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2744 TF/08 TLS error reporting now respects the incoming_interface and
2745 incoming_port log selectors.
2747 TF/09 Produce a more useful error message if an SMTP transport's hosts
2748 setting expands to an empty string.
2750 NM/06 Bugzilla 744: EXPN did not work under TLS.
2751 Patch provided by Phil Pennock.
2753 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2754 Patch provided by Richard Godbee.
2756 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2757 acl_smtp_notquit, added index entry.
2759 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2760 Patch provided by Eugene Bujak.
2762 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2763 accept(). Patch provided by Maxim Dounin.
2765 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2766 Patch provided by Phil Pennock.
2768 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2770 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2771 Patch provided by Brad "anomie" Jorsch.
2773 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2774 Patch provided by Dean Brooks.
2776 TK/05 Add native DKIM support (does not depend on external libraries).
2778 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2779 Patch provided by Graeme Fowler.
2781 NM/16 Bugzilla 851: Documentation example syntax fix.
2783 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2785 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2788 NM/19 Bugzilla 745: TLS version reporting.
2789 Patch provided by Phil Pennock.
2791 NM/20 Bugzilla 167: bool: condition support.
2792 Patch provided by Phil Pennock.
2794 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2795 clients. Patch provided by Phil Pennock.
2797 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2798 Patch provided by Brad "anomie" Jorsch.
2800 NM/23 Bugzilla 687: Fix misparses in eximstats.
2801 Patch provided by Heiko Schlittermann.
2803 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2804 Patch provided by Heiko Schlittermann.
2806 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2807 plus update to original patch.
2809 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2811 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2812 Patch provided by David Brownlee.
2814 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2816 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2818 NM/30 Bugzilla 888: TLS documentation bugfixes.
2820 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2822 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2823 Unlike the original bugzilla I have changed all shell scripts in src tree.
2825 NM/33 Bugzilla 898: Transport filter timeout fix.
2826 Patch by Todd Rinaldo.
2828 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2829 Patch by Serge Demonchaux.
2831 NM/35 Bugzilla 39: Base64 decode bug fixes.
2832 Patch by Jakob Hirsch.
2834 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2836 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2838 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2840 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2846 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2847 ALT-N's libdkim that I have put here:
2848 http://duncanthrax.net/exim-experimental/
2850 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2851 were called 'true' and it seems that C99 defines that as a reserved
2852 keyword to be used with 'bool' variable types. That means you could
2853 not include C99-style headers which use bools without triggering
2854 build errors in sieve.c.
2856 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2857 as mailq or other aliases. Changed the --help handling significantly
2858 to do whats expected. exim_usage() emits usage/help information.
2860 SC/01 Added the -bylocaldomain option to eximstats.
2862 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2864 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2866 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2872 PH/01 Another patch from the Sieve maintainer.
2874 PH/02 When an IPv6 address is converted to a string for single-key lookup
2875 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2876 dots are used instead of colons so that keys in lsearch files need not
2877 contain colons. This was done some time before quoting was made available
2878 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2879 (notated using the quote facility) so as to distinguish them from IPv4
2880 keys. This meant that lookups for IP addresses in host lists did not work
2881 for iplsearch lookups.
2883 This has been fixed by arranging for IPv6 addresses to be expressed with
2884 colons if the lookup type is iplsearch. This is not incompatible, because
2885 previously such lookups could never work.
2887 The situation is now rather anomalous, since one *can* have colons in
2888 ordinary lsearch keys. However, making the change in all cases is
2889 incompatible and would probably break a number of configurations.
2891 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2894 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2895 conversion specification without a maximum field width, thereby enabling
2896 a rogue spamd server to cause a buffer overflow. While nobody in their
2897 right mind would setup Exim to query an untrusted spamd server, an
2898 attacker that gains access to a server running spamd could potentially
2899 exploit this vulnerability to run arbitrary code as the Exim user.
2901 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2902 $primary_hostname instead of what libspf2 thinks the hosts name is.
2904 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2905 a directory entry by the name of the lookup key. Previously, if a
2906 symlink pointed to a non-existing file or a file in a directory that
2907 Exim lacked permissions to read, a lookup for a key matching that
2908 symlink would fail. Now it is enough that a matching directory entry
2909 exists, symlink or not. (Bugzilla 503.)
2911 PH/03 The body_linecount and body_zerocount variables are now exported in the
2914 PH/04 Added the $dnslist_matched variable.
2916 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2917 This means they are set thereafter only if the connection becomes
2920 PH/06 Added the client_condition to authenticators so that some can be skipped
2921 by clients under certain conditions.
2923 PH/07 The error message for a badly-placed control=no_multiline_responses left
2924 "_responses" off the end of the name.
2926 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2928 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2929 (without spaces) instead of just copying the configuration text.
2931 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2933 PH/11 Added $max_received_linelength.
2935 PH/12 Added +ignore_defer and +include_defer to host lists.
2937 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2938 way in which PCRE > 7.0 is built.
2940 PH/14 Implemented queue_only_load_latch.
2942 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2943 MAIL command. The effect was to mangle the value on 64-bit systems.
2945 PH/16 Another patch from the Sieve maintainer.
2947 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2949 PH/18 If a system quota error occurred while trying to create the file for
2950 a maildir delivery, the message "Mailbox is full" was not appended to the
2951 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2952 only to a quota excession during the actual writing of the file.
2954 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2955 characters?) which causes problems in log lines. The DN values are now
2956 passed through string_printing() before being added to log lines.
2958 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2959 and InterBase are left for another time.)
2961 PH/21 Added message_body_newlines option.
2963 PH/22 Guard against possible overflow in moan_check_errorcopy().
2965 PH/23 POSIX allows open() to be a macro; guard against that.
2967 PH/24 If the recipient of an error message contained an @ in the local part
2968 (suitably quoted, of course), incorrect values were put in $domain and
2969 $local_part during the evaluation of errors_copy.
2975 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2976 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2979 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2980 issue a MAIL command.
2982 PH/02 In an ACL statement such as
2984 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2986 if a client was not listed at all, or was listed with a value other than
2987 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2988 the condition was not true (as it should be), so access was not denied.
2989 The bug was that the ! inversion was incorrectly passed on to the second
2990 item. This has been fixed.
2992 PH/03 Added additional dnslists conditions == and =& which are different from
2993 = and & when the dns lookup returns more than one IP address.
2995 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2996 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2998 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2999 FSYNC, which compiles an option called disable_fsync that allows for
3000 bypassing fsync(). The documentation is heavily laced with warnings.
3002 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3004 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3005 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3006 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3007 including adding "make clean"; (3) Added -fPIC when compiling the test
3008 dynamically loaded module, to get rid of a warning.
3010 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3011 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3012 = 0s. The bug is otherwise harmless.
3014 PH/07 There was a bug in the dovecot authenticator such that the value of
3015 $auth1 could be overwritten, and so not correctly preserved, after a
3016 successful authentication. This usually meant that the value preserved by
3017 the server_setid option was incorrect.
3019 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3021 PH/09 Installed PCRE release 7.0.
3023 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3024 run for batched SMTP input. It is now run at the start of every message
3025 in the batch. While fixing this I discovered that the process information
3026 (output by running exiwhat) was not always getting set for -bs and -bS
3027 input. This is fixed, and it now also says "batched" for BSMTP.
3029 PH/11 Added control=no_pipelining.
3031 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3032 patch, slightly modified), and move the expansion of helo_data till after
3033 the connection is made in the smtp transport (so it can use these
3036 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3038 PH/14 Added log_selector = +pid.
3040 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3042 PH/16 Add ${if forany and ${if forall.
3044 PH/17 Added dsn_from option to vary the From: line in DSNs.
3046 PH/18 Flush SMTP output before performing a callout, unless control =
3047 no_callout_flush is set.
3049 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3050 was true (the default) a successful delivery failed to delete the retry
3051 item, thus causing premature timeout of the address. The bug is now
3054 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3056 PH/21 Long custom messages for fakedefer and fakereject are now split up
3057 into multiline responses in the same way that messages for "deny" and
3058 other ACL rejections are.
3060 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3061 with slight modification.
3063 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3064 draft, changing the syntax and factoring some duplicate code".
3066 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3067 for deliveries of the second and subsequent messages over the same SMTP
3070 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3071 ${reduce, with only minor "tidies".
3073 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3075 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3076 expansion side effects.
3078 PH/27 When a message times out after an over-quota error from an Exim-imposed
3079 quota, the bounce message says "mailbox is full". This message was not
3080 being given when it was a system quota that was exceeded. It now should
3083 MH/03 Made $recipients available in local_scan(). local_scan() already has
3084 better access to the recipient list through recipients_list[], but
3085 $recipients can be useful in postmaster-provided expansion strings.
3087 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3088 in the case of a MAIL command with additional options following the
3089 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3090 were accidentally chopped off.
3092 PH/29 SMTP synchronization checks are implemented when a command is read -
3093 there is a check that no more input is waiting when there shouldn't be
3094 any. However, for some commands, a delay in an ACL can mean that it is
3095 some time before the response is written. In this time, more input might
3096 arrive, invalidly. So now there are extra checks after an ACL has run for
3097 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3098 pipelining has not been advertised.
3100 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3102 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3103 returned in the SMTP response when smtp_return_error_details was false.
3104 This has been fixed.
3106 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3107 instead of the C I/O that was originally supplied, because problems were
3108 reported on Solaris.
3110 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3111 Exim which did not show up earlier: it was assuming that a call to
3112 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3113 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3114 was a macro that became an assignment, so it seemed to work. This has
3115 changed to a proper function call with a void return, hence the compile
3116 error. Exim's code has been fixed.
3118 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3121 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3123 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3125 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3126 because it turns out to be too big or there is a timeout) there is no
3127 "Completed" line in the log. When some messages of this type were
3128 selected by exigrep, they were listed as "not completed". Others were
3129 picked up by some special patterns. I have improved the selection
3130 criteria to be more general.
3132 PH/38 The host_find_failed option in the manualroute router can now be set
3133 to "ignore", to completely ignore a host whose IP address cannot be
3134 found. If all hosts are ignored, the behaviour is controlled by the new
3135 host_all_ignored option.
3137 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3138 homing or because of multiple MX records with /mx) generated more than
3139 one IP address, and the following item turned out to be the local host,
3140 all the secondary addresses of the first item were incorrectly removed
3141 from the list, along with the local host and any following hosts (which
3142 is what is supposed to happen).
3144 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3145 whoever called Exim into the -H file. In the case of the daemon it was
3146 behaving confusingly. When first started, it used values for whoever
3147 started the daemon, but after a SIGHUP it used the Exim user (because it
3148 calls itself on a restart). I have changed the code so that it now always
3151 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3152 message are rejected with the same error (e.g. no authentication or bad
3153 sender address), and a DATA command is nevertheless sent (as can happen
3154 with PIPELINING or a stupid MUA), the error message that was given to the
3155 RCPT commands is included in the rejection of the DATA command. This is
3156 intended to be helpful for MUAs that show only the final error to their
3159 PH/42 Another patch from the Sieve maintainer.
3161 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3162 Eximstats - Fixed some broken HTML links and added missing column headers
3164 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3165 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3167 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3169 PH/43 Yet another patch from the Sieve maintainer.
3171 PH/44 I found a way to check for a TCP/IP connection going away before sending
3172 the response to the final '.' that terminates a message, but only in the
3173 case where the client has not sent further data following the '.'
3174 (unfortunately, this is allowed). However, in many cases there won't be
3175 any further data because there won't be any more messages to send. A call
3176 to select() can be used: if it shows that the input is "ready", there is
3177 either input waiting, or the socket has been closed. An attempt to read
3178 the next input character can distinguish the two cases. Previously, Exim
3179 would have sent an OK response which the client would never have see.
3180 This could lead to message repetition. This fix should cure that, at
3181 least in a lot of common cases.
3183 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3184 advertised in response to EHLO.
3190 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3191 fixed by 4.65/MH/01 (is this a record?) are fixed:
3193 (i) An empty string was always treated as zero by the numeric comparison
3194 operators. This behaviour has been restored.
3196 (ii) It is documented that the numeric comparison operators always treat
3197 their arguments as decimal numbers. This was broken in that numbers
3198 starting with 0 were being interpreted as octal.
3200 While fixing these problems I realized that there was another issue that
3201 hadn't been noticed. Values of message_size_limit (both the global option
3202 and the transport option) were treated as octal if they started with 0.
3203 The documentation was vague. These values are now always treated as
3204 decimal, and I will make that clear in the documentation.
3210 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3211 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3214 MH/01 Don't check that the operands of numeric comparison operators are
3215 integers when their expansion is in "skipping" mode (fixes bug
3216 introduced by 4.64-PH/07).
3218 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3219 child addresses, Exim now panics and dies. Previously, because the count
3220 is held in a short int, deliveries were likely to be lost. As such a
3221 large number of recipients for a single message is ridiculous
3222 (performance will be very, very poor), I have chosen to impose a limit
3223 rather than extend the field.
3229 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3230 leftover -K file (the existence of which was triggered by #402).
3231 While we were at it, introduced process PID as part of the -K
3232 filename. This should rule out race conditions when creating
3235 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3236 processing considerably. Previous code took too long for large mails,
3237 triggering a timeout which in turn triggers #401.
3239 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3240 in the DK code in transports.c. sendfile() is not really portable,
3241 hence the _LINUX specificness.
3243 TF/01 In the add_headers option to the mail command in an Exim filter,
3244 there was a bug that Exim would claim a syntax error in any
3245 header after the first one which had an odd number of characters
3248 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3249 callout verification, Exim cached a "reject" for the entire domain. This
3250 is correct for most verifications, but it is not correct for a recipient
3251 verification with use_sender or use_postmaster set, because in that case
3252 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3253 case of MAIL FROM:<> rejection from other early rejections (e.g.
3254 rejection of HELO). When verifying a recipient using a non-null MAIL
3255 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3256 Whatever the result of the callout, the value of the domain cache is
3257 left unchanged (for any other kind of callout, getting as far as trying
3258 RCPT means that the domain itself is ok).
3260 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3263 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3264 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3265 session, the daemon ignores ECONNECT errors and logs others; it now
3266 ignores EPIPE as well.
3268 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3269 (quoted-printable decoding).
3271 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3272 later the small subsequent patch to fix an introduced bug.
3274 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3276 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3278 PH/08 An error is now given if message_size_limit is specified negative.
3280 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3281 to be given (somewhat) arbitrary names.
3283 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3286 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3287 miscellaneous code fixes
3289 PH/10 Added the log_reject_target ACL modifier to specify where to log
3292 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3293 hostname. This is wrong, because it relates to the incoming message (and
3294 probably the interface on which it is arriving) and not to the outgoing
3295 callout (which could be using a different interface). This has been
3296 changed to use the value of the helo_data option from the smtp transport
3297 instead - this is what is used when a message is actually being sent. If
3298 there is no remote transport (possible with a router that sets up host
3299 addresses), $smtp_active_hostname is used.
3301 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3302 tweaks were necessary in order to get it to work (see also 21 below):
3303 (a) The code assumed that strncpy() returns a negative number on buffer
3304 overflow, which isn't the case. Replaced with Exim's string_format()
3306 (b) There were several signed/unsigned issues. I just did the minimum
3307 hacking in of casts. There is scope for a larger refactoring.
3308 (c) The code used strcasecmp() which is not a standard C function.
3309 Replaced with Exim's strcmpic() function.
3310 (d) The code set only $1; it now sets $auth1 as well.
3311 (e) A simple test gave the error "authentication client didn't specify
3312 service in request". It would seem that Dovecot has changed its
3313 interface. Fortunately there's a specification; I followed it and
3314 changed what the client sends and it appears to be working now.
3316 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3319 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3320 address A is aliased to B and C, where B exists and C does not. Without
3321 -v the output is "A verified" because verification stops after a
3322 successful redirection if more than one address is generated. However,
3323 with -v the child addresses are also verified. Exim was outputting "A
3324 failed to verify" and then showing the successful verification for C,
3325 with its parentage. It now outputs "B failed to verify", showing B's
3326 parentage before showing the successful verification of C.
3328 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3329 look up a TXT record in a specific list after matching in a combined
3332 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3333 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3334 they consult the DNS. I had assumed they would set it the way they
3335 wanted; and indeed my experiments on Linux seem to show that in some
3336 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3337 To be on the safe side, however, I have now made the interface to
3338 host_find_byname() similar to host_find_bydns(), with an argument
3339 containing the DNS resolver options. The host_find_byname() function now
3340 sets these options at its start, just as host_find_bydns() does. The smtp
3341 transport options dns_qualify_single and dns_search_parents are passed to
3342 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3343 of host_find_byname() use the default settings of RES_DEFNAMES
3344 (qualify_single) but not RES_DNSRCH (search_parents).
3346 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3347 spool_read_header() do less string testing, by means of a preliminary
3348 switch on the second character of optional "-foo" lines. (This is
3349 overdue, caused by the large number of possibilities that now exist.
3350 Originally there were few.) While I was there, I also converted the
3351 str(n)cmp tests so they don't re-test the leading "-" and the first
3352 character, in the hope this might squeeze out yet more improvement.
3354 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3355 flag allowing group syntax was set by the header_syntax check but not
3356 turned off, possible causing trouble later; (2) The flag was not being
3357 set at all for the header_verify test, causing "group"-style headers to
3358 be rejected. I have now set it in this case, and also caused header_
3359 verify to ignore an empty address taken from a group. While doing this, I
3360 came across some other cases where the code for allowing group syntax
3361 while scanning a header line wasn't quite right (mostly, not resetting
3362 the flag correctly in the right place). These bugs could have caused
3363 trouble for malformed header lines. I hope it is now all correct.
3365 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3366 with the "reply" argument non-NULL. The code, however (which originally
3367 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3368 but it didn't always do it. This confused somebody who was copying the
3369 code for some other use. I have removed all the tests.
3371 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3372 feature that was used to support insecure browsers during the U.S. crypto
3373 embargo. It requires special client support, and Exim is probably the
3374 only MTA that supported it -- and would never use it because real RSA is
3375 always available. This code has been removed, because it had the bad
3376 effect of slowing Exim down by computing (never used) parameters for the
3377 RSA_EXPORT functionality.
3379 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3380 authenticator to fail if there's a tab character in the incoming data
3381 (there should never be unless someone is messing about, as it's supposed
3382 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3383 if the connection is using TLS or if the remote IP is the same as the
3384 local IP, and the "valid-client-cert option" if a client certificate has
3387 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3388 authenticators. This can be used for authorization after authentication
3389 succeeds. (In the case of plaintext, it servers for both authentication
3392 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3393 if any retry times were supplied.
3395 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3396 connection, where there is no client IP address to check. In this
3397 situation, the verify now always succeeds.
3399 PH/25 Applied John Jetmore's -Mset patch.
3401 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3403 PH/27 In a string expansion for a processed (not raw) header when multiple
3404 headers of the same name were present, leading whitespace was being
3405 removed from all of them, but trailing whitespace was being removed only
3406 from the last one. Now trailing whitespace is removed from each header
3407 before concatenation. Completely empty headers in a concatenation (as
3408 before) are ignored.
3410 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3411 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3413 PH/29 [Removed. This was a change that I later backed out, and forgot to
3414 correct the ChangeLog entry (that I had efficiently created) before
3415 committing the later change.]
3417 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3418 address errors (4xx response to RCPT) over the same connection as other
3419 messages routed to the same hosts. Such deliveries are always "forced",
3420 so retry times are not inspected. This resulted in far too many retries
3421 for the affected addresses. The effect occurred only when there were more
3422 hosts than the hosts_max_try setting in the smtp transport when it had
3423 the 4xx errors. Those hosts that it had tried were not added to the list
3424 of hosts for which the message was waiting, so if all were tried, there
3425 was no problem. Two fixes have been applied:
3427 (i) If there are any address or message errors in an SMTP delivery, none
3428 of the hosts (tried or untried) are now added to the list of hosts
3429 for which the message is waiting, so the message should not be a
3430 candidate for sending over the same connection that was used for a
3431 successful delivery of some other message. This seems entirely
3432 reasonable: after all the message is NOT "waiting for some host".
3433 This is so "obvious" that I'm not sure why it wasn't done
3434 previously. Hope I haven't missed anything, but it can't do any
3435 harm, as the worst effect is to miss an optimization.
3437 (ii) If, despite (i), such a delivery is accidentally attempted, the
3438 routing retry time is respected, so at least it doesn't keep
3439 hammering the server.
3441 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3442 in ${readsocket because some servers need this prod.
3444 PH/32 Added some extra debug output when updating a wait-xxx database.
3446 PH/33 The hint "could be header name not terminated by colon", which has been
3447 given for certain expansion errors for a long time, was not being given
3448 for the ${if def:h_colon_omitted{... case.
3450 PH/34 The spec says: "With one important exception, whenever a domain list is
3451 being scanned, $domain contains the subject domain." There was at least
3452 one case where this was not true.
3454 PH/35 The error "getsockname() failed: connection reset by peer" was being
3455 written to the panic log as well as the main log, but it isn't really
3456 panic-worthy as it just means the connection died rather early on. I have
3457 removed the panic log writing for the ECONNRESET error when getsockname()
3460 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3461 runs only) independently of the message's sender address. This meant
3462 that, if the 4xx error was in fact related to the sender, a different
3463 message to the same recipient with a different sender could confuse
3464 things. In particular, this can happen when sending to a greylisting
3465 server, but other circumstances could also provoke similar problems.
3466 I have changed the default so that the retry time for these errors is now
3467 based a combination of the sender and recipient addresses. This change
3468 can be overridden by setting address_retry_include_sender=false in the
3471 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3472 remote server are returned as part of bounce messages. This was not
3473 happening for LMTP over a pipe (the lmtp transport), but now it is the
3474 same for both kinds of LMTP.
3476 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3477 in header lines that were in fact CNAMEs. This is no longer the case.
3479 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3480 and queue runs started by the daemon processed all messages. This has
3481 been fixed so that -R and -S can now usefully be given with -q<time>.
3483 PH/40 Import PCRE release 6.7 (fixes some bugs).
3485 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3487 PH/42 Give an error if -q is specified more than once.
3489 PH/43 Renamed the variables $interface_address and $interface_port as
3490 $received_ip_address and $received_port, to make it clear that these
3491 values apply to message reception, and not to the outgoing interface when
3492 a message is delivered. (The old names remain recognized, of course.)
3494 PH/44 There was no timeout on the connect() call when using a Unix domain
3495 socket in the ${readsocket expansion. There now is.
3497 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3498 be meaningful with "accept".
3500 SC/01 Eximstats V1.43
3501 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3503 SC/02 Eximstats V1.44
3504 Use a glob alias rather than an array ref in the generated
3505 parser. This improves both readability and performance.
3507 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3508 Collect SpamAssassin and rejection statistics.
3509 Don't display local sender or destination tables unless
3510 there is data to show.
3511 Added average volumes into the top table text output.
3513 SC/04 Eximstats V1.46
3514 Collect data on the number of addresses (recipients)
3515 as well as the number of messages.
3517 SC/05 Eximstats V1.47
3518 Added 'Message too big' to the list of mail rejection
3519 reasons (thanks to Marco Gaiarin).
3521 SC/06 Eximstats V1.48
3522 Mainlog lines which have GMT offsets and are too short to
3523 have a flag are now skipped.
3525 SC/07 Eximstats V1.49 (Alain Williams)
3526 Added the -emptyok flag.
3528 SC/08 Eximstats V1.50
3529 Fixes for obtaining the IP address from reject messages.
3531 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3532 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3533 whitespace changes from 4.64-PH/27
3535 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3538 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3539 are found, allow negative numbers in numeric criteria)
3541 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3543 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3544 to match changes made in 4.64-PH/43
3546 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3548 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3549 "domain" is actually an IP address, and force a failure. This locks out
3550 those revolvers/nameservers that support "A-for-A" lookups, in
3551 contravention of the specifications.
3553 PH/48 When a host name was looked up from an IP address, and the subsequent
3554 forward lookup of the name timed out, the host name was left in
3555 $sender_host_name, contrary to the specification.
3557 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3558 restricted to single-key lookups, Exim was not diagnosing an error if
3559 * or *@ was used with a query-style lookup.
3561 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3563 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3564 long ago, but noone interested enough thought of it. Let's just say that
3565 the "1.1" means that there are some new functions that weren't there at
3566 some point in the past.
3568 PH/51 Error processing for expansion failure of helo_data from an smtp
3569 transport during callout processing was broken.
3571 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3572 tested/used via the -bh/-bhc/-bs options.
3574 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3575 bug, fixed in subsequent PCRE releases).
3577 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3578 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3580 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3586 SC/01 Use a glob alias rather than an array ref in eximstats generated
3587 parser. This improves both readability and performance.
3589 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3590 Don't display local sender or destination tables in eximstats unless
3591 there is data to show.
3592 Added average volumes into the eximstats top table text output.
3594 SC/03 Collect data on the number of addresses (recipients) as well
3595 as the number of messages in eximstats.
3597 TF/01 Correct an error in the documentation for the redirect router. Exim
3598 does (usually) call initgroups() when daemonizing.
3600 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3601 with consistent privilege compared to when running as a daemon.
3603 TF/03 Note in the spec that $authenticated_id is not set for local
3604 submissions from trusted users.
3606 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3607 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3609 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3610 by adding some example configuration directives to the default
3611 configuration file. A little bit of work is required to uncomment the
3612 directives and define how usernames and passwords are checked, but
3613 there is now a framework to start from.
3615 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3616 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3617 without this. I don't know how relevant this is to other LDAP libraries.
3619 PH/02 Add the verb name to the "unknown ACL verb" error.
3621 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3623 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3625 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3626 directory not expanded when it should be if an expanded home directory
3627 was set for the address (which is overridden by the transport).
3629 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3632 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3633 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3634 because it is too late at that time, and has no effect.
3636 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3637 security issue with \' (bugzilla #107). I could not use the
3638 PQescapeStringConn() function, because it needs a PGconn value as one of
3641 PH/08 When testing addresses using -bt, indicate those final addresses that
3642 are duplicates that would not cause an additional delivery. At least one
3643 person was confused, thinking that -bt output corresponded to deliveries.
3644 (Suppressing duplicates isn't a good idea as you lose the information
3645 about possibly different redirections that led to the duplicates.)
3647 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3648 systems where poll() doesn't work, in particular OS X.
3650 PH/10 Added more information to debugging output for retry time not reached.
3652 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3653 operations in malware.c.
3655 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3658 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3659 syslog_duplication set false, log lines that would normally be written
3660 both the the main log and to the reject log were not written to syslog at
3663 PH/14 In the default configuration, change the use of "message" in ACL warn
3664 statements to "add_header".
3666 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3667 not followed by a command (e.g. "seen endif").
3669 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3670 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3673 PH/17 Added extra conditions to the default value of delay_warning_condition
3677 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3678 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3679 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3682 The Auto-Submitted: and various List- headers are standardised, whereas I
3683 don't think Precedence: ever was.
3685 PH/18 Refactored debugging code in route_finduser() to show more information,
3686 in particular, the error code if getpwnam() issues one.
3688 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3689 This is apparently needed in addition to the PH/07 change above to avoid
3690 any possible encoding problems.
3692 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3693 but not after initializing Perl.
3695 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3696 output them only if debugging. By default they are written stderr,
3697 apparently, which is not desirable.
3699 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3702 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3705 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3707 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3708 authenticated or an ident call has been made. Suppress the default
3709 values for $authenticated_id and $authenticated_sender (but permit -oMai
3710 and -oMas) when testing with -bh.
3712 PH/24 Re-jigged the order of the tests in the default configuration so that the
3713 tests for valid domains and recipients precede the DNS black list and CSA
3714 tests, on the grounds that those ones are more expensive.
3716 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3717 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3718 This bug exists in every version of Exim that I still have, right back to
3721 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3722 However, an attempt to turn on case-sensitivity in a regex key by
3723 including (?-i) didn't work because the subject string was already
3724 lowercased, and the effects were non-intuitive. It turns out that a
3725 one-line patch can be used to allow (?-i) to work as expected.
3731 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3732 other effects) broke the use of negated acl sub-conditions.
3734 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3737 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3738 "Deny" causes Exim to reject the incoming connection with a 554 error.
3739 Unfortunately, if there is a major crisis, such as a disk failure,
3740 tcp-wrappers gives "deny", whereas what one would like would be some
3741 kind of temporary error. A kludge has been added to help with this.
3742 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3743 554 error is used if errno is still zero or contains ENOENT (which occurs
3744 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3747 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3749 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3750 errors. Otherwise a message that provokes a temporary error (when other
3751 messages do not) can cause a whole host to time out.
3753 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3754 addresses were routed directly to files or pipes from a redirect router.
3755 File deliveries just didn't batch; pipe deliveries might have suffered
3758 PH/06 A failure to get a lock for a hints database would erroneously always say
3759 "Failed to get write lock", even when it was really a read lock.
3761 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3762 of 0600. This has been changed to use the value of the lockfile_mode
3763 option (which defaults to 0600).
3765 PH/08 Applied small patch from the Sieve maintainer.
3767 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3768 folder from quota calculations, a direct delivery into this folder messed
3769 up the contents of the maildirsize file. This was because the regex was
3770 used only to exclude .Trash (or whatever) when the size of the mailbox
3771 was calculated. There was no check that a delivery was happening into an
3772 excluded directory. This bug has been fixed by ignoring all quota
3773 processing for deliveries into excluded directories.
3775 PH/10 Added the maildirfolder_create_regex option to appendfile.
3781 PH/01 The code for finding all the local interface addresses on a FreeBSD
3782 system running IPv6 was broken. This may well have applied to all BSD
3783 systems, as well as to others that have similar system calls. The broken
3784 code found IPv4 interfaces correctly, but gave incorrect values for the
3785 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3786 that it would not match correctly against @[] and not recognize the IPv6
3789 PH/02 The ipliteral router was not recognizing addresses of the form user@
3790 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3792 PH/03 Added disable_ipv6.
3794 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3795 decoded form, because it is most often used to construct To: headers
3796 lines in autoreplies, and the decoded form may well be syntactically
3797 invalid. However, $reply_address has leading white space removed, and all
3798 newlines turned into spaces so that the autoreply transport does not
3801 PH/05 If group was specified without a user on a router, and no group or user
3802 was specified on a transport, the group from the router was ignored.
3804 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3805 for visible compile-time settings that can be used to change these
3806 numbers, for those that want even more. Backwards compatibility with old
3807 spool files has been maintained. However, going back to a previous Exim
3808 release will lost any variables that are in spool files.
3810 PH/07 Two small changes when running in the test harness: increase delay when
3811 passing a TCP/IP connection to a new process, in case the original
3812 process has to generate a bounce, and remove special handling of
3813 127.0.0.2 (sic), which is no longer necessary.
3815 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3816 be the same on different OS.
3818 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3821 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3822 whether --show-vars was specified or not
3824 JJ/02 exipick: Added support for new ACL variable spool format introduced
3827 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3828 syntactically invalid From: or Reply-to: line, and a filter used this to
3829 generate an autoreply, and therefore failed to obtain an address for the
3830 autoreply, Exim could try to deliver to a non-existent relative file
3831 name, causing unrelated and misleading errors. What now happens is that
3832 it logs this as a hard delivery error, but does not attempt to create a
3835 PH/11 The exinext utility has a -C option for testing purposes, but although
3836 the given file was scanned by exinext itself; it wasn't being passed on
3837 when Exim was called.
3839 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3840 an end-of-file indication when reading a command response.
3842 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3843 compiled. In many other places in Exim, IPv6 addresses are always
3844 recognized, so I have changed this. It also means that IPv4 domain
3845 literals of the form [IPV4:n.n.n.n] are now always recognized.
3847 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3848 used if the router is not running as root, for example, when verifying at
3849 ACL time, or when using -bh. The debugging output from this situation was
3850 non-existent - all you got was a failure to exec. I have made two
3853 (a) Failures to set uid/gid, the current directory, or a process leader
3854 in a subprocess such as that created by queryprogram now generate
3855 suitable debugging output when -d is set.
3857 (b) The queryprogram router detects when it is not running as root,
3858 outputs suitable debugging information if -d is set, and then runs
3859 the subprocess without attempting to change uid/gid.
3861 PH/15 Minor change to Makefile for building test_host (undocumented testing
3864 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3865 additional section of a DNS packet that returns MX or SRV records.
3866 Instead, it always explicitly searches for A/AAAA records. This avoids
3867 major problems that occur when a DNS server includes only records of one
3868 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3869 fixed another bug: if SRV records were looked up and the corresponding
3870 address records were *not* found in the additional section, the port
3871 values from the SRV records were lost.
3873 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3874 using the correct key (the original address) when searching the retry
3875 rules in order to find which one to use for generating the retry hint.
3877 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3878 adding the default one. Similarly, if it contains a Reply-To: header, the
3879 errors_reply_to option, if set, is not used.
3881 PH/19 When calculating a retry time, Exim used to measure the "time since
3882 failure" by looking at the "first failed" field in the retry record. Now
3883 it does not use this if it is later than than the arrival time of the
3884 message. Instead it uses the arrival time. This makes for better
3885 behaviour in cases where some deliveries succeed, thus re-setting the
3886 "first failed" field. An example is a quota failure for a huge message
3887 when small messages continue to be delivered. Without this change, the
3888 "time since failure" will always be short, possible causing more frequent
3889 delivery attempts for the huge message than are intended.
3890 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3892 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3893 $1, $2, $3) because the numerical variables can be reset during some
3894 expansion items (e.g. "match"), thereby losing the authentication data.
3896 PH/21 Make -bV show the size of off_t variables so that the test suite can
3897 decide whether to run tests for quotas > 2G.
3899 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3900 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3901 filecount value is greater than 2G or if a quota value is greater than 2G
3902 on a system where the size of off_t is not greater than 4, a panic error
3905 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3906 never match. The debug and -bh output now contains an explicit error
3907 message indicating a malformed IPv4 address or mask.
3909 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3910 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3911 PH/23 above applies.
3913 PH/25 Do not write to syslog when running in the test harness. The only
3914 occasion when this arises is a failure to open the main or panic logs
3915 (for which there is an explicit test).
3917 PH/26 Added the /no_tell option to "control=freeze".
3919 PH/27 If a host name lookup failed very early in a connection, for example, if
3920 the IP address matched host_lookup and the reverse lookup yielded a name
3921 that did not have a forward lookup, an error message of the form "no IP
3922 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3923 could be logged. Now it outputs the IP address instead of "NULL".
3925 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3926 allows the sender and the authenticated sender to be set when
3927 submitting a message from within Exim. Since child_open_exim() is
3928 documented for local_scan(), the new function should be too.
3930 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3931 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3932 results in an empty string is now treated as unset.
3934 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3936 PH/31 Added sender_verify_fail logging option.
3938 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3939 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3940 tidied the source and removed it altogether.
3942 PH/33 When a queue run was abandoned because the load average was too high, a
3943 log line was always written; now it is written only if the queue_run log
3944 selector is set. In addition, the log line for abandonment now contains
3945 information about the queue run such as the pid. This is always present
3946 in "start" and "stop" lines but was omitted from the "abandon" line.
3948 PH/34 Omit spaces between a header name and the colon in the error message that
3949 is given when verify = headers_syntax fails (if there are lots of them,
3950 the message gets confusing).
3952 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3953 names, as there are now some PTR records that contain slashes. This check
3954 is only to protect against broken name servers that fall over on strange
3955 characters, so the fact that it applies to all lookups doesn't matter.
3957 PH/36 Now that the new test suite is complete, we can remove some of the
3958 special code in Exim that was needed for the old test suite. For example,
3959 sorting DNS records because real resolvers return them in an arbitrary
3960 order. The new test suite's fake resolver always returns records in the
3963 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3964 bounces) except when queue_only is set, to avoid logging races between
3965 the different processes.
3967 PH/38 Panic-die if .include specifies a non-absolute path.
3969 PH/39 A tweak to the "H" retry rule from its user.
3971 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3972 a label. They prevented compilation on older perls.
3974 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3975 a warning to be raised on newish perls.
3977 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3978 on queue. Changes to match documented behaviour of showing count of
3979 messages matching specified criteria.
3981 PH/40 Changed the default ident timeout from 30s to 5s.
3983 PH/41 Added support for the use of login_cap features, on those BSD systems
3984 that have them, for controlling the resources used by pipe deliveries.
3986 PH/42 The content-scanning code uses fopen() to create files in which to put
3987 message data. Previously it was not paying any attention to the mode of
3988 the files. Exim runs with umask(0) because the rest of the code creates
3989 files with open(), and sets the required mode explicitly. Thus, these
3990 files were ending up world-writeable. This was not a big issue, because,
3991 being within the spool directory, they were not world-accessible. I have
3992 created a function called modefopen, which takes an additional mode
3993 argument. It sets umask(777), creates the file, chmods it to the required
3994 mode, then resets the umask. All the relevant calls to fopen() in the
3995 content scanning code have been changed to use this function.
3997 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3998 to 24 hours. This avoids potential overflow problems when processing G
3999 and H retry rules. I suspect nobody ever tinkers with this value.
4001 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4003 PH/45 When the plaintext authenticator is running as a client, the server's
4004 challenges are checked to ensure they are valid base64 strings. By
4005 default, the authentication attempt is cancelled if an invalid string is
4006 received. Setting client_ignore_invalid_base64 true ignores these errors.
4007 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4008 they are received. Thus, the responses can be made to depend on the
4009 challenges. If an invalid string is ignored, an empty string is placed in
4012 PH/46 Messages that are created by the autoreply transport now contains a
4013 References: header, in accordance with RFCs 2822 and 3834.
4015 PH/47 Added authenticated_sender_force to the smtp transport.
4017 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4019 PH/49 Installed latest patch from the Sieve maintainer.
4021 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4022 was also set, the appendfile transport was unnecessarily scanning a
4023 directory of message files (e.g. for maildir delivery) to find the count
4024 of files (along with the size), even though it did not need this
4025 information. It now does the scan only if it needs to find either the
4026 size of the count of files.
4028 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4030 PH/52 Two bugs concerned with error handling when the smtp transport is
4033 (i) Exim was not creating retry information for temporary errors given
4034 for individual recipients after the DATA command when the smtp transport
4035 was used in LMTP mode. This meant that they could be retried too
4036 frequently, and not timed out correctly.
4038 (ii) Exim was setting the flag that allows error details to be returned
4039 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4040 recipients that were returned after the DATA command.
4042 PH/53 This is related to PH/52, but is more general: for any failing address,
4043 when detailed error information was permitted to be returned to the
4044 sender, but the error was temporary, then after the final timeout, only
4045 "retry timeout exceeded" was returned. Now it returns the full error as
4046 well as "retry timeout exceeded".
4048 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4049 do this, and (what is worse) MTAs that accept it.
4051 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4052 will now be deprecated.
4054 PH/56 New os.c-cygwin from the Cygwin maintainer.
4056 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4057 formats (previously only available in exim formats via -bpr, -bpru,
4058 and -bpra. Now also available in native and exiqgrep formats)
4060 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4061 with very large, slow to parse queues
4063 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4065 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4067 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4068 responses to authentication challenges, though it was showing the
4069 challenges; (ii) I've removed the CR characters from the debug output for
4072 PH/58 Allow for the insertion of a newline as well as a space when a string
4073 is turned into more than one encoded-word during RFC 2047 encoding. The
4074 Sieve code now uses this.
4076 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4077 data_4xx, lost_connection, tls_required.
4079 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4080 message was being sent as an SMTP response.
4082 PH/61 Add -l and -k options to exicyclog.
4084 PH/62 When verifying, if an address was redirected to one new address, so that
4085 verification continued, and the new address failed or deferred after
4086 having set something in $address_data, the value of $address_data was not
4087 passed back to the ACL. This was different to the case when no
4088 redirection occurred. The value is now passed back in both cases.
4090 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4091 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4092 use login_cap.h, so on its own it isn't the distinguishing feature. The
4093 new name refers directly to the setclassresources() function.
4095 PH/65 Added configuration files for NetBSD3.
4097 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4099 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4100 is preferred over IPv4.
4102 PH/68 The bounce_return_message and bounce_return_body options were not being
4103 honoured for bounces generated during the reception of non-SMTP messages.
4104 In particular, this applied to messages rejected by the ACL. This bug has
4105 been fixed. However, if bounce_return_message is true and bounce_return_
4106 body is false, the headers that are returned for a non-SMTP message
4107 include only those that have been read before the error was detected.
4108 (In the case of an ACL rejection, they have all been read.)
4110 PH/69 The HTML version of the specification is now built in a directory called
4111 spec_html instead of spec.html, because the latter looks like a path with
4112 a MIME-type, and this confuses some software.
4114 PH/70 Catch two compiler warnings in sieve.c.
4116 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4117 function verify_get_ident() calls ip_connect() to connect a socket, but
4118 if the "connect()" function timed out, ip_connect() used to close the
4119 socket. However, verify_get_ident() also closes the socket later, and in
4120 between Exim writes to the log, which may get opened at this point. When
4121 the socket was closed in ip_connect(), the log could get the same file
4122 descriptor number as the socket. This naturally causes chaos. The fix is
4123 not to close the socket in ip_connect(); the socket should be closed by
4124 the function that creates it. There was only one place in the code where
4125 this was missing, in the iplookup router, which I don't think anybody now
4126 uses, but I've fixed it anyway.
4128 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4129 well as to direct DNS lookups. Otherwise the handling of names in host
4130 lists is inconsistent and therefore confusing.
4136 PH/01 Two changes to the default runtime configuration:
4138 (1) Move the checks for relay_from_hosts and authenticated clients from
4139 after to before the (commented out) DNS black list checks.
4141 (2) Add control=submission to the relay_from_hosts and authenticated
4142 clients checks, on the grounds that messages accepted by these
4143 statements are most likely to be submissions.
4145 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4147 (1) Generate an error if the third argument for the ${prvs expansion is
4150 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4153 (3) Reset the variables that are obtained from the first argument of
4154 ${prvscheck and used in the second argument before leaving the code,
4155 because their memory is reclaimed, so using them afterwards may do
4158 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4159 one (it's much easier than Tom thought :-).
4161 (5) Because of (4), we can now allow for the use of $prvscheck_result
4162 inside the third argument.
4164 PH/03 For some reason, the default setting of PATH when running a command from
4165 a pipe transport was just "/usr/bin". I have changed it to
4168 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4169 anything to be listed in the output from -bV.
4171 PH/05 When a filter generated an autoreply, the entire To: header line was
4172 quoted in the delivery log line, like this:
4174 => >A.N.Other <ano@some.domain> <original@ddress> ...
4176 This has been changed so that it extracts the operative address. There
4177 may be more than one such address. If so, they are comma-separated, like
4180 => >ano@some.domain,ona@other.domain <original@ddress> ...
4182 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4183 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4184 not being looked up in the rDNS to get a host name, Exim was showing the
4185 IP address twice in Received: lines, even though the IP addresses were
4186 identical. For example:
4188 Received: from [1.2.3.4] (helo=[1.2.3.4])
4190 However, if the real host name was known, it was omitting the HELO data
4191 if it matched the actual IP address. This has been tidied up so that it
4192 doesn't show the same IP address twice.
4194 PH/07 When both +timestamp and +memory debugging was on, the value given by
4195 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4196 called by the debug printing, thereby overwriting the timestamp buffer.
4197 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4199 PH/08 When the original message was included in an autoreply transport, it
4200 always said "this is a copy of the message, including all the headers",
4201 even if body_only or headers_only was set. It now gives an appropriate
4204 PH/09 Applied a patch from the Sieve maintainer which:
4206 o fixes some comments
4207 o adds the (disabled) notify extension core
4208 o adds some debug output for the result of if/elsif tests
4209 o points to the current vacation draft in the documentation
4210 and documents the missing references header update
4214 o fixes a bug in processing the envelope test (when testing
4215 multiple envelope elements, the last element determined the
4218 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4219 Electronic Mail") by including:
4221 Auto-submitted: auto-generated
4223 in the messages that it generates (bounce messages and others, such as
4224 warnings). In the case of bounce messages for non-SMTP messages, there was
4225 also a typo: it was using "Auto_submitted" (underscore instead of
4226 hyphen). Since every message generated by Exim is necessarily in response
4227 to another message, thes have all been changed to:
4229 Auto-Submitted: auto-replied
4231 in accordance with these statements in the RFC:
4233 The auto-replied keyword:
4235 - SHOULD be used on messages sent in direct response to another
4236 message by an automatic process,
4238 - MUST NOT be used on manually-generated messages,
4240 - MAY be used on Delivery Status Notifications (DSNs) and Message
4241 Disposition Notifications (MDNs),
4243 - MUST NOT be used on messages generated by automatic or periodic
4244 processes, except for messages which are automatic responses to
4247 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4248 to the default Received: header definition.
4250 PH/12 Added log selector acl_warn_skipped (default on).
4252 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4253 variables because (a) they are in the wrong storage pool and (b) even if
4254 they were copied, it wouldn't work properly because of the caching.
4256 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4257 checking when decoding. Apparently there are clients that generate
4258 overlong encoded strings. Why am I not surprised?
4260 PH/15 If the first argument of "${if match_address" was not empty, but did not
4261 contain an "@" character, Exim crashed. Now it writes a panic log message
4262 and treats the condition as false.
4264 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4266 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4267 extension "envelope-auth". The code is finished and in agreement with
4268 other implementations, but there is no documentation so far and in fact,
4269 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4270 not changing the active code.
4272 Print executed "if" and "elsif" statements when debugging is used. This
4273 helps a great deal to understand what a filter does.
4275 Document more things not specified clearly in RFC3028. I had all this
4276 sorted out, when out of a sudden new issues came to my mind. Oops."
4278 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4281 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4282 canonical form (as documented). However, after a host name lookup from
4283 the IP address, check_host() was doing a simple string comparison with
4284 addresses acquired from the DNS when checking that the found name did
4285 have the original IP as one of its addresses. Since any found IPv6
4286 addresses are likely to be in abbreviated form, the comparison could
4287 fail. Luckily, there already exists a function for doing the comparison
4288 by converting both addresses to binary, so now that is used instead of
4289 the text comparison.
4291 PH/20 There was another similar case to PH/19, when a complete host name was
4292 given in a host list; looking up its IP address could give an abbreviated
4293 form, whereas the current host's name might or might not be abbreviated.
4294 The same fix has been applied.
4300 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4301 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4304 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4305 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4307 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4309 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4310 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4311 Also, for "Auto-Submitted", treat anything other than "no" as
4312 non-personal, in accordance with RFC 3834. (Previously it treated
4313 anything starting "auto-" as non-personal.)
4315 TF/01 The control=submission/name=... option had a problem with syntax
4316 errors if the name included a slash character. The /name= option
4317 now slurps the rest of the string, so it can include any characters
4318 but it must come last in the list of options (after /sender_retain
4321 PH/05 Some modifications to the interface to the fake nameserver for the new
4329 TK/01 Added the "success_on_redirect" address verification option. See
4330 NewStuff for rationale and an example.
4332 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4334 PH/02 Patch to exigrep to allow it to work on syslog lines.
4336 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4337 fread() to skip over the body file's header line, because in Cygwin the
4338 header line is locked and is inaccessible.
4340 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4341 co-exist for some time) to make it clear that it is the Exim ID that is
4342 referenced, not the Message-ID: header line.
4344 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4345 string_format() function, because snprintf() does not exist on all
4348 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4349 ${stat: expansion item.
4351 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4352 protocol synchronization error", to keep the pedants happy.
4354 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4355 well as for IRIX systems, when gcc is being used. See the host.c source
4358 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4360 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4363 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4364 transport and to the smtp transport in LMTP mode.
4366 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4368 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4369 run a filter in a subprocess. This could lead to confusion in subsequent
4370 lookups in the parent process. There should also be a search_tidyup() at
4371 the end of the subprocess.
4373 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4374 only if the host matched helo_try_verify_hosts, which caused the
4375 verification to occur when the EHLO/HELO command was issued. The ACL just
4376 tested the remembered result. Now, if a previous verification attempt has
4377 not happened, "verify = helo" does it there and then.
4379 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4381 TK/03 Fix log output including CR from clamd.
4383 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4384 exist provoked a memory error which could cause a segfault.
4386 PH/15 Installed PCRE 6.2
4388 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4390 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4391 of the problem. Specifically, suggested +O2 rather than +O1 for the
4394 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4396 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4397 record, the port was not correctly logged when the outgoing_port log
4398 selector was set (it logged the transort's default port).
4400 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4401 fallback_hosts, and "hosts" in the smtp transport.
4403 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4404 host errors such as "Connection refused".
4406 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4407 authentication with radiusclient 0.4.9:
4409 - Error returned from rc_read_config was caught wrongly
4410 - Username/password not passed on to radius server due to wrong length.
4412 The presumption is that some radiusclient API changes for 4.51/PH/17
4413 were not taken care of correctly. The code is still untested by me (my
4414 Linux distribution still has 0.3.2 of radiusclient), but it was
4415 contributed by a Radius user.
4417 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4418 expanding the "port" option of the smtp transport.
4420 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4421 while reading a MIME header. Thanks to Tom Hughes for a patch.
4423 PH/24 Include config.h inside local_scan.h so that configuration settings are
4426 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4427 that in an ACL for RCPT (for example), you can examine exactly what was
4430 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4431 commands, but it was not correctly comparing the address with the actual
4432 client host address. Thus, it would show the EHLO address in Received:
4433 header lines when this was not necessary.
4435 PH/27 Added the % operator to ${eval:}.
4437 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4438 it should be ignoring failures (because with -C, for example, it has lost
4439 privilege). It wasn't ignoring creation failures other than "already
4442 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4445 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4446 because an input line was too long, either on its own, or by virtue of
4447 too many continuations, the temporary file was not being removed, and the
4448 return code was incorrect.
4450 PH/31 Missing "BOOL" in function definition in filtertest.c.
4452 PH/32 Applied Sieve patches from the maintainer.
4454 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4456 PH/33 Added "verify = not_blind".
4458 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4459 Local/Makefile (with some defaults set). These are used in built scripts
4460 such as exicyclog, but they have never been used in the exim_install
4461 script (though there are many overriding facilities there). I have
4462 arranged that the exim_install script now takes note of these two
4465 PH/35 Installed configuration files for Dragonfly.
4467 PH/36 When a locally submitted message by a trusted user did not contain a
4468 From: header, and the sender address was obtained from -f or from an SMTP
4469 MAIL command, and the trusted user did not use -F to supply a sender
4470 name, $originator_name was incorrectly used when constructing a From:
4471 header. Furthermore, $originator_name was used for submission mode
4472 messages from external hosts without From: headers in a similar way,
4473 which is clearly wrong.
4475 PH/37 Added control=suppress_local_fixups.
4477 PH/38 When log_selector = +received_sender was set, and the addition of the
4478 sender made the log line's construction buffer exactly full, or one byte
4479 less than full, an overflow happened when the terminating "\n" was
4482 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4483 when the result of a list match is failure because a DNS lookup failed.
4485 PH/40 RM_COMMAND is now used in the building process.
4487 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4488 the "build-* directories that it finds.
4490 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4491 address in a domain literal was a prefix of an interface address.
4493 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4494 when verifying a sender address, unless rewrite_headers is false.
4496 PH/44 Wrote a long comment about why errors_to addresses are verified as
4497 recipients, not senders.
4499 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4500 the ratelimit ACL was added.
4502 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4504 PH/46 Added extra information about PostgreSQL errors to the error string.
4506 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4507 avoiding the need to install special zones in a real server. This is
4508 backwards compatible; if it can't find the fake resolver, it drops back.
4509 Thus, both old and new test suites can be run.
4511 TF/02 Added util/ratelimit.pl
4513 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4514 clock is set back in time.
4516 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4517 Candler <B.Candler@pobox.com>.
4519 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4520 OK for addresses that are the result of redirections.
4522 PH/48 A number of further additions for the benefit of the new test suite,
4523 including a fake gethostbyname() that interfaces to the fake DNS resolver
4526 TF/06 The fix for widen_domains has also been applied to qualify_single and
4527 search_parents which are the other dnslookup options that can cause
4530 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4533 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4535 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4536 service" if it encountered an MX record with an empty target hostname.
4537 The message is now "an MX or SRV record indicated no SMTP service".
4539 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4540 if the DNS of the sending site is misconfigured. This is quite a
4541 common situation. This change restores the behaviour of treating a
4542 helo verification defer as a failure.
4544 PH/51 If self=fail was set on a router, the bounce message did not include the
4545 actual error message.
4551 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4553 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4554 command itself ended in error, the underlying message about the transport
4555 filter timeout was being overwritten with the pipe command error. Now the
4556 underlying error message should be appended to the second error message.
4558 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4560 PH/02 Reduce the amount of output that "make" produces by default. Full output
4561 can still be requested.
4563 PH/03 The warning log line about a condition test deferring for a "warn" verb
4564 was being output only once per connection, rather than after each
4565 occurrence (because it was using the same function as for successful
4566 "warn" verbs). This seems wrong, so I have changed it.
4568 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4569 it should not have, which might have caused a crash in the right
4570 circumstances, but probably never did.
4572 PH/04 Installed a modified version of Tony Finch's patch to make submission
4573 mode fix the return path as well as the Sender: header line, and to
4574 add a /name= option so that you can make the user's friendly name appear
4577 TF/03 Added the control = fakedefer ACL modifier.
4579 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4580 Mark Lowes for thorough testing.
4582 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4584 TK/03 Merged latest SRS patch from Miles Wilton.
4586 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4587 with the definition in sysexits.h (which is #included earlier).
4588 Fortunately, Exim does not actually use EX_OK. The code used to try to
4589 preserve the sysexits.h value, by assuming that macro definitions were
4590 scanned for macro replacements. I have been disabused of this notion,
4591 so now the code just undefines EX_OK before #including unistd.h.
4593 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4594 in the smtp transport. When a block could not be written in a single
4595 write() function, the timeout was being re-applied to each part-write.
4596 This seems wrong - if the receiver was accepting one byte at a time it
4597 would take for ever. The timeout is now adjusted when this happens. It
4598 doesn't have to be particularly precise.
4600 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4601 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4603 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4604 without a domain if the check to <postmaster@domain> fails.
4606 SC/01 Eximstats: added -xls and the ability to specify output files
4607 (patch written by Frank Heydlauf).
4609 SC/02 Eximstats: use FileHandles for outputting results.
4611 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4613 SC/04 Eximstats: fixed display of large numbers with -nvr option
4615 SC/05 Eximstats: fixed merging of reports with empty tables.
4617 SC/06 Eximstats: added the -include_original_destination flag
4619 SC/07 Eximstats: removed tabs and trailing whitespace.
4621 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4623 TK/06 MBOX spool code: Add real "From " MBOX separator line
4624 so the .eml file is really in mbox format (even though
4625 most programs do not really care). Patch from Alex Miller.
4627 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4628 The latter is generated from $received_to and is only set if the
4629 message has one envelope recipient. SA can use these headers,
4630 obviously out-of-the-box. Patch from Alex Miller.
4632 PH/08 The ${def test on a variable was returning false if the variable's
4633 value was "0", contrary to what the specification has always said!
4634 The result should be true unless the variable is empty.
4636 PH/09 The syntax error of a character other than { following "${if
4637 def:variable_name" (after optional whitespace) was not being diagnosed.
4638 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4639 accidental colon was present, for example, could give incorrect results.
4641 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4642 result is used (not including appendfile, where other changes are about
4645 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4646 This involved changing a lot of size variables from int to off_t. It
4647 should work with maildirs and everything.
4649 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4650 spamd dying while we are connected to it.
4652 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4655 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4656 function for BDB 4.3.
4658 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4660 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4661 the sources for PCRE have been re-arranged and more files are now
4664 PH/15 The code I had for printing potentially long long variables in PH/11
4665 above was not the best (it lost precision). The length of off_t variables
4666 is now inspected at build time, and an appropriate printing format (%ld
4667 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4668 to be "long long int" or "long int". This is needed for the internal
4669 formatting function string_vformat().
4671 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4672 the configuration file to be ":syslog", then the script "guesses" where
4673 the logs files are, rather than using the compiled in default. In our
4674 case the guess is not the same as the compiled default, so the script
4675 suddenly stopped working when I started to use syslog. The patch checks
4676 to see if log_file_path is "". If so, it attempts to read it from exim
4677 with no configuration file to get the compiled in version, before it
4678 falls back to the previous guessing code."
4680 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4681 implementing BATV in an Exim configuration. See NewStuff for the gory
4684 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4685 Makefile that are specific to HP-UX.
4687 PH/18 If the "use_postmaster" option was set for a recipient callout together
4688 with the "random" option, the postmaster address was used as the MAIL
4689 FROM address for the random test, but not for the subsequent recipient
4690 test. It is now used for both.
4692 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4693 patch removes a few documentation additions to RFC 3028, because the
4694 latest draft now contains them. It adds the new en;ascii-case comparator
4695 and a new error check for 8bit text in MIME parts. Comparator and
4696 require names are now matched exactly. I enabled the subaddress
4697 extension, but it is not well tested yet (read: it works for me)."
4699 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4700 rework some of the code of TK/09 above to avoid the hardwired use of
4701 "%lld" and "long long". Replaced the call to snprintf() with a call to
4704 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4705 records point to non-existent hosts", "retry timeout exceeded", and
4706 "retry time not reached for any host after a long failure period".
4708 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4709 experimental DomainKeys support:
4711 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4712 (2) On an error such as an illegally used "control", the wrong name for
4713 the control was given.
4715 These problems did NOT occur unless DomainKeys support was compiled.
4717 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4719 PH/24 Added ${if match_ip condition.
4721 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4722 databases so that it will be absolutely obvious if a crash occurs in the
4723 DB library. This is a regular occurrence (often caused by mis-matched
4726 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4727 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4728 picked up on a user's system that detects such things. There doesn't seem
4729 to be a gcc warning option for this - only an attribute that has to be
4730 put on the function's prototype. It seems that in Fedora Core 4 they have
4731 set this on a number of new functions. No doubt there will be more in due
4734 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4735 specify a transport. However, if an address that was verified by such a
4736 router was the subject of a callout, Exim crashed because it tried to
4737 read the rcpt_include_affixes from the non-existent transport. Now it
4738 just assumes that the setting of that option is false. This bug was
4739 introduced by 4.51/PH/31.
4741 PH/28 Changed -d+all to exclude +memory, because that information is very
4742 rarely of interest, but it makes the output a lot bigger. People tend to
4743 do -d+all out of habit.
4745 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4746 code in os-type was giving problems when libc.so lives in lib64, like on
4749 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4750 aren't the modern standard, and it seems that some systems' include files
4751 don't always have them. Exim was already checking for some of the newer
4752 ones like T_AAAA, and defining it itself. I've added checks for all the
4753 record types that Exim uses.
4755 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4756 not automatically generating a new one, as it is supposed to. This
4757 prevented TLS from working. If the file did exist, but contained invalid
4758 data, a new version was generated, as expected. It was only the case of a
4759 non-existent file that was broken.
4761 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4762 with a change in libdomainkeys > 0.64.
4764 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4765 from DNS. If the selector record carries the flag, it now has
4766 precedence over the domain-wide flag.
4768 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4770 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4771 the use of a port name that isn't defined in /etc/services) occurred, the
4772 message was deferred as in a normal delivery, and thus remained on the
4773 spool, instead of being failed because of the mua_wrapper setting. This
4774 is now fixed, and I tidied up some of the mua_wrapper messages at the
4777 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4778 the messages in a hash of arrays rather than using individual hashes.
4779 This is a bit cleaner and results in dramatic memory savings, albeit
4780 at a slight CPU cost.
4782 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4783 as requested by Marc Sherman.
4785 SC/10 Eximstats: added histograms for user specified patterns as requested
4788 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4790 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4791 fopen() in the content-scanning modules that did not already have it.
4797 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4798 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4800 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4802 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4804 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4805 to test Sieve filters that use "vacation".
4807 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4808 that changes the way the GnuTLS parameters are stored in the cache file.
4809 The new format can be generated externally. For backward compatibility,
4810 if the data in the cache doesn't make sense, Exim assumes it has read an
4811 old-format file, and it generates new data and writes a new file. This
4812 means that you can't go back to an older release without removing the
4815 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4816 work if there are any delivery delays because "one_time" forces the
4817 parent to be marked "delivered", so its unseen clone is never tried
4818 again. For this reason, Exim now forbids the simultaneous setting of
4821 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4822 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4823 entry if you want to know the details. The fix, however, neglected to
4824 consider the case when local delivery batching is involved. The test for
4825 "previously delivered" was not happening when checking to see if an
4826 address could be batched with a previous (undelivered) one; under
4827 certain circumstances this could lead to multiple deliveries to the same
4830 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4831 in its include files, and this causes problems building Exim.
4833 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4834 header_syntax) but Exim was just ignoring anything given after a slash.
4835 In particular, this caused confusion with an attempt to use "verify =
4836 reverse_host_lookup/defer_ok". An error is now given when options are
4837 supplied for verify items that do not have them. (Maybe reverse_host_
4838 lookup should have a defer_ok option, but that's a different point.)
4840 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4841 defined by RFC 821) to 2048, because there were problems with some AUTH
4842 commands, and RFC 1869 says the size should be increased for extended
4843 SMTP commands that take arguments.
4845 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4848 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4849 "unknown" error; now it says that the functionality isn't in the binary.
4851 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4852 an address' error message when a string expansion fails (syntax or
4853 whatever). Otherwise the password may appear in the log. Following change
4854 PH/42 below, there is no longer a chance of it appearing in a bounce
4857 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4859 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4860 of its addresses was ever tried. (Bugzilla bug #2.)
4862 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4863 the result incorrectly in the debug output. (It correctly added a newline
4864 to what was transported.)
4866 TF/01 Added $received_time.
4868 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4869 commented out examples of how to interface to a virus scanner and to
4870 SpamAssassin. Also added commented examples of av_scanner and
4871 spamd_address settings.
4873 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4874 and controls are allowed in which ACLs. There were a couple of minor
4875 errors. Some of the entries in the conditions table (which is a table of
4876 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4877 negation of where the condition IS allowed.
4879 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4881 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4882 header file does not have a version number, so I've had to invent a new
4883 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4884 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4885 radiusclient), but it was contributed by a Radius user.
4887 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4888 files or directories, for OpenSSL.
4890 PH/19 When an Exim process that is running as root has to create an Exim log
4891 file, it does so in a subprocess that runs as exim:exim so as to get the
4892 ownership right at creation (otherwise, other Exim processes might see
4893 the file with the wrong ownership). There was no test for failure of this
4894 fork() call, which would lead to the process getting stuck as it waited
4895 for a non-existent subprocess. Forks do occasionally fail when resources
4896 run out. I reviewed all the other calls to fork(); they all seem to check
4899 PH/20 When checking for unexpected SMTP input at connect time (before writing
4900 the banner), Exim was not dealing correctly with a non-positive return
4901 from the read() function. If the client had disconnected by this time,
4902 the result was a log entry for a synchronization error with an empty
4903 string after "input=" when read() returned zero. If read() returned -1
4904 (an event I could not check), uninitialized data bytes were printed.
4905 There were reports of junk text (parts of files, etc) appearing after
4908 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4910 PH/22 Added support for macro redefinition, and (re)definition in between
4911 driver and ACL definitions.
4913 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4914 forgetting to use the resulting value; it was using the unexpanded value.
4916 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4917 hadn't been configured. The fix is from Juergen Kreileder, who
4918 understands it better than I do:
4920 "Here's what I see happening with three configured cyrus_sasl
4921 authenticators configured (plain, login, cram-md5):
4923 On startup auth_cyrus_sasl_init() gets called for each of these.
4924 This means three calls to sasl_listmech() without a specified mech_list.
4925 => SASL tests which mechs of all available mechs actually work
4926 => three warnings about OTP not working
4927 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4929 With the patch, sasl_listmech() also gets called three times. But now
4930 SASL's mech_list option is set to the server_mech specified in the the
4931 authenticator. Or in other words, the answer from sasl_listmech()
4932 gets limited to just the mech you're testing for (which is different
4934 => the return list contains just 'plain' or 'login', 'cram-md5' or
4935 nothing depending on the value of ob->server_mech.
4937 I've just tested the patch: Authentication still works fine,
4938 unavailable mechs specified in the exim configuration are still
4939 caught, and the auth.log warnings about OTP are gone."
4941 PH/25 When debugging is enabled, the contents of the command line are added
4942 to the debugging output, even when log_selector=+arguments is not
4945 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4946 answer is "GNU", and only if the return is "GNU/something" is the answer
4949 PH/27 $acl_verify_message is now set immediately after the failure of a
4950 verification in an ACL, and so is available in subsequent modifiers. In
4951 particular, the message can be preserved by coding like this:
4953 warn !verify = sender
4954 set acl_m0 = $acl_verify_message
4956 Previously, $acl_verify_message was set only while expanding "message"
4957 and "log_message" when a very denied access.
4959 PH/28 Modified OS/os.c-Linux with
4961 -#ifndef OS_LOAD_AVERAGE
4962 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4964 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4965 nomenclature these days.)
4967 PH/29 Installed patch from the Sieve maintainer that adds the options
4968 sieve_useraddress and sieve_subaddress to the redirect router.
4970 PH/30 In these circumstances:
4971 . Two addresses routed to the same list of hosts;
4972 . First host does not offer TLS;
4973 . First host accepts first address;
4974 . First host gives temporary error to second address;
4975 . Second host offers TLS and a TLS session is established;
4976 . Second host accepts second address.
4977 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4978 and peerdn, if requested) that were in fact used only for the second
4981 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4982 attention to any local part prefix or suffix that was matched by the
4983 router that accepted the address. It now behaves in the same way as it
4984 does for delivery: the affixes are removed from the local part unless
4985 rcpt_include_affixes is set on the transport.
4987 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4988 timeout during the DATA phase of an incoming message.
4990 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4991 applied a patch sanctioned by the Sieve maintainer.
4993 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4994 the uid or gid is negative. A case of a negative gid caused this to be
4995 noticed. The fix allows for either to be negative.
4997 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4998 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5001 PH/36 The change PH/12 above was broken. Fixed it.
5003 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5004 the grounds that routing the same address twice would always produce the
5005 same answer. This might have been true once, but it is certainly no
5006 longer true now. Routing a child address may depend on the previous
5007 routing that produced that child. Some complicated redirection strategies
5008 went wrong when messages had multiple recipients, and made Exim's
5009 behaviour dependent on the order in which the addresses were given.
5011 I have moved the duplicate checking until after the routing is complete.
5012 Exim scans the addresses that are assigned to local and remote
5013 transports, and removes any duplicates. This means that more work will be
5014 done, as duplicates will always all be routed, but duplicates are
5015 presumably rare, so I don't expect this is of any significance.
5017 For deliveries to pipes, files, and autoreplies, the duplicate checking
5018 still happens during the routing process, since they are not going to be
5021 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5022 It corrects a timeout issue with spamd. This is Ian's comment: "The
5023 background is that sometimes spamd either never reads data from a
5024 connection it has accepted, or it never writes response data. The exiscan
5025 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5026 blindly assumes that writes won't block so it may never time out."
5028 PH/39 Allow G after quota size as well as K and M.
5030 PH/40 The value set for $authenticated_id in an authenticator may not contain
5031 binary zeroes or newlines because the value is written to log lines and
5032 to spool files. There was no check on this. Now the value is run through
5033 the string_printing() function so that such characters are converted to
5034 printable escape sequences.
5036 PH/41 $message_linecount is a new variable that contains the total number of
5037 lines in the message. Compare $body_linecount, which is the count for the
5040 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5041 bounce and delay warning messages, except in certain special cases, which
5044 (a) An SMTP error message from a remote host;
5045 (b) A message specified in a :fail: redirection;
5046 (c) A message specified in a "fail" command in a system filter;
5047 (d) A message specified in a FAIL return from the queryprogram router;
5048 (e) A message specified by the cannot_route_message router option.
5050 In these cases only, Exim does include the error details in bounce and
5051 warning messages. There are also a few cases where bland messages such
5052 as "unrouteable address" or "local delivery error" are given.
5054 PH/43 $value is now also set for the "else" part of a ${run expansion.
5056 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5057 being worked on, but at least Exim now implements the latest version to
5060 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5061 process to complete was treated as a delivery failure, a timeout while
5062 writing the message to the pipe was logged, but erroneously treated as a
5063 successful delivery. Such timeouts include transport filter timeouts. For
5064 consistency with the overall process timeout, these timeouts are now
5065 treated as errors, giving rise to delivery failures by default. However,
5066 there is now a new Boolean option for the pipe transport called
5067 timeout_defer, which, if set TRUE, converts the failures into defers for
5068 both kinds of timeout. A transport filter timeout is now identified in
5071 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5072 systems where "make" and "gmake" are different, calling "gmake" at top
5073 level broke things. I've arranged for the value of $(MAKE) to be passed
5074 from the Makefile to this script so that it can call the same version of
5078 A note about Exim versions 4.44 and 4.50
5079 ----------------------------------------
5081 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5082 changes of various kinds. As a consequence, a big documentation update was
5083 needed. This delayed the release for rather longer than seemed good, especially
5084 in the light of a couple of (minor) security issues. Therefore, the changes
5085 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5086 release. So 4.44 and 4.50 are in effect two different branches that both start
5089 I have left the 4.50 change log unchanged; it contains all the changes since
5090 4.43. The change log for 4.44 is below; many of its items are identical to
5091 those for 4.50. This seems to be the most sensible way to preserve the
5092 historical information.
5098 1. Minor wording change to the doc/README.SIEVE file.
5100 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5101 computation of the current number of files was incorrect.
5103 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5104 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5107 4. Give more explanation in the error message when the command for a transport
5108 filter fails to execute.
5110 5. There are several places where Exim runs a non-Exim command in a
5111 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5112 was being done only for the command run by the queryprogram router. It is
5113 now done for all such subprocesses. The other cases are: ${run, transport
5114 filters, and the commands run by the lmtp and pipe transports.
5116 6. Added CONFIGURE_GROUP build-time option.
5118 7. Some older OS have a limit of 256 on the maximum number of file
5119 descriptors. Exim was using setrlimit() to set 1000 as a large value
5120 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5121 systems. I've change it so that if it can't get 1000, it tries for 256.
5123 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5124 was an oversight, and furthermore, ever since the addition of extra
5125 controls (e.g. 4.43/32), the checks on when to allow different forms of
5126 "control" were broken. There should now be diagnostics for all cases when a
5127 control that does not make sense is encountered.
5129 9. Added the /retain_sender option to "control=submission".
5131 10. $recipients is now available in the predata ACL (oversight).
5133 11. Tidy the search cache before the fork to do a delivery from a message
5134 received from the command line. Otherwise the child will trigger a lookup
5135 failure and thereby defer the delivery if it tries to use (for example) a
5136 cached ldap connection that the parent has called unbind on.
5138 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5139 of $address_data from the recipient verification was clobbered by the
5140 sender verification.
5142 13. The value of address_data from a sender verification is now available in
5143 $sender_address_data in subsequent conditions in the ACL statement.
5145 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5147 15. Added a new option "connect=<time>" to callout options, to set a different
5150 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5151 was its contents. (It was OK if the option was not defined at all.)
5153 17. A "Completed" log line is now written for messages that are removed from
5154 the spool by the -Mrm option.
5156 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5157 information about exactly what failed.
5159 19. Added -dd to debug only the daemon process.
5161 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5162 handles timeouts, both on the server side and network timeouts. Renamed the
5163 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5165 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5166 It is now set to "smtps".
5168 22. $host_address is now set to the target address during the checking of
5169 ignore_target_hosts.
5171 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5172 being passed; this would have caused $sender_host_name to have been used if
5173 matching the list had actually called for a host name (not very likely,
5174 since this list is usually IP addresses). A host name is now passed as
5177 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5178 code that specifies a non-restarting handler (typically sigaction() in
5179 modern systems) in an attempt to fix a rare and obscure crash bug.
5181 25. Narrowed the window for a race in the daemon that could cause it to ignore
5182 SIGCHLD signals. This is not a major problem, because they are used only to
5183 wake it up if nothing else does.
5185 26. A malformed maildirsize file could cause Exim to calculate negative values
5186 for the mailbox size or file count. Odd effects could occur as a result.
5187 The maildirsize information is now recalculated if the size or filecount
5190 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5191 support for a long time. Removed HAVE_SYS_VFS_H.
5193 28. Installed the latest version of exipick from John Jetmore.
5195 29. In an address list, if the pattern was not a regular expression, an empty
5196 subject address (from a bounce message) matched only if the pattern was an
5197 empty string. Non-empty patterns were not even tested. This was the wrong
5198 because it is perfectly reasonable to use an empty address as part of a
5199 database query. An empty address is now tested by patterns that are
5200 lookups. However, all the other forms of pattern expect the subject to
5201 contain a local part and a domain, and therefore, for them, an empty
5202 address still always fails if the pattern is not itself empty.
5204 30. Exim went into a mad DNS loop when attempting to do a callout where the
5205 host was specified on an smtp transport, and looking it up yielded more
5206 than one IP address.
5208 31. Re-factored the code for checking spool and log partition space into a
5209 function that finds that data and another that does the check. The former
5210 is then used to implement four new variables: $spool_space, $log_space,
5211 $spool_inodes, and $log_inodes.
5213 32. The RFC2047 encoding function was originally intended for short strings
5214 such as real names; it was not keeping to the 75-character limit for
5215 encoded words that the RFC imposes. It now respects the limit, and
5216 generates multiple encoded words if necessary. To be on the safe side, I
5217 have increased the buffer size for the ${rfc2047: expansion operator from
5220 33. It is now permitted to omit both strings after an "if" condition; if the
5221 condition is true, the result is "true". As before, when the second string
5222 is omitted, a false condition yields an empty string. This makes it less
5223 cumbersome to write custom ACL and router conditions.
5225 34. Failure to deliver a bounce message always caused it to be frozen, even if
5226 there was an errors_to setting on the router. The errors_to setting is now
5229 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5230 canonical form (fully expanded) before being placed in
5231 $sender_host_address.
5233 36. The table in the code that translates DNS record types into text (T_A to
5234 "A" for instance) was missing entries for NS and CNAME. It is just possible
5235 that this could have caused confusion if both these types were looked up
5236 for the same domain, because the text type is used as part of Exim's
5237 per-process caching. But the chance of anyone hitting this buglet seems
5240 37. The dnsdb lookup has been extended in a number of ways.
5242 (1) There is a new type, "zns", which walks up the domain tree until it
5243 finds some nameserver records. It should be used with care.
5245 (2) There is a new type, "mxh", which is like "mx" except that it returns
5246 just the host names, not the priorities.
5248 (3) It is now possible to give a list of domains (or IP addresses) to be
5249 looked up. The behaviour when one of the lookups defers can be
5250 controlled by a keyword.
5252 (4) It is now possible to specify the separator character for use when
5253 multiple records are returned.
5255 38. The dnslists ACL condition has been extended: it is now possible to supply
5256 a list of IP addresses and/or domains to be looked up in a particular DNS
5259 39. Added log_selector=+queue_time_overall.
5261 40. When running the queue in the test harness, wait just a tad after forking a
5262 delivery process, to get repeatability of debugging output.
5264 41. Include certificate and key file names in error message when GnuTLS fails
5265 to set them up, because the GnuTLS error message doesn't include the name
5266 of the failing file when there is a problem reading it.
5268 42. Allow both -bf and -bF in the same test run.
5270 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5272 44. The "Exiscan patch" is now merged into the mainline Exim source.
5274 45. Sometimes the final signoff response after QUIT could fail to get
5275 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5276 < 0 before doing a fflush(). This bug looks as though it goes back to the
5277 introduction of TLS in release 3.20, but "sometimes" must have been rare
5278 because the tests only now provoked it.
5280 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5281 (this can affect the format of dates).
5283 47. exim_tidydb, when checking for the continued existence of a message for
5284 which it has found a message-specific retry record, was not finding
5285 messages that were in split spool directories. Consequently, it was
5286 deleting retry records that should have stayed in existence.
5288 48. Steve fixed some bugs in eximstats.
5290 49. The SPA authentication driver was not abandoning authentication and moving
5291 on to the next authenticator when an expansion was forced to fail,
5292 contradicting the general specification for all authenticators. Instead it
5293 was generating a temporary error. It now behaves as specified.
5295 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5296 (the order specifies the preference for clients). The order is now AES256,
5297 AES128, 3DES, ARCFOUR128.
5299 51. Small patch to Sieve code - explicitly set From: when generating an
5302 52. Exim crashed if a remote delivery caused a very long error message to be
5303 recorded - for instance if somebody sent an entire SpamAssassin report back
5304 as a large number of 550 error lines. This bug was coincidentally fixed by
5305 increasing the size of one of Exim's internal buffers (big_buffer) that
5306 happened as part of the Exiscan merge. However, to be on the safe side, I
5307 have made the code more robust (and fixed the comments that describe what
5310 53. Now that there can be additional text after "Completed" in log lines (if
5311 the queue_time_overall log selector is set), a one-byte patch to exigrep
5312 was needed to allow it to recognize "Completed" as not the last thing in
5315 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5316 patch that reportedly fixes this has been added. I am not expert enough to
5317 create a test for it. This is what the patch creator wrote:
5319 "I found a little strange behaviour of ldap code when working with
5320 Windows 2003 AD Domain, where users was placed in more than one
5321 Organization Units. When I tried to give exim partial DN, the exit code
5322 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5323 But simultaneously result of request was absolutely normal ldap result,
5324 so I produce this patch..."
5326 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5327 REFERENCE, so I have modified the code to exclude the patch when that macro
5330 55. Some experimental protocols are using DNS PTR records for new purposes. The
5331 keys for these records are domain names, not reversed IP addresses. The
5332 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5333 leaves it alone. Component reversal etc. now happens only for IP addresses.
5336 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5338 57. Double the size of the debug message buffer (to 2048) so that more of very
5339 long debug lines gets shown.
5341 58. The exicyclog utility now does better if the number of log files to keep
5342 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5344 59. Two changes related to the smtp_active_hostname option:
5346 (1) $smtp_active_hostname is now available as a variable.
5347 (2) The default for smtp_banner uses $smtp_active_hostname instead
5348 of $primary_hostname.
5350 60. The host_aton() function is supposed to be passed a string that is known
5351 to be a valid IP address. However, in the case of IPv6 addresses, it was
5352 not checking this. This is a hostage to fortune. Exim now panics and dies
5353 if the condition is not met. A case was found where this could be provoked
5354 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5355 components; fortuitously, this particular loophole had already been fixed
5356 by change 4.50/55 above.
5358 If there are any other similar loopholes, the new check in host_aton()
5359 itself should stop them being exploited. The report I received stated that
5360 data on the command line could provoke the exploit when Exim was running as
5361 exim, but did not say which command line option was involved. All I could
5362 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5363 running as the user.
5366 61. There was a buffer overflow vulnerability in the SPA authentication code
5367 (which came originally from the Samba project). I have added a test to the
5368 spa_base64_to_bits() function which I hope fixes it.
5371 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5372 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5374 63. The daemon start-up calls getloadavg() while still root for those OS that
5375 need the first call to be done as root, but it missed one case: when
5376 deliver_queue_load_max is set with deliver_drop_privilege. This is
5377 necessary for the benefit of the queue runner, because there is no re-exec
5378 when deliver_drop_privilege is set.
5380 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5381 This has been fixed.
5383 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5384 was in use, was not putting the data itself into the right store pool;
5385 consequently, it could be overwritten for a subsequent message in the same
5386 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5389 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5391 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5392 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5393 the function were treating the return as a boolean value, which happened to
5394 work because 0=false and not-0=true, but is not correct code.
5396 68. The host_aton() function was not handling scoped IPv6 addresses (those
5397 with, for example, "%eth0" on the end) correctly.
5399 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5400 negated items (that is, ~something) in unsigned ints. Some compilers
5401 apparently mutter when there is no cast.
5403 70. If an address verification called from an ACL failed, and did not produce a
5404 user-specific message (i.e. there was only a "system" message), nothing was
5405 put in $acl_verify_message. In this situation, it now puts the system
5408 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5409 session; change 4.31/43 added the unwanted input to the log line - except
5410 that it did not do this in the start of session case. It now does.
5412 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5413 This is wrong and can cause the other end to generate a synchronization
5414 error if it is another Exim or anything else that does the synchronization
5415 check. A QUIT command is no longer sent after a timeout.
5417 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5418 during host lookups.
5420 74. The defer_ok option of callout verification was not working if it was used
5421 when verifying addresses in header lines, that is, for this case:
5423 verify = header_sender/callout=defer_ok
5425 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5426 those file descriptors could be used for SMTP connections. If anything
5427 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5428 could be sent to the SMTP client, causing chaos. The daemon now opens
5429 stdin, stdout, and stderr to /dev/null when it puts itself into the
5432 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5433 log by default. The user can override this with suitable Perl magic.
5435 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5436 the log message when discard triggers, was not working for the DATA ACL or
5437 for the non-SMTP ACL.
5439 78. Error message wording change in sieve.c.
5441 79. If smtp_accept_max_per_host was set, the number of connections could be
5442 restricted to fewer than expected, because the daemon was trying to set up
5443 a new connection before checking whether the processes handling previous
5444 connections had finished. The check for completed processes is now done
5445 earlier. On busy systems, this bug wouldn't be noticed because something
5446 else would have woken the daemon, and it would have reaped the completed
5449 80. If a message was submitted locally by a user whose login name contained one
5450 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5451 It caused a spool format error. I have fixed the spool reading code. A
5452 related problem was that the "from" clause in the Received: line became
5453 illegal because of the space(s). It is now covered by ${quote_local_part.
5455 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5458 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5460 83. Patch from Sieve maintainer to fix unterminated string problem in
5461 "vacation" handling.
5463 84. Some minor changes to the Linux configuration files to help with other
5464 OS variants using glibc.
5466 85. One more patch for Sieve to update vacation handling to latest spec.
5469 ----------------------------------------------------
5470 See the note above about the 4.44 and 4.50 releases.
5471 ----------------------------------------------------
5477 1. Change 4.43/35 introduced a bug that caused file counts to be
5478 incorrectly computed when quota_filecount was set in an appendfile
5481 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5482 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5485 3. Give more explanation in the error message when the command for a transport
5486 filter fails to execute.
5488 4. There are several places where Exim runs a non-Exim command in a
5489 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5490 was being done only for the command run by the queryprogram router. It is
5491 now done for all such subprocesses. The other cases are: ${run, transport
5492 filters, and the commands run by the lmtp and pipe transports.
5494 5. Some older OS have a limit of 256 on the maximum number of file
5495 descriptors. Exim was using setrlimit() to set 1000 as a large value
5496 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5497 systems. I've change it so that if it can't get 1000, it tries for 256.
5499 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5500 was an oversight, and furthermore, ever since the addition of extra
5501 controls (e.g. 4.43/32), the checks on when to allow different forms of
5502 "control" were broken. There should now be diagnostics for all cases when a
5503 control that does not make sense is encountered.
5505 7. $recipients is now available in the predata ACL (oversight).
5507 8. Tidy the search cache before the fork to do a delivery from a message
5508 received from the command line. Otherwise the child will trigger a lookup
5509 failure and thereby defer the delivery if it tries to use (for example) a
5510 cached ldap connection that the parent has called unbind on.
5512 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5513 of $address_data from the recipient verification was clobbered by the
5514 sender verification.
5516 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5517 was its contents. (It was OK if the option was not defined at all.)
5519 11. A "Completed" log line is now written for messages that are removed from
5520 the spool by the -Mrm option.
5522 12. $host_address is now set to the target address during the checking of
5523 ignore_target_hosts.
5525 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5526 being passed; this would have caused $sender_host_name to have been used if
5527 matching the list had actually called for a host name (not very likely,
5528 since this list is usually IP addresses). A host name is now passed as
5531 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5532 code that specifies a non-restarting handler (typically sigaction() in
5533 modern systems) in an attempt to fix a rare and obscure crash bug.
5535 15. Narrowed the window for a race in the daemon that could cause it to ignore
5536 SIGCHLD signals. This is not a major problem, because they are used only to
5537 wake it up if nothing else does.
5539 16. A malformed maildirsize file could cause Exim to calculate negative values
5540 for the mailbox size or file count. Odd effects could occur as a result.
5541 The maildirsize information is now recalculated if the size or filecount
5544 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5545 support for a long time. Removed HAVE_SYS_VFS_H.
5547 18. Updated exipick to current release from John Jetmore.
5549 19. Allow an empty sender to be matched against a lookup in an address list.
5550 Previously the only cases considered were a regular expression, or an
5553 20. Exim went into a mad DNS lookup loop when doing a callout where the
5554 host was specified on the transport, if the DNS lookup yielded more than
5557 21. The RFC2047 encoding function was originally intended for short strings
5558 such as real names; it was not keeping to the 75-character limit for
5559 encoded words that the RFC imposes. It now respects the limit, and
5560 generates multiple encoded words if necessary. To be on the safe side, I
5561 have increased the buffer size for the ${rfc2047: expansion operator from
5564 22. Failure to deliver a bounce message always caused it to be frozen, even if
5565 there was an errors_to setting on the router. The errors_to setting is now
5568 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5569 canonical form (fully expanded) before being placed in
5570 $sender_host_address.
5572 24. Updated eximstats to version 1.33
5574 25. Include certificate and key file names in error message when GnuTLS fails
5575 to set them up, because the GnuTLS error message doesn't include the name
5576 of the failing file when there is a problem reading it.
5578 26. Expand error message when OpenSSL has problems setting up cert/key files.
5581 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5582 (this can affect the format of dates).
5584 28. exim_tidydb, when checking for the continued existence of a message for
5585 which it has found a message-specific retry record, was not finding
5586 messages that were in split spool directories. Consequently, it was
5587 deleting retry records that should have stayed in existence.
5589 29. eximstats updated to version 1.35
5590 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5591 1.35 - bugfix such that pie charts by volume are generated correctly
5593 30. The SPA authentication driver was not abandoning authentication and moving
5594 on to the next authenticator when an expansion was forced to fail,
5595 contradicting the general specification for all authenticators. Instead it
5596 was generating a temporary error. It now behaves as specified.
5598 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5599 (the order specifies the preference for clients). The order is now AES256,
5600 AES128, 3DES, ARCFOUR128.
5602 31. Small patch to Sieve code - explicitly set From: when generating an
5605 32. Exim crashed if a remote delivery caused a very long error message to be
5606 recorded - for instance if somebody sent an entire SpamAssassin report back
5607 as a large number of 550 error lines. This bug was coincidentally fixed by
5608 increasing the size of one of Exim's internal buffers (big_buffer) that
5609 happened as part of the Exiscan merge. However, to be on the safe side, I
5610 have made the code more robust (and fixed the comments that describe what
5613 33. Some experimental protocols are using DNS PTR records for new purposes. The
5614 keys for these records are domain names, not reversed IP addresses. The
5615 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5616 leaves it alone. Component reversal etc. now happens only for IP addresses.
5619 34. The host_aton() function is supposed to be passed a string that is known
5620 to be a valid IP address. However, in the case of IPv6 addresses, it was
5621 not checking this. This is a hostage to fortune. Exim now panics and dies
5622 if the condition is not met. A case was found where this could be provoked
5623 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5624 components; fortuitously, this particular loophole had already been fixed
5625 by change 4.50/55 or 4.44/33 above.
5627 If there are any other similar loopholes, the new check in host_aton()
5628 itself should stop them being exploited. The report I received stated that
5629 data on the command line could provoke the exploit when Exim was running as
5630 exim, but did not say which command line option was involved. All I could
5631 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5632 running as the user.
5635 35. There was a buffer overflow vulnerability in the SPA authentication code
5636 (which came originally from the Samba project). I have added a test to the
5637 spa_base64_to_bits() function which I hope fixes it.
5640 36. The daemon start-up calls getloadavg() while still root for those OS that
5641 need the first call to be done as root, but it missed one case: when
5642 deliver_queue_load_max is set with deliver_drop_privilege. This is
5643 necessary for the benefit of the queue runner, because there is no re-exec
5644 when deliver_drop_privilege is set.
5646 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5647 was in use, was not putting the data itself into the right store pool;
5648 consequently, it could be overwritten for a subsequent message in the same
5649 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5652 38. Sometimes the final signoff response after QUIT could fail to get
5653 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5654 < 0 before doing a fflush(). This bug looks as though it goes back to the
5655 introduction of TLS in release 3.20, but "sometimes" must have been rare
5656 because the tests only now provoked it.
5662 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5663 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5664 Now it returns an integer. A number of calls were still expecting a T/F
5665 return. Fortuitously, in all cases, the tests worked in OK situations,
5666 which is the norm. However, things would have gone wrong on any write
5667 failures on the smtp file descriptor. This function is used when sending
5668 messages over SMTP and also when doing verify callouts.
5670 2. When Exim is called to do synchronous delivery of a locally submitted
5671 message (the -odf or -odi options), it no longer closes stderr before doing
5674 3. Implemented the mua_wrapper option.
5676 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5678 5. Implemented the functions header_remove(), header_testname(),
5679 header_add_at_position(), and receive_remove_recipient(), and exported them
5682 6. If an ACL "warn" statement specified the addition of headers, Exim already
5683 inserted X-ACL-Warn: at the start if there was no header name. However, it
5684 was not making this test for the second and subsequent header lines if
5685 there were newlines in the string. This meant that an invalid header could
5686 be inserted if Exim was badly configured.
5688 7. Allow an ACL "warn" statement to add header lines at the start or after all
5689 the Received: headers, as well as at the end.
5691 8. Added the rcpt_4xx retry error code.
5693 9. Added postmaster_mailfrom=xxx to callout verification option.
5695 10. Added mailfrom=xxxx to the callout verification option, for verify=
5698 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5699 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5701 12. Inserted some casts to stop certain compilers warning when using pointer
5702 differences as field lengths or precisions in printf-type calls (mostly
5703 affecting debugging statements).
5705 13. Added optional readline() support for -be (dynamically loaded).
5707 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5708 same clock tick as a message's arrival, so that its received time was the
5709 same as the "first fail" time on the retry record, and that message
5710 remained on the queue past the ultimate address timeout, every queue runner
5711 would try a delivery (because it was past the ultimate address timeout) but
5712 after another failure, the ultimate address timeout, which should have then
5713 bounced the address, did not kick in. This was a "< instead of <=" error;
5714 in most cases the first failure would have been in the next clock tick
5715 after the received time, and all would be well.
5717 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5718 being recognized when the domain list was tested by the match_domain
5719 condition in an expansion string.
5721 16. Added the ${str2b64: operator.
5723 17. Exim was always calling setrlimit() to set a large limit for the number of
5724 processes, without checking whether the existing limit was already
5725 adequate. (It did check for the limit on file descriptors.) Furthermore,
5726 errors from getrlimit() and setrlimit() were being ignored. Now they are
5727 logged to the main and panic logs, but Exim does carry on, to try to do its
5728 job under whatever limits there are.
5730 18. Imported PCRE 5.0.
5732 19. Trivial typo in log message " temporarily refused connection" (the leading
5735 20. If the log selector return_path_on_delivery was set and an address was
5736 redirected to /dev/null, the delivery process crashed because it assumed
5737 that a return path would always be set for a "successful" delivery. In this
5738 case, the whole delivery is bypassed as an optimization, and therefore no
5741 21. Internal re-arrangement: the function for sending a challenge and reading
5742 a response while authentication was assuming a zero-terminated challenge
5743 string. It's now changed to take a pointer and a length, to allow for
5744 binary data in such strings.
5746 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5748 23. Exim was not respecting finduser_retries when seeking the login of the
5749 uid under which it was called; it was always trying 10 times. (The default
5750 setting of finduser_retries is zero.) Also, it was sleeping after the final
5751 failure, which is pointless.
5753 24. Implemented tls_on_connect_ports.
5755 25. Implemented acl_smtp_predata.
5757 26. If the domain in control=submission is set empty, Exim assumes that the
5758 authenticated id is a complete email address when it generates From: or
5759 Sender: header lines.
5761 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5762 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5763 chown and chgrp in /bin and hostname in /usr/bin.
5765 28. Exim was keeping the "process log" file open after each use, just as it
5766 does for the main log. This opens the possibility of it remaining open for
5767 long periods when the USR1 signal hits a daemon. Occasional processlog
5768 errors were reported, that could have been caused by this. Anyway, it seems
5769 much more sensible not to leave this file open at all, so that is what now
5772 29. The long-running daemon process does not normally write to the log once it
5773 has entered its main loop, and it closes the log before doing so. This is
5774 so that log files can straightforwardly be renamed and moved. However,
5775 there are a couple of unusual error situations where the daemon does write
5776 log entries, and I had neglected to close the log afterwards.
5778 30. The text of an SMTP error response that was received during a remote
5779 delivery was being truncated at 512 bytes. This is too short for some of
5780 the long messages that one sometimes sees. I've increased the limit to
5783 31. It is now possible to make retry rules that apply only when a message has a
5784 specific sender, in particular, an empty sender.
5786 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5787 it possible to be selective about when SMTP synchronization is enforced.
5789 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5791 32. Implemented hosts_connection_nolog.
5793 33. Added an ACL for QUIT.
5795 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5798 35. Added mailbox_size and mailbox_filecount to appendfile.
5800 36. Added control = no_multiline_responses to ACLs.
5802 37. There was a bug in the logic of the code that waits for the clock to tick
5803 in the case where the clock went backwards by a substantial amount such
5804 that the microsecond fraction of "now" was more than the microsecond
5805 fraction of "then" (but the whole seconds number was less).
5807 38. Added support for the libradius Radius client library this is found on
5808 FreeBSD (previously only the radiusclient library was supported).
5814 1. When certain lookups returned multiple values in the form name=value, the
5815 quoting of the values was not always being done properly. Specifically:
5816 (a) If the value started with a double quote, but contained no whitespace,
5818 (b) If the value contained whitespace other than a space character (i.e.
5819 tabs or newlines or carriage returns) it was not quoted.
5820 This fix has been applied to the mysql and pgsql lookups by writing a
5821 separate quoting function and calling it from the lookup code. The fix
5822 should probably also be applied to nisplus, ibase and oracle lookups, but
5823 since I cannot test any of those, I have not disturbed their existing code.
5825 2. A hit in the callout cache for a specific address caused a log line with no
5826 reason for rejecting RCPT. Now it says "Previous (cached) callout
5827 verification failure".
5829 3. There was an off-by-one bug in the queryprogram router. An over-long
5830 return line was truncated at 256 instead of 255 characters, thereby
5831 overflowing its buffer with the terminating zero. As well as fixing this, I
5832 have increased the buffer size to 1024 (and made a note to document this).
5834 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5835 when Exim is waiting for an SMTP response from a remote server, Exim
5836 restarts its select() call on the socket, thereby resetting its timeout.
5837 This is not a problem when such interrupts are rare. Somebody set up a cron
5838 job to run exiwhat every 2 minutes, which is less than the normal select()
5839 timeout (5 or 10 minutes). This meant that the select() timeout never
5840 kicked in because it was always reset. I have fixed this by comparing the
5841 time when an interrupt arrives with the time at the start of the first call
5842 to select(). If more time than the timeout has elapsed, the interrupt is
5843 treated as a timeout.
5845 5. Some internal re-factoring in preparation for the addition of Sieve
5846 extensions (by MH). In particular, the "personal" test is moved to a
5847 separate function, and given an option for scanning Cc: and Bcc: (which is
5848 not set for Exim filters).
5850 6. When Exim created an email address using the login of the caller as the
5851 local part (e.g. when creating a From: or Sender: header line), it was not
5852 quoting the local part when it contained special characters such as @.
5854 7. Installed new OpenBSD configuration files.
5856 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5857 try to make them clearer.
5859 9. Callout options, other than the timeout value, were being ignored when
5860 verifying sender addresses in header lines. For example, when using
5862 verify = header_sender/callout=no_cache
5864 the cache was (incorrectly) being used.
5866 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5867 only the Cygwin environment.
5869 11. When return_path_on_delivery was set as a log selector, if different remote
5870 addresses in the same message used different return paths and parallel
5871 remote delivery occurred, the wrong values would sometimes be logged.
5872 (Whenever a remote delivery process finished, the return path value from
5873 the most recently started remote delivery process was logged.)
5875 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5876 lines when AUTH and/or TLS are in use. This is the "received protocol"
5877 field. Exim used to use "asmtp" for authenticated SMTP, without any
5878 indication (in the protocol name) for TLS use. Now it follows the RFC and
5879 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5880 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5881 names appear in log lines as well as in Received: header lines.
5883 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5884 extensions, and comparison tests, and to fix some bugs.
5886 14. Changes to the "personal" filter test:
5888 (1) The test was buggy in that it was just doing the equivalent of
5889 "contains" tests on header lines. For example, if a user's address was
5890 anne@some.where, the "personal" test would incorrectly be true for
5892 To: susanne@some.where
5894 This test is now done by extracting each address from the header in turn,
5895 and checking the entire address. Other tests that are part of "personal"
5896 are now done using regular expressions (for example, to check local parts
5897 of addresses in From: header lines).
5899 (2) The list of non-personal local parts in From: addresses has been
5900 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5901 taken from the Sieve specification recommendations.
5903 (3) If the message contains any header line starting with "List-" it is
5904 treated as non-personal.
5906 (4) The test for "circular" in the Subject: header line has been removed
5907 because it now seems ill-conceived.
5909 15. Minor typos in src/EDITME comments corrected.
5911 16. Installed latest exipick from John Jetmore.
5913 17. If headers_add on a router specified a text string that was too long for
5914 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5915 of string_sprintf() is now avoided.
5917 18. $message_body_size was not set (it was always zero) when running the DATA
5918 ACL and the local_scan() function.
5920 19. For the "mail" command in an Exim filter, no default was being set for
5921 the once_repeat time, causing a random time value to be used if "once" was
5922 specified. (If the value happened to be <= 0, no repeat happened.) The
5923 default is now 0s, meaning "never repeat". The "vacation" command was OK
5924 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5925 (I found it when inspecting the code).
5927 20. There is now an overall timeout for performing a callout verification. It
5928 defaults to 4 times the callout timeout, which applies to individual SMTP
5929 commands during the callout. The overall timeout applies when there is more
5930 than one host that can be tried. The timeout is checked before trying the
5931 next host. This prevents very long delays if there are a large number of
5932 hosts and all are timing out (e.g. when the network connections are timing
5933 out). The value of the overall timeout can be changed by specifying an
5934 additional sub-option for "callout", called "maxwait". For example:
5936 verify = sender/callout=5s,maxwait=20s
5938 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5939 to the end before writing, but this should make it even safer).
5941 22. Exim was forgetting that it had advertised PIPELINING for the second and
5942 subsequent messages on an SMTP connection. It was also not resetting its
5943 memory on STARTTLS and an internal HELO.
5945 23. When Exim logs an SMTP synchronization error within a session, it now
5946 records whether PIPELINING has been advertised or not.
5948 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5949 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5950 rather than long int.
5952 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5954 26. Added the never_mail option to autoreply.
5960 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5961 crash if the getsockname() call failed; this can happen if a connection is
5962 closed very soon after it is established. The problem was simply in the
5963 order in which certain operations were done, causing Exim to try to write
5964 to the SMTP stream before it had set up the file descriptor. The bug has
5965 been fixed by making things happen in the correct order.
5971 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5972 before the connection was closed, thus losing the rejection response.
5974 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5975 some early Solaris releases, but causes trouble in current releases where
5976 socklen_t is defined.
5978 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5981 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5984 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5985 that releasing the top of it at the end releases what was used for sub-
5986 expansions (unless the block got too big). However, discard this block if
5987 the first thing is a variable or header, so that we can use its block when
5988 it is dynamic (useful for very large $message_headers, for example).
5990 6. Lookups now cache *every* query, not just the most recent. A new, separate
5991 store pool is used for this. It can be recovered when all lookup caches are
5992 flushed. Lookups now release memory at the end of their result strings.
5993 This has involved some general refactoring of the lookup sources.
5995 7. Some code has been added to the store_xxx() functions to reduce the amount
5996 of flapping under certain conditions.
5998 8. log_incoming_interface used to affect only the <= reception log lines. Now
5999 it causes the local interface and port to be added to several more SMTP log
6000 lines, for example "SMTP connection from", and rejection lines.
6002 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6004 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6006 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6007 could be overwritten at the end of the current message (or the start of a
6008 new message if it was set in a HELO ACL). The value is now preserved for
6009 the duration of the SMTP connection.
6011 12. If a transport had a headers_rewrite setting, and a matching header line
6012 contained an unqualified address, that address was qualified, even if it
6013 did not match any rewriting rules. The underlying bug was that the values
6014 of the flags that permit the existence of unqualified sender and recipient
6015 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6016 non-local messages, and by -bnq for local messages) were not being
6017 preserved with the message after it was received.
6019 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6020 "next input=" as part of the text comprising the host identity instead of
6021 the correct text. The code was using the same buffer for two different
6022 strings. However, depending on which order the printing function evaluated
6023 its arguments, the bug did not always show up. Under Linux, for example, my
6024 test suite worked just fine.
6026 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6027 doesn't work with some older versions of Perl. It has been changed to "my",
6028 which in any case is probably the better facility to use.
6030 15. A really picky compiler found some instances of statements for creating
6031 error messages that either had too many or two few arguments for the format
6034 16. The size of the buffer for calls to the DNS resolver has been increased
6035 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6036 for addresses that have a lot of PTR records. This alleviates a problem; it
6037 does not fully solve it.
6039 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6040 buffer now truncates the list and logs the incident, which is the same
6041 action as happens when Exim is looking up a host name and its aliases.
6042 Previously in this situation something unpredictable would happen;
6043 sometimes it was "internal error: store_reset failed".
6045 18. If a server dropped the connection unexpectedly when an Exim client was
6046 using GnuTLS and trying to read a response, the client delivery process
6047 crashed while trying to generate an error log message.
6049 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6050 string, for example:
6052 warn message = H1: something\nH2: something
6054 the text was added as a single header line from Exim's point of view
6055 though it ended up OK in the delivered message. However, searching for the
6056 second and subsequent header lines using $h_h2: did not work. This has been
6057 fixed. Similarly, if a system filter added multiple headers in this way,
6058 the routers could not see them.
6060 20. Expanded the error message when iplsearch is called with an invalid key to
6061 suggest using net-iplsearch in a host list.
6063 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6064 ACLs are no longer actually imposed (and a message to that effect is
6067 22. If a "gecos" field in a passwd entry contained escaped characters, in
6068 particular, if it contained a \" sequence, Exim got it wrong when building
6069 a From: or a Sender: header from that name. A second bug also caused
6070 incorrect handling when an unquoted " was present following a character
6071 that needed quoting.
6073 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6074 was not being matched caselessly.
6076 24. Arranged for all hyphens in the exim.8 source to be escaped with
6079 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6080 properly. Recipient callout cache records were still being keyed to include
6081 the sender, even when use_sender was set false. This led to far more
6082 callouts that were necessary. The sender is no longer included in the key
6083 when use_sender is false.
6085 26. Added "control = submission" modifier to ACLs.
6087 27. Added the ${base62d: operator to decode base 62 numbers.
6089 28. dnsdb lookups can now access SRV records.
6091 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6092 the configuration file.
6094 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6095 (-v) mode. This makes the output for a verbose queue run more intelligible.
6097 31. Added a use_postmaster feature to recipient callouts.
6099 32. Added the $body_zerocount variable, containing the number of binary zero
6100 bytes in the message body.
6102 33. The time of last modification of the "new" subdirectory is now used as the
6103 "mailbox time last read" when there is a quota error for a maildir
6106 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6108 35. Added +ignore_unknown as a special item in host lists.
6110 36. Code for decoding IPv6 addresses in host lists is now included, even if
6111 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6112 address was recognized as an IP address, but was then not correctly decoded
6113 into binary, causing unexpected and incorrect effects when compared with
6120 1. Very minor rewording of debugging text in manualroute to say "list of
6121 hosts" instead of "hostlist".
6123 2. If verify=header_syntax was set, and a header line with an unqualified
6124 address (no domain) and a large number of spaces between the end of the
6125 name and the colon was received, the reception process suffered a buffer
6126 overflow, and (when I tested it) crashed. This was caused by some obsolete
6127 code that should have been removed. The fix is to remove it!
6129 3. When running in the test harness, delay a bit after writing a bounce
6130 message to get a bit more predictability in the log output.
6132 4. Added a call to search_tidyup() just before forking a reception process. In
6133 theory, someone could use a lookup in the expansion of smtp_accept_max_
6134 per_host which, without the tidyup, could leave open a database connection.
6136 5. Added the variables $recipient_data and $sender_data which get set from a
6137 lookup success in an ACL "recipients" or "senders" condition, or a router
6138 "senders" option, similar to $domain_data and $local_part_data.
6140 6. Moved the writing of debug_print from before to after the "senders" test
6143 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6144 problems for message scanning, either using a data ACL, or using
6145 local_scan() because the Received: header was not generated till after they
6146 were called (in order to set the time as the time of reception completion).
6147 I have revised the way this works. The header is now generated after the
6148 body is received, but before the ACL or local_scan() are called. After they
6149 are run, the timestamp in the header is updated.
6155 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6156 before starting a queue runner without re-exec. This happened only when
6157 deliver_drop_privilege was set or when the Exim user was set to root. The
6158 effect of the bug was that timeouts during subsequent deliveries caused
6159 crashes instead of being properly handled. The handler is now left at its
6160 default (and expected) setting.
6162 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6163 message, again when deliver_drop_privilege is set or Exim is run as root.
6164 The bug described in (1) was not present in this case, but the tidying up
6165 of the other signals was missing. I have made the two cases consistent.
6167 3. The ignore_target_hosts setting on a manualroute router was being ignored
6168 for hosts that were looked up using the /MX notation.
6170 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6173 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6174 operated on the sender address. After changing the $sender_address to <>
6175 for the sender address verify, Exim was re-instated it as the original
6176 (before rewriting) address, but remembering that it had rewritten it, so it
6177 wasn't rewriting it again. This bug also had the effect of breaking the
6178 sender address verification caching when the sender address was rewritten.
6180 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6181 This has been changed so that if the ip literal address matches
6182 ignore_target_hosts, the router declines.
6184 7. Added expansion conditions match_domain, match_address, and match_local_
6185 part (NOT match_host).
6187 8. The placeholder for the Received: header didn't have a length field set.
6189 9. Added code to Exim itself and to exim_lock to test for a specific race
6190 condition that could lead to file corruption when using MBX delivery. The
6191 issue is with the lockfile that is created in /tmp. If this file is removed
6192 after a process has opened it but before that process has acquired a lock,
6193 there is the potential for a second process to recreate the file and also
6194 acquire a lock. This could lead to two Exim processes writing to the file
6195 at the same time. The added code performs the same test as UW imapd; it
6196 checks after acquiring the lock that its file descriptor still refers to
6197 the same named file.
6199 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6200 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6203 11. Added the smtp_active_hostname option. If used, this will typically be made
6204 to depend on the incoming interface address. Because $interface_address is
6205 not set up until the daemon has forked a reception process, error responses
6206 that can happen earlier (such as "too many connections") no longer contain
6209 12. If an expansion in a condition on a "warn" statement fails because a lookup
6210 defers, the "warn" statement is abandoned, and the next ACL statement is
6211 processed. Previously this caused the whole ACL to be aborted.
6213 13. Added the iplsearch lookup type.
6215 14. Added ident_timeout as a log selector.
6217 15. Added tls_certificate_verified as a log selector.
6219 16. Added a global option tls_require_ciphers (compare the smtp transport
6220 option of the same name). This controls incoming TLS connections.
6222 17. I finally figured out how to make tls_require_ciphers do a similar thing
6223 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6224 before starting the TLS session.
6226 18. Tabs are now shown as \t in -bP output.
6228 19. If the log selector return_path_on_delivery was set, Exim crashed when
6229 bouncing a message because it had too many Received: header lines.
6231 20. If two routers both had headers_remove settings, and the first one included
6232 a superfluous trailing colon, the final name in the first list and the
6233 first name in the second list were incorrectly joined into one item (with a
6234 colon in the middle).
6240 1. Added -C and -D options to the exinext utility, mainly to make it easier
6241 to include in the automated testing, but these could be helpful when
6242 multiple configurations are in use.
6244 2. The exinext utility was not formatting the output nicely when there was
6245 an alternate port involved in the retry record key, nor when there was a
6246 message id as well (for retries that were specific to a specific message
6247 and a specific host). It was also confused by IPv6 addresses, because of
6248 the additional colons they contain. I have fixed the IPv4 problem, and
6249 patched it up to do a reasonable job for IPv6.
6251 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6252 delivery, the log line now contains "pipelined" if PIPELINING was used.
6254 4. An SMTP transport process used to panic and die if the bind() call to set
6255 an explicit outgoing interface failed. This has been changed; it is now
6256 treated in the same way as a connect() failure.
6258 5. A reference to $sender_host_name in the part of a conditional expansion
6259 that was being skipped was still causing a DNS lookup. This no longer
6262 6. The def: expansion condition was not recognizing references to header lines
6263 that used bh_ and bheader_.
6265 7. Added the _cache feature to named lists.
6267 8. The code for checking quota_filecount in the appendfile transport was
6268 allowing one more file than it should have been.
6270 9. For compatibility with Sendmail, the command line option
6278 and sets the incoming protocol and host name (for trusted callers). The
6279 host name and its colon can be omitted when only the protocol is to be set.
6280 Note the Exim already has two private options, -pd and -ps, that refer to
6281 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6282 "s", but I don't think that's a major issue.
6284 10. A number of refactoring changes to the code, none of which should affect
6287 (a) The number of logging options was getting close to filling up the
6288 32-bit word that was used as a bit map. I have split them into two classes:
6289 those that are passed in the argument to log_write(), and those that are
6290 only ever tested independently outside of that function. These are now in
6291 separate 32-bit words, so there is plenty of room for expansion again.
6292 There is no change in the user interface or the logging behaviour.
6294 (b) When building, for example, log lines, the code previously used a
6295 macro that called string_cat() twice, in order to add two strings. This is
6296 not really sufficiently general. Furthermore, there was one instance where
6297 it was actually wrong because one of the argument was used twice, and in
6298 one call a function was used. (As it happened, calling the function twice
6299 did not affect the overall behaviour.) The macro has been replaced by a
6300 function that can join an arbitrary number of extra strings onto a growing
6303 (c) The code for expansion conditions now uses a table and a binary chop
6304 instead of a serial search (which was left over from when there were very
6305 few conditions). Also, it now recognizes conditions like "pam" even when
6306 the relevant support is not compiled in: a suitably worded error message is
6307 given if an attempt is made to use such a condition.
6309 11. Added ${time_interval:xxxxx}.
6311 12. A bug was causing one of the ddress fields not to be passed back correctly
6312 from remote delivery subprocesses. The field in question was not being
6313 subsequently used, so this caused to problems in practice.
6315 13. Added new log selectors queue_time and deliver_time.
6317 14. Might have fixed a bug in maildirsizefile handling that threw up
6318 "unexpected character" debug warnings, and recalculated the data
6319 unnecessarily. In any case, I expanded the warning message to give more
6322 15. Added the message "Restricted characters in address" to the statements in
6323 the default ACL that block characters like @ and % in local parts.
6325 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6326 Three changes have been made:
6328 (a) There was a serious bug; a negative response to MAIL caused the whole
6329 recipient domain to be cached as invalid, thereby blocking all messages
6330 to all local parts at the same domain, from all senders. This bug has
6331 been fixed. The domain is no longer cached after a negative response to
6332 MAIL if the sender used is not empty.
6334 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6337 (c) A new callout option, "use_sender" has been added for people who want
6338 the modified behaviour.
6344 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6347 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6348 indeed breaks things for older releases.
6350 3. Added additional logging to the case where there is a problem reading data
6351 from a filter that is running in a subprocess using a pipe, in order to
6352 try to track down a specific problem.
6354 4. Testing facility fudge: when running in the test harness and attempting
6355 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6356 getting "No route to host". Convert this to a timeout.
6358 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6361 6. Some OS don't have socklen_t but use size_t instead. This affects the
6362 fifth argument of getsockopt() amongst other things. This is now
6363 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6364 can be set for individual OS. I have set it for SunOS5, OSF1, and
6365 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6366 some earlier ones do not.
6368 7. Change 4.30/15 was not doing the test caselessly.
6370 8. The standard form for an IPv6 address literal was being rejected by address
6371 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6372 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6373 this, as well as the form without the "IPv6" on the front (but only when
6374 address literals are enabled, of course).
6376 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6378 10. Exim crashed if a message with an empty sender address specified by -f
6379 encountered a router with an errors_to setting. This could be provoked only
6380 by a command such as
6384 where an empty string was supplied; "<>" did not hit this bug.
6386 11. Installed PCRE release 4.5.
6388 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6389 remained set. It is now erased.
6391 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6392 times from message ids (which are base 36 rather than the normal 62).
6394 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6395 were being counted as actual protocol errors, and logged if the log
6396 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6397 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6398 and DATA following a set of rejected RCPTs do not count as protocol errors.
6399 In other words, Exim assumes they were pipelined, though this may not
6400 actually be the case. Of course, in all cases the client gets an
6401 appropriate error code.
6403 15. If a lookup fails in an ACL condition, a message about the failure may
6404 be available; it is used if testing the ACL cannot continue, because most
6405 such messages specify what the cause of the deferral is. However, some
6406 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6407 that caused an old message to be retained and used if a later statement
6408 caused a defer, replacing the real cause of the deferral.
6410 16. If an IP address had so many PTR records that the DNS lookup buffer
6411 was not large enough to hold them, Exim could crash while trying to process
6412 the truncated data. It now detects and logs this case.
6414 17. Further to 4.21/58, another change has been made: if (and only if) the
6415 first line of a message (the first header line) ends with CRLF, a bare LF
6416 in a subsequent header line has a space inserted after it, so as not to
6417 terminate the header.
6419 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6420 unnecessarily, used atoi() instead of strtol(), and didn't check the
6421 termination when getting file sizes from file names by regex.
6423 19. Completely re-implemented the support for maildirsize files, in the light
6424 of a number of problems with the previous contributed implementation
6425 (4.30/29). In particular:
6427 . If the quota is zero, the maildirsize file is maintained, but no quota is
6430 . If the maildir directory does not exist, it is created before any attempt
6431 to write a maildirsize file.
6433 . The quota value in the file is just a cache; if the quota is changed in
6434 the transport, the new value overrides.
6436 . A regular expression is available for excluding directories from the
6439 20. The autoreply transport checks the characters in options that define the
6440 message's headers; it allows continued headers, but it was checking with
6441 isspace() after an embedded newline instead of explicitly looking for a
6444 21. If all the "regular" hosts to which an address was routed had passed their
6445 expiry times, and had not reached their retry times, the address was
6446 bounced, even if fallback hosts were defined. Now Exim should go on to try
6449 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6450 equivalent code in the SMTP transport. Some hosts send humungous responses
6451 to HELO/EHLO, more than 1024 it seems.
6453 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6454 gives compiler warnings in some environments. I've now done it "properly",
6457 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6458 (because of problems with the built-in one) was declared to return uschar *
6459 instead of char *, causing compiler failure.
6461 25. Fixed a file descriptor leak when processing alias/forward files.
6463 26. Fixed a minor format string issue in dbfn.c.
6465 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6467 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6468 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6471 29. When a sender address is verified, it is cached, to save repeating the test
6472 when there is more than one recipient in a message. However, when the
6473 verification involves a callout, it is possible for different callout
6474 options to be set for different recipients. It is too complicated to keep
6475 track of this in the cache, so now Exim always runs a verification when a
6476 callout is required, relying on the callout cache for the optimization.
6477 The overhead is duplication of the address routing, but this should not be
6480 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6481 to be verified with callout=postmaster, and the main callout worked but the
6482 postmaster check failed, the verification correctly failed. However, if a
6483 subsequent RCPT command asked for sender verification *without* the
6484 postmaster check, incorrect caching caused this verification also to fail,
6487 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6488 it was not caching the DNS options (qualify_single, search_parents) that
6489 were used when the lookup failed. A subsequent lookup with different
6490 options therefore always gave the same answer, though there were cases
6491 where it should not have. (Example: a "domains = !$mx_any" option on a
6492 dnslookup router: the "domains" option is always processed without any
6493 widening, but the router might have qualify_single set.) Now Exim uses the
6494 cached value only when the same options are set.
6496 32. Added John Jetmore's "exipick" utility to the distribution.
6498 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6499 than a timeout (e.g. a certificate is required, and is not provided), an
6500 Exim server now closes the connection immediately. Previously it waited for
6501 the client to close - but if the client is SSL, it seems that they each
6502 wait for each other, leading to a delay before one of them times out.
6504 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6505 maintained 0.8.x compatibility because I don't think many are using it, and
6506 it is clearly obsolete.
6508 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6511 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6512 client certificate was expired. A simple patch fixes this, though I don't
6513 understand the full logic of why the verify callback is called multiple
6516 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6517 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6518 which causes problems with some clients (such as the Certicom SSL Plus
6519 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6520 disables the coutermeasure allowing Eudora to connect."
6522 38. Exim was not checking that a write() to a log file succeeded. This could
6523 lead to Bad Things if a log got too big, in particular if it hit a file
6524 size limit. Exim now panics and dies if it cannot write to a log file, just
6525 as it does if it cannot open a log file.
6527 39. Modified OS/Makefile-Linux so that it now contains
6529 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6531 The two -D definitions ensure that Exim is compiled with large file
6532 support, which makes it possible to handle log files that are bigger than
6535 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6536 instance) a domain was checked against a named list that involved a lookup,
6537 causing $domain_data to be set, then another domain was checked against the
6538 same list, then the first domain was re-checked, the value of $domain_data
6539 after the final check could be wrong. In particular, if the second check
6540 failed, it could be set empty. This bug probably also applied to
6543 41. The strip_trailing_dot option was not being applied to the address given
6544 with the -f command-line option.
6546 42. The code for reading a message's header from the spool was incrementing
6547 $received_count, but never initializing it. This meant that the value was
6548 incorrect (doubled) while delivering a message in the same process in which
6549 it was received. In the most common configuration of Exim, this never
6550 happens - a fresh exec is done - but it can happen when
6551 deliver_drop_privilege is set.
6553 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6554 it now includes up to 150 characters of the unexpected data in the log
6557 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6558 and building data strings. The size of both of these buffers was 10 000
6559 bytes - far larger than anybody would *ever* want, thought I. Needless to
6560 say, somebody hit the limit. I have increased the maximum line length to
6561 20 000 and the maximum data length of concatenated lines to 100 000. I have
6562 also fixed two bugs, because there was no checking on these buffers. Tsk,
6563 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6564 buffer is too small.
6566 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6567 lsearch lookups. Now it does.
6569 46. When parsing a route_list item in a manualroute router, a fixed-length
6570 buffer was used for the list of hosts. I made this 1024 bytes long,
6571 thinking that nobody would ever have a list of hosts that long. Wrong.
6572 Somebody had a whole pile of complicated expansion conditions, and the
6573 string was silently truncated, leading to an expansion error. It turns out
6574 that it is easier to change to an unlimited length (owing to other changes
6575 that have happened since this code was originally written) than to build
6576 structure for giving a limitation error. The length of the item that
6577 expands into the list of hosts is now unlimited.
6579 47. The lsearch lookup could not handle data where the length of text line was
6580 more than 4095 characters. Such lines were truncated, leading to shortened
6581 data being returned. It should now handle lines of any length.
6583 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6584 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6587 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6588 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6589 no longer happens in comment lines. A list of replacements is now placed
6590 at the head of all of the source files, except those whose only change is
6591 to replace PERL_COMMAND in the very first #! line.
6593 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6594 messages on the queue, with a bottom-up merge sort, using code contributed
6595 by Michael Haardt. This should make operations like -bp somewhat faster on
6596 large queues. It won't affect queue runners, except when queue_run_in_order
6599 51. Installed eximstats 1.31 in the distribution.
6601 52. Added support for SRV lookups to the dnslookup router.
6603 53. If an ACL referred to $message_body or $message_body_end, the value was not
6604 reset for any messages that followed in the same SMTP session.
6606 54. The store-handling optimization for building very long strings was not
6607 differentiating between the different store pools. I don't think this
6608 actually made any difference in practice, but I've tidied it.
6610 55. While running the routers to verify a sender address, $sender_address
6611 was still set to the sender address. This is wrong, because when routing to
6612 send a bounce to the sender, it would be empty. Therefore, I have changed
6613 it so that, while verifying a sender address, $sender_address is set to <>.
6614 (There is no change to what happens when verifying a recipient address.)
6616 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6617 target A or AAAA records (if not already returned) without resetting the
6618 qualify_single or search_parents options of the DNS resolver. These are
6619 inappropriate in this case because the targets of MX and SRV records must
6620 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6621 target that, when qualified, matched something in the local domain. These
6622 two options are now turned off when doing these lookups.
6624 57. It seems that at least some releases of Reiserfs (which does not have the
6625 concept of a fixed number of inodes) returns zero and not -1 for the
6626 number of available inodes. This interacted badly with check_spool_inodes,
6627 which assumed that -1 was the "no such thing" setting. What I have done is
6628 to check that the total number of inodes is greater than zero before doing
6629 the test of how many are available.
6631 58. When a "warn" ACL statement has a log_message modifier, the message is
6632 remembered, and not repeated. This is to avoid a lot of repetition when a
6633 message has many recipients that cause the same warning to be written.
6634 However, Exim was preserving the list of already written lines for an
6635 entire SMTP session, which doesn't seem right. The memory is now reset if a
6636 new message is started.
6638 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6639 debugging output unless log_rewrite was also set.
6641 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6642 of (int)(handle) when we know that handle contains (void *)(-1).
6644 61. The Exim daemon panic-logs an error return when it closes the incoming
6645 connection. However "connection reset by peer" seems to be common, and
6646 isn't really an error worthy of noting specially, so that particular error
6649 62. When Exim is trying to find all the local interfaces, it used to panic and
6650 die if the ioctl to get the interface flags failed. However, it seems that
6651 on at least one OS (Solaris 9) it is possible to have an interface that is
6652 included in the list of interfaces, but for which you get a failure error
6653 for this call. This happens when the interface is not "plumbed" into a
6654 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6655 failure of the "get flags" call assumes that the interface is down.
6657 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6658 makes life easier for people who are doing arithmetic on fields extracted
6659 from dates, where you often get leading zeros that should not be
6660 interpreted as octal.
6662 64. Added qualify_domain to the redirect router, to override the global
6665 65. If a pathologically long header line contained very many addresses (the
6666 report of this problem mentioned 10 000) and each of them was rewritten,
6667 Exim could use up a very large amount of memory. (It kept on making new
6668 copies of the header line as it rewrote, and never released the old ones.)
6669 At the expense of a bit more processing, the header rewriting function has
6670 been changed so that it no longer eats memory in this way.
6672 66. The generation of the Received: header has been moved from the time that a
6673 message starts to be received, to the time that it finishes. The timestamp
6674 in the Received: header should now be very close to that of the <= log
6675 line. There are two side-effects of this change:
6677 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6678 logged header lines no longer include the local Received: line, because
6679 it has not yet been created. The same applies to a copy of the message
6680 that is returned to a non-SMTP sender when a message is rejected.
6682 (b) When a filter file is tested using -bf, no additional Received: header
6683 is added to the test message. After some thought, I decided that this
6686 This change does not affect the value of $received_for. It is still set
6687 after address rewriting, but before local_scan() is called.
6689 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6691 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6692 gave an unhelpful panic error message, and a defer error. I have managed to
6693 change this behaviour so that it now rejects any supplied certificate,
6694 which seems right, as the list of acceptable certificates is empty.
6696 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6697 gave an unhelpful defer error. I have not managed to make this reject any
6698 supplied certificates, but the error message it gives is "no certificate
6699 supplied", which is not helpful.
6701 70. exigrep's output now also includes lines that are not associated with any
6702 message, but which match the given pattern. Implemented by a patch from
6703 Martin Sluka, which also tidied up the Perl a bit.
6705 71. Recipient callout verification, like sender verification, was using <> in
6706 the MAIL FROM command. This isn't really the right thing, since the actual
6707 sender may affect whether the remote host accepts the recipient or not. I
6708 have changed it to use the actual sender in the callout; this means that
6709 the cache record is now keyed on a recipient/sender pair, not just the
6710 recipient address. There doesn't seem to be a real danger of callout loops,
6711 since a callout by the remote host to check the sender would use <>.
6712 [SEE ABOVE: changed after hitting problems.]
6714 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6715 temporary errors. However, in the case of such a code being given after
6716 the end of a data transmission (i.e. after ".") Exim was failing to write
6717 a retry record for the message. (Yes, there was some broken host that was
6718 actually sending 8xx at this point.)
6720 73. An unknown lookup type in a host list could cause Exim to panic-die when
6721 the list was checked. (An example that provoked this was putting <; in the
6722 middle of a list instead of at the start.) If this happened during a DATA
6723 ACL check, a -D file could be left lying around. This kind of configuration
6724 error no longer causes Exim to die; instead it causes a defer error. The
6725 incident is still logged to the main and panic logs.
6727 74. Buglet left over from Exim 3 conversion. The message "too many messages
6728 in one connection" was written to the rejectlog but not the mainlog, except
6729 when address rewriting (yes!) was being logged.
6731 75. Added write_rejectlog option.
6733 76. When a system filter was run not as root (that is, when system_filter_user
6734 was set), the values of the $n variables were not being returned to the
6735 main process; thus, they were not subsequently available in the $sn
6738 77. Added +return_path_on_delivery log selector.
6740 78. A connection timeout was being treated differently from recipients deferred
6741 when testing hosts_max_try with a message that was older than the host's
6742 retry timeout. (The host should not be counted, thus allowing all hosts to
6743 be tried at least once before bouncing.) This may have been the cause of an
6744 occasionally reported bug whereby a message would remain on the queue
6745 longer than the retry timeout, but would be bounced if a delivery was
6746 forced. I say "may" because I never totally pinned down the problem;
6747 setting up timeout/retry tests is difficult. See also the next item.
6749 79. The ultimate address timeout was not being applied to errors that involved
6750 a combination of host plus message (for example, a timeout on a MAIL
6751 command). When an address resolved to a number of possible hosts, and they
6752 were not all tried for each delivery (e.g. because of hosts_max_try), a
6753 message could remain on the queue longer than the retry timeout.
6755 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6758 81. Fixed an obscure SMTP outgoing bug which required at least the following
6759 conditions: (a) there was another message waiting for the same server;
6760 (b) the server returned 5xx to all RCPT commands in the first message so
6761 that the message was not completed; (c) the server dropped the connection
6762 or gave a negative response to the RSET that Exim sends to abort the
6763 transaction. The observed case was a dropped connection after DATA that had
6764 been sent in pipelining mode. That is, the server had advertised PIPELINING
6765 but was not implementing it correctly. The effect of the bug was incorrect
6766 behaviour, such as trying another host, and this could lead to a crash.
6772 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6773 and daemon.c were passed as pointers to ints; they should have been
6774 pointers to socklen_t variables (which are typically unsigned ints).
6776 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6779 3. Fixed a really odd bug that affected only the testing scheme; patching a
6780 certain fixed string in the binary changed the value of another string that
6781 happened to be identical to the end of the original first string.
6783 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6784 name", it returns that address as the IP address. On some operating
6785 systems (e.g. Solaris), it also passes back the IP address string as the
6786 "host name". However, on others (e.g. Linux), it passes back an empty
6787 string. Exim wasn't checking for this, and was changing the host name to an
6788 empty string, assuming it had been canonicalized.
6790 5. Although rare, it is permitted to have more than one PTR record for a given
6791 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6792 all the names associated with an address, because they do in Solaris.
6793 However, it seems that they do not in Linux for data that comes from the
6794 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6795 I found this out when I moved to a new Linux workstation and tried to run
6796 the Exim test suite.
6798 To get round this problem I have changed the code so that it now does its
6799 own call to the DNS to look up PTR records when searching for a host name.
6800 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6801 addresses that are only in /etc/hosts are still found.
6803 This behaviour is, however, controlled by an option called host_lookup_
6804 order, which defaults to "bydns:byaddr". If people want to use the other
6805 order, or indeed, just use one or the other means of lookup, they can
6806 specify it in this variable.
6808 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6809 some operating systems, this comes back from gethostbyaddr() as an empty
6810 string, and this is what Exim used to test for. However, it seems that in
6811 other systems, "." is yielded. Exim now tests for this case too.
6813 7. The values of check_spool_space and check_log_space are now held internally
6814 as a number of kilobytes instead of an absolute number of bytes. If a
6815 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6816 kilobyte. This means that much larger values can be stored.
6818 8. Exim monitor: an attempt to get the action menu when not actually pointing
6819 at a message produces an empty menu entitled "No message selected". This
6820 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6821 no entries in it ("Shell widget menu has zero width and/or height"). So I
6822 have added a single, blank menu entry in this case.
6824 9. Added ${quote_local_part.
6826 10. MIME decoding is now applied to the contents of Subject: header lines when
6829 11. Now that a reference to $sender_host_address automatically causes a reverse
6830 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6831 host lookup before query-style lookups in lists that might use this
6832 variable. This has therefore been abolished, and the "net-" prefix is no
6833 longer necessary for query-style lookups.
6835 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6836 have been a typo for LFLAGS, so it has been changed.
6838 13. The install script calls Exim with "-C /dev/null" in order to find the
6839 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6840 to be output. However, since Exim outputs its version number before the
6841 error, it didn't break the script. It just looked ugly. I fixed this by
6842 always allowing "-C /dev/null" if the caller is root.
6844 14. Ignore overlarge ACL variable number when reading spool file - insurance
6845 against a later release with more variables having written the file.
6847 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6848 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6849 this, as well as the form without the "IPv6" on the front.
6851 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6852 OS/Makefile-Darwin file.
6854 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6855 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6857 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6858 variable. However, some OS use ints for inodes. I've added cast to long int
6859 to get rid of the compiler warning.
6861 19. I had forgotten to lock out "/../" in configuration file names when
6862 ALT_CONFIG_PREFIX was set.
6864 20. Routers used for verification do not need to specify transports. However,
6865 if such a router generated a host list, and callout was configured, Exim
6866 crashed, because it could not find a port number from the (non-existent)
6867 transport. It now assumes port 25 in this circumstance.
6869 21. Added the -t option to exigrep.
6871 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6872 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6873 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6874 them is set, LOOKUP_LSEARCH is forced.
6876 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6878 24. Added sender and host information to the "rejected by local_scan()" log
6879 line; previously there was no indication of these.
6881 25. Added .include_if_exists.
6883 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6884 Linux. It turns out that not all file systems support this. Apparently some
6885 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6886 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6887 not supported on the file descriptor, is now ignored when Exim is trying to
6888 sync a directory. This applies only to Linux.
6890 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6892 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6893 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6896 29. Added support for maildirsize files from supplied patch (modified a bit).
6898 30. The use of :fail: followed by an empty string could lead Exim to respond to
6899 sender verification failures with (e.g.):
6901 550 Verification failed for <xxx>
6902 550 Sender verify failed
6904 where the first response line was missing the '-' that indicates it is not
6905 the final line of the response.
6907 31. The loop for finding the name of the user that called Exim had a hardwired
6908 limit of 10; it now uses the value of finduser_retries, which is used for
6909 all other user lookups.
6911 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6914 33. Exim was neglecting to zero errno before one call of strtol() when
6915 expanding a string and expecting an integer value. On some systems this
6916 resulted in spurious "integer overflow" errors. Also, it was casting the
6917 result into an int without checking.
6919 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6920 did not work. The code looks as if it has *never* worked, though it appears
6921 to have been documented since at least release 1.62. I have made it work.
6923 35. The "timeout_DNS" error in retry rules, also documented since at least
6924 1.62, also never worked. As it isn't clear exactly what this means, and
6925 clearly it isn't a major issue, I have abolished the feature by treating it
6926 as "timeout", and writing a warning to the main and panic logs.
6928 36. The display of retry rules for -brt wasn't always showing the error code
6931 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6932 timeout_connect_A, timeout_connect_MX.
6934 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6935 to the empty sender.
6937 39. The daemon was not analysing the content of -oX till after it had closed
6938 stderr and disconnected from the controlling terminal. This meant that any
6939 syntax errors were only noted on the panic log, and the return code from
6940 the command was 0. By re-arranging the code a little, I've made the
6941 decoding happen first, so such errors now appear on stderr, and the return
6942 code is 1. However, the actual setting up of the sockets still happens in
6943 the disconnected process, so errors there are still only recorded on the
6946 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6947 connections (as happens on some IP stacks) was logged at start up time as
6948 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6949 it from "IPv6 and IPv4", which means that two separate sockets are being
6952 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6953 says whether AF_INET or AF_INET6 was passed as an argument.
6955 42. Exiwhat output was messed up when time zones were included in log
6958 43. Exiwhat now gives more information about the daemon's listening ports,
6959 and whether -tls-on-connect was used.
6961 44. The "port" option of the smtp transport is now expanded.
6963 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6964 silently ignored. Now an error message is written to the main and panic
6967 46. There's a new ACL modifier called "logwrite" which writes to a log file
6968 as soon as it is encountered.
6970 47. Added $local_user_uid and $local_user_gid at routing time.
6972 48. Exim crashed when trying to verify a sender address that was being
6975 49. Exim was recognizing only a space character after ".include". It now also
6976 recognizes a tab character.
6978 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6979 extracting the relevant information from the specification. The man page no
6980 longer contains scrambled data for the -d option, and I've added a section
6981 at the front about calling Exim under different names.
6983 51. Added "extra_headers" argument to the "mail" command in filter files.
6985 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6988 53. Installed eximstats 1.29.
6990 54. Added transport_filter_timeout as a generic transport option.
6992 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6993 Cc: header lines. This was required by RFC 822, but it not required by RFC
6996 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6997 incoming messages that did not have them. Now it does so only if the
6998 message originates locally, that is, if there is no associated remote host
6999 address. When Resent- header lines are present, this applies to the Resent-
7000 lines rather than the non-Resent- lines.
7002 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7003 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7005 58. Messages for configuration errors now include the name of the main
7006 configuration file - useful now that there may be more than one file in a
7007 list (.included file names were always shown).
7009 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7010 for those rare installations that do not start the daemon as root or run it
7011 setuid root. I've cut out the call to initgroups() if the daemon is not
7014 60. The Exim user and group can now be bound into the binary as text strings
7015 that are looked up at the start of Exim's processing.
7017 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7019 62. Added $mailstore_basename variable.
7021 63. Installed patch to sieve.c from Michael Haardt.
7023 64. When Exim failed to open the panic log after failing to open the main log,
7024 the original message it was trying to log was written to stderr and debug
7025 output, but if they were not available (the usual case in production), it
7026 was lost. Now it is written to syslog before the two lines that record the
7027 failures to open the logs.
7029 65. Users' Exim filters run in subprocesses under the user's uid. It is
7030 possible for a "deliver" command or an alias in a "personal" command to
7031 provoke an address rewrite. If logging of address rewriting is configured,
7032 this fails because the process is not running as root or exim. There may be
7033 a better way of dealing with this, but for the moment (because 4.30 needs
7034 to be released), I have disabled address rewrite logging when running a
7035 filter in a non-root, non-exim process.
7041 1. The buildconfig auxiliary program wasn't quoting the value set for
7042 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7043 not defined. This bug was masked in 4.22 by the effect that was fixed in
7046 2. Some messages that were rejected after a message id was allocated were
7047 shown as "incomplete" by exigrep. It no longer does this for messages that
7048 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7050 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7051 have allow_domain_literals set, the ID did not get logged in the <= line.
7052 Domain literals are now always recognized in Message-ID: header lines.
7054 4. The first argument for a ${extract expansion item is the key name or field
7055 number. Leading and trailing spaces in this item were not being ignored,
7056 causing some misleading effects.
7058 5. When deliver_drop_privilege was set, single queue runner processes started
7059 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7060 same command in the process it spins off) were not dropping privilege.
7062 6. When the daemon running as "exim" started a queue runner, it always
7063 re-executed Exim in the spun-off process. This is a waste of effort when
7064 deliver_drop_privilege is set. The new process now just calls the
7065 queue-runner function directly.
7071 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7074 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7075 ignored. Though the use of -oP was forcing the writing of a pid file, it
7076 was always written to the default place.
7078 3. If the message "no IP address found for host xxxx" is generated during
7079 incoming verification, it is now followed by identification of the incoming
7080 connection (so you can more easily find what provoked it).
7082 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7084 5. Added some features to "harden" Exim a bit more against certain attacks:
7086 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7087 be put in Local/Makefile. This is like the never_users runtime option,
7088 but it cannot be overridden. The default setting is "root".
7090 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7091 prefix string with which any file named in a -C command line option
7094 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7095 is retained for -C and -D only if the caller of Exim is root. Without
7096 it, the exim user may also use -C and -D and retain privilege.
7098 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7099 command line option is disabled.
7101 6. Macro names set by the -D option must start with an upper case letter, just
7102 like macro names defined in the configuration file.
7104 7. Added "dereference=" facility to LDAP.
7106 8. Two instances of the typo "uknown" in the source files are fixed.
7108 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7109 the Configure-Makefile script screwed up while processing it.
7111 10. Incorporated PCRE 4.4.
7113 11. The SMTP synchronization check was not operating right at the start of an
7114 SMTP session. For example, it could not catch a HELO sent before the client
7115 waited for the greeting. There is now a check for outstanding input at the
7116 point when the greeting is written. Because of the duplex, asynchronous
7117 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7118 way, but not yet received, when the check is performed.
7120 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7121 on TCP/IP sockets, because this apparently causes some broken clients to
7124 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7125 unchanged) from the Cygwin maintainer.
7127 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7128 was supported instead of testing for mbx. Effectively a typo.
7130 15. The spa authenticator server code was not checking that the input it
7131 received was valid base64.
7133 16. The debug output line for the "set" modifier in ACLs was not showing the
7134 name of the variable that was being set.
7136 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7138 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7139 lookup on its own. Something else was needed to trigger the lookup. For
7140 example, a match in host_lookup or the need for a host name in a host list.
7141 Now, if $sender_host_name is referenced and the host name has not yet been
7142 looked up, a lookup is performed. If the lookup fails, the variable remains
7143 empty, and $host_lookup_failed is set to "1".
7145 19. Added "eqi" as a case-independent comparison operator.
7147 20. The saslauthd authentication condition could segfault if neither service
7148 nor realm was specified.
7150 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7151 error message that was logged was misleading, and incoming SMTP
7152 connections were dropped. The message is now more accurate, and temporary
7153 errors are given to SMTP connections.
7155 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7156 (except RSET) with a 421 error, until QUIT is received. However, it was
7157 failing to send a response to QUIT.
7159 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7160 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7161 ACL is now run after the helo_try_verify_hosts code.
7163 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7164 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7165 case-independent, so other case variants are also recognized). Apparently
7166 some systems use these upper case variants.
7168 25. If more than two messages were waiting for the same host, and a transport
7169 filter was specified for the transport, Exim sent two messages over the
7170 same TCP/IP connection, and then failed with "socket operation on non-
7171 socket" when it tried to send the third.
7173 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7175 27. The extern definition of crypt16() in expand.c was not being excluded when
7176 the OS had its own crypt16() function.
7178 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7179 as a preferred synonym for return_size_limit, both as an option and as an
7182 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7184 30. Changed the default configuration ACL to relax the local part checking rule
7185 for addresses that are not in any local domains. For these addresses,
7186 slashes and pipe symbols are allowed within local parts, but the sequence
7187 /../ is explicitly forbidden.
7189 31. SPA server authentication was not clearing the challenge buffer before
7192 32. log_message in a "warn" ACL statement was writing to the reject log as
7193 well as to the main log, which contradicts the documentation and doesn't
7194 seem right (because no rejection is happening). So I have stopped it.
7196 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7197 However, I am unable to do any testing of this.
7199 34. Fixed an infelicity in the appendfile transport. When checking directories
7200 for a mailbox, to see if any needed to be created, it was accidentally
7201 using path names with one or more superfluous leading slashes; tracing
7202 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7204 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7205 contents are added to the log line that is written for every discarded
7206 recipient. (Previously a log_message setting was ignored.)
7208 36. The ${quote: operator now quotes the string if it is empty.
7210 37. The install script runs exim in order to find its version number. If for
7211 some reason other than non-existence or emptiness, which it checks, it
7212 could not run './exim', it was installing it with an empty version number,
7213 i.e. as "exim-". This error state is now caught, and the installation is
7216 38. An argument was missing from the function that creates an error message
7217 when Exim fails to connect to the socket for saslauthd authentication.
7218 This could cause Exim to crash, or give a corrupted message.
7220 39. Added isip, isip4, and isip6 to ${if conditions.
7222 40. The ACL variables $acl_xx are now saved with the message, and can be
7223 accessed later in routers, transports, and filters.
7225 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7226 strings in the file are not string-expanded.
7228 42. If a MAIL command specified a SIZE value that was too large to fit into an
7229 int variable, the check against message_size_limit failed. Such values are
7230 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7231 day this will have to be increased, but I don't think I want to be around
7232 when emails are that large.
7239 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7240 iconv() is not standard in FreeBSD.
7242 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7243 IPv6 enabled. The observed symptom was a segmentation fault on return from
7244 the function os_common_find_running_interfaces() in src/os.c.
7246 3. In the check_special_case() function in daemon.c I had used "errno" as an
7247 argument name, which causes warnings on some systems. This was basically a
7248 typo, since it was named "eno" in the comments!
7250 4. The code that waits for the clock to tick (at a resolution of some fraction
7251 of a second) so as to ensure message-id uniqueness was always waiting for
7252 at least one whole tick, when it could have waited for less. [This is
7253 almost certainly not relevant at current processor speeds, where it is
7254 unlikely to ever wait at all. But we try to future-proof.]
7256 5. The function that sleeps for a time interval that includes fractions of a
7257 second contained a race. It did not block SIGALRM between setting the
7258 timer, and suspending (a couple of lines later). If the interval was short
7259 and the sigsuspend() was delayed until after it had expired, the suspension
7260 never ended. On busy systems this could lead to processes getting stuck for
7263 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7264 process, before it forks any delivery processes. The open lookup caching
7265 mechanism meant that the open file or database connection was passed into
7266 the delivery process. The problem was that delivery processes always tidy
7267 up cached lookup data. This could cause a problem for the next delivery
7268 process started by the queue runner, because the external queue runner
7269 process does not know about the closure. So the next delivery process
7270 still has data in the lookup cache. In the case of a file lookup, there was
7271 no problem because closing a file descriptor in a subprocess doesn't affect
7272 the parent. However, if the lookup was caching a connection to a database,
7273 the connection was closed, and the second delivery process was likely to
7274 see errors such as "PGSQL: query failed: server closed the connection
7275 unexpectedly". The problem has been fixed by closing all cached lookups
7276 in a queue runner before running a delivery process.
7278 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7279 seem to have the "const" qualifier which it has on other OS. I've
7282 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7283 *of the same type* (client or server) with the same public name that an
7284 error should be diagnosed.
7286 9. When Exim looked up a host name for an IP address, but failed to find the
7287 original IP address when looking up the host name (a safety check), it
7288 output the message "<ip address> does not match any IP for NULL", which was
7289 confusing, to say the least. The bug was that the host name should have
7290 appeared instead of "NULL".
7292 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7293 user that is built into the binary, and the -C or -D options is used, root
7294 privilege is dropped before the configuration file is read. In addition,
7295 logging is switched to stderr instead of the normal log files. If the
7296 configuration then re-defines the Exim user, the unprivileged environment
7297 is probably not what is expected, so Exim logs a panic warning message (but
7300 However, if deliver_drop_privilege is set, the unprivileged state may well
7301 be exactly what is intended, so the warning has been cut out in that case,
7302 and Exim is allowed to try to write to its normal log files.
7308 1. smtp_return_error_details was not giving details for temporary sender
7309 or receiver verification errors.
7311 2. Diagnose a configuration error if two authenticators have the same public
7314 3. Exim used not to create the message log file for a message until the first
7315 delivery attempt. This could be confusing when incoming messages were held
7316 for policy or load reasons. The message log file is now created at the time
7317 the message is received, and an initial "Received" line is written to it.
7319 4. The automatically generated man page for command line options had a minor
7320 bug that caused no ill effects; however, a more serious problem was that
7321 the procedure for building the man page automatically didn't always
7322 operate. Consequently, release 4.20 contains an out-of-date version. This
7323 shouldn't happen again.
7325 5. When building Exim with embedded Perl support, the script that builds the
7326 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7327 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7329 6. The freeze_tell option was not being used for messages that were frozen on
7330 arrival, either by an ACL or by local_scan().
7332 7. Added the smtp_incomplete_transaction log selector.
7334 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7335 was accepting AUTH without a new EHLO.
7337 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7338 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7341 10. Logging of TCP/IP connections (when configured) now happens in the main
7342 daemon process instead of the child process, so that the TCP/IP connection
7343 count is more accurate (but it can never be perfect).
7345 11. The use of "drop" in a nested ACL was not being handled correctly in the
7346 outer ACL. Now, if condition failure induced by the nested "drop" causes
7347 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7348 or "require"), the connection is dropped.
7350 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7351 that yield "discard" can now be used with an "accept" or a "discard" verb,
7352 but an error is generated for any others (because I can't see a useful way
7353 to define what should happen).
7355 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7356 are now processed in the same way as lines in the Exim configuration file.
7357 In particular, continuation lines are supported.
7359 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7361 15. Added -ti meaning -t -i.
7363 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7364 domains, and warn by logging if others are found.
7366 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7367 returned from GIFCONF when Exim is trying to find the list of interfaces on
7368 a host. The code in os.c has been modified to copy each ifreq to an aligned
7369 structure in all cases.
7371 Also, in some cases, the returned ifreq's were being copied to a 'struct
7372 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7373 means the last couple of bytes of an IPv6 address could be chopped if the
7374 ifreq contained only a normal sockaddr (14 bytes storage).
7376 18. Named domain lists were not supported in the hosts_treat_as_local option.
7377 An entry such as +xxxx was not recognized, and was treated as a literal
7380 19. Ensure that header lines added by a DATA ACL are included in the reject log
7381 if the ACL subsequently rejects the message.
7383 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7384 MD5 (which is deprecated).
7386 21. When testing a filter file using -bf, Exim was writing a message when it
7387 took the sender from a "From " line in the message, but it was not doing so
7388 when it took $return_path from a Return-Path: header line. It now does.
7390 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7391 with a valid header line field name (a series of printing characters
7392 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7394 23. Changed "disc" in the source to "disk" to conform to the documentation and
7395 the book and for uniformity.
7397 24. Ignore Sendmail's -Ooption=value command line item.
7399 25. When execve() failed while trying to run a command in a pipe transport,
7400 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7401 could be confused with a return value of 69 from the command itself. This
7402 has been changed to 127, the value the shell returns if it is asked to run
7403 a non-existent command. The wording for the related log line suggests a
7404 non-existent command as the problem.
7406 26. If received_header_text expands to an empty string, do not add a Received:
7407 header line to the message. (Well, it adds a token one on the spool, but
7408 marks it "old" so that it doesn't get used or transmitted.)
7410 27. Installed eximstats 1.28 (addition of -nt option).
7412 28. There was no check for failure on the call to getsockname() in the daemon
7413 code. This can fail if there is a shortage of resources on the system, with
7414 ENOMEM, for example. A temporary error is now given on failure.
7416 29. Contrary to the C standard, it seems that in some environments, the
7417 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7418 program. Exim now does this explicitly; it affects the formatting of
7419 timestamps using strftime().
7421 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7422 complaints. I've now initialized all the variables, to avoid this.
7424 32. Header lines added by a system filter were not being "seen" during
7425 transport-time rewrites.
7427 33. The info_callback() function passed to OpenSSL is set up with type void
7428 (*)(SSL *, int, int), as described somewhere. However, when calling the
7429 function (actually a macro) that sets it up, the type void(*)() is
7430 expected. I've put in a cast to prevent warnings from picky compilers.
7432 34. If a DNS black list lookup found a CNAME record, but there were no A
7433 records associated with the domain it pointed at, Exim crashed.
7435 35. If a DNS black list lookup returned more than one A record, Exim ignored
7436 all but the first. It now scans all returned addresses if a particular IP
7437 value is being sought. In this situation, the contents of the
7438 $dnslist_value variable are a list of all the addresses, separated by a
7441 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7442 to accept a host name and all its aliases if the forward lookup for any of
7443 them yielded the IP address of the incoming connection. Now it accepts only
7444 those names whose forward lookup yields the correct IP address. Any other
7445 names are discarded. This closes a loophole whereby a rogue DNS
7446 administrator could create reverse DNS records to break through a
7447 wildcarded host restriction in an ACL.
7449 37. If a user filter or a system filter that ran in a subprocess used any of
7450 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7451 the wrong values were passed to the pipe command ($thisaddress had the
7452 value of $0, $0 had the value of $1, etc). This bug was introduced by
7453 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7455 38. Improved the line breaking for long SMTP error messages from ACLs.
7456 Previously, if there was no break point between 40 and 75 characters, Exim
7457 left the rest of the message alone. Two changes have been made: (a) I've
7458 reduced the minimum length to 35 characters; (b) if it can't find a break
7459 point between 35 and 75 characters, it looks ahead and uses the first one
7460 that it finds. This may give the occasional overlong line, but at least the
7461 remaining text gets split now.
7463 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7464 file descriptors might be low, and that setting 1000 would always raise it.
7465 It turns out that in some environments, the limit is already over 1000 and
7466 that lowering it causes trouble. So now Exim takes care not to decrease it.
7468 40. When delivering a message, the value of $return_path is set to $sender_
7469 address at the start of routing (routers may change the value). By an
7470 oversight, this default was not being set up when an address was tested by
7471 -bt or -bv, which affected the outcome if any router or filter referred to
7474 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7475 log only when "message" or "log_message" are set. However, if one of the
7476 conditions was an address verification, or a call to a nested ACL, the
7477 messages generated by the underlying test were being passed through. This
7478 no longer happens. The underlying message is available in $acl_verify_
7479 message for both "message" and "log_message" expansions, so it can be
7480 passed through if needed.
7482 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7483 new expansion $bh_ to give the encoded byte string without charset
7484 translation. Translation happens only if iconv() is available; HAVE_ICONV
7485 indicates this at build time. HEADERS_CHARSET gives the charset to
7486 translate to; headers_charset can change it in the configuration, and
7487 "headers charset" can change it in an individual filter file.
7489 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7490 that creates RFC 2047 encoded "words" labels them as that charset instead
7491 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7492 expansion operator; (ii) when Exim creates a From: line for a local
7493 message; (iii) when a header line is rewritten to include a "phrase" part.
7495 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7496 buggy, causing it to skip the first lines of messages whose message ID
7497 ended in 'D'. This would not have bitten before Exim release 4.14, because
7498 message IDs were unlikely to end in 'D' before then. The effect was to have
7499 incorrect size information for certain domains.
7501 45. #include "config.h" was missing at the start of the crypt16.c module. This
7502 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7505 46. If there was a timeout during a "random" callout check, Exim treated it as
7506 a failure of the random address, and carried on sending RSET and the real
7507 address. If the delay was just some slowness somewhere, the response to the
7508 original RCPT would be taken as a response to RSET and so on, causing
7509 mayhem of various kinds.
7511 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7512 when I implemented it. It didn't allow for the fact that some option values
7513 may legitimately be negative (e.g. size_addition), and it didn't even do
7514 the right test for positive values.
7516 48. Domain names in DNS records are case-independent. Exim always looks them up
7517 in lower case. Some resolvers return domain names in exactly the case they
7518 appear in the zone file, that is, they may contain uppercase letters. Not
7519 all resolvers do this - some return always lower case. Exim was treating a
7520 change of case by a resolver as a change of domain, similar to a widening
7521 of a domain abbreviation. This triggered its re-routing code and so it was
7522 trying to route what was effectively the same domain again. This normally
7523 caused routing to fail (because the router wouldn't handle the domain
7524 twice). Now Exim checks for this case specially, and just changes the
7525 casing of the domain that it ultimately uses when it transmits the message
7528 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7531 50. If a filter generated a file delivery with a non-absolute name (possible if
7532 no home directory exists for the router), the forbid_file option was not
7535 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7536 the existing equality matching.
7538 52. Exim was using ints instead of ino_t variables in some places where it was
7539 dealing with inode numbers.
7541 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7542 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7543 called TMPDIR, and if it finds it is different, it changes its value.
7545 54. The smtp_printf() function is now made available to local_scan() so
7546 additional output lines can be written before returning. There is also an
7547 smtp_fflush() function to enable the detection of a dropped connection.
7548 The variables smtp_input and smtp_batched_input are exported to
7551 55. Changed the default runtime configuration: the message "Unknown user"
7552 has been removed from the ACL, and instead placed on the localuser router,
7553 using the cannot_route_message feature. This means that any verification
7554 failures that generate their own messages won't get overridden. Similarly,
7555 the "Unrouteable address" message that was in the ACL for unverifiable
7556 relay addresses has also been removed.
7558 56. Added hosts_avoid_esmtp to the smtp transport.
7560 57. The exicyclog script was not checking for the esoteric option
7561 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7562 will work only if exicyclog is run under the appropriate euid.
7564 58. Following a discussion on the list, the rules by which Exim recognises line
7565 endings on incoming messages have been changed. The -dropcr and drop_cr
7566 options are now no-ops, retained only for backwards compatibility. The
7567 following line terminators are recognized: LF CRLF CR. However, special
7568 processing applies to CR:
7570 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7571 nor a local message in the state where . is a terminator.
7573 (ii) If a bare CR is encountered in a header line, an extra space is added
7574 after the line terminator so as not to end the header. The reasoning
7575 behind this is that bare CRs in header lines are most likely either
7576 to be mistakes, or people trying to play silly games.
7578 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7579 was being incorrectly given as 18 bytes larger than it should have been.
7580 This is a VOB (very old bug).
7582 60. This may never have affected anything current, but just in case it has:
7583 When the local host is found other than at the start of a list of hosts,
7584 the local host, those with the same MX, and any that follow, are discarded.
7585 When the list in question was part of a longer list of hosts, the following
7586 hosts (not currently being processed) were also being discarded. This no
7587 longer happens. I'm not sure if this situation could ever has previously
7590 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7593 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7594 header. This is something that is recommended in a new Internet Draft, and
7595 is something that is documented as being done by Sendmail. There are two
7596 possible values. For messages generated by the autoreply transport, Exim
7599 Auto-Submitted: auto-replied
7601 whereas for all other generated messages (e.g. bounces) it adds
7603 Auto-Submitted: auto-generated
7605 63. The "personal" condition in filters now includes a test for the
7606 Auto-Submitted: header. If it contains the string "auto-" the message it
7607 not considered personal.
7609 64. Added rcpt_include_affixes as a generic transport option.
7611 65. Added queue_only_override (default true).
7613 66. Added the syslog_duplication option.
7615 67. If what should have been the first header line of a message consisted of
7616 a space followed by a colon, Exim was mis-interpreting it as a header line.
7617 It isn't of course - it is syntactically invalid and should therefore be
7618 treated as the start of the message body. The misbehaviour could have
7619 caused a number of strange effects, including loss of data in subsequent
7620 header lines, and spool format errors.
7622 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7623 client host had authenticated. This control can now be exercised by an ACL
7624 for more flexibility.
7626 69. By default, callouts do not happen when testing with -bh. There is now a
7627 variant, -bhc, which does actually run the callout code, including
7628 consulting and updating the callout cache.
7630 70. Added support for saslauthd authentication, courtesy of Alexander
7633 71. If statvfs() failed on the spool or log directories while checking their
7634 size for availability, Exim confusingly gave the error "space shortage".
7635 Furthermore, in debugging mode it crashed with a floating point exception.
7636 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7637 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7638 serious problem, Exim now writes to the main and panic logs when this
7639 happens, with details of the failure. It then refuses to accept the
7640 incoming message, giving the message "spool directory problem" or "log
7641 directory problem" with a 421 code for SMTP messages.
7643 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7644 0, 1, and 2 exist, because some OS complain for execs without them (see
7645 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7646 descriptors. However, the code omitted to check that the open succeeded,
7647 causing mysterious errors if for some reason the permissions on /dev/null
7648 got screwed. Now Exim writes a message to the main and panic logs, and
7649 bombs out if it can't open /dev/null.
7651 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7652 interact so that it is all more flexible. It is supposed to remain
7653 backwards compatible. Also added extra_local_interfaces.
7655 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7656 to bomb out with an assertion failure - to the client this appears as a
7657 connection drop. This problem occurs in the part of the code that was taken
7658 from the Samba project. Fortunately, the assertion is in a very simple
7659 function, so I have fixed this by reproducing the function inline in the
7660 one place where it is called, and arranging for authentication to fail
7661 instead of killing the process with assert().
7663 75. The SPA client code was not working when the server requested OEM rather
7664 than Unicode encoding.
7666 76. Added code to make require_files with a specific uid setting more usable in
7667 the case where statting the file as root fails - usually a non-root-mounted
7668 NFS file system. When this happens and the failure is EACCES, Exim now
7669 forks a subprocess and does the per-uid checking as the relevant uid.
7671 77. Added process_log_path.
7673 78. If log_file_path was not explicitly set, a setting of check_log_space or
7674 check_log_inodes was ignored.
7676 79. If a space check for the spool or log partitions fails, the incident is now
7677 logged. Of course, in the latter case the data may get lost...
7679 80. Added the %p formatting code to string_format() so that it can be used to
7680 print addresses in debug_print(). Adjusted all the address printing in the
7681 debugging in store.c to use %p rather than %d.
7683 81. There was a concern that a line of code in smtp_in.c could overflow a
7684 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7685 initially expressed, the concern was not well-founded, because trailing
7686 spaces are removed early. However, if the trailing spaces were followed by
7687 a NULL, they did not get removed, so the overflow was possible. Two fixes
7690 (a) I re-wrote the offending code in a cleaner fashion.
7691 (b) If an incoming SMTP command contains a NULL character, it is rejected
7694 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7695 runs initgroups(), so that if the Exim user is in any additional groups,
7696 they will be used during message reception.
7702 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob