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 - paths used by single-key lookups
97 Previously this was permitted.
99 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
100 adjusted the size of a major service buffer; this failed because the
101 buffer was in use at the time. Change to a compile-time increase in the
102 buffer size, when this authenticator is compiled into exim.
104 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
105 previous fast-mode was untenable in the face of glibs using mmap to
106 support larger malloc requests.
108 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
109 New values supported, if defined on system where compiled:
110 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
111 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
113 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
114 running a limited number of proceses in parallel, a benefit is gained. The
115 amount varies with the platform hardware and load. The use of the option
116 queue_run_in_order means we cannot do this, as ordering becomes
119 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
120 had introduced a string-copy (for ensuring NUL-termination) which was not
121 appropriate for that case, which can include embedded NUL bytes in the
122 block of data. Investigation showed the copy to actually be needless, the
123 data being length-specified.
125 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
126 done during a receiving connection, and both used TLS, global info was
127 used rather than per-connection info for tracking the state of data
128 queued for transmission. This could result in a connection hang.
130 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
131 Previously, when delivering serveral messages down a single connection
132 only the first would provide a SIZE. This was due to the size information
133 not being properly tracked.
135 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
136 TAI (at 37 seconds currently), pretend to be in UTC for time-related
137 expansion and logging. Previously, spurious values such as a future
138 minute could be seen.
140 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
141 it could crash from a null-deref. This could also affect the
142 ${addresses: } operator and ${readsock } item.
144 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
145 message following a mime one, the variable was not reset.
147 JH/30 When an pipelined-connect fails at the first response, assume incorrect
148 cached capability (perhaps the peer reneged?) and immediately retry in
151 JH/31 Fix spurious detection of timeout while writing to transport filter.
153 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
154 an attempt to copy the string was made before checking it.
156 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
157 taint of the lookup key was maintained; we now regard the presence in the
158 filesystem as sufficient validation.
160 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
161 argument is supplied.
163 JH/35 The dsearch lookup now requires that the directory is an absolute path.
164 Previously this was not checked, and nonempty relative paths made an
165 access under Exim's current working directory.
167 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
168 Previously no event was raised.
170 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
171 parameter supplied by the sender MAIL FROM command. Previously it was
172 ignored, and only the check_spool_space option value for the required
175 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
176 the size of the signing public-key. Previously it was instead giving
177 the size of the signature hash.
179 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
180 the default. See the (new) dkim_verify_min_keysizes option.
182 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
183 and an ACL use a lookup for checking either the local_part or domain,
184 stale data could be accessed. Ensure that variable references are
185 dropped between messages.
191 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
192 side, to decode a TLS 1.3 packet capture.
194 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
195 Previously the default library behaviour applied, sending two, each in
198 JH/03 Debug output for ACL now gives the config file name and line number for
201 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
203 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
205 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
206 buffer overrun for (non-chunking) other transports.
208 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
209 TLS1.3, means that a server rejecting a client certificate is not visible
210 to the client until the first read of encrypted data (typically the
211 response to EHLO). Add detection for that case and treat it as a failed
212 TLS connection attempt, so that the normal retry-in-clear can work (if
213 suitably configured).
215 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
216 and/or domain. Found and fixed by Jason Betts.
218 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
219 configuration). If a CNAME target was not a wellformed name pattern, a
222 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
223 the OS reports them interleaved with other addresses.
225 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
226 used both for input and for a verify callout, both encrypted, SMTP
227 responses being sent by the server could be lost. This resulted in
228 dropped connections and sometimes bounces generated by a peer sending
231 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
232 string. Previously it was possible to cause undefined behaviour in a
233 library routine (usually a crash). Found by "zerons".
235 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
238 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
239 API was removed, so update to use the newer ones.
241 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
242 any timeout set, is taking a long time. Previously we would hang on to a
243 rotated logfile "forever" if the input was arriving with long gaps
244 (a previous attempt to fix addressed lack, for a long time, of initial
247 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
248 shared (NFS) environment. The length of the tempfile name is now
249 4 + 16 ("hdr.$message_exim_id") which might break on file
250 systems which restrict the file name length to lower values.
253 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
254 shared (NFS) environment.
256 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
257 did for all versions <4.90). Notably -M, -m, --invert, -I may be
260 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
261 on some platforms for bit 31.
263 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
264 to changes apparently associated with TLS1.3 handling some of the APIs
265 previously used were either nonfunctional or inappropriate. Strings
266 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
267 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
268 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
269 This affects log line X= elements, the $tls_{in,out}_cipher variables,
270 and the use of specific cipher names in the encrypted= ACL condition.
272 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
274 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
275 verification result was not updated unless hosts_require_ocsp applied.
277 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
278 queue_list_requires_admin set to false, non-admin users were denied the
281 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
282 directory-of-certs mode. Previously they were advertised despite the
285 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
286 A single TCP connection by a client will now hold a TLS connection open
287 for multiple message deliveries, by default. Previoud the default was to
290 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
291 default. If built with the facility, DANE will be used. The facility
292 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
294 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
295 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
296 must be defined and you must still, unless you define DISABLE_TLS, manage
297 the the include-dir and library-file requirements that go with that
298 choice. Non-TLS builds are still supported.
300 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
303 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
304 default. If the platform supports and has the facility enabled, it will
305 be requested on all coneections.
307 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
308 controlled by the build-time option SUPPORT_PIPE_CONNECT.
310 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
312 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
313 success-DSN messages. Previously the From: header was always the default
314 one for these; the option was ignored.
316 JH/28 Fix the timeout on smtp response to apply to the whole response.
317 Previously it was reset for every read, so a teergrubing peer sending
318 single bytes within the time limit could extend the connection for a
319 long time. Credit to Qualsys Security Advisory Team for the discovery.
321 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
322 delivery address, which leaked information of the results of local
323 forwarding. Change to the original envelope recipient address, per
326 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
327 requested. Previously not bounce was generated and a log entry of
328 error ignored was made.
330 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
332 JH/32 Introduce a general tainting mechanism for values read from the input
333 channel, and values derived from them. Refuse to expand any tainted
334 values, to catch one form of exploit.
336 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
337 was unused and the unexpanded text used for the test. Found and
338 fixed by Ruben Jenster.
340 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
341 an attempt to use a TLS library read routine dereffed a nul pointer,
344 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
345 them in our smtp response.
347 JH/36 Have the general router option retry_use_local_part default to true when
348 any of the restrictive preconditions are set (to anything). Previously it
349 was only for check_local user. The change removes one item of manual
350 configuration which is required for proper retries when a remote router
351 handles a subset of addresses for a domain.
353 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
354 link count into consideration.
356 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
357 caused the extension of big_buffer, the following lines were ignored.
359 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
360 accordance with RFC 2308. Previously there was no expiry, so a longlived
361 receive process (eg. due to ACL delays) versus a short SOA value could
364 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
366 JH/39 Promote DMARC support to mainline.
368 JH/40 Bug 2452: Add a References: header to DSNs.
370 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
371 parameters. The relevant library call is documented as "Deprecated: This
372 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
373 3.6.0, DH parameters are negotiated following RFC7919."
375 HS/06 Change the default of dnssec_request_domains to "*"
377 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
378 carried on and emitted a BDAT command, even when PIPELINING was not
381 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
382 buffer was used for the filename, resulting in a trap when tainted
383 arguments (eg. $domain) were used.
385 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
386 recommended to avoid a possible server-load attack. The feature can be
387 re-enabled via the openssl_options main cofiguration option.
389 JH/45 local_scan API: documented the current smtp_printf() call. This changed
390 for version 4.90 - adding a "more data" boolean to the arguments.
391 Bumped the ABI version number also, this having been missed previously;
392 release versions 4.90 to 4.92.3 inclusive were effectively broken in
393 respect of usage of smtp_printf() by either local_scan code or libraries
394 accessed via the ${dlfunc } expansion item. Both will need coding
395 adjustment for any calls to smtp_printf() to match the new function
396 signature; a FALSE value for the new argument is always safe.
398 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
399 the file-offset (which the Linux syscall does, and exim expects); this
400 resulted in an indefinite loop.
402 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
403 to do ARC verification. The Authentication-Results: header line added
404 by the configuration then had no ARC item.
410 JH/01 Remove code calling the customisable local_scan function, unless a new
411 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
413 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
414 non-signal-safe functions being used.
416 JH/03 Bug 2269: When presented with a received message having a stupidly large
417 number of DKIM-Signature headers, disable DKIM verification to avoid
418 a resource-consumption attack. The limit is set at twenty.
420 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
421 report of oldest_pass in ${authres } in consequence, and separate out
422 some descriptions of reasons for verification fail.
424 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
425 files in the spool were present and unlocked. A queue-runner could spot
426 them, resulting in a duplicate delivery. Fix that by doing the unlock
427 after the unlink. Investigation by Tim Stewart. Take the opportunity to
428 add more error-checking on spoolfile handling while that code is being
431 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
432 No known attacks, no CVE, this is defensive hardening.
434 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
435 a queue-runner could start a delivery while other operations were ongoing.
436 Cutthrough delivery was a common victim, resulting in duplicate delivery.
437 Found and investigated by Tim Stewart. Fix by using the open message data
438 file handle rather than opening another, and not locally closing it (which
439 releases a lock) for that case, while creating the temporary .eml format
440 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
442 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
443 $sender_verify_failure/$recipient_verify_failure to "random".
445 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
448 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
449 Previously this would segfault.
451 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
454 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
455 like zero, since the resolver should be doing this for us, But we need one
456 as a CNAME but no MX presence gets the CNAME returned; we need to check
457 that doesn't point to an MX to declare it "no MX returned" rather than
458 "error, loop". A new main option is added so the older capability of
459 following some limited number of chain links is maintained.
461 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
463 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
464 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
465 now-deprecated earlier definitions used only specified the range up to TLS
466 1.2 (in the older-version library docs).
468 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
470 JH/15 Rework TLS client-side context management. Stop using a global, and
471 explicitly pass a context around. This enables future use of TLS for
472 connections to service-daemons (eg. malware scanning) while a client smtp
473 connection is using TLS; with cutthrough connections this is quite likely.
475 JH/16 Fix ARC verification to do AS checks in reverse order.
477 JH/17 Support a "tls" option on the ${readsocket } expansion item.
479 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
480 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
481 Previously the "utf8" would be re-prepended for every additional message.
483 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
484 Previously thery were accepted, resulting in issues when attempting to
485 forward messages to a non-supporting MTA.
487 PP/02 Let -n work with printing macros too, not just options.
489 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
490 one parent address was copied, and bogus data was used at delivery-logging
491 time. Either a crash (after delivery) or bogus log data could result.
492 Discovery and analysis by Tim Stewart.
494 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
495 Previously if the string ended mid-character, we did not insert the
496 promised '?' replacement.
498 PP/04 Documentation: current string operators work on bytes, not codepoints.
500 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
501 should pack well giving a smaller memory footprint so better caching and
502 therefore performance. Group the declarations where this can't be done so
503 that the byte-sized flag variables are not interspersed among pointer
504 variables, giving a better chance of good packing by the compiler.
506 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
507 non-null, to avoid issues with sites running BATV. Previously reports were
508 sent with an empty envelope sender so looked like bounces.
510 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
511 The ignore_error flag wasn't being returned from the filter subprocess so
512 was not set for later routers. Investigation and fix by Matthias Kurz.
514 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
515 and a msg:complete for the whole, when a message is manually removed using
516 -Mrm. Developement by Matthias Kurz, hacked on by JH.
518 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
519 a "Gnu special" function, asprintf() in the DB utility binary builds; I
520 hope that is portable enough.
522 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
523 requiring a known-CA anchor certificate; make it now rely entirely on the
524 TLSA as an anchor. Checking the name on the leaf cert against the name
525 on the A-record for the host is still done for TA (but not for EE mode).
527 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
528 included in delivery lines for non-proxied connections, when compiled with
529 SUPPORT_SOCKS and running with proxy logging enabled.
531 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
532 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
533 move the existing event to fire before the normal logging of message
534 failure so that custom logging is bracketed by normal logging.
536 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
537 msg:fail:internal event. Developement by Matthias Kurz.
539 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
540 far too small for todays use of crypto signatures stored there. Go all
541 the way to the max DNS message size of 64kB, even though this might be
542 overmuch for IOT constrained device use.
544 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
545 copy a string over itself. The library routine is documented as not
546 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
548 JH/32 For main options check_spool_space and check_inode_space, where the
549 platform supports 64b integers, support more than the previous 2^31 kB
550 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
551 the previous G, M, k.
553 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
554 $authenticated_fail_id variable on authentication failure. Previously
557 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
558 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
559 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
560 GNUTLS_SEC_PARAM_MEDIUM.
562 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
563 side. Previously we would continue as if no SNI had been received.
565 JH/36 Harden the handling of string-lists. When a list consisted of a sole
566 "<" character, which should be a list-separator specification, we walked
567 off past the nul-terimation.
569 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
570 causes) even when the retry time is not yet met. Previously they were
571 not, meaning that when (say) an account was over-quota and temp-rejecting,
572 and multiple senders' messages were queued, only one sender would get
573 notified on each configured delay_warning cycle.
575 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
577 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
578 cipher-suites, an error can be left on the stack even for a succeeding
579 accept; this results in impossible error messages when a later operation
582 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
583 return error codes indicating retry. Under TLS1.3 this becomes required.
585 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
586 it only wrote the new authenticators, resulting in a lack of tracking of
587 peer changes of ESMTP extensions until the next cache flush.
589 JH/41 Fix the loop reading a message header line to check for integer overflow,
590 and more-often against header_maxsize. Previously a crafted message could
591 induce a crash of the recive process; now the message is cleanly rejected.
593 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
594 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
600 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
601 When redis_servers is set to a list of > 1 element, and the Redis servers
602 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
603 case of MOVED into a DEFER case instead, thus moving the query onto the
604 next server in the list. For a cluster of N elements, all N servers must
605 be defined in redis_servers.
607 GF/02 Catch and remove uninitialized value warning in exiqsumm
608 Check for existence of @ARGV before looking at $ARGV[0]
610 JH/01 Replace the store_release() internal interface with store_newblock(),
611 which internalises the check required to safely use the old one, plus
612 the allocate and data copy operations duplicated in both (!) of the
613 extant use locations.
615 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
616 modifier. This matches the restriction on the commandline.
618 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
619 Previously only the last row was returned.
621 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
622 we assumed that tags in the header were well-formed, and parsed the
623 element content after inspecting only the first char of the tag.
624 Assumptions at that stage could crash the receive process on malformed
627 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
628 While running the DKIM ACL we operate on the Permanent memory pool so that
629 variables created with "set" persist to the DATA ACL. Also (at any time)
630 DNS lookups that fail create cache records using the Permanent pool. But
631 expansions release any allocations made on the current pool - so a dnsdb
632 lookup expansion done in the DKIM ACL releases the memory used for the
633 DNS negative-cache, and bad things result. Solution is to switch to the
634 Main pool for expansions.
635 While we're in that code, add checks on the DNS cache during store_reset,
636 active in the testsuite.
637 Problem spotted, and debugging aided, by Wolfgang Breyha.
639 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
640 When none of the hosts presented to a transport match an already-open
641 connection, close it and proceed with the list. Previously we would
642 queue the message. Spotted by Lena with Yahoo, probably involving
645 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
646 Previously a spurious "250 OK id=" response was appended to the proper
649 JH/08 The "support for" informational output now, which built with Content
650 Scanning support, has a line for the malware scanner interfaces compiled
651 in. Interface can be individually included or not at build time.
653 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
654 by the template makefile "src/EDITME". The "STREAM" support for an older
655 ClamAV interface method is removed.
657 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
658 rows affected is given instead).
660 JH/11 The runtime Berkeley DB library version is now additionally output by
661 "exim -d -bV". Previously only the compile-time version was shown.
663 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
664 SMTP connection. Previously, when one had more receipients than the
665 first, an abortive onward connection was made. Move to full support for
666 multiple onward connections in sequence, handling cutthrough connection
667 for all multi-message initiating connections.
669 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
670 routers. Previously, a multi-recipient message would fail to match the
671 onward-connection opened for the first recipient, and cause its closure.
673 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
674 a timeout on read on a GnuTLS initiating connection, resulting in the
675 initiating connection being dropped. This mattered most when the callout
676 was marked defer_ok. Fix to keep the two timeout-detection methods
679 JH/15 Relax results from ACL control request to enable cutthrough, in
680 unsupported situations, from error to silently (except under debug)
681 ignoring. This covers use with PRDR, frozen messages, queue-only and
684 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
686 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
687 metadata, resulting in a crash in free().
689 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
690 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
691 Broken also in d185889f4, with init system revamp.
693 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
694 Previously we abruptly closed the connection after reading a malware-
695 found indication; now we go on to read the "scan ok" response line,
698 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
699 ACL. Previously, a crash would result.
701 JH/19 Speed up macro lookups during configuration file read, by skipping non-
702 macro text after a replacement (previously it was only once per line) and
703 by skipping builtin macros when searching for an uppercase lead character.
705 JH/20 DANE support moved from Experimental to mainline. The Makefile control
706 for the build is renamed.
708 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
709 was allocated for every new TLS startup, meaning one per message. Fix
710 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
712 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
713 reported the original. Fix to report (as far as possible) the ACL
714 result replacing the original.
716 JH/23 Fix memory leak during multi-message connections using STARTTLS under
717 OpenSSL. Certificate information is loaded for every new TLS startup,
718 and the resources needed to be freed.
720 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
722 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
725 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
726 DATA response info to the (existing) per-recipient response info for
727 the "C=" log element. It can have useful tracking info from the
728 destination system. Patch from Simon Arlott.
730 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
731 length value. Previously this would segfault.
733 HS/02 Support Avast multiline protoocol, this allows passing flags to
734 newer versions of the scanner.
736 JH/28 Ensure that variables possibly set during message acceptance are marked
737 dead before release of memory in the daemon loop. This stops complaints
738 about them when the debug_store option is enabled. Discovered specifically
739 for sender_rate_period, but applies to a whole set of variables.
740 Do the same for the queue-runner and queue-list loops, for variables set
741 from spool message files. Do the same for the SMTP per-message loop, for
742 certain variables indirectly set in ACL operations.
744 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
745 as a multi-recipient message from a mailinglist manager). The coding had
746 an arbitrary cutoff number of characters while checking for more input;
747 enforced by writing a NUL into the buffer. This corrupted long / fast
748 input. The problem was exposed more widely when more pipelineing of SMTP
749 responses was introduced, and one Exim system was feeding another.
750 The symptom is log complaints of SMTP syntax error (NUL chars) on the
751 receiving system, and refused recipients seen by the sending system
752 (propating to people being dropped from mailing lists).
753 Discovered and pinpointed by David Carter.
755 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
756 replaced by the ${authresults } expansion.
758 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
760 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
761 allows proper process termination in container environments.
763 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
764 Previously the "final dot" had a newline after it; ensure it is CR,LF.
766 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
767 and "err_perm", deprecated since 4.83 when the RFC-defined words
768 "temperror" and "permerror" were introduced.
770 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
771 transport-filters or DKIM-signing. The restriction was lost in the
772 consolidation of verify-callout and delivery SMTP handling.
773 Extend the restriction to also cover ARC-signing.
775 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
776 in defer=pass mode supply a 450 to the initiator. Previously the message
779 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
780 tls_require_ciphers is used as before.
782 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
783 "pass_unscanned". Only tmpfails from the scanner are written to
784 the paniclog, as they may require admin intervention (permission
785 denied, license issues). Other scanner errors (like decompression
786 bombs) do not cause a paniclog entry.
788 JH/36 Fix reinitialisation of DKIM logging variable between messages.
789 Previously it was possible to log spurious information in receive log
792 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
793 triggered odd behaviour from Outlook Express clients.
795 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
798 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
799 since the IETF WG has not yet settled on that versus the original
800 "bare" representation.
802 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
803 Previously the millisecond value corrupted the output.
804 Fix also for syslog_pid=no and log_selector +pid, for which the pid
805 corrupted the output.
811 JH/01 Rework error string handling in TLS interface so that the caller in
812 more cases is responsible for logging. This permits library-sourced
813 string to be attached to addresses during delivery, and collapses
814 pairs of long lines into single ones.
816 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
817 during configuration. Wildcards are allowed and expanded.
819 JH/02 Rework error string handling in DKIM to pass more info back to callers.
820 This permits better logging.
822 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
823 do not close it down and have the child transport start it up again on
824 the passed-on TCP connection. Instead, proxy the child (and any
825 subsequent ones) for TLS via a unix-domain socket channel. Logging is
826 affected: the continued delivery log lines do not have any DNSSEC, TLS
827 Certificate or OCSP information. TLS cipher information is still logged.
829 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
830 identical IP addresses on different listening ports. Will also affect
833 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
834 add noisy ifdef guards to special-case this sillyness.
835 Patch from Bernd Kuhls.
837 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
838 than 255 are no longer allowed.
840 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
841 Disable the session-cache too, which might reduce our load. Since we
842 currrectly use a new context for every connection, both as server and
843 client, there is no benefit for these.
844 GnuTLS appears to not support tickets server-side by default (we don't
845 call gnutls_session_ticket_enable_server()) but client side is enabled
846 by default on recent versions (3.1.3 +) unless the PFS priority string
849 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
850 <https://reproducible-builds.org/specs/source-date-epoch/>.
852 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
853 the check for any unsuccessful recipients did not notice the limit, and
854 erroneously found still-pending ones.
856 JH/08 Pipeline CHUNKING command and data together, on kernels that support
857 MSG_MORE. Only in-clear (not on TLS connections).
859 JH/09 Avoid using a temporary file during transport using dkim. Unless a
860 transport-filter is involved we can buffer the headers in memory for
861 creating the signature, and read the spool data file once for the
862 signature and again for transmission.
864 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
865 4.77 as the kernel support then wasn't solid, having issues in 64bit
866 mode. Now, it's been long enough. Add support for FreeBSD also.
868 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
869 case where the routing stage had gathered several addresses to send to
870 a host before calling the transport for the first, we previously failed
871 to close down TLS in the old transport process before passing the TCP
872 connection to the new process. The new one sent a STARTTLS command
873 which naturally failed, giving a failed delivery and bloating the retry
874 database. Investigation and fix prototype from Wolfgang Breyha.
876 JH/12 Fix check on SMTP command input synchronisation. Previously there were
877 false-negatives in the check that the sender had not preempted a response
878 or prompt from Exim (running as a server), due to that code's lack of
879 awareness of the SMTP input buffering.
881 PP/04 Add commandline_checks_require_admin option.
882 Exim drops privileges sanely, various checks such as -be aren't a
883 security problem, as long as you trust local users with access to their
884 own account. When invoked by services which pass untrusted data to
885 Exim, this might be an issue. Set this option in main configuration
886 AND make fixes to the calling application, such as using `--` to stop
889 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
890 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
891 responses to those, into a single TLS record each way (this usually means
892 a single packet). As a side issue, smtp_enforce_sync now works on TLS
895 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
896 affects you only if you're dancing at the edge of the param size limits.
897 If you are, and this message makes sense to you, then: raise the
898 configured limit or use OpenSSL 1.1. Nothing we can do for older
901 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
902 cmdline element to get the documented default one. Previously it was
905 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
906 are unsuable for non-message-specific reasons - eg. connection timeout,
907 banner-time rejection.
909 JH/16 Fix logging of delivery remote port, when specified by router, under
912 PP/06 Repair manualroute's ability to take options in any order, even if one
913 is the name of a transport.
916 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
918 JH/17 Change the list-building routines interface to use the expanding-string
919 triplet model, for better allocation and copying behaviour.
921 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
922 Previously it was constructed the first time a possibly-matching string
923 was met in the configuration file input during startup; now it is done
926 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
927 compatible one, to avoid the (poorly documented) possibility of a config
928 file in the working directory redirecting the DB files, possibly correpting
929 some existing file. CVE-2017-10140 assigned for BDB.
931 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
932 cache-hot. Previously, although the result was properly cached, the
933 initial verify call returned a defer.
935 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
936 the main verify for receipient in uncached-mode.
938 JH/22 Retire historical build files to an "unsupported" subdir. These are
939 defined as "ones for which we have no current evidence of testing".
941 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
942 if present. Previously it was ignored.
944 JH/24 Start using specified-initialisers in C structure init coding. This is
945 a C99 feature (it's 2017, so now considered safe).
947 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
948 if was a fixed-sized field and bitmask ops via macros; it is now more
951 PP/07 GitHub PR 56: Apply MariaDB build fix.
952 Patch provided by Jaroslav Škarvada.
954 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
955 during Coverity cleanups [4.87 JH/47]
956 Diagnosis and fix provided by Michael Fischer v. Mollard.
958 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
959 the right size to place the terminating semicolon on its own folded
960 line, the header hash was calculated to an incorrect value thanks to
961 the (relaxed) space the fold became.
963 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
964 and confused the parent.
966 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
967 which could crash as a result. This could lead to undeliverable messages.
969 JH/28 Logging: "next input sent too soon" now shows where input was truncated
972 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
973 matters on fast-turnover and PID-randomising systems, which were getting
974 out-of-order delivery.
976 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
977 a possibly-overlapping copy. The symptom was that "Remote host closed
978 connection in response to HELO" was logged instead of the actual 4xx
981 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
982 Previously only that bufferd was discarded, resulting in SYMTP command
985 JH/32 DKIM: when a message has multiple signatures matching an identity given
986 in dkim_verify_signers, run the dkim acl once for each. Previously only
987 one run was done. Bug 2189.
989 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
990 "panic the current process" to "deliberately defer". The panic log is
991 still written with the problem list name; the mail and reject logs now
992 get a temp-reject line for the message that was being handled, saying
993 something like "domains check lookup or other defer". The SMTP 451
994 message is still "Temporary local problem".
996 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
997 A crafted sequence of BDAT commands could result in in-use memory beeing
998 freed. CVE-2017-16943.
1000 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1001 from SMTP input. Previously it was always done; now only done for DATA
1002 and not BDAT commands. CVE-2017-16944.
1004 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1005 to the message (such as an overlong header line). Previously this was
1006 not done and we did not exit BDAT mode. Followon from the previous item
1007 though a different problem.
1013 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1014 than -2003 did; needs libidn2 in addition to libidn.
1016 JH/02 The path option on a pipe transport is now expanded before use.
1018 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1019 Patch provided by "Björn", documentation fix added too.
1021 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1022 missing a wire-to-host endian conversion.
1024 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1025 close after a BDAT command line could be taken as a following command,
1026 giving a synch failure. Fix by only checking for synch immediately
1027 before acknowledging the chunk.
1029 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1030 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1031 macro. Patches provided by Josh Soref.
1033 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1034 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1035 by IANA as a service extension. However, John Klensin suggests that we
1038 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1039 the dkim code may be unix-mode line endings rather than smtp wire-format
1040 CRLF, so prepend a CR to any bare LF.
1042 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1043 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1045 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1046 the first were themselves being wrongly included in the feed into dkim
1047 processing; with most chunk sizes in use this resulted in an incorrect
1048 body hash calculated value.
1050 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1051 DKIM signature block, for verification. Although advised against by
1052 standards it is specifically not ruled illegal.
1054 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1056 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1057 missing a body hash (the bh= tag).
1059 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1060 It seems that HAProxy sends the Proxy Protocol information in clear and
1061 only then does a TLS startup, so do the same.
1063 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1064 TCP connections (such as for Spamd) unless the daemon successfully set
1065 Fast Open mode on its listening sockets. This fixes breakage seen on
1066 too-old kernels or those not configured for Fast Open, at the cost of
1067 requiring both directions being enabled for TFO, and TFO never being used
1068 by non-daemon-related Exim processes.
1070 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1071 endings, at least on the first header line. Try to canonify any that get
1072 past that check, despite the cost.
1074 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1075 now limited to an arbitrary five deep, while parsing addresses with the
1076 strip_excess_angle_brackets option enabled.
1078 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1079 instead leave the unprompted TLS handshake in socket buffer for the
1080 TLS library to consume.
1082 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1084 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1086 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1087 region in one operation, for speed, and this leaves assigned pointers
1088 dangling. Add checks run only under the testsuite which checks all
1089 variables at a store-reset and panics on a dangling pointer; add code
1090 explicitly nulling out all the variables discovered. Fixes one known
1091 bug: a transport crash, where a dangling pointer for $sending_ip_address
1092 originally assigned in a verify callout, is re-used.
1094 PP/06 Drop '.' from @INC in various Perl scripts.
1096 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1098 PP/08 Reduce a number of compilation warnings under clang; building with
1099 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1100 should be warning-free.
1102 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1104 HS/01 Fix portability problems introduced by PP/08 for platforms where
1105 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1107 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1108 chunk. This allows us to accept broken chunked messages. We need a more
1109 general solution here.
1111 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1112 already-broken messages in the queue.
1114 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1116 JH/19 Fix reference counting bug in routing-generated-address tracking.
1122 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1123 supports it and a size is available (ie. the sending peer gave us one).
1125 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1126 years of being deprecated). The replacements are the ACLs
1127 acl_smtp_mime and acl_not_smtp_mime.
1129 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1130 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1131 or even an in-clear connection were permitted. Now, if the host lookup
1132 was dnssec and dane was requested then the host is only used if the
1133 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1134 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1135 if one fails this test.
1136 This means that a poorly-configured remote DNS will make it incommunicado;
1137 but it protects against a DNS-interception attack on it.
1139 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1140 non-noisy when a race steals the message being considered.
1142 JH/05 If main configuration option tls_certificate is unset, generate a
1143 self-signed certificate for inbound TLS connections.
1145 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1146 in rewrites and routers.
1148 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1149 and logged a warning sing 4.83; now they are a configuration file error.
1151 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1152 (lacking @domain). Apply the same qualification processing as RCPT.
1154 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1156 JH/10 Support ${sha256:} applied to a string (as well as the previous
1159 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1160 a cutthrough deliver is pending, as we always want to make a connection.
1161 This also avoids re-routing the message when later placing the cutthrough
1162 connection after a verify cache hit.
1163 Do not update it with the verify result either.
1165 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1166 when routing results in more than one destination address.
1168 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1169 signing (which inhibits the cutthrough capability). Previously only
1170 the presence of an option was tested; now an expansion evaluating as
1171 empty is permissible (obviously it should depend only on data available
1172 when the cutthrough connection is made).
1174 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1175 the relevant preceding SMTP command did not note the pipelining mode.
1177 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1178 Previously they were not counted.
1180 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1181 as one having no matching records. Previously we deferred the message
1182 that needed the lookup.
1184 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1185 distinguished as "(=".
1187 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1188 for missing MX records. Previously it only worked for missing A records.
1190 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1192 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1193 after the data-go-ahead and data-ack. Patch from Jason Betts.
1195 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1196 even for a "none" policy. Patch from Tony Meyer.
1198 JH/22 Fix continued use of a connection for further deliveries. If a port was
1199 specified by a router, it must also match for the delivery to be
1202 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1203 When one of the recipients of a message was unsuitable for the connection
1204 (has no matching addresses), we lost track of needing to mark it
1205 deferred. As a result mail would be lost.
1207 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1209 JH/25 Decoding ACL controls is now done using a binary search; the source code
1210 takes up less space and should be simpler to maintain. Merge the ACL
1211 condition decode tables also, with similar effect.
1213 JH/26 Fix problem with one_time used on a redirect router which returned the
1214 parent address unchanged. A retry would see the parent address marked as
1215 delivered, so not attempt the (identical) child. As a result mail would
1218 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1219 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1220 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1223 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1224 Main config options check_{log,spool}_{inodes,space} are now
1225 100 inodes, 10MB unless set otherwise in the configuration.
1227 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1228 Previously it only applied to the main-section connection policy
1231 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1233 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1234 by me. Added RFC7919 DH primes as an alternative.
1236 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1237 are not in the system include path.
1239 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1240 GnuTLS, when a session startup failed (eg because the client disconnected)
1241 Exim did stdio operations after fclose. This was exposed by a recent
1242 change which nulled out the file handle after the fclose.
1244 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1245 signed directly by the cert-signing cert, rather than an intermediate
1246 OCSP-signing cert. This is the model used by LetsEncrypt.
1248 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1250 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1251 an incoming connection.
1253 HS/02 Bug 1802: Do not half-close the connection after sending a request
1256 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1257 fallback to "prime256v1".
1259 JH/34 SECURITY: Use proper copy of DATA command in error message.
1260 Could leak key material. Remotely exploitable. CVE-2016-9963.
1266 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1267 and 3.4.4 - once the server is enabled to respond to an OCSP request
1268 it does even when not requested, resulting in a stapling non-aware
1269 client dropping the TLS connection.
1271 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1272 support variable-length bit vectors. No functional change.
1274 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1275 The I= interface field on outgoing lines is now after the H= remote
1276 host field, same as incoming lines. There is a separate
1277 outgoing_interface log selector which allows you to disable the
1280 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1281 If not running log_selector +smtp_connection the mainlog would be held
1282 open indefinitely after a "too many connections" event, including to a
1283 deleted file after a log rotate. Leave the per net connection logging
1284 leaving it open for efficiency as that will be quickly detected by the
1285 check on the next write.
1287 HS/01 Bug 1671: Fix post transport crash.
1288 Processing the wait-<transport> messages could crash the delivery
1289 process if the message IDs didn't exist for some reason. When
1290 using 'split_spool_directory=yes' the construction of the spool
1291 file name failed already, exposing the same netto behaviour.
1293 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1294 mime_regex ACL conditions.
1296 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1297 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1298 to HELO, local diagnostic string.
1300 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1301 log line to debug. Even when configured with a tls authenticator many
1302 client connections are expected to not authenticate in this way, so
1303 an authenticate fail is not an error.
1305 HS/02 Add the Exim version string to the process info. This way exiwhat
1306 gives some more detail about the running daemon.
1308 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1309 matter for fast-change records such as DNSBLs.
1311 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1312 retry record, even if constant. There may be multiple transports with
1313 different interface settings and the retry behaviour needs to be kept
1316 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1318 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1320 JH/10 Bug 840: fix log_defer_output option of pipe transport
1322 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1323 hosts_randomize. This matters a lot when combined with mua_wrapper.
1325 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1326 ${quote_pgsql:<string>} operator.
1328 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1330 JH/14 Change header file name for freeradius-client. Relevant if compiling
1331 with Radius support; from the Gentoo tree and checked under Fedora.
1333 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1335 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1336 option transport_filter, meaning no filtering.
1338 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1340 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1341 defaults to "*" (all hosts). The variable is now available when not built
1342 with TLS, default unset, mainly to enable keeping the testsuite sane.
1343 If a server certificate is not supplied (via tls_certificate) an error is
1344 logged, and clients will find TLS connections fail on startup. Presumably
1345 they will retry in-clear.
1346 Packagers of Exim are strongly encouraged to create a server certificate
1347 at installation time.
1349 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1350 with the $config_file variable.
1352 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1353 in transport context, after the attempt, and per-recipient. The latter type
1354 is per host attempted. The event data is the error message, and the errno
1355 information encodes the lookup type (A vs. MX) used for the (first) host,
1356 and the trailing two digits of the smtp 4xx response.
1358 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1359 to write to mainlog (or rejectlog, paniclog) in the window between file
1360 creation and permissions/ownership being changed. Particularly affects
1361 installations where exicyclog is run as root, rather than exim user;
1362 result is that the running daemon panics and dies.
1364 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1366 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1367 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1368 "pri" and "weight". Note that the previous implicit priority given by the
1369 list order is no longer honoured.
1371 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1372 for DKIM processing.
1374 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1375 by defining SUPPORT_SOCKS.
1377 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1378 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1379 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1380 variables are renamed to proxy_{local,external}_{address,port}.
1382 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1383 for a build by defining SUPPORT_I18N
1385 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1386 of the query string, and make ${quote_redis:} do that quoting.
1388 JH/29 Move Events support from Experimental to mainline, enabled by default
1389 and removable for a build by defining DISABLE_EVENT.
1391 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1393 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1394 cached by the daemon.
1396 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1397 by defining LOOKUP_REDIS. The libhiredis library is required.
1399 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1400 keys are given for lookup.
1402 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1403 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1404 only supported when built with TLS support. The PolarSSL SHA routines
1405 are still used when the TLS library is too old for convenient support.
1407 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1408 openssl_options), for security. OpenSSL forces this from version 1.1.0
1409 server-side so match that on older versions.
1411 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1412 allocation for $value could be released as the expansion processing
1413 concluded, but leaving the global pointer active for it.
1415 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1416 and to use the domains and local_parts ACL conditions.
1418 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1419 incorrectly not doubled on cutthrough transmission, hence seen as a
1420 body-termination at the receiving system - resulting in truncated mails.
1421 Commonly the sender saw a TCP-level error, and retransmitted the message
1422 via the normal store-and-forward channel. This could result in duplicates
1423 received - but deduplicating mailstores were liable to retain only the
1424 initial truncated version.
1426 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1428 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1430 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1431 we're in there, support oversigning also; bug 1309.
1433 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1435 HS/04 Add support for keep_environment and add_environment options.
1437 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1438 either intentional arithmetic overflow during PRNG, or testing config-
1441 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1442 delivery resulted in actual delivery. Cancel cutthrough before DATA
1445 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1446 rejects a recipient: pass the reject to the originator.
1448 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1449 Many were false-positives and ignorable, but it's worth fixing the
1452 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1453 for the new environment-manipulation done at startup. Move the routines
1454 from being local to tls.c to being global via the os.c file.
1456 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1457 an extract embedded as result-arg for a map, the first arg for extract
1458 is unavailable so we cannot tell if this is a numbered or keyed
1459 extraction. Accept either.
1465 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1468 JH/02 The smtp transport option "multi_domain" is now expanded.
1470 JH/03 The smtp transport now requests PRDR by default, if the server offers
1473 JH/04 Certificate name checking on server certificates, when exim is a client,
1474 is now done by default. The transport option tls_verify_cert_hostnames
1475 can be used to disable this per-host. The build option
1476 EXPERIMENTAL_CERTNAMES is withdrawn.
1478 JH/05 The value of the tls_verify_certificates smtp transport and main options
1479 default to the word "system" to access the system default CA bundle.
1480 For GnuTLS, only version 3.0.20 or later.
1482 JH/06 Verification of the server certificate for a TLS connection is now tried
1483 (but not required) by default. The verification status is now logged by
1484 default, for both outbound TLS and client-certificate supplying inbound
1487 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1490 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1491 Status Notification (bounce) messages are now MIME format per RFC 3464.
1492 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1493 under the control of the dsn_advertise_hosts option, and routers may
1494 have a dsn_lasthop option.
1496 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1497 default, modifiable by a malware= option. The list separator for
1498 the options can now be changed in the usual way. Bug 68.
1500 JH/10 The smtp_receive_timeout main option is now expanded before use.
1502 JH/11 The incoming_interface log option now also enables logging of the
1503 local interface on delivery outgoing connections.
1505 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1506 if the interface and destination host and port all match.
1508 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1511 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1512 Patch from Andrew Lewis.
1514 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1515 now supports optional time-restrictions, weighting, and priority
1516 modifiers per server. Patch originally by <rommer@active.by>.
1518 JH/16 The spamd_address main option now supports a mixed list of local
1519 and remote servers. Remote servers can be IPv6 addresses, and
1520 specify a port-range.
1522 JH/17 Bug 68: The spamd_address main option now supports an optional
1523 timeout value per server.
1525 JH/18 Bug 1581: Router and transport options headers_add/remove can
1526 now have the list separator specified.
1528 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1531 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1534 JH/21 Support for the A6 type of dns record is withdrawn.
1536 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1537 rather than the verbs used.
1539 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1540 from 255 to 1024 chars.
1542 JH/24 Verification callouts now attempt to use TLS by default.
1544 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1545 are generic router options now. The defaults didn't change.
1547 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1548 Original patch from Alexander Shikoff, worked over by JH.
1550 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1551 files if ZCAT_COMMAND is not executable.
1553 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1555 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1557 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1558 Normally benign, it bites when the pair was led to by a CNAME;
1559 modern usage is to not canonicalize the domain to a CNAME target
1560 (and we were inconsistent anyway for A-only vs AAAA+A).
1562 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1564 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1565 when evaluating $sender_host_dnssec.
1567 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1568 $sender_helo_dnssec variable.
1570 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1572 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1574 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1576 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1577 documented as working, but never had. Support all but $spam_report.
1579 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1580 added for tls authenticator.
1582 HS/03 Add perl_taintmode main config option
1588 TL/01 When running the test suite, the README says that variables such as
1589 no_msglog_check are global and can be placed anywhere in a specific
1590 test's script, however it was observed that placement needed to be near
1591 the beginning for it to behave that way. Changed the runtest perl
1592 script to read through the entire script once to detect and set these
1593 variables, reset to the beginning of the script, and then run through
1594 the script parsing/test process like normal.
1596 TL/02 The BSD's have an arc4random API. One of the functions to induce
1597 adding randomness was arc4random_stir(), but it has been removed in
1598 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1599 function when detected.
1601 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1602 cause callback expansion.
1604 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1605 syntax errors in an expansion can be treated as a string instead of
1606 logging or causing an error, due to the internal use of bool_lax
1607 instead of bool when processing it.
1609 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1610 server certificates when making smtp deliveries.
1612 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1614 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1616 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1618 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1619 Merged patch from Sebastian Wiedenroth.
1621 JH/05 Fix results-pipe from transport process. Several recipients, combined
1622 with certificate use, exposed issues where response data items split
1623 over buffer boundaries were not parsed properly. This eventually
1624 resulted in duplicates being sent. This issue only became common enough
1625 to notice due to the introduction of connection certificate information,
1626 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1628 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1629 size buffer was used, resulting in syntax errors when an expansion
1632 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1633 version 3.3.6 or later.
1635 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1636 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1637 both become "event_action", the variables become $event_name, $event_data
1638 and $event_defer_errno. There is a new variable $verify_mode, usable in
1639 routers, transports and related events. The tls:cert event is now also
1640 raised for inbound connections, if the main configuration event_action
1643 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1644 early OCSP support, but no stapling (appears to be less than 1.0.0).
1646 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1647 server certificate names available under the smtp transport option
1648 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1651 JH/10 Time-related extraction expansions from certificates now use the main
1652 option "timezone" setting for output formatting, and are consistent
1653 between OpenSSL and GnuTLS compilations. Bug 1541.
1655 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1656 encoded parameter in the incoming message. Bug 1558.
1658 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1659 include certificate info, eximon was claiming there were spoolfile
1662 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1664 JH/14 Log delivery-related information more consistently, using the sequence
1665 "H=<name> [<ip>]" wherever possible.
1667 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1668 are problematic for Debian distribution, omit them from the release
1671 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1673 JH/16 Fix string representation of time values on 64bit time_t architectures.
1676 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1677 output list separator was used.
1682 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1683 checkers that were complaining about end of non-void function with no
1686 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1687 This was a regression introduced in 4.83 by another bugfix.
1689 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1691 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1692 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1698 TF/01 Correctly close the server side of TLS when forking for delivery.
1700 When a message was received over SMTP with TLS, Exim failed to clear up
1701 the incoming connection properly after forking off the child process to
1702 deliver the message. In some situations the subsequent outgoing
1703 delivery connection happened to have the same fd number as the incoming
1704 connection previously had. Exim would try to use TLS and fail, logging
1705 a "Bad file descriptor" error.
1707 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1708 utilities have not been installed.
1710 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1711 temporary space as the ACL may create new global variables.
1713 TL/01 LDAP support uses per connection or global context settings, depending
1714 upon the detected version of the libraries at build time.
1716 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1717 to extract and use the src ip:port in logging and expansions as if it
1718 were a direct connection from the outside internet. PPv2 support was
1719 updated based on HAProxy spec change in May 2014.
1721 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1723 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1724 Properly escape header and check for NULL return.
1726 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1729 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1731 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1732 characters in header names, implemented as a verify condition.
1733 Contributed by Michael Fischer v. Mollard.
1735 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1736 results permerror and temperror. Previous values are deprecated but
1737 still accepted. In a future release, err_perm and err_temp will be
1738 completely removed, which will be a backward incompatibility if the
1739 ACL tests for either of these two old results. Patch contributed by
1740 user bes-internal on the mailing list.
1742 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1744 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1745 selectors, in both main and reject logs.
1747 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1750 JH/07 Add malware type "sock" for talking to simple daemon.
1752 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1754 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1755 routers/transports under cutthrough routing.
1757 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1758 numbers. Touch up "bool" conditional to keep the same definition.
1760 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1762 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1764 JH/12 Expand items in router/transport headers_add or headers_remove lists
1765 individually rather than the list as a whole. Bug 1452.
1767 Required for reasonable handling of multiple headers_ options when
1768 they may be empty; requires that headers_remove items with embedded
1769 colons must have them doubled (or the list-separator changed).
1771 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1772 view the policy declared in the DMARC record. Currently, $dmarc_status
1773 is a combined value of both the record presence and the result of the
1776 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1778 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1779 dnslookup router and the smtp transport (applying to the forward
1782 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1783 of ldap servers used for a specific lookup. Patch provided by Heiko
1786 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1787 New variable $lookup_dnssec_authenticated for observability.
1789 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1790 Patch submitted by Lars Timman.
1792 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1794 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1795 Requires trusted mode and valid format message id, aborts otherwise.
1796 Patch contributed by Heiko Schlichting.
1798 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1799 certextract with support for various fields. Bug 1358.
1801 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1802 is requested by default, modifiable by smtp transport option
1805 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1806 operate on certificate variables to give certificate fingerprints
1807 Also new ${sha256:cert_variable}.
1809 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1811 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1812 Christian Aistleitner.
1814 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1816 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1817 file. Patch from Wolfgang Breyha.
1819 JH/25 Expand the coverage of the delivery $host and $host_address to
1820 client authenticators run in verify callout. Bug 1476.
1822 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1823 align with daemon_smtp_ports. Bug 72.
1825 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1826 support and error reporting did not work properly.
1828 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1829 and is readable. Patch from Andrew Colin Kissa.
1831 TL/14 Enhance documentation of ${run expansion and how it parses the
1832 commandline after expansion, particularly in the case when an
1833 unquoted variable expansion results in an empty value.
1835 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1837 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1838 by renaming away from it. Addresses GH issue 15, reported by
1841 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1843 TL/15 SECURITY: prevent double expansion in math comparison functions
1844 (can expand unsanitized data). Not remotely exploitable.
1851 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1853 PP/02 Make -n do something, by making it not do something.
1854 When combined with -bP, the name of an option is not output.
1856 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1859 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1860 $sender_host_name and config options to manage this, and basic check
1863 PP/05 DSCP support for outbound connections and control modifier for inbound.
1865 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1866 (Only plugin which currently uses this is kerberos4, which nobody should
1867 be using, but we should make it available and other future plugins might
1868 conceivably use it, even though it would break NAT; stuff *should* be
1869 using channel bindings instead).
1871 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1872 name; added for Sendmail compatibility; requires admin caller.
1873 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1874 just ignore it); requires trusted caller.
1875 Also parse but ignore: -Ac -Am -X<logfile>
1878 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1880 TL/02 Add +smtp_confirmation as a default logging option.
1882 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1883 Patch by Magnus Holmgren from 2007-02-20.
1885 TL/04 Bugzilla 1281 - Spec typo.
1886 Bugzilla 1283 - Spec typo.
1887 Bugzilla 1290 - Spec grammar fixes.
1889 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1891 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1893 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1894 mailing list by by Dmitry Isaikin.
1896 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1898 JH/02 Support "G" suffix to numbers in ${if comparisons.
1900 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1902 NM/01 Bugzilla 1197 - Spec typo
1903 Bugzilla 1196 - Spec examples corrections
1905 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1907 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1908 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1911 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1912 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1914 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1915 "acl {{name}{arg}...}", and optional args on acl condition
1918 JH/05 Permit multiple router/transport headers_add/remove lines.
1920 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1922 JH/07 Avoid using a waiting database for a single-message-only transport.
1923 Performance patch from Paul Fisher. Bugzilla 1262.
1925 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1928 JH/09 Add $headers_added variable, with content from use of ACL modifier
1929 add_header (but not yet added to the message). Bugzilla 199.
1931 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1932 Pulled from Bugzilla 817 by Wolfgang Breyha.
1934 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1936 (nb: this is the same fix as in Exim 4.80.1)
1938 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1941 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1944 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1946 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1947 advertise SMTP AUTH mechanism to us, instead of a generic
1948 protocol violation error. Also, make Exim more robust to bad
1949 data from the Dovecot auth socket.
1951 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1953 When a queue runner is handling a message, Exim first routes the
1954 recipient addresses, during which it prunes them based on the retry
1955 hints database. After that it attempts to deliver the message to
1956 any remaining recipients. It then updates the hints database using
1959 So if a recipient address works intermittently, it can get repeatedly
1960 deferred at routing time. The retry hints record remains fresh so the
1961 address never reaches the final cutoff time.
1963 This is a fairly common occurrence when a user is bumping up against
1964 their storage quota. Exim had some logic in its local delivery code
1965 to deal with this. However it did not apply to per-recipient defers
1966 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1968 This change adds a proper retry rule check during routing so that the
1969 final cutoff time is checked against the message's age. We only do
1970 this check if there is an address retry record and there is not a
1971 domain retry record; this implies that previous attempts to handle
1972 the address had the retry_use_local_parts option turned on. We use
1973 this as an approximation for the destination being like a local
1974 delivery, as in LMTP.
1976 I suspect this new check makes the old local delivery cutoff check
1977 redundant, but I have not verified this so I left the code in place.
1979 TF/02 Correct gecos expansion when From: is a prefix of the username.
1981 Test 0254 submits a message to Exim with the header
1985 When I ran the test suite under the user fanf2, Exim expanded
1986 the header to contain my full name, whereas it should have added
1987 a Resent-Sender: header. It erroneously treats any prefix of the
1988 username as equal to the username.
1990 This change corrects that bug.
1992 GF/01 DCC debug and logging tidyup
1993 Error conditions log to paniclog rather than rejectlog.
1994 Debug lines prefixed by "DCC: " to remove any ambiguity.
1996 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1998 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1999 Bug spotted by Jeremy Harris; was flawed since initial commit.
2000 Would have resulted in OCSP responses post-SNI triggering an Exim
2001 NULL dereference and crash.
2003 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2005 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2006 Bug detection, analysis and fix by Samuel Thibault.
2007 Bugzilla 1331, Debian bug #698092.
2009 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2011 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2012 Server implementation by Todd Lyons, client by JH.
2013 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2014 config variable "prdr_enable" controls whether the server
2015 advertises the facility. If the client requests PRDR a new
2016 acl_data_smtp_prdr ACL is called once for each recipient, after
2017 the body content is received and before the acl_smtp_data ACL.
2018 The client is controlled by both of: a hosts_try_prdr option
2019 on the smtp transport, and the server advertisement.
2020 Default client logging of deliveries and rejections involving
2021 PRDR are flagged with the string "PRDR".
2023 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2024 fclose(). Diagnosis by Todd Lyons.
2026 PP/17 Update configure.default to handle IPv6 localhost better.
2027 Patch by Alain Williams (plus minor tweaks).
2030 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2031 This is now consistent with GnuTLS, and is now documented: the
2032 previous undocumented portable approach to treating the option as
2033 unset was to force an expansion failure. That still works, and
2034 an empty string is now equivalent.
2036 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2037 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2038 not performing validation itself.
2040 PP/20 Added force_command boolean option to pipe transport.
2041 Patch from Nick Koston, of cPanel Inc.
2043 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2046 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2048 PP/21 Fix eximon continuous updating with timestamped log-files.
2049 Broken in a format-string cleanup in 4.80, missed when I repaired the
2050 other false fix of the same issue.
2051 Report and fix from Heiko Schlichting.
2054 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2055 Report from Prashanth Katuri.
2057 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2058 It's SecureTransport, so affects any MacOS clients which use the
2059 system-integrated TLS libraries, including email clients.
2061 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2062 using a MIME ACL for non-SMTP local injection.
2063 Report and assistance in diagnosis by Warren Baker.
2065 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2067 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2069 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2070 last id that failed so it may be referenced in subsequent ACL's.
2072 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2075 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2076 ldap library initialization, allowing self-signed CA's to be
2077 used. Also properly sets require_cert option later in code by
2078 using NULL (global ldap config) instead of ldap handle (per
2079 session). Bug diagnosis and testing by alxgomz.
2081 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2082 the src/util/ subdirectory.
2084 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2085 renamed to Transport Post Delivery Action by Jeremy Harris, as
2088 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2089 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2090 redis_servers = needs to be configured which will be used by the redis
2091 lookup. Patch from Warren Baker, of The Packet Hub.
2093 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2095 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2096 hostname or reverse DNS when processing a host list. Used suggestions
2097 from multiple comments on this bug.
2099 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2101 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2102 Missed a few lines, added it to make the runtest require no keyboard
2105 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2106 contains upper case chars. Make router use caseful_local_part.
2108 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2109 support when GnuTLS has been built with p11-kit.
2115 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2117 This, or similar/improved, will also be change PP/11 of 4.82.
2123 PP/01 Handle short writes when writing local log-files.
2124 In practice, only affects FreeBSD (8 onwards).
2125 Bugzilla 1053, with thanks to Dmitry Isaikin.
2127 NM/01 Bugzilla 949 - Documentation tweak
2129 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2132 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2134 PP/02 Implemented gsasl authenticator.
2136 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2138 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2139 `pkg-config foo` for cflags/libs.
2141 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2142 with rest of GSASL and with heimdal_gssapi.
2144 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2145 `pkg-config foo` for cflags/libs for the TLS implementation.
2147 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2148 properties get this fed in as external SSF. A number of robustness
2149 and debugging improvements to the cyrus_sasl authenticator.
2151 PP/08 cyrus_sasl server now expands the server_realm option.
2153 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2154 Patch by Jeremy Harris.
2156 PP/10 Added dbmjz lookup type.
2158 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2160 PP/12 MAIL args handles TAB as well as SP, for better interop with
2161 non-compliant senders.
2162 Analysis and variant patch by Todd Lyons.
2164 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2165 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2166 Patch from Dirk Mueller <dmueller@suse.com>
2168 PP/13 tls_peerdn now print-escaped for spool files.
2169 Observed some $tls_peerdn in wild which contained \n, which resulted
2170 in spool file corruption.
2172 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2173 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2174 or write after TLS renegotiation, which otherwise led to messages
2177 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2178 as a tracking header (ie: a signed header comes before the signature).
2179 Patch from Wolfgang Breyha.
2181 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2182 comma-sep list; embedded commas doubled.
2184 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2186 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2188 Report and patch from Dmitry Banschikov.
2190 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2191 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2192 failures. We appear to now support TLS1.1+ with Exim.
2194 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2195 lets Exim select keys and certificates based upon TLS SNI from client.
2196 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2197 before an outbound SMTP session. New log_selector, +tls_sni.
2199 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2200 NULL dereference. Report and patch from Alun Jones.
2202 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2203 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2204 Not seeing resolver debug output on NetBSD, but suspect this is a
2205 resolver implementation change.
2207 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2208 Left warnings. Added "eximon gdb" invocation mode.
2210 PP/21 Defaulting "accept_8bitmime" to true, not false.
2212 PP/22 Added -bw for inetd wait mode support.
2214 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2215 locate the relevant includes and libraries. Made this the default.
2217 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2218 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2220 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2221 This may cause build issues on older platforms.
2223 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2224 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2225 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2226 Added SNI support via GnuTLS too.
2227 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2229 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2231 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2232 Applied second patch from Janne, implementing suggestion to default
2233 multiple-strings-in-record handling to match SPF spec.
2235 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2237 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2238 read-only, out of scope).
2239 Patch from Wolfgang Breyha, report from Stuart Northfield.
2241 PP/29 Fix three issues highlighted by clang analyser static analysis.
2242 Only crash-plausible issue would require the Cambridge-specific
2243 iplookup router and a misconfiguration.
2244 Report from Marcin Mirosław.
2246 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2248 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2249 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2250 As part of this, removing so much warning spew let me fix some minor
2251 real issues in debug logging.
2253 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2254 assignment on my part. Fixed.
2256 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2257 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2258 Janne Snabb (who went above and beyond: thank you).
2260 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2261 string otherwise requires a connection and a bunch more work and it's
2262 relatively easy to get wrong. Should also expose TLS library linkage
2265 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2266 64-bit ${eval} (JH/03).
2268 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2269 GNU libc to support some of the 64-bit stuff, should not lead to
2270 conflicts. Defined before os.h is pulled in, so if a given platform
2271 needs to override this, it can.
2273 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2274 protection layer was required, which is not implemented.
2275 Bugzilla 1254, patch from Wolfgang Breyha.
2277 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2278 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2279 tls_dhparam take prime identifiers. Also unbreak combination of
2280 OpenSSL+DH_params+TLSSNI.
2282 PP/39 Disable SSLv2 by default in OpenSSL support.
2288 PP/01 Solaris build fix for Oracle's LDAP libraries.
2289 Bugzilla 1109, patch from Stephen Usher.
2291 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2293 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2296 TF/02 Fix a couple more cases where we did not log the error message
2297 when unlink() failed. See also change 4.74-TF/03.
2299 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2300 lock up or crash if it happened to be inside a call to libc when it
2301 got a SIGUSR1 from exiwhat.
2303 The SIGUSR1 handler appends the current process status to the process
2304 log which is later printed by exiwhat. It used to use the general
2305 purpose logging code to do this, but several functions it calls are
2306 not safe for signals.
2308 The new output code in the SIGUSR1 handler is specific to the process
2309 log, and simple enough that it's easy to inspect for signal safety.
2310 Removing some special cases also simplifies the general logging code.
2311 Removing the spurious timestamps from the process log simplifies
2314 TF/04 Improved ratelimit ACL condition.
2316 The /noupdate option has been deprecated in favour of /readonly which
2317 has clearer semantics. The /leaky, /strict, and /readonly update modes
2318 are mutually exclusive. The update mode is no longer included in the
2319 database key; it just determines when the database is updated. (This
2320 means that when you upgrade Exim will forget old rate measurements.)
2322 Exim now checks that the per_* options are used with an update mode that
2323 makes sense for the current ACL. For example, when Exim is processing a
2324 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2325 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2326 must specify per_mail/readonly. If you omit the update mode it defaults to
2327 /leaky where that makes sense (as before) or /readonly where required.
2329 The /noupdate option is now undocumented but still supported for
2330 backwards compatibility. It is equivalent to /readonly except that in
2331 ACLs where /readonly is required you may specify /leaky/noupdate or
2332 /strict/noupdate which are treated the same as /readonly.
2334 A useful new feature is the /count= option. This is a generalization
2335 of the per_byte option, so that you can measure the throughput of other
2336 aggregate values. For example, the per_byte option is now equivalent
2337 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2339 The per_rcpt option has been generalized using the /count= mechanism
2340 (though it's more complicated than the per_byte equivalence). When it is
2341 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2342 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2343 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2344 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2345 Note that using per_rcpt with a non-readonly update mode in more than
2346 one ACL will cause the recipients to be double-counted. (The per_mail
2347 and per_byte options don't have this problem.)
2349 The handling of very low rates has changed slightly. If the computed rate
2350 is less than the event's count (usually one) then this event is the first
2351 after a long gap. In this case the rate is set to the same as this event's
2352 count, so that the first message of a spam run is counted properly.
2354 The major new feature is a mechanism for counting the rate of unique
2355 events. The new per_addr option counts the number of different
2356 recipients that someone has sent messages to in the last time period. It
2357 behaves like per_rcpt if all the recipient addresses are different, but
2358 duplicate recipient addresses do not increase the measured rate. Like
2359 the /count= option this is a general mechanism, so the per_addr option
2360 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2361 example, measure the rate that a client uses different sender addresses
2362 with the options per_mail/unique=$sender_address. There are further
2363 details in the main documentation.
2365 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2367 TF/06 Removed a few PCRE remnants.
2369 TF/07 Automatically extract Exim's version number from tags in the git
2370 repository when doing development or release builds.
2372 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2373 Bugzilla 879. Patch from Paul Fisher.
2375 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2376 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2379 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2381 PP/05 Variable $av_failed, true if the AV scanner deferred.
2382 Bugzilla 1078. Patch from John Horne.
2384 PP/06 Stop make process more reliably on build failure.
2385 Bugzilla 1087. Patch from Heiko Schlittermann.
2387 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2388 Bugzilla 1089. Patch from Heiko Schlittermann.
2390 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2391 Bugzilla 1131. Patch from Holger Weiß.
2393 PP/09 Handle IPv6 addresses with SPF.
2394 Bugzilla 860. Patch from Wolfgang Breyha.
2396 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2398 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2401 PP/11 match_* no longer expand right-hand-side by default.
2402 New compile-time build option, EXPAND_LISTMATCH_RHS.
2403 New expansion conditions, "inlist", "inlisti".
2405 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2407 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2409 PP/14 fix log_write() format string regression from TF/03.
2410 Bugzilla 1152. Patch from Dmitry Isaikin.
2416 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2418 PP/02 Harmonised TLS library version reporting; only show if debugging.
2419 Layout now matches that introduced for other libraries in 4.74 PP/03.
2421 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2423 PP/04 New "dns_use_edns0" global option.
2425 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2428 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2429 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2431 TK/01 Updated PolarSSL code to 0.14.2.
2432 Bugzilla 1097. Patch from Andreas Metzler.
2434 PP/07 Catch divide-by-zero in ${eval:...}.
2435 Fixes bugzilla 1102.
2437 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2440 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2441 format-string attack -- SECURITY: remote arbitrary code execution.
2443 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2444 time unintentionally subject to list matching rules, letting the header
2445 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2446 arbitrary string expansion). This allowed for information disclosure.
2448 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2449 INT_MIN/-1 -- value coerced to INT_MAX.
2455 NM/01 Workaround for PCRE version dependency in version reporting
2458 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2459 This fixes portability to compilers other than gcc, notably
2460 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2462 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2463 makefiles for portability to HP-UX and POSIX correctness.
2465 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2466 Also via indented variable definition in the Makefile.
2467 (Debugging by Oliver Heesakkers).
2469 PP/02 Restore caching of spamd results with expanded spamd_address.
2470 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2472 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2473 Improves build reliability. Fix from: Frank Elsner
2475 NM/02 Fix wide character breakage in the rfc2047 coding
2476 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2478 NM/03 Allow underscore in dnslist lookups
2479 Fixes bug 1026. Patch from Graeme Fowler
2481 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2482 Code patches from Adam Ciarcinski of NetBSD.
2484 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2487 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2488 is logged, to avoid truncation. Patch from John Horne.
2490 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2491 Patch from Jakob Hirsch.
2493 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2494 SQL string expansion failure details.
2495 Patch from Andrey Oktyabrski.
2497 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2498 Patch from Simon Arlott.
2500 PP/09 New lookups functionality failed to compile on old gcc which rejects
2501 extern declarations in function scope.
2502 Patch from Oliver Fleischmann
2504 PP/10 Use sig_atomic_t for flags set from signal handlers.
2505 Check getgroups() return and improve debugging.
2506 Fixed developed for diagnosis in bug 927 (which turned out to be
2509 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2510 Patch from Mark Zealey.
2512 PP/12 Bugzilla 1056: Improved spamd server selection.
2513 Patch from Mark Zealey.
2515 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2516 Based on patch from Heiko Schlittermann.
2518 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2519 Patch from Uwe Doering, sign-off by Michael Haardt.
2521 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2522 variable declaration deep within a block. Bug and patch from
2525 PP/15 lookups-Makefile IRIX compatibility coercion.
2527 PP/16 Make DISABLE_DKIM build knob functional.
2529 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2530 Patch by Simon Arlott
2532 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2533 variable argument macros. Our copy now differs from upstream.
2539 TF/01 Failure to get a lock on a hints database can have serious
2540 consequences so log it to the panic log.
2542 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2543 controlled using the smtp_confirmation log selector.
2545 TF/03 Include the error message when we fail to unlink a spool file.
2547 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2548 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2549 for maintaining out-of-tree patches for some time.
2551 PP/01 Bugzilla 139: Documentation and portability issues.
2552 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2553 Handle per-OS dynamic-module compilation flags.
2555 PP/02 Let /dev/null have normal permissions.
2556 The 4.73 fixes were a little too stringent and complained about the
2557 permissions on /dev/null. Exempt it from some checks.
2558 Reported by Andreas M. Kirchwitz.
2560 PP/03 Report version information for many libraries, including
2561 Exim version information for dynamically loaded libraries. Created
2562 version.h, now support a version extension string for distributors
2563 who patch heavily. Dynamic module ABI change.
2565 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2566 privilege escalation vulnerability whereby the Exim run-time user
2567 can cause root to append content of the attacker's choosing to
2570 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2573 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2574 If dropping privileges for untrusted macros, we disabled normal logging
2575 on the basis that it would fail; for the Exim run-time user, this is not
2576 the case, and it resulted in successful deliveries going unlogged.
2577 Fixed. Reported by Andreas Metzler.
2583 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2584 only prepend for the Resent-* case. Fixes regression introduced in
2585 Exim 4.70 by NM/22 for Bugzilla 607.
2587 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2588 increasing numbers of administrators be bitten by this.
2590 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2592 PP/03 Bugzilla 994: added openssl_options main configuration option.
2594 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2596 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2598 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2600 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2601 without a peer certificate, leading to a segfault because of an
2602 assumption that peers always have certificates. Be a little more
2603 paranoid. Problem reported by Martin Tscholak.
2605 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2606 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2607 NB: ClamAV planning to remove STREAM in "middle of 2010".
2608 CL also introduces -bmalware, various -d+acl logging additions and
2609 more caution in buffer sizes.
2611 PP/09 Implemented reverse_ip expansion operator.
2613 PP/10 Bugzilla 937: provide a "debug" ACL control.
2615 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2617 PP/12 Bugzilla 973: Implement --version.
2619 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2621 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2623 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2625 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2626 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2627 ignore trailing whitespace.
2629 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2631 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2634 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2635 Notification from Dr Andrew Aitchison.
2637 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2638 ExtendedDetectionInfo response format.
2639 Notification from John Horne.
2641 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2644 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2645 XSL and documented dependency on system catalogs, with examples of how
2648 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2651 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2652 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2653 configuration file which is writeable by the Exim user or group.
2655 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2656 of configuration files to cover files specified with the -C option if
2657 they are going to be used with root privileges, not just the default
2660 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2661 option (effectively making it always true).
2663 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2664 files to be used while preserving root privileges.
2666 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2667 that rogue child processes cannot use them.
2669 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2670 run-time user, instead of root.
2672 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2673 Exim run-time user without dropping privileges.
2675 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2676 result string, instead of calling string_vformat() twice with the same
2679 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2680 for other users. Others should always drop root privileges if they use
2681 -C on the command line, even for a whitelisted configure file.
2683 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2685 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2691 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2692 $data_path, and $header_path variables; fixed documentation bugs and
2695 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2696 exipick to access non-standard spools, including the "frozen" queue
2699 NM/01 Bugzilla 965: Support mysql stored procedures.
2700 Patch from Alain Williams
2702 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2704 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2705 Patch from Andreas Metzler
2707 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2708 Patch from Kirill Miazine
2710 NM/05 Bugzilla 671: Added umask to procmail example.
2712 JJ/03 installed exipick 20100323.0, fixing doc bug
2714 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2715 directory. Notification and patch from Dan Rosenberg.
2717 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2719 TK/02 Improve log output when DKIM signing operation fails.
2721 MH/01 Treat the transport option dkim_domain as a colon separated
2722 list, not as a single string, and sign the message with each element,
2723 omitting multiple occurences of the same signer.
2725 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2726 Bugzilla 985, 986. Patch by Simon Arlott
2728 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2729 Patch by Simon Arlott
2731 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2732 MBX locking. Notification from Dan Rosenberg.
2738 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2740 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2742 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2744 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2746 NM/04 Bugzilla 915: Flag broken perl installation during build.
2752 TK/01 Added patch by Johannes Berg that expands the main option
2753 "spamd_address" if it starts with a dollar sign.
2755 TK/02 Write list of recipients to X-Envelope-Sender header when building
2756 the mbox-format spool file for content scanning (suggested by Jakob
2759 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2760 (http://www.dcc-servers.net/) support via dccifd. Activated by
2761 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2763 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2764 by Mark Daniel Reidel <mr@df.eu>.
2766 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2767 When building exim an external PCRE library is now needed -
2768 PCRE is a system library on the majority of modern systems.
2769 See entry on PCRE_LIBS in EDITME file.
2771 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2772 conversation. Added nologin parameter to request.
2773 Patch contributed by Kirill Miazine.
2775 TF/01 Do not log submission mode rewrites if they do not change the address.
2777 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2779 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2780 log files in place. Contributed by Roberto Lima.
2782 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2784 TF/03 Bugzilla 615: When checking the local_parts router precondition
2785 after a local_part_suffix or local_part_prefix option, Exim now
2786 does not use the address's named list lookup cache, since this
2787 contains cached lookups for the whole local part.
2789 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2790 Robert Millan. Documentation is in experimental-spec.txt.
2792 TF/04 Bugzilla 668: Fix parallel build (make -j).
2794 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2796 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2797 Patch provided by Jan Srzednicki.
2799 TF/05 Leading white space used to be stripped from $spam_report which
2800 wrecked the formatting. Now it is preserved.
2802 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2803 that they are available at delivery time.
2805 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2807 TF/08 TLS error reporting now respects the incoming_interface and
2808 incoming_port log selectors.
2810 TF/09 Produce a more useful error message if an SMTP transport's hosts
2811 setting expands to an empty string.
2813 NM/06 Bugzilla 744: EXPN did not work under TLS.
2814 Patch provided by Phil Pennock.
2816 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2817 Patch provided by Richard Godbee.
2819 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2820 acl_smtp_notquit, added index entry.
2822 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2823 Patch provided by Eugene Bujak.
2825 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2826 accept(). Patch provided by Maxim Dounin.
2828 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2829 Patch provided by Phil Pennock.
2831 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2833 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2834 Patch provided by Brad "anomie" Jorsch.
2836 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2837 Patch provided by Dean Brooks.
2839 TK/05 Add native DKIM support (does not depend on external libraries).
2841 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2842 Patch provided by Graeme Fowler.
2844 NM/16 Bugzilla 851: Documentation example syntax fix.
2846 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2848 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2851 NM/19 Bugzilla 745: TLS version reporting.
2852 Patch provided by Phil Pennock.
2854 NM/20 Bugzilla 167: bool: condition support.
2855 Patch provided by Phil Pennock.
2857 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2858 clients. Patch provided by Phil Pennock.
2860 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2861 Patch provided by Brad "anomie" Jorsch.
2863 NM/23 Bugzilla 687: Fix misparses in eximstats.
2864 Patch provided by Heiko Schlittermann.
2866 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2867 Patch provided by Heiko Schlittermann.
2869 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2870 plus update to original patch.
2872 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2874 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2875 Patch provided by David Brownlee.
2877 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2879 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2881 NM/30 Bugzilla 888: TLS documentation bugfixes.
2883 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2885 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2886 Unlike the original bugzilla I have changed all shell scripts in src tree.
2888 NM/33 Bugzilla 898: Transport filter timeout fix.
2889 Patch by Todd Rinaldo.
2891 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2892 Patch by Serge Demonchaux.
2894 NM/35 Bugzilla 39: Base64 decode bug fixes.
2895 Patch by Jakob Hirsch.
2897 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2899 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2901 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2903 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2909 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2910 ALT-N's libdkim that I have put here:
2911 http://duncanthrax.net/exim-experimental/
2913 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2914 were called 'true' and it seems that C99 defines that as a reserved
2915 keyword to be used with 'bool' variable types. That means you could
2916 not include C99-style headers which use bools without triggering
2917 build errors in sieve.c.
2919 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2920 as mailq or other aliases. Changed the --help handling significantly
2921 to do whats expected. exim_usage() emits usage/help information.
2923 SC/01 Added the -bylocaldomain option to eximstats.
2925 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2927 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2929 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2935 PH/01 Another patch from the Sieve maintainer.
2937 PH/02 When an IPv6 address is converted to a string for single-key lookup
2938 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2939 dots are used instead of colons so that keys in lsearch files need not
2940 contain colons. This was done some time before quoting was made available
2941 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2942 (notated using the quote facility) so as to distinguish them from IPv4
2943 keys. This meant that lookups for IP addresses in host lists did not work
2944 for iplsearch lookups.
2946 This has been fixed by arranging for IPv6 addresses to be expressed with
2947 colons if the lookup type is iplsearch. This is not incompatible, because
2948 previously such lookups could never work.
2950 The situation is now rather anomalous, since one *can* have colons in
2951 ordinary lsearch keys. However, making the change in all cases is
2952 incompatible and would probably break a number of configurations.
2954 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2957 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2958 conversion specification without a maximum field width, thereby enabling
2959 a rogue spamd server to cause a buffer overflow. While nobody in their
2960 right mind would setup Exim to query an untrusted spamd server, an
2961 attacker that gains access to a server running spamd could potentially
2962 exploit this vulnerability to run arbitrary code as the Exim user.
2964 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2965 $primary_hostname instead of what libspf2 thinks the hosts name is.
2967 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2968 a directory entry by the name of the lookup key. Previously, if a
2969 symlink pointed to a non-existing file or a file in a directory that
2970 Exim lacked permissions to read, a lookup for a key matching that
2971 symlink would fail. Now it is enough that a matching directory entry
2972 exists, symlink or not. (Bugzilla 503.)
2974 PH/03 The body_linecount and body_zerocount variables are now exported in the
2977 PH/04 Added the $dnslist_matched variable.
2979 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2980 This means they are set thereafter only if the connection becomes
2983 PH/06 Added the client_condition to authenticators so that some can be skipped
2984 by clients under certain conditions.
2986 PH/07 The error message for a badly-placed control=no_multiline_responses left
2987 "_responses" off the end of the name.
2989 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2991 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2992 (without spaces) instead of just copying the configuration text.
2994 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2996 PH/11 Added $max_received_linelength.
2998 PH/12 Added +ignore_defer and +include_defer to host lists.
3000 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3001 way in which PCRE > 7.0 is built.
3003 PH/14 Implemented queue_only_load_latch.
3005 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3006 MAIL command. The effect was to mangle the value on 64-bit systems.
3008 PH/16 Another patch from the Sieve maintainer.
3010 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3012 PH/18 If a system quota error occurred while trying to create the file for
3013 a maildir delivery, the message "Mailbox is full" was not appended to the
3014 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3015 only to a quota excession during the actual writing of the file.
3017 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3018 characters?) which causes problems in log lines. The DN values are now
3019 passed through string_printing() before being added to log lines.
3021 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3022 and InterBase are left for another time.)
3024 PH/21 Added message_body_newlines option.
3026 PH/22 Guard against possible overflow in moan_check_errorcopy().
3028 PH/23 POSIX allows open() to be a macro; guard against that.
3030 PH/24 If the recipient of an error message contained an @ in the local part
3031 (suitably quoted, of course), incorrect values were put in $domain and
3032 $local_part during the evaluation of errors_copy.
3038 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3039 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3042 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3043 issue a MAIL command.
3045 PH/02 In an ACL statement such as
3047 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3049 if a client was not listed at all, or was listed with a value other than
3050 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3051 the condition was not true (as it should be), so access was not denied.
3052 The bug was that the ! inversion was incorrectly passed on to the second
3053 item. This has been fixed.
3055 PH/03 Added additional dnslists conditions == and =& which are different from
3056 = and & when the dns lookup returns more than one IP address.
3058 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3059 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3061 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3062 FSYNC, which compiles an option called disable_fsync that allows for
3063 bypassing fsync(). The documentation is heavily laced with warnings.
3065 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3067 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3068 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3069 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3070 including adding "make clean"; (3) Added -fPIC when compiling the test
3071 dynamically loaded module, to get rid of a warning.
3073 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3074 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3075 = 0s. The bug is otherwise harmless.
3077 PH/07 There was a bug in the dovecot authenticator such that the value of
3078 $auth1 could be overwritten, and so not correctly preserved, after a
3079 successful authentication. This usually meant that the value preserved by
3080 the server_setid option was incorrect.
3082 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3084 PH/09 Installed PCRE release 7.0.
3086 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3087 run for batched SMTP input. It is now run at the start of every message
3088 in the batch. While fixing this I discovered that the process information
3089 (output by running exiwhat) was not always getting set for -bs and -bS
3090 input. This is fixed, and it now also says "batched" for BSMTP.
3092 PH/11 Added control=no_pipelining.
3094 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3095 patch, slightly modified), and move the expansion of helo_data till after
3096 the connection is made in the smtp transport (so it can use these
3099 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3101 PH/14 Added log_selector = +pid.
3103 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3105 PH/16 Add ${if forany and ${if forall.
3107 PH/17 Added dsn_from option to vary the From: line in DSNs.
3109 PH/18 Flush SMTP output before performing a callout, unless control =
3110 no_callout_flush is set.
3112 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3113 was true (the default) a successful delivery failed to delete the retry
3114 item, thus causing premature timeout of the address. The bug is now
3117 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3119 PH/21 Long custom messages for fakedefer and fakereject are now split up
3120 into multiline responses in the same way that messages for "deny" and
3121 other ACL rejections are.
3123 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3124 with slight modification.
3126 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3127 draft, changing the syntax and factoring some duplicate code".
3129 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3130 for deliveries of the second and subsequent messages over the same SMTP
3133 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3134 ${reduce, with only minor "tidies".
3136 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3138 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3139 expansion side effects.
3141 PH/27 When a message times out after an over-quota error from an Exim-imposed
3142 quota, the bounce message says "mailbox is full". This message was not
3143 being given when it was a system quota that was exceeded. It now should
3146 MH/03 Made $recipients available in local_scan(). local_scan() already has
3147 better access to the recipient list through recipients_list[], but
3148 $recipients can be useful in postmaster-provided expansion strings.
3150 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3151 in the case of a MAIL command with additional options following the
3152 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3153 were accidentally chopped off.
3155 PH/29 SMTP synchronization checks are implemented when a command is read -
3156 there is a check that no more input is waiting when there shouldn't be
3157 any. However, for some commands, a delay in an ACL can mean that it is
3158 some time before the response is written. In this time, more input might
3159 arrive, invalidly. So now there are extra checks after an ACL has run for
3160 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3161 pipelining has not been advertised.
3163 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3165 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3166 returned in the SMTP response when smtp_return_error_details was false.
3167 This has been fixed.
3169 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3170 instead of the C I/O that was originally supplied, because problems were
3171 reported on Solaris.
3173 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3174 Exim which did not show up earlier: it was assuming that a call to
3175 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3176 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3177 was a macro that became an assignment, so it seemed to work. This has
3178 changed to a proper function call with a void return, hence the compile
3179 error. Exim's code has been fixed.
3181 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3184 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3186 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3188 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3189 because it turns out to be too big or there is a timeout) there is no
3190 "Completed" line in the log. When some messages of this type were
3191 selected by exigrep, they were listed as "not completed". Others were
3192 picked up by some special patterns. I have improved the selection
3193 criteria to be more general.
3195 PH/38 The host_find_failed option in the manualroute router can now be set
3196 to "ignore", to completely ignore a host whose IP address cannot be
3197 found. If all hosts are ignored, the behaviour is controlled by the new
3198 host_all_ignored option.
3200 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3201 homing or because of multiple MX records with /mx) generated more than
3202 one IP address, and the following item turned out to be the local host,
3203 all the secondary addresses of the first item were incorrectly removed
3204 from the list, along with the local host and any following hosts (which
3205 is what is supposed to happen).
3207 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3208 whoever called Exim into the -H file. In the case of the daemon it was
3209 behaving confusingly. When first started, it used values for whoever
3210 started the daemon, but after a SIGHUP it used the Exim user (because it
3211 calls itself on a restart). I have changed the code so that it now always
3214 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3215 message are rejected with the same error (e.g. no authentication or bad
3216 sender address), and a DATA command is nevertheless sent (as can happen
3217 with PIPELINING or a stupid MUA), the error message that was given to the
3218 RCPT commands is included in the rejection of the DATA command. This is
3219 intended to be helpful for MUAs that show only the final error to their
3222 PH/42 Another patch from the Sieve maintainer.
3224 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3225 Eximstats - Fixed some broken HTML links and added missing column headers
3227 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3228 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3230 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3232 PH/43 Yet another patch from the Sieve maintainer.
3234 PH/44 I found a way to check for a TCP/IP connection going away before sending
3235 the response to the final '.' that terminates a message, but only in the
3236 case where the client has not sent further data following the '.'
3237 (unfortunately, this is allowed). However, in many cases there won't be
3238 any further data because there won't be any more messages to send. A call
3239 to select() can be used: if it shows that the input is "ready", there is
3240 either input waiting, or the socket has been closed. An attempt to read
3241 the next input character can distinguish the two cases. Previously, Exim
3242 would have sent an OK response which the client would never have see.
3243 This could lead to message repetition. This fix should cure that, at
3244 least in a lot of common cases.
3246 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3247 advertised in response to EHLO.
3253 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3254 fixed by 4.65/MH/01 (is this a record?) are fixed:
3256 (i) An empty string was always treated as zero by the numeric comparison
3257 operators. This behaviour has been restored.
3259 (ii) It is documented that the numeric comparison operators always treat
3260 their arguments as decimal numbers. This was broken in that numbers
3261 starting with 0 were being interpreted as octal.
3263 While fixing these problems I realized that there was another issue that
3264 hadn't been noticed. Values of message_size_limit (both the global option
3265 and the transport option) were treated as octal if they started with 0.
3266 The documentation was vague. These values are now always treated as
3267 decimal, and I will make that clear in the documentation.
3273 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3274 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3277 MH/01 Don't check that the operands of numeric comparison operators are
3278 integers when their expansion is in "skipping" mode (fixes bug
3279 introduced by 4.64-PH/07).
3281 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3282 child addresses, Exim now panics and dies. Previously, because the count
3283 is held in a short int, deliveries were likely to be lost. As such a
3284 large number of recipients for a single message is ridiculous
3285 (performance will be very, very poor), I have chosen to impose a limit
3286 rather than extend the field.
3292 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3293 leftover -K file (the existence of which was triggered by #402).
3294 While we were at it, introduced process PID as part of the -K
3295 filename. This should rule out race conditions when creating
3298 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3299 processing considerably. Previous code took too long for large mails,
3300 triggering a timeout which in turn triggers #401.
3302 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3303 in the DK code in transports.c. sendfile() is not really portable,
3304 hence the _LINUX specificness.
3306 TF/01 In the add_headers option to the mail command in an Exim filter,
3307 there was a bug that Exim would claim a syntax error in any
3308 header after the first one which had an odd number of characters
3311 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3312 callout verification, Exim cached a "reject" for the entire domain. This
3313 is correct for most verifications, but it is not correct for a recipient
3314 verification with use_sender or use_postmaster set, because in that case
3315 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3316 case of MAIL FROM:<> rejection from other early rejections (e.g.
3317 rejection of HELO). When verifying a recipient using a non-null MAIL
3318 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3319 Whatever the result of the callout, the value of the domain cache is
3320 left unchanged (for any other kind of callout, getting as far as trying
3321 RCPT means that the domain itself is ok).
3323 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3326 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3327 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3328 session, the daemon ignores ECONNECT errors and logs others; it now
3329 ignores EPIPE as well.
3331 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3332 (quoted-printable decoding).
3334 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3335 later the small subsequent patch to fix an introduced bug.
3337 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3339 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3341 PH/08 An error is now given if message_size_limit is specified negative.
3343 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3344 to be given (somewhat) arbitrary names.
3346 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3349 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3350 miscellaneous code fixes
3352 PH/10 Added the log_reject_target ACL modifier to specify where to log
3355 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3356 hostname. This is wrong, because it relates to the incoming message (and
3357 probably the interface on which it is arriving) and not to the outgoing
3358 callout (which could be using a different interface). This has been
3359 changed to use the value of the helo_data option from the smtp transport
3360 instead - this is what is used when a message is actually being sent. If
3361 there is no remote transport (possible with a router that sets up host
3362 addresses), $smtp_active_hostname is used.
3364 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3365 tweaks were necessary in order to get it to work (see also 21 below):
3366 (a) The code assumed that strncpy() returns a negative number on buffer
3367 overflow, which isn't the case. Replaced with Exim's string_format()
3369 (b) There were several signed/unsigned issues. I just did the minimum
3370 hacking in of casts. There is scope for a larger refactoring.
3371 (c) The code used strcasecmp() which is not a standard C function.
3372 Replaced with Exim's strcmpic() function.
3373 (d) The code set only $1; it now sets $auth1 as well.
3374 (e) A simple test gave the error "authentication client didn't specify
3375 service in request". It would seem that Dovecot has changed its
3376 interface. Fortunately there's a specification; I followed it and
3377 changed what the client sends and it appears to be working now.
3379 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3382 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3383 address A is aliased to B and C, where B exists and C does not. Without
3384 -v the output is "A verified" because verification stops after a
3385 successful redirection if more than one address is generated. However,
3386 with -v the child addresses are also verified. Exim was outputting "A
3387 failed to verify" and then showing the successful verification for C,
3388 with its parentage. It now outputs "B failed to verify", showing B's
3389 parentage before showing the successful verification of C.
3391 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3392 look up a TXT record in a specific list after matching in a combined
3395 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3396 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3397 they consult the DNS. I had assumed they would set it the way they
3398 wanted; and indeed my experiments on Linux seem to show that in some
3399 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3400 To be on the safe side, however, I have now made the interface to
3401 host_find_byname() similar to host_find_bydns(), with an argument
3402 containing the DNS resolver options. The host_find_byname() function now
3403 sets these options at its start, just as host_find_bydns() does. The smtp
3404 transport options dns_qualify_single and dns_search_parents are passed to
3405 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3406 of host_find_byname() use the default settings of RES_DEFNAMES
3407 (qualify_single) but not RES_DNSRCH (search_parents).
3409 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3410 spool_read_header() do less string testing, by means of a preliminary
3411 switch on the second character of optional "-foo" lines. (This is
3412 overdue, caused by the large number of possibilities that now exist.
3413 Originally there were few.) While I was there, I also converted the
3414 str(n)cmp tests so they don't re-test the leading "-" and the first
3415 character, in the hope this might squeeze out yet more improvement.
3417 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3418 flag allowing group syntax was set by the header_syntax check but not
3419 turned off, possible causing trouble later; (2) The flag was not being
3420 set at all for the header_verify test, causing "group"-style headers to
3421 be rejected. I have now set it in this case, and also caused header_
3422 verify to ignore an empty address taken from a group. While doing this, I
3423 came across some other cases where the code for allowing group syntax
3424 while scanning a header line wasn't quite right (mostly, not resetting
3425 the flag correctly in the right place). These bugs could have caused
3426 trouble for malformed header lines. I hope it is now all correct.
3428 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3429 with the "reply" argument non-NULL. The code, however (which originally
3430 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3431 but it didn't always do it. This confused somebody who was copying the
3432 code for some other use. I have removed all the tests.
3434 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3435 feature that was used to support insecure browsers during the U.S. crypto
3436 embargo. It requires special client support, and Exim is probably the
3437 only MTA that supported it -- and would never use it because real RSA is
3438 always available. This code has been removed, because it had the bad
3439 effect of slowing Exim down by computing (never used) parameters for the
3440 RSA_EXPORT functionality.
3442 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3443 authenticator to fail if there's a tab character in the incoming data
3444 (there should never be unless someone is messing about, as it's supposed
3445 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3446 if the connection is using TLS or if the remote IP is the same as the
3447 local IP, and the "valid-client-cert option" if a client certificate has
3450 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3451 authenticators. This can be used for authorization after authentication
3452 succeeds. (In the case of plaintext, it servers for both authentication
3455 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3456 if any retry times were supplied.
3458 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3459 connection, where there is no client IP address to check. In this
3460 situation, the verify now always succeeds.
3462 PH/25 Applied John Jetmore's -Mset patch.
3464 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3466 PH/27 In a string expansion for a processed (not raw) header when multiple
3467 headers of the same name were present, leading whitespace was being
3468 removed from all of them, but trailing whitespace was being removed only
3469 from the last one. Now trailing whitespace is removed from each header
3470 before concatenation. Completely empty headers in a concatenation (as
3471 before) are ignored.
3473 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3474 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3476 PH/29 [Removed. This was a change that I later backed out, and forgot to
3477 correct the ChangeLog entry (that I had efficiently created) before
3478 committing the later change.]
3480 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3481 address errors (4xx response to RCPT) over the same connection as other
3482 messages routed to the same hosts. Such deliveries are always "forced",
3483 so retry times are not inspected. This resulted in far too many retries
3484 for the affected addresses. The effect occurred only when there were more
3485 hosts than the hosts_max_try setting in the smtp transport when it had
3486 the 4xx errors. Those hosts that it had tried were not added to the list
3487 of hosts for which the message was waiting, so if all were tried, there
3488 was no problem. Two fixes have been applied:
3490 (i) If there are any address or message errors in an SMTP delivery, none
3491 of the hosts (tried or untried) are now added to the list of hosts
3492 for which the message is waiting, so the message should not be a
3493 candidate for sending over the same connection that was used for a
3494 successful delivery of some other message. This seems entirely
3495 reasonable: after all the message is NOT "waiting for some host".
3496 This is so "obvious" that I'm not sure why it wasn't done
3497 previously. Hope I haven't missed anything, but it can't do any
3498 harm, as the worst effect is to miss an optimization.
3500 (ii) If, despite (i), such a delivery is accidentally attempted, the
3501 routing retry time is respected, so at least it doesn't keep
3502 hammering the server.
3504 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3505 in ${readsocket because some servers need this prod.
3507 PH/32 Added some extra debug output when updating a wait-xxx database.
3509 PH/33 The hint "could be header name not terminated by colon", which has been
3510 given for certain expansion errors for a long time, was not being given
3511 for the ${if def:h_colon_omitted{... case.
3513 PH/34 The spec says: "With one important exception, whenever a domain list is
3514 being scanned, $domain contains the subject domain." There was at least
3515 one case where this was not true.
3517 PH/35 The error "getsockname() failed: connection reset by peer" was being
3518 written to the panic log as well as the main log, but it isn't really
3519 panic-worthy as it just means the connection died rather early on. I have
3520 removed the panic log writing for the ECONNRESET error when getsockname()
3523 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3524 runs only) independently of the message's sender address. This meant
3525 that, if the 4xx error was in fact related to the sender, a different
3526 message to the same recipient with a different sender could confuse
3527 things. In particular, this can happen when sending to a greylisting
3528 server, but other circumstances could also provoke similar problems.
3529 I have changed the default so that the retry time for these errors is now
3530 based a combination of the sender and recipient addresses. This change
3531 can be overridden by setting address_retry_include_sender=false in the
3534 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3535 remote server are returned as part of bounce messages. This was not
3536 happening for LMTP over a pipe (the lmtp transport), but now it is the
3537 same for both kinds of LMTP.
3539 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3540 in header lines that were in fact CNAMEs. This is no longer the case.
3542 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3543 and queue runs started by the daemon processed all messages. This has
3544 been fixed so that -R and -S can now usefully be given with -q<time>.
3546 PH/40 Import PCRE release 6.7 (fixes some bugs).
3548 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3550 PH/42 Give an error if -q is specified more than once.
3552 PH/43 Renamed the variables $interface_address and $interface_port as
3553 $received_ip_address and $received_port, to make it clear that these
3554 values apply to message reception, and not to the outgoing interface when
3555 a message is delivered. (The old names remain recognized, of course.)
3557 PH/44 There was no timeout on the connect() call when using a Unix domain
3558 socket in the ${readsocket expansion. There now is.
3560 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3561 be meaningful with "accept".
3563 SC/01 Eximstats V1.43
3564 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3566 SC/02 Eximstats V1.44
3567 Use a glob alias rather than an array ref in the generated
3568 parser. This improves both readability and performance.
3570 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3571 Collect SpamAssassin and rejection statistics.
3572 Don't display local sender or destination tables unless
3573 there is data to show.
3574 Added average volumes into the top table text output.
3576 SC/04 Eximstats V1.46
3577 Collect data on the number of addresses (recipients)
3578 as well as the number of messages.
3580 SC/05 Eximstats V1.47
3581 Added 'Message too big' to the list of mail rejection
3582 reasons (thanks to Marco Gaiarin).
3584 SC/06 Eximstats V1.48
3585 Mainlog lines which have GMT offsets and are too short to
3586 have a flag are now skipped.
3588 SC/07 Eximstats V1.49 (Alain Williams)
3589 Added the -emptyok flag.
3591 SC/08 Eximstats V1.50
3592 Fixes for obtaining the IP address from reject messages.
3594 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3595 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3596 whitespace changes from 4.64-PH/27
3598 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3601 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3602 are found, allow negative numbers in numeric criteria)
3604 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3606 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3607 to match changes made in 4.64-PH/43
3609 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3611 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3612 "domain" is actually an IP address, and force a failure. This locks out
3613 those revolvers/nameservers that support "A-for-A" lookups, in
3614 contravention of the specifications.
3616 PH/48 When a host name was looked up from an IP address, and the subsequent
3617 forward lookup of the name timed out, the host name was left in
3618 $sender_host_name, contrary to the specification.
3620 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3621 restricted to single-key lookups, Exim was not diagnosing an error if
3622 * or *@ was used with a query-style lookup.
3624 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3626 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3627 long ago, but noone interested enough thought of it. Let's just say that
3628 the "1.1" means that there are some new functions that weren't there at
3629 some point in the past.
3631 PH/51 Error processing for expansion failure of helo_data from an smtp
3632 transport during callout processing was broken.
3634 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3635 tested/used via the -bh/-bhc/-bs options.
3637 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3638 bug, fixed in subsequent PCRE releases).
3640 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3641 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3643 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3649 SC/01 Use a glob alias rather than an array ref in eximstats generated
3650 parser. This improves both readability and performance.
3652 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3653 Don't display local sender or destination tables in eximstats unless
3654 there is data to show.
3655 Added average volumes into the eximstats top table text output.
3657 SC/03 Collect data on the number of addresses (recipients) as well
3658 as the number of messages in eximstats.
3660 TF/01 Correct an error in the documentation for the redirect router. Exim
3661 does (usually) call initgroups() when daemonizing.
3663 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3664 with consistent privilege compared to when running as a daemon.
3666 TF/03 Note in the spec that $authenticated_id is not set for local
3667 submissions from trusted users.
3669 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3670 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3672 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3673 by adding some example configuration directives to the default
3674 configuration file. A little bit of work is required to uncomment the
3675 directives and define how usernames and passwords are checked, but
3676 there is now a framework to start from.
3678 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3679 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3680 without this. I don't know how relevant this is to other LDAP libraries.
3682 PH/02 Add the verb name to the "unknown ACL verb" error.
3684 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3686 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3688 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3689 directory not expanded when it should be if an expanded home directory
3690 was set for the address (which is overridden by the transport).
3692 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3695 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3696 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3697 because it is too late at that time, and has no effect.
3699 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3700 security issue with \' (bugzilla #107). I could not use the
3701 PQescapeStringConn() function, because it needs a PGconn value as one of
3704 PH/08 When testing addresses using -bt, indicate those final addresses that
3705 are duplicates that would not cause an additional delivery. At least one
3706 person was confused, thinking that -bt output corresponded to deliveries.
3707 (Suppressing duplicates isn't a good idea as you lose the information
3708 about possibly different redirections that led to the duplicates.)
3710 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3711 systems where poll() doesn't work, in particular OS X.
3713 PH/10 Added more information to debugging output for retry time not reached.
3715 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3716 operations in malware.c.
3718 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3721 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3722 syslog_duplication set false, log lines that would normally be written
3723 both the the main log and to the reject log were not written to syslog at
3726 PH/14 In the default configuration, change the use of "message" in ACL warn
3727 statements to "add_header".
3729 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3730 not followed by a command (e.g. "seen endif").
3732 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3733 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3736 PH/17 Added extra conditions to the default value of delay_warning_condition
3740 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3741 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3742 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3745 The Auto-Submitted: and various List- headers are standardised, whereas I
3746 don't think Precedence: ever was.
3748 PH/18 Refactored debugging code in route_finduser() to show more information,
3749 in particular, the error code if getpwnam() issues one.
3751 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3752 This is apparently needed in addition to the PH/07 change above to avoid
3753 any possible encoding problems.
3755 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3756 but not after initializing Perl.
3758 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3759 output them only if debugging. By default they are written stderr,
3760 apparently, which is not desirable.
3762 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3765 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3768 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3770 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3771 authenticated or an ident call has been made. Suppress the default
3772 values for $authenticated_id and $authenticated_sender (but permit -oMai
3773 and -oMas) when testing with -bh.
3775 PH/24 Re-jigged the order of the tests in the default configuration so that the
3776 tests for valid domains and recipients precede the DNS black list and CSA
3777 tests, on the grounds that those ones are more expensive.
3779 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3780 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3781 This bug exists in every version of Exim that I still have, right back to
3784 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3785 However, an attempt to turn on case-sensitivity in a regex key by
3786 including (?-i) didn't work because the subject string was already
3787 lowercased, and the effects were non-intuitive. It turns out that a
3788 one-line patch can be used to allow (?-i) to work as expected.
3794 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3795 other effects) broke the use of negated acl sub-conditions.
3797 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3800 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3801 "Deny" causes Exim to reject the incoming connection with a 554 error.
3802 Unfortunately, if there is a major crisis, such as a disk failure,
3803 tcp-wrappers gives "deny", whereas what one would like would be some
3804 kind of temporary error. A kludge has been added to help with this.
3805 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3806 554 error is used if errno is still zero or contains ENOENT (which occurs
3807 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3810 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3812 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3813 errors. Otherwise a message that provokes a temporary error (when other
3814 messages do not) can cause a whole host to time out.
3816 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3817 addresses were routed directly to files or pipes from a redirect router.
3818 File deliveries just didn't batch; pipe deliveries might have suffered
3821 PH/06 A failure to get a lock for a hints database would erroneously always say
3822 "Failed to get write lock", even when it was really a read lock.
3824 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3825 of 0600. This has been changed to use the value of the lockfile_mode
3826 option (which defaults to 0600).
3828 PH/08 Applied small patch from the Sieve maintainer.
3830 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3831 folder from quota calculations, a direct delivery into this folder messed
3832 up the contents of the maildirsize file. This was because the regex was
3833 used only to exclude .Trash (or whatever) when the size of the mailbox
3834 was calculated. There was no check that a delivery was happening into an
3835 excluded directory. This bug has been fixed by ignoring all quota
3836 processing for deliveries into excluded directories.
3838 PH/10 Added the maildirfolder_create_regex option to appendfile.
3844 PH/01 The code for finding all the local interface addresses on a FreeBSD
3845 system running IPv6 was broken. This may well have applied to all BSD
3846 systems, as well as to others that have similar system calls. The broken
3847 code found IPv4 interfaces correctly, but gave incorrect values for the
3848 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3849 that it would not match correctly against @[] and not recognize the IPv6
3852 PH/02 The ipliteral router was not recognizing addresses of the form user@
3853 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3855 PH/03 Added disable_ipv6.
3857 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3858 decoded form, because it is most often used to construct To: headers
3859 lines in autoreplies, and the decoded form may well be syntactically
3860 invalid. However, $reply_address has leading white space removed, and all
3861 newlines turned into spaces so that the autoreply transport does not
3864 PH/05 If group was specified without a user on a router, and no group or user
3865 was specified on a transport, the group from the router was ignored.
3867 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3868 for visible compile-time settings that can be used to change these
3869 numbers, for those that want even more. Backwards compatibility with old
3870 spool files has been maintained. However, going back to a previous Exim
3871 release will lost any variables that are in spool files.
3873 PH/07 Two small changes when running in the test harness: increase delay when
3874 passing a TCP/IP connection to a new process, in case the original
3875 process has to generate a bounce, and remove special handling of
3876 127.0.0.2 (sic), which is no longer necessary.
3878 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3879 be the same on different OS.
3881 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3884 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3885 whether --show-vars was specified or not
3887 JJ/02 exipick: Added support for new ACL variable spool format introduced
3890 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3891 syntactically invalid From: or Reply-to: line, and a filter used this to
3892 generate an autoreply, and therefore failed to obtain an address for the
3893 autoreply, Exim could try to deliver to a non-existent relative file
3894 name, causing unrelated and misleading errors. What now happens is that
3895 it logs this as a hard delivery error, but does not attempt to create a
3898 PH/11 The exinext utility has a -C option for testing purposes, but although
3899 the given file was scanned by exinext itself; it wasn't being passed on
3900 when Exim was called.
3902 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3903 an end-of-file indication when reading a command response.
3905 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3906 compiled. In many other places in Exim, IPv6 addresses are always
3907 recognized, so I have changed this. It also means that IPv4 domain
3908 literals of the form [IPV4:n.n.n.n] are now always recognized.
3910 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3911 used if the router is not running as root, for example, when verifying at
3912 ACL time, or when using -bh. The debugging output from this situation was
3913 non-existent - all you got was a failure to exec. I have made two
3916 (a) Failures to set uid/gid, the current directory, or a process leader
3917 in a subprocess such as that created by queryprogram now generate
3918 suitable debugging output when -d is set.
3920 (b) The queryprogram router detects when it is not running as root,
3921 outputs suitable debugging information if -d is set, and then runs
3922 the subprocess without attempting to change uid/gid.
3924 PH/15 Minor change to Makefile for building test_host (undocumented testing
3927 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3928 additional section of a DNS packet that returns MX or SRV records.
3929 Instead, it always explicitly searches for A/AAAA records. This avoids
3930 major problems that occur when a DNS server includes only records of one
3931 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3932 fixed another bug: if SRV records were looked up and the corresponding
3933 address records were *not* found in the additional section, the port
3934 values from the SRV records were lost.
3936 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3937 using the correct key (the original address) when searching the retry
3938 rules in order to find which one to use for generating the retry hint.
3940 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3941 adding the default one. Similarly, if it contains a Reply-To: header, the
3942 errors_reply_to option, if set, is not used.
3944 PH/19 When calculating a retry time, Exim used to measure the "time since
3945 failure" by looking at the "first failed" field in the retry record. Now
3946 it does not use this if it is later than than the arrival time of the
3947 message. Instead it uses the arrival time. This makes for better
3948 behaviour in cases where some deliveries succeed, thus re-setting the
3949 "first failed" field. An example is a quota failure for a huge message
3950 when small messages continue to be delivered. Without this change, the
3951 "time since failure" will always be short, possible causing more frequent
3952 delivery attempts for the huge message than are intended.
3953 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3955 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3956 $1, $2, $3) because the numerical variables can be reset during some
3957 expansion items (e.g. "match"), thereby losing the authentication data.
3959 PH/21 Make -bV show the size of off_t variables so that the test suite can
3960 decide whether to run tests for quotas > 2G.
3962 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3963 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3964 filecount value is greater than 2G or if a quota value is greater than 2G
3965 on a system where the size of off_t is not greater than 4, a panic error
3968 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3969 never match. The debug and -bh output now contains an explicit error
3970 message indicating a malformed IPv4 address or mask.
3972 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3973 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3974 PH/23 above applies.
3976 PH/25 Do not write to syslog when running in the test harness. The only
3977 occasion when this arises is a failure to open the main or panic logs
3978 (for which there is an explicit test).
3980 PH/26 Added the /no_tell option to "control=freeze".
3982 PH/27 If a host name lookup failed very early in a connection, for example, if
3983 the IP address matched host_lookup and the reverse lookup yielded a name
3984 that did not have a forward lookup, an error message of the form "no IP
3985 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3986 could be logged. Now it outputs the IP address instead of "NULL".
3988 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3989 allows the sender and the authenticated sender to be set when
3990 submitting a message from within Exim. Since child_open_exim() is
3991 documented for local_scan(), the new function should be too.
3993 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3994 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3995 results in an empty string is now treated as unset.
3997 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3999 PH/31 Added sender_verify_fail logging option.
4001 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4002 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4003 tidied the source and removed it altogether.
4005 PH/33 When a queue run was abandoned because the load average was too high, a
4006 log line was always written; now it is written only if the queue_run log
4007 selector is set. In addition, the log line for abandonment now contains
4008 information about the queue run such as the pid. This is always present
4009 in "start" and "stop" lines but was omitted from the "abandon" line.
4011 PH/34 Omit spaces between a header name and the colon in the error message that
4012 is given when verify = headers_syntax fails (if there are lots of them,
4013 the message gets confusing).
4015 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4016 names, as there are now some PTR records that contain slashes. This check
4017 is only to protect against broken name servers that fall over on strange
4018 characters, so the fact that it applies to all lookups doesn't matter.
4020 PH/36 Now that the new test suite is complete, we can remove some of the
4021 special code in Exim that was needed for the old test suite. For example,
4022 sorting DNS records because real resolvers return them in an arbitrary
4023 order. The new test suite's fake resolver always returns records in the
4026 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4027 bounces) except when queue_only is set, to avoid logging races between
4028 the different processes.
4030 PH/38 Panic-die if .include specifies a non-absolute path.
4032 PH/39 A tweak to the "H" retry rule from its user.
4034 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4035 a label. They prevented compilation on older perls.
4037 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4038 a warning to be raised on newish perls.
4040 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4041 on queue. Changes to match documented behaviour of showing count of
4042 messages matching specified criteria.
4044 PH/40 Changed the default ident timeout from 30s to 5s.
4046 PH/41 Added support for the use of login_cap features, on those BSD systems
4047 that have them, for controlling the resources used by pipe deliveries.
4049 PH/42 The content-scanning code uses fopen() to create files in which to put
4050 message data. Previously it was not paying any attention to the mode of
4051 the files. Exim runs with umask(0) because the rest of the code creates
4052 files with open(), and sets the required mode explicitly. Thus, these
4053 files were ending up world-writeable. This was not a big issue, because,
4054 being within the spool directory, they were not world-accessible. I have
4055 created a function called modefopen, which takes an additional mode
4056 argument. It sets umask(777), creates the file, chmods it to the required
4057 mode, then resets the umask. All the relevant calls to fopen() in the
4058 content scanning code have been changed to use this function.
4060 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4061 to 24 hours. This avoids potential overflow problems when processing G
4062 and H retry rules. I suspect nobody ever tinkers with this value.
4064 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4066 PH/45 When the plaintext authenticator is running as a client, the server's
4067 challenges are checked to ensure they are valid base64 strings. By
4068 default, the authentication attempt is cancelled if an invalid string is
4069 received. Setting client_ignore_invalid_base64 true ignores these errors.
4070 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4071 they are received. Thus, the responses can be made to depend on the
4072 challenges. If an invalid string is ignored, an empty string is placed in
4075 PH/46 Messages that are created by the autoreply transport now contains a
4076 References: header, in accordance with RFCs 2822 and 3834.
4078 PH/47 Added authenticated_sender_force to the smtp transport.
4080 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4082 PH/49 Installed latest patch from the Sieve maintainer.
4084 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4085 was also set, the appendfile transport was unnecessarily scanning a
4086 directory of message files (e.g. for maildir delivery) to find the count
4087 of files (along with the size), even though it did not need this
4088 information. It now does the scan only if it needs to find either the
4089 size of the count of files.
4091 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4093 PH/52 Two bugs concerned with error handling when the smtp transport is
4096 (i) Exim was not creating retry information for temporary errors given
4097 for individual recipients after the DATA command when the smtp transport
4098 was used in LMTP mode. This meant that they could be retried too
4099 frequently, and not timed out correctly.
4101 (ii) Exim was setting the flag that allows error details to be returned
4102 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4103 recipients that were returned after the DATA command.
4105 PH/53 This is related to PH/52, but is more general: for any failing address,
4106 when detailed error information was permitted to be returned to the
4107 sender, but the error was temporary, then after the final timeout, only
4108 "retry timeout exceeded" was returned. Now it returns the full error as
4109 well as "retry timeout exceeded".
4111 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4112 do this, and (what is worse) MTAs that accept it.
4114 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4115 will now be deprecated.
4117 PH/56 New os.c-cygwin from the Cygwin maintainer.
4119 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4120 formats (previously only available in exim formats via -bpr, -bpru,
4121 and -bpra. Now also available in native and exiqgrep formats)
4123 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4124 with very large, slow to parse queues
4126 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4128 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4130 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4131 responses to authentication challenges, though it was showing the
4132 challenges; (ii) I've removed the CR characters from the debug output for
4135 PH/58 Allow for the insertion of a newline as well as a space when a string
4136 is turned into more than one encoded-word during RFC 2047 encoding. The
4137 Sieve code now uses this.
4139 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4140 data_4xx, lost_connection, tls_required.
4142 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4143 message was being sent as an SMTP response.
4145 PH/61 Add -l and -k options to exicyclog.
4147 PH/62 When verifying, if an address was redirected to one new address, so that
4148 verification continued, and the new address failed or deferred after
4149 having set something in $address_data, the value of $address_data was not
4150 passed back to the ACL. This was different to the case when no
4151 redirection occurred. The value is now passed back in both cases.
4153 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4154 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4155 use login_cap.h, so on its own it isn't the distinguishing feature. The
4156 new name refers directly to the setclassresources() function.
4158 PH/65 Added configuration files for NetBSD3.
4160 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4162 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4163 is preferred over IPv4.
4165 PH/68 The bounce_return_message and bounce_return_body options were not being
4166 honoured for bounces generated during the reception of non-SMTP messages.
4167 In particular, this applied to messages rejected by the ACL. This bug has
4168 been fixed. However, if bounce_return_message is true and bounce_return_
4169 body is false, the headers that are returned for a non-SMTP message
4170 include only those that have been read before the error was detected.
4171 (In the case of an ACL rejection, they have all been read.)
4173 PH/69 The HTML version of the specification is now built in a directory called
4174 spec_html instead of spec.html, because the latter looks like a path with
4175 a MIME-type, and this confuses some software.
4177 PH/70 Catch two compiler warnings in sieve.c.
4179 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4180 function verify_get_ident() calls ip_connect() to connect a socket, but
4181 if the "connect()" function timed out, ip_connect() used to close the
4182 socket. However, verify_get_ident() also closes the socket later, and in
4183 between Exim writes to the log, which may get opened at this point. When
4184 the socket was closed in ip_connect(), the log could get the same file
4185 descriptor number as the socket. This naturally causes chaos. The fix is
4186 not to close the socket in ip_connect(); the socket should be closed by
4187 the function that creates it. There was only one place in the code where
4188 this was missing, in the iplookup router, which I don't think anybody now
4189 uses, but I've fixed it anyway.
4191 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4192 well as to direct DNS lookups. Otherwise the handling of names in host
4193 lists is inconsistent and therefore confusing.
4199 PH/01 Two changes to the default runtime configuration:
4201 (1) Move the checks for relay_from_hosts and authenticated clients from
4202 after to before the (commented out) DNS black list checks.
4204 (2) Add control=submission to the relay_from_hosts and authenticated
4205 clients checks, on the grounds that messages accepted by these
4206 statements are most likely to be submissions.
4208 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4210 (1) Generate an error if the third argument for the ${prvs expansion is
4213 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4216 (3) Reset the variables that are obtained from the first argument of
4217 ${prvscheck and used in the second argument before leaving the code,
4218 because their memory is reclaimed, so using them afterwards may do
4221 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4222 one (it's much easier than Tom thought :-).
4224 (5) Because of (4), we can now allow for the use of $prvscheck_result
4225 inside the third argument.
4227 PH/03 For some reason, the default setting of PATH when running a command from
4228 a pipe transport was just "/usr/bin". I have changed it to
4231 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4232 anything to be listed in the output from -bV.
4234 PH/05 When a filter generated an autoreply, the entire To: header line was
4235 quoted in the delivery log line, like this:
4237 => >A.N.Other <ano@some.domain> <original@ddress> ...
4239 This has been changed so that it extracts the operative address. There
4240 may be more than one such address. If so, they are comma-separated, like
4243 => >ano@some.domain,ona@other.domain <original@ddress> ...
4245 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4246 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4247 not being looked up in the rDNS to get a host name, Exim was showing the
4248 IP address twice in Received: lines, even though the IP addresses were
4249 identical. For example:
4251 Received: from [1.2.3.4] (helo=[1.2.3.4])
4253 However, if the real host name was known, it was omitting the HELO data
4254 if it matched the actual IP address. This has been tidied up so that it
4255 doesn't show the same IP address twice.
4257 PH/07 When both +timestamp and +memory debugging was on, the value given by
4258 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4259 called by the debug printing, thereby overwriting the timestamp buffer.
4260 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4262 PH/08 When the original message was included in an autoreply transport, it
4263 always said "this is a copy of the message, including all the headers",
4264 even if body_only or headers_only was set. It now gives an appropriate
4267 PH/09 Applied a patch from the Sieve maintainer which:
4269 o fixes some comments
4270 o adds the (disabled) notify extension core
4271 o adds some debug output for the result of if/elsif tests
4272 o points to the current vacation draft in the documentation
4273 and documents the missing references header update
4277 o fixes a bug in processing the envelope test (when testing
4278 multiple envelope elements, the last element determined the
4281 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4282 Electronic Mail") by including:
4284 Auto-submitted: auto-generated
4286 in the messages that it generates (bounce messages and others, such as
4287 warnings). In the case of bounce messages for non-SMTP messages, there was
4288 also a typo: it was using "Auto_submitted" (underscore instead of
4289 hyphen). Since every message generated by Exim is necessarily in response
4290 to another message, thes have all been changed to:
4292 Auto-Submitted: auto-replied
4294 in accordance with these statements in the RFC:
4296 The auto-replied keyword:
4298 - SHOULD be used on messages sent in direct response to another
4299 message by an automatic process,
4301 - MUST NOT be used on manually-generated messages,
4303 - MAY be used on Delivery Status Notifications (DSNs) and Message
4304 Disposition Notifications (MDNs),
4306 - MUST NOT be used on messages generated by automatic or periodic
4307 processes, except for messages which are automatic responses to
4310 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4311 to the default Received: header definition.
4313 PH/12 Added log selector acl_warn_skipped (default on).
4315 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4316 variables because (a) they are in the wrong storage pool and (b) even if
4317 they were copied, it wouldn't work properly because of the caching.
4319 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4320 checking when decoding. Apparently there are clients that generate
4321 overlong encoded strings. Why am I not surprised?
4323 PH/15 If the first argument of "${if match_address" was not empty, but did not
4324 contain an "@" character, Exim crashed. Now it writes a panic log message
4325 and treats the condition as false.
4327 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4329 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4330 extension "envelope-auth". The code is finished and in agreement with
4331 other implementations, but there is no documentation so far and in fact,
4332 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4333 not changing the active code.
4335 Print executed "if" and "elsif" statements when debugging is used. This
4336 helps a great deal to understand what a filter does.
4338 Document more things not specified clearly in RFC3028. I had all this
4339 sorted out, when out of a sudden new issues came to my mind. Oops."
4341 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4344 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4345 canonical form (as documented). However, after a host name lookup from
4346 the IP address, check_host() was doing a simple string comparison with
4347 addresses acquired from the DNS when checking that the found name did
4348 have the original IP as one of its addresses. Since any found IPv6
4349 addresses are likely to be in abbreviated form, the comparison could
4350 fail. Luckily, there already exists a function for doing the comparison
4351 by converting both addresses to binary, so now that is used instead of
4352 the text comparison.
4354 PH/20 There was another similar case to PH/19, when a complete host name was
4355 given in a host list; looking up its IP address could give an abbreviated
4356 form, whereas the current host's name might or might not be abbreviated.
4357 The same fix has been applied.
4363 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4364 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4367 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4368 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4370 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4372 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4373 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4374 Also, for "Auto-Submitted", treat anything other than "no" as
4375 non-personal, in accordance with RFC 3834. (Previously it treated
4376 anything starting "auto-" as non-personal.)
4378 TF/01 The control=submission/name=... option had a problem with syntax
4379 errors if the name included a slash character. The /name= option
4380 now slurps the rest of the string, so it can include any characters
4381 but it must come last in the list of options (after /sender_retain
4384 PH/05 Some modifications to the interface to the fake nameserver for the new
4392 TK/01 Added the "success_on_redirect" address verification option. See
4393 NewStuff for rationale and an example.
4395 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4397 PH/02 Patch to exigrep to allow it to work on syslog lines.
4399 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4400 fread() to skip over the body file's header line, because in Cygwin the
4401 header line is locked and is inaccessible.
4403 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4404 co-exist for some time) to make it clear that it is the Exim ID that is
4405 referenced, not the Message-ID: header line.
4407 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4408 string_format() function, because snprintf() does not exist on all
4411 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4412 ${stat: expansion item.
4414 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4415 protocol synchronization error", to keep the pedants happy.
4417 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4418 well as for IRIX systems, when gcc is being used. See the host.c source
4421 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4423 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4426 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4427 transport and to the smtp transport in LMTP mode.
4429 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4431 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4432 run a filter in a subprocess. This could lead to confusion in subsequent
4433 lookups in the parent process. There should also be a search_tidyup() at
4434 the end of the subprocess.
4436 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4437 only if the host matched helo_try_verify_hosts, which caused the
4438 verification to occur when the EHLO/HELO command was issued. The ACL just
4439 tested the remembered result. Now, if a previous verification attempt has
4440 not happened, "verify = helo" does it there and then.
4442 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4444 TK/03 Fix log output including CR from clamd.
4446 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4447 exist provoked a memory error which could cause a segfault.
4449 PH/15 Installed PCRE 6.2
4451 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4453 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4454 of the problem. Specifically, suggested +O2 rather than +O1 for the
4457 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4459 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4460 record, the port was not correctly logged when the outgoing_port log
4461 selector was set (it logged the transort's default port).
4463 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4464 fallback_hosts, and "hosts" in the smtp transport.
4466 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4467 host errors such as "Connection refused".
4469 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4470 authentication with radiusclient 0.4.9:
4472 - Error returned from rc_read_config was caught wrongly
4473 - Username/password not passed on to radius server due to wrong length.
4475 The presumption is that some radiusclient API changes for 4.51/PH/17
4476 were not taken care of correctly. The code is still untested by me (my
4477 Linux distribution still has 0.3.2 of radiusclient), but it was
4478 contributed by a Radius user.
4480 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4481 expanding the "port" option of the smtp transport.
4483 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4484 while reading a MIME header. Thanks to Tom Hughes for a patch.
4486 PH/24 Include config.h inside local_scan.h so that configuration settings are
4489 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4490 that in an ACL for RCPT (for example), you can examine exactly what was
4493 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4494 commands, but it was not correctly comparing the address with the actual
4495 client host address. Thus, it would show the EHLO address in Received:
4496 header lines when this was not necessary.
4498 PH/27 Added the % operator to ${eval:}.
4500 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4501 it should be ignoring failures (because with -C, for example, it has lost
4502 privilege). It wasn't ignoring creation failures other than "already
4505 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4508 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4509 because an input line was too long, either on its own, or by virtue of
4510 too many continuations, the temporary file was not being removed, and the
4511 return code was incorrect.
4513 PH/31 Missing "BOOL" in function definition in filtertest.c.
4515 PH/32 Applied Sieve patches from the maintainer.
4517 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4519 PH/33 Added "verify = not_blind".
4521 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4522 Local/Makefile (with some defaults set). These are used in built scripts
4523 such as exicyclog, but they have never been used in the exim_install
4524 script (though there are many overriding facilities there). I have
4525 arranged that the exim_install script now takes note of these two
4528 PH/35 Installed configuration files for Dragonfly.
4530 PH/36 When a locally submitted message by a trusted user did not contain a
4531 From: header, and the sender address was obtained from -f or from an SMTP
4532 MAIL command, and the trusted user did not use -F to supply a sender
4533 name, $originator_name was incorrectly used when constructing a From:
4534 header. Furthermore, $originator_name was used for submission mode
4535 messages from external hosts without From: headers in a similar way,
4536 which is clearly wrong.
4538 PH/37 Added control=suppress_local_fixups.
4540 PH/38 When log_selector = +received_sender was set, and the addition of the
4541 sender made the log line's construction buffer exactly full, or one byte
4542 less than full, an overflow happened when the terminating "\n" was
4545 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4546 when the result of a list match is failure because a DNS lookup failed.
4548 PH/40 RM_COMMAND is now used in the building process.
4550 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4551 the "build-* directories that it finds.
4553 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4554 address in a domain literal was a prefix of an interface address.
4556 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4557 when verifying a sender address, unless rewrite_headers is false.
4559 PH/44 Wrote a long comment about why errors_to addresses are verified as
4560 recipients, not senders.
4562 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4563 the ratelimit ACL was added.
4565 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4567 PH/46 Added extra information about PostgreSQL errors to the error string.
4569 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4570 avoiding the need to install special zones in a real server. This is
4571 backwards compatible; if it can't find the fake resolver, it drops back.
4572 Thus, both old and new test suites can be run.
4574 TF/02 Added util/ratelimit.pl
4576 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4577 clock is set back in time.
4579 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4580 Candler <B.Candler@pobox.com>.
4582 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4583 OK for addresses that are the result of redirections.
4585 PH/48 A number of further additions for the benefit of the new test suite,
4586 including a fake gethostbyname() that interfaces to the fake DNS resolver
4589 TF/06 The fix for widen_domains has also been applied to qualify_single and
4590 search_parents which are the other dnslookup options that can cause
4593 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4596 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4598 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4599 service" if it encountered an MX record with an empty target hostname.
4600 The message is now "an MX or SRV record indicated no SMTP service".
4602 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4603 if the DNS of the sending site is misconfigured. This is quite a
4604 common situation. This change restores the behaviour of treating a
4605 helo verification defer as a failure.
4607 PH/51 If self=fail was set on a router, the bounce message did not include the
4608 actual error message.
4614 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4616 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4617 command itself ended in error, the underlying message about the transport
4618 filter timeout was being overwritten with the pipe command error. Now the
4619 underlying error message should be appended to the second error message.
4621 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4623 PH/02 Reduce the amount of output that "make" produces by default. Full output
4624 can still be requested.
4626 PH/03 The warning log line about a condition test deferring for a "warn" verb
4627 was being output only once per connection, rather than after each
4628 occurrence (because it was using the same function as for successful
4629 "warn" verbs). This seems wrong, so I have changed it.
4631 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4632 it should not have, which might have caused a crash in the right
4633 circumstances, but probably never did.
4635 PH/04 Installed a modified version of Tony Finch's patch to make submission
4636 mode fix the return path as well as the Sender: header line, and to
4637 add a /name= option so that you can make the user's friendly name appear
4640 TF/03 Added the control = fakedefer ACL modifier.
4642 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4643 Mark Lowes for thorough testing.
4645 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4647 TK/03 Merged latest SRS patch from Miles Wilton.
4649 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4650 with the definition in sysexits.h (which is #included earlier).
4651 Fortunately, Exim does not actually use EX_OK. The code used to try to
4652 preserve the sysexits.h value, by assuming that macro definitions were
4653 scanned for macro replacements. I have been disabused of this notion,
4654 so now the code just undefines EX_OK before #including unistd.h.
4656 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4657 in the smtp transport. When a block could not be written in a single
4658 write() function, the timeout was being re-applied to each part-write.
4659 This seems wrong - if the receiver was accepting one byte at a time it
4660 would take for ever. The timeout is now adjusted when this happens. It
4661 doesn't have to be particularly precise.
4663 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4664 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4666 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4667 without a domain if the check to <postmaster@domain> fails.
4669 SC/01 Eximstats: added -xls and the ability to specify output files
4670 (patch written by Frank Heydlauf).
4672 SC/02 Eximstats: use FileHandles for outputting results.
4674 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4676 SC/04 Eximstats: fixed display of large numbers with -nvr option
4678 SC/05 Eximstats: fixed merging of reports with empty tables.
4680 SC/06 Eximstats: added the -include_original_destination flag
4682 SC/07 Eximstats: removed tabs and trailing whitespace.
4684 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4686 TK/06 MBOX spool code: Add real "From " MBOX separator line
4687 so the .eml file is really in mbox format (even though
4688 most programs do not really care). Patch from Alex Miller.
4690 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4691 The latter is generated from $received_to and is only set if the
4692 message has one envelope recipient. SA can use these headers,
4693 obviously out-of-the-box. Patch from Alex Miller.
4695 PH/08 The ${def test on a variable was returning false if the variable's
4696 value was "0", contrary to what the specification has always said!
4697 The result should be true unless the variable is empty.
4699 PH/09 The syntax error of a character other than { following "${if
4700 def:variable_name" (after optional whitespace) was not being diagnosed.
4701 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4702 accidental colon was present, for example, could give incorrect results.
4704 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4705 result is used (not including appendfile, where other changes are about
4708 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4709 This involved changing a lot of size variables from int to off_t. It
4710 should work with maildirs and everything.
4712 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4713 spamd dying while we are connected to it.
4715 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4718 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4719 function for BDB 4.3.
4721 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4723 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4724 the sources for PCRE have been re-arranged and more files are now
4727 PH/15 The code I had for printing potentially long long variables in PH/11
4728 above was not the best (it lost precision). The length of off_t variables
4729 is now inspected at build time, and an appropriate printing format (%ld
4730 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4731 to be "long long int" or "long int". This is needed for the internal
4732 formatting function string_vformat().
4734 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4735 the configuration file to be ":syslog", then the script "guesses" where
4736 the logs files are, rather than using the compiled in default. In our
4737 case the guess is not the same as the compiled default, so the script
4738 suddenly stopped working when I started to use syslog. The patch checks
4739 to see if log_file_path is "". If so, it attempts to read it from exim
4740 with no configuration file to get the compiled in version, before it
4741 falls back to the previous guessing code."
4743 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4744 implementing BATV in an Exim configuration. See NewStuff for the gory
4747 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4748 Makefile that are specific to HP-UX.
4750 PH/18 If the "use_postmaster" option was set for a recipient callout together
4751 with the "random" option, the postmaster address was used as the MAIL
4752 FROM address for the random test, but not for the subsequent recipient
4753 test. It is now used for both.
4755 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4756 patch removes a few documentation additions to RFC 3028, because the
4757 latest draft now contains them. It adds the new en;ascii-case comparator
4758 and a new error check for 8bit text in MIME parts. Comparator and
4759 require names are now matched exactly. I enabled the subaddress
4760 extension, but it is not well tested yet (read: it works for me)."
4762 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4763 rework some of the code of TK/09 above to avoid the hardwired use of
4764 "%lld" and "long long". Replaced the call to snprintf() with a call to
4767 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4768 records point to non-existent hosts", "retry timeout exceeded", and
4769 "retry time not reached for any host after a long failure period".
4771 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4772 experimental DomainKeys support:
4774 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4775 (2) On an error such as an illegally used "control", the wrong name for
4776 the control was given.
4778 These problems did NOT occur unless DomainKeys support was compiled.
4780 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4782 PH/24 Added ${if match_ip condition.
4784 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4785 databases so that it will be absolutely obvious if a crash occurs in the
4786 DB library. This is a regular occurrence (often caused by mis-matched
4789 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4790 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4791 picked up on a user's system that detects such things. There doesn't seem
4792 to be a gcc warning option for this - only an attribute that has to be
4793 put on the function's prototype. It seems that in Fedora Core 4 they have
4794 set this on a number of new functions. No doubt there will be more in due
4797 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4798 specify a transport. However, if an address that was verified by such a
4799 router was the subject of a callout, Exim crashed because it tried to
4800 read the rcpt_include_affixes from the non-existent transport. Now it
4801 just assumes that the setting of that option is false. This bug was
4802 introduced by 4.51/PH/31.
4804 PH/28 Changed -d+all to exclude +memory, because that information is very
4805 rarely of interest, but it makes the output a lot bigger. People tend to
4806 do -d+all out of habit.
4808 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4809 code in os-type was giving problems when libc.so lives in lib64, like on
4812 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4813 aren't the modern standard, and it seems that some systems' include files
4814 don't always have them. Exim was already checking for some of the newer
4815 ones like T_AAAA, and defining it itself. I've added checks for all the
4816 record types that Exim uses.
4818 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4819 not automatically generating a new one, as it is supposed to. This
4820 prevented TLS from working. If the file did exist, but contained invalid
4821 data, a new version was generated, as expected. It was only the case of a
4822 non-existent file that was broken.
4824 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4825 with a change in libdomainkeys > 0.64.
4827 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4828 from DNS. If the selector record carries the flag, it now has
4829 precedence over the domain-wide flag.
4831 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4833 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4834 the use of a port name that isn't defined in /etc/services) occurred, the
4835 message was deferred as in a normal delivery, and thus remained on the
4836 spool, instead of being failed because of the mua_wrapper setting. This
4837 is now fixed, and I tidied up some of the mua_wrapper messages at the
4840 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4841 the messages in a hash of arrays rather than using individual hashes.
4842 This is a bit cleaner and results in dramatic memory savings, albeit
4843 at a slight CPU cost.
4845 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4846 as requested by Marc Sherman.
4848 SC/10 Eximstats: added histograms for user specified patterns as requested
4851 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4853 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4854 fopen() in the content-scanning modules that did not already have it.
4860 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4861 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4863 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4865 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4867 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4868 to test Sieve filters that use "vacation".
4870 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4871 that changes the way the GnuTLS parameters are stored in the cache file.
4872 The new format can be generated externally. For backward compatibility,
4873 if the data in the cache doesn't make sense, Exim assumes it has read an
4874 old-format file, and it generates new data and writes a new file. This
4875 means that you can't go back to an older release without removing the
4878 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4879 work if there are any delivery delays because "one_time" forces the
4880 parent to be marked "delivered", so its unseen clone is never tried
4881 again. For this reason, Exim now forbids the simultaneous setting of
4884 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4885 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4886 entry if you want to know the details. The fix, however, neglected to
4887 consider the case when local delivery batching is involved. The test for
4888 "previously delivered" was not happening when checking to see if an
4889 address could be batched with a previous (undelivered) one; under
4890 certain circumstances this could lead to multiple deliveries to the same
4893 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4894 in its include files, and this causes problems building Exim.
4896 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4897 header_syntax) but Exim was just ignoring anything given after a slash.
4898 In particular, this caused confusion with an attempt to use "verify =
4899 reverse_host_lookup/defer_ok". An error is now given when options are
4900 supplied for verify items that do not have them. (Maybe reverse_host_
4901 lookup should have a defer_ok option, but that's a different point.)
4903 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4904 defined by RFC 821) to 2048, because there were problems with some AUTH
4905 commands, and RFC 1869 says the size should be increased for extended
4906 SMTP commands that take arguments.
4908 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4911 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4912 "unknown" error; now it says that the functionality isn't in the binary.
4914 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4915 an address' error message when a string expansion fails (syntax or
4916 whatever). Otherwise the password may appear in the log. Following change
4917 PH/42 below, there is no longer a chance of it appearing in a bounce
4920 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4922 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4923 of its addresses was ever tried. (Bugzilla bug #2.)
4925 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4926 the result incorrectly in the debug output. (It correctly added a newline
4927 to what was transported.)
4929 TF/01 Added $received_time.
4931 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4932 commented out examples of how to interface to a virus scanner and to
4933 SpamAssassin. Also added commented examples of av_scanner and
4934 spamd_address settings.
4936 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4937 and controls are allowed in which ACLs. There were a couple of minor
4938 errors. Some of the entries in the conditions table (which is a table of
4939 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4940 negation of where the condition IS allowed.
4942 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4944 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4945 header file does not have a version number, so I've had to invent a new
4946 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4947 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4948 radiusclient), but it was contributed by a Radius user.
4950 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4951 files or directories, for OpenSSL.
4953 PH/19 When an Exim process that is running as root has to create an Exim log
4954 file, it does so in a subprocess that runs as exim:exim so as to get the
4955 ownership right at creation (otherwise, other Exim processes might see
4956 the file with the wrong ownership). There was no test for failure of this
4957 fork() call, which would lead to the process getting stuck as it waited
4958 for a non-existent subprocess. Forks do occasionally fail when resources
4959 run out. I reviewed all the other calls to fork(); they all seem to check
4962 PH/20 When checking for unexpected SMTP input at connect time (before writing
4963 the banner), Exim was not dealing correctly with a non-positive return
4964 from the read() function. If the client had disconnected by this time,
4965 the result was a log entry for a synchronization error with an empty
4966 string after "input=" when read() returned zero. If read() returned -1
4967 (an event I could not check), uninitialized data bytes were printed.
4968 There were reports of junk text (parts of files, etc) appearing after
4971 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4973 PH/22 Added support for macro redefinition, and (re)definition in between
4974 driver and ACL definitions.
4976 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4977 forgetting to use the resulting value; it was using the unexpanded value.
4979 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4980 hadn't been configured. The fix is from Juergen Kreileder, who
4981 understands it better than I do:
4983 "Here's what I see happening with three configured cyrus_sasl
4984 authenticators configured (plain, login, cram-md5):
4986 On startup auth_cyrus_sasl_init() gets called for each of these.
4987 This means three calls to sasl_listmech() without a specified mech_list.
4988 => SASL tests which mechs of all available mechs actually work
4989 => three warnings about OTP not working
4990 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4992 With the patch, sasl_listmech() also gets called three times. But now
4993 SASL's mech_list option is set to the server_mech specified in the the
4994 authenticator. Or in other words, the answer from sasl_listmech()
4995 gets limited to just the mech you're testing for (which is different
4997 => the return list contains just 'plain' or 'login', 'cram-md5' or
4998 nothing depending on the value of ob->server_mech.
5000 I've just tested the patch: Authentication still works fine,
5001 unavailable mechs specified in the exim configuration are still
5002 caught, and the auth.log warnings about OTP are gone."
5004 PH/25 When debugging is enabled, the contents of the command line are added
5005 to the debugging output, even when log_selector=+arguments is not
5008 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5009 answer is "GNU", and only if the return is "GNU/something" is the answer
5012 PH/27 $acl_verify_message is now set immediately after the failure of a
5013 verification in an ACL, and so is available in subsequent modifiers. In
5014 particular, the message can be preserved by coding like this:
5016 warn !verify = sender
5017 set acl_m0 = $acl_verify_message
5019 Previously, $acl_verify_message was set only while expanding "message"
5020 and "log_message" when a very denied access.
5022 PH/28 Modified OS/os.c-Linux with
5024 -#ifndef OS_LOAD_AVERAGE
5025 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5027 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5028 nomenclature these days.)
5030 PH/29 Installed patch from the Sieve maintainer that adds the options
5031 sieve_useraddress and sieve_subaddress to the redirect router.
5033 PH/30 In these circumstances:
5034 . Two addresses routed to the same list of hosts;
5035 . First host does not offer TLS;
5036 . First host accepts first address;
5037 . First host gives temporary error to second address;
5038 . Second host offers TLS and a TLS session is established;
5039 . Second host accepts second address.
5040 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5041 and peerdn, if requested) that were in fact used only for the second
5044 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5045 attention to any local part prefix or suffix that was matched by the
5046 router that accepted the address. It now behaves in the same way as it
5047 does for delivery: the affixes are removed from the local part unless
5048 rcpt_include_affixes is set on the transport.
5050 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5051 timeout during the DATA phase of an incoming message.
5053 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5054 applied a patch sanctioned by the Sieve maintainer.
5056 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5057 the uid or gid is negative. A case of a negative gid caused this to be
5058 noticed. The fix allows for either to be negative.
5060 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5061 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5064 PH/36 The change PH/12 above was broken. Fixed it.
5066 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5067 the grounds that routing the same address twice would always produce the
5068 same answer. This might have been true once, but it is certainly no
5069 longer true now. Routing a child address may depend on the previous
5070 routing that produced that child. Some complicated redirection strategies
5071 went wrong when messages had multiple recipients, and made Exim's
5072 behaviour dependent on the order in which the addresses were given.
5074 I have moved the duplicate checking until after the routing is complete.
5075 Exim scans the addresses that are assigned to local and remote
5076 transports, and removes any duplicates. This means that more work will be
5077 done, as duplicates will always all be routed, but duplicates are
5078 presumably rare, so I don't expect this is of any significance.
5080 For deliveries to pipes, files, and autoreplies, the duplicate checking
5081 still happens during the routing process, since they are not going to be
5084 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5085 It corrects a timeout issue with spamd. This is Ian's comment: "The
5086 background is that sometimes spamd either never reads data from a
5087 connection it has accepted, or it never writes response data. The exiscan
5088 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5089 blindly assumes that writes won't block so it may never time out."
5091 PH/39 Allow G after quota size as well as K and M.
5093 PH/40 The value set for $authenticated_id in an authenticator may not contain
5094 binary zeroes or newlines because the value is written to log lines and
5095 to spool files. There was no check on this. Now the value is run through
5096 the string_printing() function so that such characters are converted to
5097 printable escape sequences.
5099 PH/41 $message_linecount is a new variable that contains the total number of
5100 lines in the message. Compare $body_linecount, which is the count for the
5103 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5104 bounce and delay warning messages, except in certain special cases, which
5107 (a) An SMTP error message from a remote host;
5108 (b) A message specified in a :fail: redirection;
5109 (c) A message specified in a "fail" command in a system filter;
5110 (d) A message specified in a FAIL return from the queryprogram router;
5111 (e) A message specified by the cannot_route_message router option.
5113 In these cases only, Exim does include the error details in bounce and
5114 warning messages. There are also a few cases where bland messages such
5115 as "unrouteable address" or "local delivery error" are given.
5117 PH/43 $value is now also set for the "else" part of a ${run expansion.
5119 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5120 being worked on, but at least Exim now implements the latest version to
5123 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5124 process to complete was treated as a delivery failure, a timeout while
5125 writing the message to the pipe was logged, but erroneously treated as a
5126 successful delivery. Such timeouts include transport filter timeouts. For
5127 consistency with the overall process timeout, these timeouts are now
5128 treated as errors, giving rise to delivery failures by default. However,
5129 there is now a new Boolean option for the pipe transport called
5130 timeout_defer, which, if set TRUE, converts the failures into defers for
5131 both kinds of timeout. A transport filter timeout is now identified in
5134 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5135 systems where "make" and "gmake" are different, calling "gmake" at top
5136 level broke things. I've arranged for the value of $(MAKE) to be passed
5137 from the Makefile to this script so that it can call the same version of
5141 A note about Exim versions 4.44 and 4.50
5142 ----------------------------------------
5144 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5145 changes of various kinds. As a consequence, a big documentation update was
5146 needed. This delayed the release for rather longer than seemed good, especially
5147 in the light of a couple of (minor) security issues. Therefore, the changes
5148 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5149 release. So 4.44 and 4.50 are in effect two different branches that both start
5152 I have left the 4.50 change log unchanged; it contains all the changes since
5153 4.43. The change log for 4.44 is below; many of its items are identical to
5154 those for 4.50. This seems to be the most sensible way to preserve the
5155 historical information.
5161 1. Minor wording change to the doc/README.SIEVE file.
5163 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5164 computation of the current number of files was incorrect.
5166 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5167 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5170 4. Give more explanation in the error message when the command for a transport
5171 filter fails to execute.
5173 5. There are several places where Exim runs a non-Exim command in a
5174 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5175 was being done only for the command run by the queryprogram router. It is
5176 now done for all such subprocesses. The other cases are: ${run, transport
5177 filters, and the commands run by the lmtp and pipe transports.
5179 6. Added CONFIGURE_GROUP build-time option.
5181 7. Some older OS have a limit of 256 on the maximum number of file
5182 descriptors. Exim was using setrlimit() to set 1000 as a large value
5183 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5184 systems. I've change it so that if it can't get 1000, it tries for 256.
5186 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5187 was an oversight, and furthermore, ever since the addition of extra
5188 controls (e.g. 4.43/32), the checks on when to allow different forms of
5189 "control" were broken. There should now be diagnostics for all cases when a
5190 control that does not make sense is encountered.
5192 9. Added the /retain_sender option to "control=submission".
5194 10. $recipients is now available in the predata ACL (oversight).
5196 11. Tidy the search cache before the fork to do a delivery from a message
5197 received from the command line. Otherwise the child will trigger a lookup
5198 failure and thereby defer the delivery if it tries to use (for example) a
5199 cached ldap connection that the parent has called unbind on.
5201 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5202 of $address_data from the recipient verification was clobbered by the
5203 sender verification.
5205 13. The value of address_data from a sender verification is now available in
5206 $sender_address_data in subsequent conditions in the ACL statement.
5208 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5210 15. Added a new option "connect=<time>" to callout options, to set a different
5213 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5214 was its contents. (It was OK if the option was not defined at all.)
5216 17. A "Completed" log line is now written for messages that are removed from
5217 the spool by the -Mrm option.
5219 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5220 information about exactly what failed.
5222 19. Added -dd to debug only the daemon process.
5224 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5225 handles timeouts, both on the server side and network timeouts. Renamed the
5226 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5228 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5229 It is now set to "smtps".
5231 22. $host_address is now set to the target address during the checking of
5232 ignore_target_hosts.
5234 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5235 being passed; this would have caused $sender_host_name to have been used if
5236 matching the list had actually called for a host name (not very likely,
5237 since this list is usually IP addresses). A host name is now passed as
5240 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5241 code that specifies a non-restarting handler (typically sigaction() in
5242 modern systems) in an attempt to fix a rare and obscure crash bug.
5244 25. Narrowed the window for a race in the daemon that could cause it to ignore
5245 SIGCHLD signals. This is not a major problem, because they are used only to
5246 wake it up if nothing else does.
5248 26. A malformed maildirsize file could cause Exim to calculate negative values
5249 for the mailbox size or file count. Odd effects could occur as a result.
5250 The maildirsize information is now recalculated if the size or filecount
5253 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5254 support for a long time. Removed HAVE_SYS_VFS_H.
5256 28. Installed the latest version of exipick from John Jetmore.
5258 29. In an address list, if the pattern was not a regular expression, an empty
5259 subject address (from a bounce message) matched only if the pattern was an
5260 empty string. Non-empty patterns were not even tested. This was the wrong
5261 because it is perfectly reasonable to use an empty address as part of a
5262 database query. An empty address is now tested by patterns that are
5263 lookups. However, all the other forms of pattern expect the subject to
5264 contain a local part and a domain, and therefore, for them, an empty
5265 address still always fails if the pattern is not itself empty.
5267 30. Exim went into a mad DNS loop when attempting to do a callout where the
5268 host was specified on an smtp transport, and looking it up yielded more
5269 than one IP address.
5271 31. Re-factored the code for checking spool and log partition space into a
5272 function that finds that data and another that does the check. The former
5273 is then used to implement four new variables: $spool_space, $log_space,
5274 $spool_inodes, and $log_inodes.
5276 32. The RFC2047 encoding function was originally intended for short strings
5277 such as real names; it was not keeping to the 75-character limit for
5278 encoded words that the RFC imposes. It now respects the limit, and
5279 generates multiple encoded words if necessary. To be on the safe side, I
5280 have increased the buffer size for the ${rfc2047: expansion operator from
5283 33. It is now permitted to omit both strings after an "if" condition; if the
5284 condition is true, the result is "true". As before, when the second string
5285 is omitted, a false condition yields an empty string. This makes it less
5286 cumbersome to write custom ACL and router conditions.
5288 34. Failure to deliver a bounce message always caused it to be frozen, even if
5289 there was an errors_to setting on the router. The errors_to setting is now
5292 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5293 canonical form (fully expanded) before being placed in
5294 $sender_host_address.
5296 36. The table in the code that translates DNS record types into text (T_A to
5297 "A" for instance) was missing entries for NS and CNAME. It is just possible
5298 that this could have caused confusion if both these types were looked up
5299 for the same domain, because the text type is used as part of Exim's
5300 per-process caching. But the chance of anyone hitting this buglet seems
5303 37. The dnsdb lookup has been extended in a number of ways.
5305 (1) There is a new type, "zns", which walks up the domain tree until it
5306 finds some nameserver records. It should be used with care.
5308 (2) There is a new type, "mxh", which is like "mx" except that it returns
5309 just the host names, not the priorities.
5311 (3) It is now possible to give a list of domains (or IP addresses) to be
5312 looked up. The behaviour when one of the lookups defers can be
5313 controlled by a keyword.
5315 (4) It is now possible to specify the separator character for use when
5316 multiple records are returned.
5318 38. The dnslists ACL condition has been extended: it is now possible to supply
5319 a list of IP addresses and/or domains to be looked up in a particular DNS
5322 39. Added log_selector=+queue_time_overall.
5324 40. When running the queue in the test harness, wait just a tad after forking a
5325 delivery process, to get repeatability of debugging output.
5327 41. Include certificate and key file names in error message when GnuTLS fails
5328 to set them up, because the GnuTLS error message doesn't include the name
5329 of the failing file when there is a problem reading it.
5331 42. Allow both -bf and -bF in the same test run.
5333 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5335 44. The "Exiscan patch" is now merged into the mainline Exim source.
5337 45. Sometimes the final signoff response after QUIT could fail to get
5338 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5339 < 0 before doing a fflush(). This bug looks as though it goes back to the
5340 introduction of TLS in release 3.20, but "sometimes" must have been rare
5341 because the tests only now provoked it.
5343 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5344 (this can affect the format of dates).
5346 47. exim_tidydb, when checking for the continued existence of a message for
5347 which it has found a message-specific retry record, was not finding
5348 messages that were in split spool directories. Consequently, it was
5349 deleting retry records that should have stayed in existence.
5351 48. Steve fixed some bugs in eximstats.
5353 49. The SPA authentication driver was not abandoning authentication and moving
5354 on to the next authenticator when an expansion was forced to fail,
5355 contradicting the general specification for all authenticators. Instead it
5356 was generating a temporary error. It now behaves as specified.
5358 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5359 (the order specifies the preference for clients). The order is now AES256,
5360 AES128, 3DES, ARCFOUR128.
5362 51. Small patch to Sieve code - explicitly set From: when generating an
5365 52. Exim crashed if a remote delivery caused a very long error message to be
5366 recorded - for instance if somebody sent an entire SpamAssassin report back
5367 as a large number of 550 error lines. This bug was coincidentally fixed by
5368 increasing the size of one of Exim's internal buffers (big_buffer) that
5369 happened as part of the Exiscan merge. However, to be on the safe side, I
5370 have made the code more robust (and fixed the comments that describe what
5373 53. Now that there can be additional text after "Completed" in log lines (if
5374 the queue_time_overall log selector is set), a one-byte patch to exigrep
5375 was needed to allow it to recognize "Completed" as not the last thing in
5378 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5379 patch that reportedly fixes this has been added. I am not expert enough to
5380 create a test for it. This is what the patch creator wrote:
5382 "I found a little strange behaviour of ldap code when working with
5383 Windows 2003 AD Domain, where users was placed in more than one
5384 Organization Units. When I tried to give exim partial DN, the exit code
5385 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5386 But simultaneously result of request was absolutely normal ldap result,
5387 so I produce this patch..."
5389 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5390 REFERENCE, so I have modified the code to exclude the patch when that macro
5393 55. Some experimental protocols are using DNS PTR records for new purposes. The
5394 keys for these records are domain names, not reversed IP addresses. The
5395 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5396 leaves it alone. Component reversal etc. now happens only for IP addresses.
5399 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5401 57. Double the size of the debug message buffer (to 2048) so that more of very
5402 long debug lines gets shown.
5404 58. The exicyclog utility now does better if the number of log files to keep
5405 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5407 59. Two changes related to the smtp_active_hostname option:
5409 (1) $smtp_active_hostname is now available as a variable.
5410 (2) The default for smtp_banner uses $smtp_active_hostname instead
5411 of $primary_hostname.
5413 60. The host_aton() function is supposed to be passed a string that is known
5414 to be a valid IP address. However, in the case of IPv6 addresses, it was
5415 not checking this. This is a hostage to fortune. Exim now panics and dies
5416 if the condition is not met. A case was found where this could be provoked
5417 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5418 components; fortuitously, this particular loophole had already been fixed
5419 by change 4.50/55 above.
5421 If there are any other similar loopholes, the new check in host_aton()
5422 itself should stop them being exploited. The report I received stated that
5423 data on the command line could provoke the exploit when Exim was running as
5424 exim, but did not say which command line option was involved. All I could
5425 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5426 running as the user.
5429 61. There was a buffer overflow vulnerability in the SPA authentication code
5430 (which came originally from the Samba project). I have added a test to the
5431 spa_base64_to_bits() function which I hope fixes it.
5434 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5435 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5437 63. The daemon start-up calls getloadavg() while still root for those OS that
5438 need the first call to be done as root, but it missed one case: when
5439 deliver_queue_load_max is set with deliver_drop_privilege. This is
5440 necessary for the benefit of the queue runner, because there is no re-exec
5441 when deliver_drop_privilege is set.
5443 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5444 This has been fixed.
5446 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5447 was in use, was not putting the data itself into the right store pool;
5448 consequently, it could be overwritten for a subsequent message in the same
5449 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5452 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5454 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5455 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5456 the function were treating the return as a boolean value, which happened to
5457 work because 0=false and not-0=true, but is not correct code.
5459 68. The host_aton() function was not handling scoped IPv6 addresses (those
5460 with, for example, "%eth0" on the end) correctly.
5462 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5463 negated items (that is, ~something) in unsigned ints. Some compilers
5464 apparently mutter when there is no cast.
5466 70. If an address verification called from an ACL failed, and did not produce a
5467 user-specific message (i.e. there was only a "system" message), nothing was
5468 put in $acl_verify_message. In this situation, it now puts the system
5471 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5472 session; change 4.31/43 added the unwanted input to the log line - except
5473 that it did not do this in the start of session case. It now does.
5475 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5476 This is wrong and can cause the other end to generate a synchronization
5477 error if it is another Exim or anything else that does the synchronization
5478 check. A QUIT command is no longer sent after a timeout.
5480 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5481 during host lookups.
5483 74. The defer_ok option of callout verification was not working if it was used
5484 when verifying addresses in header lines, that is, for this case:
5486 verify = header_sender/callout=defer_ok
5488 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5489 those file descriptors could be used for SMTP connections. If anything
5490 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5491 could be sent to the SMTP client, causing chaos. The daemon now opens
5492 stdin, stdout, and stderr to /dev/null when it puts itself into the
5495 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5496 log by default. The user can override this with suitable Perl magic.
5498 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5499 the log message when discard triggers, was not working for the DATA ACL or
5500 for the non-SMTP ACL.
5502 78. Error message wording change in sieve.c.
5504 79. If smtp_accept_max_per_host was set, the number of connections could be
5505 restricted to fewer than expected, because the daemon was trying to set up
5506 a new connection before checking whether the processes handling previous
5507 connections had finished. The check for completed processes is now done
5508 earlier. On busy systems, this bug wouldn't be noticed because something
5509 else would have woken the daemon, and it would have reaped the completed
5512 80. If a message was submitted locally by a user whose login name contained one
5513 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5514 It caused a spool format error. I have fixed the spool reading code. A
5515 related problem was that the "from" clause in the Received: line became
5516 illegal because of the space(s). It is now covered by ${quote_local_part.
5518 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5521 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5523 83. Patch from Sieve maintainer to fix unterminated string problem in
5524 "vacation" handling.
5526 84. Some minor changes to the Linux configuration files to help with other
5527 OS variants using glibc.
5529 85. One more patch for Sieve to update vacation handling to latest spec.
5532 ----------------------------------------------------
5533 See the note above about the 4.44 and 4.50 releases.
5534 ----------------------------------------------------
5540 1. Change 4.43/35 introduced a bug that caused file counts to be
5541 incorrectly computed when quota_filecount was set in an appendfile
5544 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5545 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5548 3. Give more explanation in the error message when the command for a transport
5549 filter fails to execute.
5551 4. There are several places where Exim runs a non-Exim command in a
5552 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5553 was being done only for the command run by the queryprogram router. It is
5554 now done for all such subprocesses. The other cases are: ${run, transport
5555 filters, and the commands run by the lmtp and pipe transports.
5557 5. Some older OS have a limit of 256 on the maximum number of file
5558 descriptors. Exim was using setrlimit() to set 1000 as a large value
5559 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5560 systems. I've change it so that if it can't get 1000, it tries for 256.
5562 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5563 was an oversight, and furthermore, ever since the addition of extra
5564 controls (e.g. 4.43/32), the checks on when to allow different forms of
5565 "control" were broken. There should now be diagnostics for all cases when a
5566 control that does not make sense is encountered.
5568 7. $recipients is now available in the predata ACL (oversight).
5570 8. Tidy the search cache before the fork to do a delivery from a message
5571 received from the command line. Otherwise the child will trigger a lookup
5572 failure and thereby defer the delivery if it tries to use (for example) a
5573 cached ldap connection that the parent has called unbind on.
5575 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5576 of $address_data from the recipient verification was clobbered by the
5577 sender verification.
5579 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5580 was its contents. (It was OK if the option was not defined at all.)
5582 11. A "Completed" log line is now written for messages that are removed from
5583 the spool by the -Mrm option.
5585 12. $host_address is now set to the target address during the checking of
5586 ignore_target_hosts.
5588 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5589 being passed; this would have caused $sender_host_name to have been used if
5590 matching the list had actually called for a host name (not very likely,
5591 since this list is usually IP addresses). A host name is now passed as
5594 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5595 code that specifies a non-restarting handler (typically sigaction() in
5596 modern systems) in an attempt to fix a rare and obscure crash bug.
5598 15. Narrowed the window for a race in the daemon that could cause it to ignore
5599 SIGCHLD signals. This is not a major problem, because they are used only to
5600 wake it up if nothing else does.
5602 16. A malformed maildirsize file could cause Exim to calculate negative values
5603 for the mailbox size or file count. Odd effects could occur as a result.
5604 The maildirsize information is now recalculated if the size or filecount
5607 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5608 support for a long time. Removed HAVE_SYS_VFS_H.
5610 18. Updated exipick to current release from John Jetmore.
5612 19. Allow an empty sender to be matched against a lookup in an address list.
5613 Previously the only cases considered were a regular expression, or an
5616 20. Exim went into a mad DNS lookup loop when doing a callout where the
5617 host was specified on the transport, if the DNS lookup yielded more than
5620 21. The RFC2047 encoding function was originally intended for short strings
5621 such as real names; it was not keeping to the 75-character limit for
5622 encoded words that the RFC imposes. It now respects the limit, and
5623 generates multiple encoded words if necessary. To be on the safe side, I
5624 have increased the buffer size for the ${rfc2047: expansion operator from
5627 22. Failure to deliver a bounce message always caused it to be frozen, even if
5628 there was an errors_to setting on the router. The errors_to setting is now
5631 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5632 canonical form (fully expanded) before being placed in
5633 $sender_host_address.
5635 24. Updated eximstats to version 1.33
5637 25. Include certificate and key file names in error message when GnuTLS fails
5638 to set them up, because the GnuTLS error message doesn't include the name
5639 of the failing file when there is a problem reading it.
5641 26. Expand error message when OpenSSL has problems setting up cert/key files.
5644 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5645 (this can affect the format of dates).
5647 28. exim_tidydb, when checking for the continued existence of a message for
5648 which it has found a message-specific retry record, was not finding
5649 messages that were in split spool directories. Consequently, it was
5650 deleting retry records that should have stayed in existence.
5652 29. eximstats updated to version 1.35
5653 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5654 1.35 - bugfix such that pie charts by volume are generated correctly
5656 30. The SPA authentication driver was not abandoning authentication and moving
5657 on to the next authenticator when an expansion was forced to fail,
5658 contradicting the general specification for all authenticators. Instead it
5659 was generating a temporary error. It now behaves as specified.
5661 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5662 (the order specifies the preference for clients). The order is now AES256,
5663 AES128, 3DES, ARCFOUR128.
5665 31. Small patch to Sieve code - explicitly set From: when generating an
5668 32. Exim crashed if a remote delivery caused a very long error message to be
5669 recorded - for instance if somebody sent an entire SpamAssassin report back
5670 as a large number of 550 error lines. This bug was coincidentally fixed by
5671 increasing the size of one of Exim's internal buffers (big_buffer) that
5672 happened as part of the Exiscan merge. However, to be on the safe side, I
5673 have made the code more robust (and fixed the comments that describe what
5676 33. Some experimental protocols are using DNS PTR records for new purposes. The
5677 keys for these records are domain names, not reversed IP addresses. The
5678 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5679 leaves it alone. Component reversal etc. now happens only for IP addresses.
5682 34. The host_aton() function is supposed to be passed a string that is known
5683 to be a valid IP address. However, in the case of IPv6 addresses, it was
5684 not checking this. This is a hostage to fortune. Exim now panics and dies
5685 if the condition is not met. A case was found where this could be provoked
5686 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5687 components; fortuitously, this particular loophole had already been fixed
5688 by change 4.50/55 or 4.44/33 above.
5690 If there are any other similar loopholes, the new check in host_aton()
5691 itself should stop them being exploited. The report I received stated that
5692 data on the command line could provoke the exploit when Exim was running as
5693 exim, but did not say which command line option was involved. All I could
5694 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5695 running as the user.
5698 35. There was a buffer overflow vulnerability in the SPA authentication code
5699 (which came originally from the Samba project). I have added a test to the
5700 spa_base64_to_bits() function which I hope fixes it.
5703 36. The daemon start-up calls getloadavg() while still root for those OS that
5704 need the first call to be done as root, but it missed one case: when
5705 deliver_queue_load_max is set with deliver_drop_privilege. This is
5706 necessary for the benefit of the queue runner, because there is no re-exec
5707 when deliver_drop_privilege is set.
5709 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5710 was in use, was not putting the data itself into the right store pool;
5711 consequently, it could be overwritten for a subsequent message in the same
5712 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5715 38. Sometimes the final signoff response after QUIT could fail to get
5716 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5717 < 0 before doing a fflush(). This bug looks as though it goes back to the
5718 introduction of TLS in release 3.20, but "sometimes" must have been rare
5719 because the tests only now provoked it.
5725 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5726 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5727 Now it returns an integer. A number of calls were still expecting a T/F
5728 return. Fortuitously, in all cases, the tests worked in OK situations,
5729 which is the norm. However, things would have gone wrong on any write
5730 failures on the smtp file descriptor. This function is used when sending
5731 messages over SMTP and also when doing verify callouts.
5733 2. When Exim is called to do synchronous delivery of a locally submitted
5734 message (the -odf or -odi options), it no longer closes stderr before doing
5737 3. Implemented the mua_wrapper option.
5739 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5741 5. Implemented the functions header_remove(), header_testname(),
5742 header_add_at_position(), and receive_remove_recipient(), and exported them
5745 6. If an ACL "warn" statement specified the addition of headers, Exim already
5746 inserted X-ACL-Warn: at the start if there was no header name. However, it
5747 was not making this test for the second and subsequent header lines if
5748 there were newlines in the string. This meant that an invalid header could
5749 be inserted if Exim was badly configured.
5751 7. Allow an ACL "warn" statement to add header lines at the start or after all
5752 the Received: headers, as well as at the end.
5754 8. Added the rcpt_4xx retry error code.
5756 9. Added postmaster_mailfrom=xxx to callout verification option.
5758 10. Added mailfrom=xxxx to the callout verification option, for verify=
5761 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5762 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5764 12. Inserted some casts to stop certain compilers warning when using pointer
5765 differences as field lengths or precisions in printf-type calls (mostly
5766 affecting debugging statements).
5768 13. Added optional readline() support for -be (dynamically loaded).
5770 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5771 same clock tick as a message's arrival, so that its received time was the
5772 same as the "first fail" time on the retry record, and that message
5773 remained on the queue past the ultimate address timeout, every queue runner
5774 would try a delivery (because it was past the ultimate address timeout) but
5775 after another failure, the ultimate address timeout, which should have then
5776 bounced the address, did not kick in. This was a "< instead of <=" error;
5777 in most cases the first failure would have been in the next clock tick
5778 after the received time, and all would be well.
5780 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5781 being recognized when the domain list was tested by the match_domain
5782 condition in an expansion string.
5784 16. Added the ${str2b64: operator.
5786 17. Exim was always calling setrlimit() to set a large limit for the number of
5787 processes, without checking whether the existing limit was already
5788 adequate. (It did check for the limit on file descriptors.) Furthermore,
5789 errors from getrlimit() and setrlimit() were being ignored. Now they are
5790 logged to the main and panic logs, but Exim does carry on, to try to do its
5791 job under whatever limits there are.
5793 18. Imported PCRE 5.0.
5795 19. Trivial typo in log message " temporarily refused connection" (the leading
5798 20. If the log selector return_path_on_delivery was set and an address was
5799 redirected to /dev/null, the delivery process crashed because it assumed
5800 that a return path would always be set for a "successful" delivery. In this
5801 case, the whole delivery is bypassed as an optimization, and therefore no
5804 21. Internal re-arrangement: the function for sending a challenge and reading
5805 a response while authentication was assuming a zero-terminated challenge
5806 string. It's now changed to take a pointer and a length, to allow for
5807 binary data in such strings.
5809 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5811 23. Exim was not respecting finduser_retries when seeking the login of the
5812 uid under which it was called; it was always trying 10 times. (The default
5813 setting of finduser_retries is zero.) Also, it was sleeping after the final
5814 failure, which is pointless.
5816 24. Implemented tls_on_connect_ports.
5818 25. Implemented acl_smtp_predata.
5820 26. If the domain in control=submission is set empty, Exim assumes that the
5821 authenticated id is a complete email address when it generates From: or
5822 Sender: header lines.
5824 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5825 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5826 chown and chgrp in /bin and hostname in /usr/bin.
5828 28. Exim was keeping the "process log" file open after each use, just as it
5829 does for the main log. This opens the possibility of it remaining open for
5830 long periods when the USR1 signal hits a daemon. Occasional processlog
5831 errors were reported, that could have been caused by this. Anyway, it seems
5832 much more sensible not to leave this file open at all, so that is what now
5835 29. The long-running daemon process does not normally write to the log once it
5836 has entered its main loop, and it closes the log before doing so. This is
5837 so that log files can straightforwardly be renamed and moved. However,
5838 there are a couple of unusual error situations where the daemon does write
5839 log entries, and I had neglected to close the log afterwards.
5841 30. The text of an SMTP error response that was received during a remote
5842 delivery was being truncated at 512 bytes. This is too short for some of
5843 the long messages that one sometimes sees. I've increased the limit to
5846 31. It is now possible to make retry rules that apply only when a message has a
5847 specific sender, in particular, an empty sender.
5849 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5850 it possible to be selective about when SMTP synchronization is enforced.
5852 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5854 32. Implemented hosts_connection_nolog.
5856 33. Added an ACL for QUIT.
5858 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5861 35. Added mailbox_size and mailbox_filecount to appendfile.
5863 36. Added control = no_multiline_responses to ACLs.
5865 37. There was a bug in the logic of the code that waits for the clock to tick
5866 in the case where the clock went backwards by a substantial amount such
5867 that the microsecond fraction of "now" was more than the microsecond
5868 fraction of "then" (but the whole seconds number was less).
5870 38. Added support for the libradius Radius client library this is found on
5871 FreeBSD (previously only the radiusclient library was supported).
5877 1. When certain lookups returned multiple values in the form name=value, the
5878 quoting of the values was not always being done properly. Specifically:
5879 (a) If the value started with a double quote, but contained no whitespace,
5881 (b) If the value contained whitespace other than a space character (i.e.
5882 tabs or newlines or carriage returns) it was not quoted.
5883 This fix has been applied to the mysql and pgsql lookups by writing a
5884 separate quoting function and calling it from the lookup code. The fix
5885 should probably also be applied to nisplus, ibase and oracle lookups, but
5886 since I cannot test any of those, I have not disturbed their existing code.
5888 2. A hit in the callout cache for a specific address caused a log line with no
5889 reason for rejecting RCPT. Now it says "Previous (cached) callout
5890 verification failure".
5892 3. There was an off-by-one bug in the queryprogram router. An over-long
5893 return line was truncated at 256 instead of 255 characters, thereby
5894 overflowing its buffer with the terminating zero. As well as fixing this, I
5895 have increased the buffer size to 1024 (and made a note to document this).
5897 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5898 when Exim is waiting for an SMTP response from a remote server, Exim
5899 restarts its select() call on the socket, thereby resetting its timeout.
5900 This is not a problem when such interrupts are rare. Somebody set up a cron
5901 job to run exiwhat every 2 minutes, which is less than the normal select()
5902 timeout (5 or 10 minutes). This meant that the select() timeout never
5903 kicked in because it was always reset. I have fixed this by comparing the
5904 time when an interrupt arrives with the time at the start of the first call
5905 to select(). If more time than the timeout has elapsed, the interrupt is
5906 treated as a timeout.
5908 5. Some internal re-factoring in preparation for the addition of Sieve
5909 extensions (by MH). In particular, the "personal" test is moved to a
5910 separate function, and given an option for scanning Cc: and Bcc: (which is
5911 not set for Exim filters).
5913 6. When Exim created an email address using the login of the caller as the
5914 local part (e.g. when creating a From: or Sender: header line), it was not
5915 quoting the local part when it contained special characters such as @.
5917 7. Installed new OpenBSD configuration files.
5919 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5920 try to make them clearer.
5922 9. Callout options, other than the timeout value, were being ignored when
5923 verifying sender addresses in header lines. For example, when using
5925 verify = header_sender/callout=no_cache
5927 the cache was (incorrectly) being used.
5929 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5930 only the Cygwin environment.
5932 11. When return_path_on_delivery was set as a log selector, if different remote
5933 addresses in the same message used different return paths and parallel
5934 remote delivery occurred, the wrong values would sometimes be logged.
5935 (Whenever a remote delivery process finished, the return path value from
5936 the most recently started remote delivery process was logged.)
5938 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5939 lines when AUTH and/or TLS are in use. This is the "received protocol"
5940 field. Exim used to use "asmtp" for authenticated SMTP, without any
5941 indication (in the protocol name) for TLS use. Now it follows the RFC and
5942 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5943 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5944 names appear in log lines as well as in Received: header lines.
5946 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5947 extensions, and comparison tests, and to fix some bugs.
5949 14. Changes to the "personal" filter test:
5951 (1) The test was buggy in that it was just doing the equivalent of
5952 "contains" tests on header lines. For example, if a user's address was
5953 anne@some.where, the "personal" test would incorrectly be true for
5955 To: susanne@some.where
5957 This test is now done by extracting each address from the header in turn,
5958 and checking the entire address. Other tests that are part of "personal"
5959 are now done using regular expressions (for example, to check local parts
5960 of addresses in From: header lines).
5962 (2) The list of non-personal local parts in From: addresses has been
5963 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5964 taken from the Sieve specification recommendations.
5966 (3) If the message contains any header line starting with "List-" it is
5967 treated as non-personal.
5969 (4) The test for "circular" in the Subject: header line has been removed
5970 because it now seems ill-conceived.
5972 15. Minor typos in src/EDITME comments corrected.
5974 16. Installed latest exipick from John Jetmore.
5976 17. If headers_add on a router specified a text string that was too long for
5977 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5978 of string_sprintf() is now avoided.
5980 18. $message_body_size was not set (it was always zero) when running the DATA
5981 ACL and the local_scan() function.
5983 19. For the "mail" command in an Exim filter, no default was being set for
5984 the once_repeat time, causing a random time value to be used if "once" was
5985 specified. (If the value happened to be <= 0, no repeat happened.) The
5986 default is now 0s, meaning "never repeat". The "vacation" command was OK
5987 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5988 (I found it when inspecting the code).
5990 20. There is now an overall timeout for performing a callout verification. It
5991 defaults to 4 times the callout timeout, which applies to individual SMTP
5992 commands during the callout. The overall timeout applies when there is more
5993 than one host that can be tried. The timeout is checked before trying the
5994 next host. This prevents very long delays if there are a large number of
5995 hosts and all are timing out (e.g. when the network connections are timing
5996 out). The value of the overall timeout can be changed by specifying an
5997 additional sub-option for "callout", called "maxwait". For example:
5999 verify = sender/callout=5s,maxwait=20s
6001 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6002 to the end before writing, but this should make it even safer).
6004 22. Exim was forgetting that it had advertised PIPELINING for the second and
6005 subsequent messages on an SMTP connection. It was also not resetting its
6006 memory on STARTTLS and an internal HELO.
6008 23. When Exim logs an SMTP synchronization error within a session, it now
6009 records whether PIPELINING has been advertised or not.
6011 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6012 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6013 rather than long int.
6015 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6017 26. Added the never_mail option to autoreply.
6023 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6024 crash if the getsockname() call failed; this can happen if a connection is
6025 closed very soon after it is established. The problem was simply in the
6026 order in which certain operations were done, causing Exim to try to write
6027 to the SMTP stream before it had set up the file descriptor. The bug has
6028 been fixed by making things happen in the correct order.
6034 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6035 before the connection was closed, thus losing the rejection response.
6037 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6038 some early Solaris releases, but causes trouble in current releases where
6039 socklen_t is defined.
6041 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6044 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6047 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6048 that releasing the top of it at the end releases what was used for sub-
6049 expansions (unless the block got too big). However, discard this block if
6050 the first thing is a variable or header, so that we can use its block when
6051 it is dynamic (useful for very large $message_headers, for example).
6053 6. Lookups now cache *every* query, not just the most recent. A new, separate
6054 store pool is used for this. It can be recovered when all lookup caches are
6055 flushed. Lookups now release memory at the end of their result strings.
6056 This has involved some general refactoring of the lookup sources.
6058 7. Some code has been added to the store_xxx() functions to reduce the amount
6059 of flapping under certain conditions.
6061 8. log_incoming_interface used to affect only the <= reception log lines. Now
6062 it causes the local interface and port to be added to several more SMTP log
6063 lines, for example "SMTP connection from", and rejection lines.
6065 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6067 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6069 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6070 could be overwritten at the end of the current message (or the start of a
6071 new message if it was set in a HELO ACL). The value is now preserved for
6072 the duration of the SMTP connection.
6074 12. If a transport had a headers_rewrite setting, and a matching header line
6075 contained an unqualified address, that address was qualified, even if it
6076 did not match any rewriting rules. The underlying bug was that the values
6077 of the flags that permit the existence of unqualified sender and recipient
6078 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6079 non-local messages, and by -bnq for local messages) were not being
6080 preserved with the message after it was received.
6082 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6083 "next input=" as part of the text comprising the host identity instead of
6084 the correct text. The code was using the same buffer for two different
6085 strings. However, depending on which order the printing function evaluated
6086 its arguments, the bug did not always show up. Under Linux, for example, my
6087 test suite worked just fine.
6089 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6090 doesn't work with some older versions of Perl. It has been changed to "my",
6091 which in any case is probably the better facility to use.
6093 15. A really picky compiler found some instances of statements for creating
6094 error messages that either had too many or two few arguments for the format
6097 16. The size of the buffer for calls to the DNS resolver has been increased
6098 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6099 for addresses that have a lot of PTR records. This alleviates a problem; it
6100 does not fully solve it.
6102 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6103 buffer now truncates the list and logs the incident, which is the same
6104 action as happens when Exim is looking up a host name and its aliases.
6105 Previously in this situation something unpredictable would happen;
6106 sometimes it was "internal error: store_reset failed".
6108 18. If a server dropped the connection unexpectedly when an Exim client was
6109 using GnuTLS and trying to read a response, the client delivery process
6110 crashed while trying to generate an error log message.
6112 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6113 string, for example:
6115 warn message = H1: something\nH2: something
6117 the text was added as a single header line from Exim's point of view
6118 though it ended up OK in the delivered message. However, searching for the
6119 second and subsequent header lines using $h_h2: did not work. This has been
6120 fixed. Similarly, if a system filter added multiple headers in this way,
6121 the routers could not see them.
6123 20. Expanded the error message when iplsearch is called with an invalid key to
6124 suggest using net-iplsearch in a host list.
6126 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6127 ACLs are no longer actually imposed (and a message to that effect is
6130 22. If a "gecos" field in a passwd entry contained escaped characters, in
6131 particular, if it contained a \" sequence, Exim got it wrong when building
6132 a From: or a Sender: header from that name. A second bug also caused
6133 incorrect handling when an unquoted " was present following a character
6134 that needed quoting.
6136 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6137 was not being matched caselessly.
6139 24. Arranged for all hyphens in the exim.8 source to be escaped with
6142 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6143 properly. Recipient callout cache records were still being keyed to include
6144 the sender, even when use_sender was set false. This led to far more
6145 callouts that were necessary. The sender is no longer included in the key
6146 when use_sender is false.
6148 26. Added "control = submission" modifier to ACLs.
6150 27. Added the ${base62d: operator to decode base 62 numbers.
6152 28. dnsdb lookups can now access SRV records.
6154 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6155 the configuration file.
6157 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6158 (-v) mode. This makes the output for a verbose queue run more intelligible.
6160 31. Added a use_postmaster feature to recipient callouts.
6162 32. Added the $body_zerocount variable, containing the number of binary zero
6163 bytes in the message body.
6165 33. The time of last modification of the "new" subdirectory is now used as the
6166 "mailbox time last read" when there is a quota error for a maildir
6169 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6171 35. Added +ignore_unknown as a special item in host lists.
6173 36. Code for decoding IPv6 addresses in host lists is now included, even if
6174 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6175 address was recognized as an IP address, but was then not correctly decoded
6176 into binary, causing unexpected and incorrect effects when compared with
6183 1. Very minor rewording of debugging text in manualroute to say "list of
6184 hosts" instead of "hostlist".
6186 2. If verify=header_syntax was set, and a header line with an unqualified
6187 address (no domain) and a large number of spaces between the end of the
6188 name and the colon was received, the reception process suffered a buffer
6189 overflow, and (when I tested it) crashed. This was caused by some obsolete
6190 code that should have been removed. The fix is to remove it!
6192 3. When running in the test harness, delay a bit after writing a bounce
6193 message to get a bit more predictability in the log output.
6195 4. Added a call to search_tidyup() just before forking a reception process. In
6196 theory, someone could use a lookup in the expansion of smtp_accept_max_
6197 per_host which, without the tidyup, could leave open a database connection.
6199 5. Added the variables $recipient_data and $sender_data which get set from a
6200 lookup success in an ACL "recipients" or "senders" condition, or a router
6201 "senders" option, similar to $domain_data and $local_part_data.
6203 6. Moved the writing of debug_print from before to after the "senders" test
6206 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6207 problems for message scanning, either using a data ACL, or using
6208 local_scan() because the Received: header was not generated till after they
6209 were called (in order to set the time as the time of reception completion).
6210 I have revised the way this works. The header is now generated after the
6211 body is received, but before the ACL or local_scan() are called. After they
6212 are run, the timestamp in the header is updated.
6218 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6219 before starting a queue runner without re-exec. This happened only when
6220 deliver_drop_privilege was set or when the Exim user was set to root. The
6221 effect of the bug was that timeouts during subsequent deliveries caused
6222 crashes instead of being properly handled. The handler is now left at its
6223 default (and expected) setting.
6225 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6226 message, again when deliver_drop_privilege is set or Exim is run as root.
6227 The bug described in (1) was not present in this case, but the tidying up
6228 of the other signals was missing. I have made the two cases consistent.
6230 3. The ignore_target_hosts setting on a manualroute router was being ignored
6231 for hosts that were looked up using the /MX notation.
6233 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6236 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6237 operated on the sender address. After changing the $sender_address to <>
6238 for the sender address verify, Exim was re-instated it as the original
6239 (before rewriting) address, but remembering that it had rewritten it, so it
6240 wasn't rewriting it again. This bug also had the effect of breaking the
6241 sender address verification caching when the sender address was rewritten.
6243 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6244 This has been changed so that if the ip literal address matches
6245 ignore_target_hosts, the router declines.
6247 7. Added expansion conditions match_domain, match_address, and match_local_
6248 part (NOT match_host).
6250 8. The placeholder for the Received: header didn't have a length field set.
6252 9. Added code to Exim itself and to exim_lock to test for a specific race
6253 condition that could lead to file corruption when using MBX delivery. The
6254 issue is with the lockfile that is created in /tmp. If this file is removed
6255 after a process has opened it but before that process has acquired a lock,
6256 there is the potential for a second process to recreate the file and also
6257 acquire a lock. This could lead to two Exim processes writing to the file
6258 at the same time. The added code performs the same test as UW imapd; it
6259 checks after acquiring the lock that its file descriptor still refers to
6260 the same named file.
6262 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6263 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6266 11. Added the smtp_active_hostname option. If used, this will typically be made
6267 to depend on the incoming interface address. Because $interface_address is
6268 not set up until the daemon has forked a reception process, error responses
6269 that can happen earlier (such as "too many connections") no longer contain
6272 12. If an expansion in a condition on a "warn" statement fails because a lookup
6273 defers, the "warn" statement is abandoned, and the next ACL statement is
6274 processed. Previously this caused the whole ACL to be aborted.
6276 13. Added the iplsearch lookup type.
6278 14. Added ident_timeout as a log selector.
6280 15. Added tls_certificate_verified as a log selector.
6282 16. Added a global option tls_require_ciphers (compare the smtp transport
6283 option of the same name). This controls incoming TLS connections.
6285 17. I finally figured out how to make tls_require_ciphers do a similar thing
6286 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6287 before starting the TLS session.
6289 18. Tabs are now shown as \t in -bP output.
6291 19. If the log selector return_path_on_delivery was set, Exim crashed when
6292 bouncing a message because it had too many Received: header lines.
6294 20. If two routers both had headers_remove settings, and the first one included
6295 a superfluous trailing colon, the final name in the first list and the
6296 first name in the second list were incorrectly joined into one item (with a
6297 colon in the middle).
6303 1. Added -C and -D options to the exinext utility, mainly to make it easier
6304 to include in the automated testing, but these could be helpful when
6305 multiple configurations are in use.
6307 2. The exinext utility was not formatting the output nicely when there was
6308 an alternate port involved in the retry record key, nor when there was a
6309 message id as well (for retries that were specific to a specific message
6310 and a specific host). It was also confused by IPv6 addresses, because of
6311 the additional colons they contain. I have fixed the IPv4 problem, and
6312 patched it up to do a reasonable job for IPv6.
6314 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6315 delivery, the log line now contains "pipelined" if PIPELINING was used.
6317 4. An SMTP transport process used to panic and die if the bind() call to set
6318 an explicit outgoing interface failed. This has been changed; it is now
6319 treated in the same way as a connect() failure.
6321 5. A reference to $sender_host_name in the part of a conditional expansion
6322 that was being skipped was still causing a DNS lookup. This no longer
6325 6. The def: expansion condition was not recognizing references to header lines
6326 that used bh_ and bheader_.
6328 7. Added the _cache feature to named lists.
6330 8. The code for checking quota_filecount in the appendfile transport was
6331 allowing one more file than it should have been.
6333 9. For compatibility with Sendmail, the command line option
6341 and sets the incoming protocol and host name (for trusted callers). The
6342 host name and its colon can be omitted when only the protocol is to be set.
6343 Note the Exim already has two private options, -pd and -ps, that refer to
6344 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6345 "s", but I don't think that's a major issue.
6347 10. A number of refactoring changes to the code, none of which should affect
6350 (a) The number of logging options was getting close to filling up the
6351 32-bit word that was used as a bit map. I have split them into two classes:
6352 those that are passed in the argument to log_write(), and those that are
6353 only ever tested independently outside of that function. These are now in
6354 separate 32-bit words, so there is plenty of room for expansion again.
6355 There is no change in the user interface or the logging behaviour.
6357 (b) When building, for example, log lines, the code previously used a
6358 macro that called string_cat() twice, in order to add two strings. This is
6359 not really sufficiently general. Furthermore, there was one instance where
6360 it was actually wrong because one of the argument was used twice, and in
6361 one call a function was used. (As it happened, calling the function twice
6362 did not affect the overall behaviour.) The macro has been replaced by a
6363 function that can join an arbitrary number of extra strings onto a growing
6366 (c) The code for expansion conditions now uses a table and a binary chop
6367 instead of a serial search (which was left over from when there were very
6368 few conditions). Also, it now recognizes conditions like "pam" even when
6369 the relevant support is not compiled in: a suitably worded error message is
6370 given if an attempt is made to use such a condition.
6372 11. Added ${time_interval:xxxxx}.
6374 12. A bug was causing one of the ddress fields not to be passed back correctly
6375 from remote delivery subprocesses. The field in question was not being
6376 subsequently used, so this caused to problems in practice.
6378 13. Added new log selectors queue_time and deliver_time.
6380 14. Might have fixed a bug in maildirsizefile handling that threw up
6381 "unexpected character" debug warnings, and recalculated the data
6382 unnecessarily. In any case, I expanded the warning message to give more
6385 15. Added the message "Restricted characters in address" to the statements in
6386 the default ACL that block characters like @ and % in local parts.
6388 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6389 Three changes have been made:
6391 (a) There was a serious bug; a negative response to MAIL caused the whole
6392 recipient domain to be cached as invalid, thereby blocking all messages
6393 to all local parts at the same domain, from all senders. This bug has
6394 been fixed. The domain is no longer cached after a negative response to
6395 MAIL if the sender used is not empty.
6397 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6400 (c) A new callout option, "use_sender" has been added for people who want
6401 the modified behaviour.
6407 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6410 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6411 indeed breaks things for older releases.
6413 3. Added additional logging to the case where there is a problem reading data
6414 from a filter that is running in a subprocess using a pipe, in order to
6415 try to track down a specific problem.
6417 4. Testing facility fudge: when running in the test harness and attempting
6418 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6419 getting "No route to host". Convert this to a timeout.
6421 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6424 6. Some OS don't have socklen_t but use size_t instead. This affects the
6425 fifth argument of getsockopt() amongst other things. This is now
6426 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6427 can be set for individual OS. I have set it for SunOS5, OSF1, and
6428 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6429 some earlier ones do not.
6431 7. Change 4.30/15 was not doing the test caselessly.
6433 8. The standard form for an IPv6 address literal was being rejected by address
6434 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6435 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6436 this, as well as the form without the "IPv6" on the front (but only when
6437 address literals are enabled, of course).
6439 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6441 10. Exim crashed if a message with an empty sender address specified by -f
6442 encountered a router with an errors_to setting. This could be provoked only
6443 by a command such as
6447 where an empty string was supplied; "<>" did not hit this bug.
6449 11. Installed PCRE release 4.5.
6451 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6452 remained set. It is now erased.
6454 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6455 times from message ids (which are base 36 rather than the normal 62).
6457 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6458 were being counted as actual protocol errors, and logged if the log
6459 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6460 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6461 and DATA following a set of rejected RCPTs do not count as protocol errors.
6462 In other words, Exim assumes they were pipelined, though this may not
6463 actually be the case. Of course, in all cases the client gets an
6464 appropriate error code.
6466 15. If a lookup fails in an ACL condition, a message about the failure may
6467 be available; it is used if testing the ACL cannot continue, because most
6468 such messages specify what the cause of the deferral is. However, some
6469 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6470 that caused an old message to be retained and used if a later statement
6471 caused a defer, replacing the real cause of the deferral.
6473 16. If an IP address had so many PTR records that the DNS lookup buffer
6474 was not large enough to hold them, Exim could crash while trying to process
6475 the truncated data. It now detects and logs this case.
6477 17. Further to 4.21/58, another change has been made: if (and only if) the
6478 first line of a message (the first header line) ends with CRLF, a bare LF
6479 in a subsequent header line has a space inserted after it, so as not to
6480 terminate the header.
6482 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6483 unnecessarily, used atoi() instead of strtol(), and didn't check the
6484 termination when getting file sizes from file names by regex.
6486 19. Completely re-implemented the support for maildirsize files, in the light
6487 of a number of problems with the previous contributed implementation
6488 (4.30/29). In particular:
6490 . If the quota is zero, the maildirsize file is maintained, but no quota is
6493 . If the maildir directory does not exist, it is created before any attempt
6494 to write a maildirsize file.
6496 . The quota value in the file is just a cache; if the quota is changed in
6497 the transport, the new value overrides.
6499 . A regular expression is available for excluding directories from the
6502 20. The autoreply transport checks the characters in options that define the
6503 message's headers; it allows continued headers, but it was checking with
6504 isspace() after an embedded newline instead of explicitly looking for a
6507 21. If all the "regular" hosts to which an address was routed had passed their
6508 expiry times, and had not reached their retry times, the address was
6509 bounced, even if fallback hosts were defined. Now Exim should go on to try
6512 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6513 equivalent code in the SMTP transport. Some hosts send humungous responses
6514 to HELO/EHLO, more than 1024 it seems.
6516 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6517 gives compiler warnings in some environments. I've now done it "properly",
6520 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6521 (because of problems with the built-in one) was declared to return uschar *
6522 instead of char *, causing compiler failure.
6524 25. Fixed a file descriptor leak when processing alias/forward files.
6526 26. Fixed a minor format string issue in dbfn.c.
6528 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6530 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6531 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6534 29. When a sender address is verified, it is cached, to save repeating the test
6535 when there is more than one recipient in a message. However, when the
6536 verification involves a callout, it is possible for different callout
6537 options to be set for different recipients. It is too complicated to keep
6538 track of this in the cache, so now Exim always runs a verification when a
6539 callout is required, relying on the callout cache for the optimization.
6540 The overhead is duplication of the address routing, but this should not be
6543 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6544 to be verified with callout=postmaster, and the main callout worked but the
6545 postmaster check failed, the verification correctly failed. However, if a
6546 subsequent RCPT command asked for sender verification *without* the
6547 postmaster check, incorrect caching caused this verification also to fail,
6550 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6551 it was not caching the DNS options (qualify_single, search_parents) that
6552 were used when the lookup failed. A subsequent lookup with different
6553 options therefore always gave the same answer, though there were cases
6554 where it should not have. (Example: a "domains = !$mx_any" option on a
6555 dnslookup router: the "domains" option is always processed without any
6556 widening, but the router might have qualify_single set.) Now Exim uses the
6557 cached value only when the same options are set.
6559 32. Added John Jetmore's "exipick" utility to the distribution.
6561 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6562 than a timeout (e.g. a certificate is required, and is not provided), an
6563 Exim server now closes the connection immediately. Previously it waited for
6564 the client to close - but if the client is SSL, it seems that they each
6565 wait for each other, leading to a delay before one of them times out.
6567 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6568 maintained 0.8.x compatibility because I don't think many are using it, and
6569 it is clearly obsolete.
6571 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6574 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6575 client certificate was expired. A simple patch fixes this, though I don't
6576 understand the full logic of why the verify callback is called multiple
6579 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6580 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6581 which causes problems with some clients (such as the Certicom SSL Plus
6582 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6583 disables the coutermeasure allowing Eudora to connect."
6585 38. Exim was not checking that a write() to a log file succeeded. This could
6586 lead to Bad Things if a log got too big, in particular if it hit a file
6587 size limit. Exim now panics and dies if it cannot write to a log file, just
6588 as it does if it cannot open a log file.
6590 39. Modified OS/Makefile-Linux so that it now contains
6592 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6594 The two -D definitions ensure that Exim is compiled with large file
6595 support, which makes it possible to handle log files that are bigger than
6598 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6599 instance) a domain was checked against a named list that involved a lookup,
6600 causing $domain_data to be set, then another domain was checked against the
6601 same list, then the first domain was re-checked, the value of $domain_data
6602 after the final check could be wrong. In particular, if the second check
6603 failed, it could be set empty. This bug probably also applied to
6606 41. The strip_trailing_dot option was not being applied to the address given
6607 with the -f command-line option.
6609 42. The code for reading a message's header from the spool was incrementing
6610 $received_count, but never initializing it. This meant that the value was
6611 incorrect (doubled) while delivering a message in the same process in which
6612 it was received. In the most common configuration of Exim, this never
6613 happens - a fresh exec is done - but it can happen when
6614 deliver_drop_privilege is set.
6616 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6617 it now includes up to 150 characters of the unexpected data in the log
6620 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6621 and building data strings. The size of both of these buffers was 10 000
6622 bytes - far larger than anybody would *ever* want, thought I. Needless to
6623 say, somebody hit the limit. I have increased the maximum line length to
6624 20 000 and the maximum data length of concatenated lines to 100 000. I have
6625 also fixed two bugs, because there was no checking on these buffers. Tsk,
6626 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6627 buffer is too small.
6629 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6630 lsearch lookups. Now it does.
6632 46. When parsing a route_list item in a manualroute router, a fixed-length
6633 buffer was used for the list of hosts. I made this 1024 bytes long,
6634 thinking that nobody would ever have a list of hosts that long. Wrong.
6635 Somebody had a whole pile of complicated expansion conditions, and the
6636 string was silently truncated, leading to an expansion error. It turns out
6637 that it is easier to change to an unlimited length (owing to other changes
6638 that have happened since this code was originally written) than to build
6639 structure for giving a limitation error. The length of the item that
6640 expands into the list of hosts is now unlimited.
6642 47. The lsearch lookup could not handle data where the length of text line was
6643 more than 4095 characters. Such lines were truncated, leading to shortened
6644 data being returned. It should now handle lines of any length.
6646 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6647 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6650 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6651 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6652 no longer happens in comment lines. A list of replacements is now placed
6653 at the head of all of the source files, except those whose only change is
6654 to replace PERL_COMMAND in the very first #! line.
6656 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6657 messages on the queue, with a bottom-up merge sort, using code contributed
6658 by Michael Haardt. This should make operations like -bp somewhat faster on
6659 large queues. It won't affect queue runners, except when queue_run_in_order
6662 51. Installed eximstats 1.31 in the distribution.
6664 52. Added support for SRV lookups to the dnslookup router.
6666 53. If an ACL referred to $message_body or $message_body_end, the value was not
6667 reset for any messages that followed in the same SMTP session.
6669 54. The store-handling optimization for building very long strings was not
6670 differentiating between the different store pools. I don't think this
6671 actually made any difference in practice, but I've tidied it.
6673 55. While running the routers to verify a sender address, $sender_address
6674 was still set to the sender address. This is wrong, because when routing to
6675 send a bounce to the sender, it would be empty. Therefore, I have changed
6676 it so that, while verifying a sender address, $sender_address is set to <>.
6677 (There is no change to what happens when verifying a recipient address.)
6679 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6680 target A or AAAA records (if not already returned) without resetting the
6681 qualify_single or search_parents options of the DNS resolver. These are
6682 inappropriate in this case because the targets of MX and SRV records must
6683 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6684 target that, when qualified, matched something in the local domain. These
6685 two options are now turned off when doing these lookups.
6687 57. It seems that at least some releases of Reiserfs (which does not have the
6688 concept of a fixed number of inodes) returns zero and not -1 for the
6689 number of available inodes. This interacted badly with check_spool_inodes,
6690 which assumed that -1 was the "no such thing" setting. What I have done is
6691 to check that the total number of inodes is greater than zero before doing
6692 the test of how many are available.
6694 58. When a "warn" ACL statement has a log_message modifier, the message is
6695 remembered, and not repeated. This is to avoid a lot of repetition when a
6696 message has many recipients that cause the same warning to be written.
6697 However, Exim was preserving the list of already written lines for an
6698 entire SMTP session, which doesn't seem right. The memory is now reset if a
6699 new message is started.
6701 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6702 debugging output unless log_rewrite was also set.
6704 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6705 of (int)(handle) when we know that handle contains (void *)(-1).
6707 61. The Exim daemon panic-logs an error return when it closes the incoming
6708 connection. However "connection reset by peer" seems to be common, and
6709 isn't really an error worthy of noting specially, so that particular error
6712 62. When Exim is trying to find all the local interfaces, it used to panic and
6713 die if the ioctl to get the interface flags failed. However, it seems that
6714 on at least one OS (Solaris 9) it is possible to have an interface that is
6715 included in the list of interfaces, but for which you get a failure error
6716 for this call. This happens when the interface is not "plumbed" into a
6717 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6718 failure of the "get flags" call assumes that the interface is down.
6720 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6721 makes life easier for people who are doing arithmetic on fields extracted
6722 from dates, where you often get leading zeros that should not be
6723 interpreted as octal.
6725 64. Added qualify_domain to the redirect router, to override the global
6728 65. If a pathologically long header line contained very many addresses (the
6729 report of this problem mentioned 10 000) and each of them was rewritten,
6730 Exim could use up a very large amount of memory. (It kept on making new
6731 copies of the header line as it rewrote, and never released the old ones.)
6732 At the expense of a bit more processing, the header rewriting function has
6733 been changed so that it no longer eats memory in this way.
6735 66. The generation of the Received: header has been moved from the time that a
6736 message starts to be received, to the time that it finishes. The timestamp
6737 in the Received: header should now be very close to that of the <= log
6738 line. There are two side-effects of this change:
6740 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6741 logged header lines no longer include the local Received: line, because
6742 it has not yet been created. The same applies to a copy of the message
6743 that is returned to a non-SMTP sender when a message is rejected.
6745 (b) When a filter file is tested using -bf, no additional Received: header
6746 is added to the test message. After some thought, I decided that this
6749 This change does not affect the value of $received_for. It is still set
6750 after address rewriting, but before local_scan() is called.
6752 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6754 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6755 gave an unhelpful panic error message, and a defer error. I have managed to
6756 change this behaviour so that it now rejects any supplied certificate,
6757 which seems right, as the list of acceptable certificates is empty.
6759 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6760 gave an unhelpful defer error. I have not managed to make this reject any
6761 supplied certificates, but the error message it gives is "no certificate
6762 supplied", which is not helpful.
6764 70. exigrep's output now also includes lines that are not associated with any
6765 message, but which match the given pattern. Implemented by a patch from
6766 Martin Sluka, which also tidied up the Perl a bit.
6768 71. Recipient callout verification, like sender verification, was using <> in
6769 the MAIL FROM command. This isn't really the right thing, since the actual
6770 sender may affect whether the remote host accepts the recipient or not. I
6771 have changed it to use the actual sender in the callout; this means that
6772 the cache record is now keyed on a recipient/sender pair, not just the
6773 recipient address. There doesn't seem to be a real danger of callout loops,
6774 since a callout by the remote host to check the sender would use <>.
6775 [SEE ABOVE: changed after hitting problems.]
6777 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6778 temporary errors. However, in the case of such a code being given after
6779 the end of a data transmission (i.e. after ".") Exim was failing to write
6780 a retry record for the message. (Yes, there was some broken host that was
6781 actually sending 8xx at this point.)
6783 73. An unknown lookup type in a host list could cause Exim to panic-die when
6784 the list was checked. (An example that provoked this was putting <; in the
6785 middle of a list instead of at the start.) If this happened during a DATA
6786 ACL check, a -D file could be left lying around. This kind of configuration
6787 error no longer causes Exim to die; instead it causes a defer error. The
6788 incident is still logged to the main and panic logs.
6790 74. Buglet left over from Exim 3 conversion. The message "too many messages
6791 in one connection" was written to the rejectlog but not the mainlog, except
6792 when address rewriting (yes!) was being logged.
6794 75. Added write_rejectlog option.
6796 76. When a system filter was run not as root (that is, when system_filter_user
6797 was set), the values of the $n variables were not being returned to the
6798 main process; thus, they were not subsequently available in the $sn
6801 77. Added +return_path_on_delivery log selector.
6803 78. A connection timeout was being treated differently from recipients deferred
6804 when testing hosts_max_try with a message that was older than the host's
6805 retry timeout. (The host should not be counted, thus allowing all hosts to
6806 be tried at least once before bouncing.) This may have been the cause of an
6807 occasionally reported bug whereby a message would remain on the queue
6808 longer than the retry timeout, but would be bounced if a delivery was
6809 forced. I say "may" because I never totally pinned down the problem;
6810 setting up timeout/retry tests is difficult. See also the next item.
6812 79. The ultimate address timeout was not being applied to errors that involved
6813 a combination of host plus message (for example, a timeout on a MAIL
6814 command). When an address resolved to a number of possible hosts, and they
6815 were not all tried for each delivery (e.g. because of hosts_max_try), a
6816 message could remain on the queue longer than the retry timeout.
6818 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6821 81. Fixed an obscure SMTP outgoing bug which required at least the following
6822 conditions: (a) there was another message waiting for the same server;
6823 (b) the server returned 5xx to all RCPT commands in the first message so
6824 that the message was not completed; (c) the server dropped the connection
6825 or gave a negative response to the RSET that Exim sends to abort the
6826 transaction. The observed case was a dropped connection after DATA that had
6827 been sent in pipelining mode. That is, the server had advertised PIPELINING
6828 but was not implementing it correctly. The effect of the bug was incorrect
6829 behaviour, such as trying another host, and this could lead to a crash.
6835 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6836 and daemon.c were passed as pointers to ints; they should have been
6837 pointers to socklen_t variables (which are typically unsigned ints).
6839 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6842 3. Fixed a really odd bug that affected only the testing scheme; patching a
6843 certain fixed string in the binary changed the value of another string that
6844 happened to be identical to the end of the original first string.
6846 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6847 name", it returns that address as the IP address. On some operating
6848 systems (e.g. Solaris), it also passes back the IP address string as the
6849 "host name". However, on others (e.g. Linux), it passes back an empty
6850 string. Exim wasn't checking for this, and was changing the host name to an
6851 empty string, assuming it had been canonicalized.
6853 5. Although rare, it is permitted to have more than one PTR record for a given
6854 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6855 all the names associated with an address, because they do in Solaris.
6856 However, it seems that they do not in Linux for data that comes from the
6857 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6858 I found this out when I moved to a new Linux workstation and tried to run
6859 the Exim test suite.
6861 To get round this problem I have changed the code so that it now does its
6862 own call to the DNS to look up PTR records when searching for a host name.
6863 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6864 addresses that are only in /etc/hosts are still found.
6866 This behaviour is, however, controlled by an option called host_lookup_
6867 order, which defaults to "bydns:byaddr". If people want to use the other
6868 order, or indeed, just use one or the other means of lookup, they can
6869 specify it in this variable.
6871 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6872 some operating systems, this comes back from gethostbyaddr() as an empty
6873 string, and this is what Exim used to test for. However, it seems that in
6874 other systems, "." is yielded. Exim now tests for this case too.
6876 7. The values of check_spool_space and check_log_space are now held internally
6877 as a number of kilobytes instead of an absolute number of bytes. If a
6878 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6879 kilobyte. This means that much larger values can be stored.
6881 8. Exim monitor: an attempt to get the action menu when not actually pointing
6882 at a message produces an empty menu entitled "No message selected". This
6883 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6884 no entries in it ("Shell widget menu has zero width and/or height"). So I
6885 have added a single, blank menu entry in this case.
6887 9. Added ${quote_local_part.
6889 10. MIME decoding is now applied to the contents of Subject: header lines when
6892 11. Now that a reference to $sender_host_address automatically causes a reverse
6893 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6894 host lookup before query-style lookups in lists that might use this
6895 variable. This has therefore been abolished, and the "net-" prefix is no
6896 longer necessary for query-style lookups.
6898 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6899 have been a typo for LFLAGS, so it has been changed.
6901 13. The install script calls Exim with "-C /dev/null" in order to find the
6902 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6903 to be output. However, since Exim outputs its version number before the
6904 error, it didn't break the script. It just looked ugly. I fixed this by
6905 always allowing "-C /dev/null" if the caller is root.
6907 14. Ignore overlarge ACL variable number when reading spool file - insurance
6908 against a later release with more variables having written the file.
6910 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6911 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6912 this, as well as the form without the "IPv6" on the front.
6914 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6915 OS/Makefile-Darwin file.
6917 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6918 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6920 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6921 variable. However, some OS use ints for inodes. I've added cast to long int
6922 to get rid of the compiler warning.
6924 19. I had forgotten to lock out "/../" in configuration file names when
6925 ALT_CONFIG_PREFIX was set.
6927 20. Routers used for verification do not need to specify transports. However,
6928 if such a router generated a host list, and callout was configured, Exim
6929 crashed, because it could not find a port number from the (non-existent)
6930 transport. It now assumes port 25 in this circumstance.
6932 21. Added the -t option to exigrep.
6934 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6935 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6936 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6937 them is set, LOOKUP_LSEARCH is forced.
6939 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6941 24. Added sender and host information to the "rejected by local_scan()" log
6942 line; previously there was no indication of these.
6944 25. Added .include_if_exists.
6946 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6947 Linux. It turns out that not all file systems support this. Apparently some
6948 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6949 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6950 not supported on the file descriptor, is now ignored when Exim is trying to
6951 sync a directory. This applies only to Linux.
6953 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6955 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6956 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6959 29. Added support for maildirsize files from supplied patch (modified a bit).
6961 30. The use of :fail: followed by an empty string could lead Exim to respond to
6962 sender verification failures with (e.g.):
6964 550 Verification failed for <xxx>
6965 550 Sender verify failed
6967 where the first response line was missing the '-' that indicates it is not
6968 the final line of the response.
6970 31. The loop for finding the name of the user that called Exim had a hardwired
6971 limit of 10; it now uses the value of finduser_retries, which is used for
6972 all other user lookups.
6974 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6977 33. Exim was neglecting to zero errno before one call of strtol() when
6978 expanding a string and expecting an integer value. On some systems this
6979 resulted in spurious "integer overflow" errors. Also, it was casting the
6980 result into an int without checking.
6982 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6983 did not work. The code looks as if it has *never* worked, though it appears
6984 to have been documented since at least release 1.62. I have made it work.
6986 35. The "timeout_DNS" error in retry rules, also documented since at least
6987 1.62, also never worked. As it isn't clear exactly what this means, and
6988 clearly it isn't a major issue, I have abolished the feature by treating it
6989 as "timeout", and writing a warning to the main and panic logs.
6991 36. The display of retry rules for -brt wasn't always showing the error code
6994 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6995 timeout_connect_A, timeout_connect_MX.
6997 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6998 to the empty sender.
7000 39. The daemon was not analysing the content of -oX till after it had closed
7001 stderr and disconnected from the controlling terminal. This meant that any
7002 syntax errors were only noted on the panic log, and the return code from
7003 the command was 0. By re-arranging the code a little, I've made the
7004 decoding happen first, so such errors now appear on stderr, and the return
7005 code is 1. However, the actual setting up of the sockets still happens in
7006 the disconnected process, so errors there are still only recorded on the
7009 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7010 connections (as happens on some IP stacks) was logged at start up time as
7011 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7012 it from "IPv6 and IPv4", which means that two separate sockets are being
7015 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7016 says whether AF_INET or AF_INET6 was passed as an argument.
7018 42. Exiwhat output was messed up when time zones were included in log
7021 43. Exiwhat now gives more information about the daemon's listening ports,
7022 and whether -tls-on-connect was used.
7024 44. The "port" option of the smtp transport is now expanded.
7026 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7027 silently ignored. Now an error message is written to the main and panic
7030 46. There's a new ACL modifier called "logwrite" which writes to a log file
7031 as soon as it is encountered.
7033 47. Added $local_user_uid and $local_user_gid at routing time.
7035 48. Exim crashed when trying to verify a sender address that was being
7038 49. Exim was recognizing only a space character after ".include". It now also
7039 recognizes a tab character.
7041 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7042 extracting the relevant information from the specification. The man page no
7043 longer contains scrambled data for the -d option, and I've added a section
7044 at the front about calling Exim under different names.
7046 51. Added "extra_headers" argument to the "mail" command in filter files.
7048 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7051 53. Installed eximstats 1.29.
7053 54. Added transport_filter_timeout as a generic transport option.
7055 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7056 Cc: header lines. This was required by RFC 822, but it not required by RFC
7059 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7060 incoming messages that did not have them. Now it does so only if the
7061 message originates locally, that is, if there is no associated remote host
7062 address. When Resent- header lines are present, this applies to the Resent-
7063 lines rather than the non-Resent- lines.
7065 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7066 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7068 58. Messages for configuration errors now include the name of the main
7069 configuration file - useful now that there may be more than one file in a
7070 list (.included file names were always shown).
7072 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7073 for those rare installations that do not start the daemon as root or run it
7074 setuid root. I've cut out the call to initgroups() if the daemon is not
7077 60. The Exim user and group can now be bound into the binary as text strings
7078 that are looked up at the start of Exim's processing.
7080 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7082 62. Added $mailstore_basename variable.
7084 63. Installed patch to sieve.c from Michael Haardt.
7086 64. When Exim failed to open the panic log after failing to open the main log,
7087 the original message it was trying to log was written to stderr and debug
7088 output, but if they were not available (the usual case in production), it
7089 was lost. Now it is written to syslog before the two lines that record the
7090 failures to open the logs.
7092 65. Users' Exim filters run in subprocesses under the user's uid. It is
7093 possible for a "deliver" command or an alias in a "personal" command to
7094 provoke an address rewrite. If logging of address rewriting is configured,
7095 this fails because the process is not running as root or exim. There may be
7096 a better way of dealing with this, but for the moment (because 4.30 needs
7097 to be released), I have disabled address rewrite logging when running a
7098 filter in a non-root, non-exim process.
7104 1. The buildconfig auxiliary program wasn't quoting the value set for
7105 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7106 not defined. This bug was masked in 4.22 by the effect that was fixed in
7109 2. Some messages that were rejected after a message id was allocated were
7110 shown as "incomplete" by exigrep. It no longer does this for messages that
7111 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7113 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7114 have allow_domain_literals set, the ID did not get logged in the <= line.
7115 Domain literals are now always recognized in Message-ID: header lines.
7117 4. The first argument for a ${extract expansion item is the key name or field
7118 number. Leading and trailing spaces in this item were not being ignored,
7119 causing some misleading effects.
7121 5. When deliver_drop_privilege was set, single queue runner processes started
7122 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7123 same command in the process it spins off) were not dropping privilege.
7125 6. When the daemon running as "exim" started a queue runner, it always
7126 re-executed Exim in the spun-off process. This is a waste of effort when
7127 deliver_drop_privilege is set. The new process now just calls the
7128 queue-runner function directly.
7134 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7137 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7138 ignored. Though the use of -oP was forcing the writing of a pid file, it
7139 was always written to the default place.
7141 3. If the message "no IP address found for host xxxx" is generated during
7142 incoming verification, it is now followed by identification of the incoming
7143 connection (so you can more easily find what provoked it).
7145 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7147 5. Added some features to "harden" Exim a bit more against certain attacks:
7149 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7150 be put in Local/Makefile. This is like the never_users runtime option,
7151 but it cannot be overridden. The default setting is "root".
7153 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7154 prefix string with which any file named in a -C command line option
7157 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7158 is retained for -C and -D only if the caller of Exim is root. Without
7159 it, the exim user may also use -C and -D and retain privilege.
7161 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7162 command line option is disabled.
7164 6. Macro names set by the -D option must start with an upper case letter, just
7165 like macro names defined in the configuration file.
7167 7. Added "dereference=" facility to LDAP.
7169 8. Two instances of the typo "uknown" in the source files are fixed.
7171 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7172 the Configure-Makefile script screwed up while processing it.
7174 10. Incorporated PCRE 4.4.
7176 11. The SMTP synchronization check was not operating right at the start of an
7177 SMTP session. For example, it could not catch a HELO sent before the client
7178 waited for the greeting. There is now a check for outstanding input at the
7179 point when the greeting is written. Because of the duplex, asynchronous
7180 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7181 way, but not yet received, when the check is performed.
7183 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7184 on TCP/IP sockets, because this apparently causes some broken clients to
7187 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7188 unchanged) from the Cygwin maintainer.
7190 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7191 was supported instead of testing for mbx. Effectively a typo.
7193 15. The spa authenticator server code was not checking that the input it
7194 received was valid base64.
7196 16. The debug output line for the "set" modifier in ACLs was not showing the
7197 name of the variable that was being set.
7199 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7201 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7202 lookup on its own. Something else was needed to trigger the lookup. For
7203 example, a match in host_lookup or the need for a host name in a host list.
7204 Now, if $sender_host_name is referenced and the host name has not yet been
7205 looked up, a lookup is performed. If the lookup fails, the variable remains
7206 empty, and $host_lookup_failed is set to "1".
7208 19. Added "eqi" as a case-independent comparison operator.
7210 20. The saslauthd authentication condition could segfault if neither service
7211 nor realm was specified.
7213 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7214 error message that was logged was misleading, and incoming SMTP
7215 connections were dropped. The message is now more accurate, and temporary
7216 errors are given to SMTP connections.
7218 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7219 (except RSET) with a 421 error, until QUIT is received. However, it was
7220 failing to send a response to QUIT.
7222 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7223 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7224 ACL is now run after the helo_try_verify_hosts code.
7226 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7227 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7228 case-independent, so other case variants are also recognized). Apparently
7229 some systems use these upper case variants.
7231 25. If more than two messages were waiting for the same host, and a transport
7232 filter was specified for the transport, Exim sent two messages over the
7233 same TCP/IP connection, and then failed with "socket operation on non-
7234 socket" when it tried to send the third.
7236 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7238 27. The extern definition of crypt16() in expand.c was not being excluded when
7239 the OS had its own crypt16() function.
7241 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7242 as a preferred synonym for return_size_limit, both as an option and as an
7245 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7247 30. Changed the default configuration ACL to relax the local part checking rule
7248 for addresses that are not in any local domains. For these addresses,
7249 slashes and pipe symbols are allowed within local parts, but the sequence
7250 /../ is explicitly forbidden.
7252 31. SPA server authentication was not clearing the challenge buffer before
7255 32. log_message in a "warn" ACL statement was writing to the reject log as
7256 well as to the main log, which contradicts the documentation and doesn't
7257 seem right (because no rejection is happening). So I have stopped it.
7259 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7260 However, I am unable to do any testing of this.
7262 34. Fixed an infelicity in the appendfile transport. When checking directories
7263 for a mailbox, to see if any needed to be created, it was accidentally
7264 using path names with one or more superfluous leading slashes; tracing
7265 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7267 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7268 contents are added to the log line that is written for every discarded
7269 recipient. (Previously a log_message setting was ignored.)
7271 36. The ${quote: operator now quotes the string if it is empty.
7273 37. The install script runs exim in order to find its version number. If for
7274 some reason other than non-existence or emptiness, which it checks, it
7275 could not run './exim', it was installing it with an empty version number,
7276 i.e. as "exim-". This error state is now caught, and the installation is
7279 38. An argument was missing from the function that creates an error message
7280 when Exim fails to connect to the socket for saslauthd authentication.
7281 This could cause Exim to crash, or give a corrupted message.
7283 39. Added isip, isip4, and isip6 to ${if conditions.
7285 40. The ACL variables $acl_xx are now saved with the message, and can be
7286 accessed later in routers, transports, and filters.
7288 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7289 strings in the file are not string-expanded.
7291 42. If a MAIL command specified a SIZE value that was too large to fit into an
7292 int variable, the check against message_size_limit failed. Such values are
7293 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7294 day this will have to be increased, but I don't think I want to be around
7295 when emails are that large.
7302 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7303 iconv() is not standard in FreeBSD.
7305 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7306 IPv6 enabled. The observed symptom was a segmentation fault on return from
7307 the function os_common_find_running_interfaces() in src/os.c.
7309 3. In the check_special_case() function in daemon.c I had used "errno" as an
7310 argument name, which causes warnings on some systems. This was basically a
7311 typo, since it was named "eno" in the comments!
7313 4. The code that waits for the clock to tick (at a resolution of some fraction
7314 of a second) so as to ensure message-id uniqueness was always waiting for
7315 at least one whole tick, when it could have waited for less. [This is
7316 almost certainly not relevant at current processor speeds, where it is
7317 unlikely to ever wait at all. But we try to future-proof.]
7319 5. The function that sleeps for a time interval that includes fractions of a
7320 second contained a race. It did not block SIGALRM between setting the
7321 timer, and suspending (a couple of lines later). If the interval was short
7322 and the sigsuspend() was delayed until after it had expired, the suspension
7323 never ended. On busy systems this could lead to processes getting stuck for
7326 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7327 process, before it forks any delivery processes. The open lookup caching
7328 mechanism meant that the open file or database connection was passed into
7329 the delivery process. The problem was that delivery processes always tidy
7330 up cached lookup data. This could cause a problem for the next delivery
7331 process started by the queue runner, because the external queue runner
7332 process does not know about the closure. So the next delivery process
7333 still has data in the lookup cache. In the case of a file lookup, there was
7334 no problem because closing a file descriptor in a subprocess doesn't affect
7335 the parent. However, if the lookup was caching a connection to a database,
7336 the connection was closed, and the second delivery process was likely to
7337 see errors such as "PGSQL: query failed: server closed the connection
7338 unexpectedly". The problem has been fixed by closing all cached lookups
7339 in a queue runner before running a delivery process.
7341 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7342 seem to have the "const" qualifier which it has on other OS. I've
7345 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7346 *of the same type* (client or server) with the same public name that an
7347 error should be diagnosed.
7349 9. When Exim looked up a host name for an IP address, but failed to find the
7350 original IP address when looking up the host name (a safety check), it
7351 output the message "<ip address> does not match any IP for NULL", which was
7352 confusing, to say the least. The bug was that the host name should have
7353 appeared instead of "NULL".
7355 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7356 user that is built into the binary, and the -C or -D options is used, root
7357 privilege is dropped before the configuration file is read. In addition,
7358 logging is switched to stderr instead of the normal log files. If the
7359 configuration then re-defines the Exim user, the unprivileged environment
7360 is probably not what is expected, so Exim logs a panic warning message (but
7363 However, if deliver_drop_privilege is set, the unprivileged state may well
7364 be exactly what is intended, so the warning has been cut out in that case,
7365 and Exim is allowed to try to write to its normal log files.
7371 1. smtp_return_error_details was not giving details for temporary sender
7372 or receiver verification errors.
7374 2. Diagnose a configuration error if two authenticators have the same public
7377 3. Exim used not to create the message log file for a message until the first
7378 delivery attempt. This could be confusing when incoming messages were held
7379 for policy or load reasons. The message log file is now created at the time
7380 the message is received, and an initial "Received" line is written to it.
7382 4. The automatically generated man page for command line options had a minor
7383 bug that caused no ill effects; however, a more serious problem was that
7384 the procedure for building the man page automatically didn't always
7385 operate. Consequently, release 4.20 contains an out-of-date version. This
7386 shouldn't happen again.
7388 5. When building Exim with embedded Perl support, the script that builds the
7389 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7390 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7392 6. The freeze_tell option was not being used for messages that were frozen on
7393 arrival, either by an ACL or by local_scan().
7395 7. Added the smtp_incomplete_transaction log selector.
7397 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7398 was accepting AUTH without a new EHLO.
7400 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7401 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7404 10. Logging of TCP/IP connections (when configured) now happens in the main
7405 daemon process instead of the child process, so that the TCP/IP connection
7406 count is more accurate (but it can never be perfect).
7408 11. The use of "drop" in a nested ACL was not being handled correctly in the
7409 outer ACL. Now, if condition failure induced by the nested "drop" causes
7410 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7411 or "require"), the connection is dropped.
7413 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7414 that yield "discard" can now be used with an "accept" or a "discard" verb,
7415 but an error is generated for any others (because I can't see a useful way
7416 to define what should happen).
7418 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7419 are now processed in the same way as lines in the Exim configuration file.
7420 In particular, continuation lines are supported.
7422 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7424 15. Added -ti meaning -t -i.
7426 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7427 domains, and warn by logging if others are found.
7429 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7430 returned from GIFCONF when Exim is trying to find the list of interfaces on
7431 a host. The code in os.c has been modified to copy each ifreq to an aligned
7432 structure in all cases.
7434 Also, in some cases, the returned ifreq's were being copied to a 'struct
7435 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7436 means the last couple of bytes of an IPv6 address could be chopped if the
7437 ifreq contained only a normal sockaddr (14 bytes storage).
7439 18. Named domain lists were not supported in the hosts_treat_as_local option.
7440 An entry such as +xxxx was not recognized, and was treated as a literal
7443 19. Ensure that header lines added by a DATA ACL are included in the reject log
7444 if the ACL subsequently rejects the message.
7446 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7447 MD5 (which is deprecated).
7449 21. When testing a filter file using -bf, Exim was writing a message when it
7450 took the sender from a "From " line in the message, but it was not doing so
7451 when it took $return_path from a Return-Path: header line. It now does.
7453 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7454 with a valid header line field name (a series of printing characters
7455 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7457 23. Changed "disc" in the source to "disk" to conform to the documentation and
7458 the book and for uniformity.
7460 24. Ignore Sendmail's -Ooption=value command line item.
7462 25. When execve() failed while trying to run a command in a pipe transport,
7463 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7464 could be confused with a return value of 69 from the command itself. This
7465 has been changed to 127, the value the shell returns if it is asked to run
7466 a non-existent command. The wording for the related log line suggests a
7467 non-existent command as the problem.
7469 26. If received_header_text expands to an empty string, do not add a Received:
7470 header line to the message. (Well, it adds a token one on the spool, but
7471 marks it "old" so that it doesn't get used or transmitted.)
7473 27. Installed eximstats 1.28 (addition of -nt option).
7475 28. There was no check for failure on the call to getsockname() in the daemon
7476 code. This can fail if there is a shortage of resources on the system, with
7477 ENOMEM, for example. A temporary error is now given on failure.
7479 29. Contrary to the C standard, it seems that in some environments, the
7480 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7481 program. Exim now does this explicitly; it affects the formatting of
7482 timestamps using strftime().
7484 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7485 complaints. I've now initialized all the variables, to avoid this.
7487 32. Header lines added by a system filter were not being "seen" during
7488 transport-time rewrites.
7490 33. The info_callback() function passed to OpenSSL is set up with type void
7491 (*)(SSL *, int, int), as described somewhere. However, when calling the
7492 function (actually a macro) that sets it up, the type void(*)() is
7493 expected. I've put in a cast to prevent warnings from picky compilers.
7495 34. If a DNS black list lookup found a CNAME record, but there were no A
7496 records associated with the domain it pointed at, Exim crashed.
7498 35. If a DNS black list lookup returned more than one A record, Exim ignored
7499 all but the first. It now scans all returned addresses if a particular IP
7500 value is being sought. In this situation, the contents of the
7501 $dnslist_value variable are a list of all the addresses, separated by a
7504 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7505 to accept a host name and all its aliases if the forward lookup for any of
7506 them yielded the IP address of the incoming connection. Now it accepts only
7507 those names whose forward lookup yields the correct IP address. Any other
7508 names are discarded. This closes a loophole whereby a rogue DNS
7509 administrator could create reverse DNS records to break through a
7510 wildcarded host restriction in an ACL.
7512 37. If a user filter or a system filter that ran in a subprocess used any of
7513 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7514 the wrong values were passed to the pipe command ($thisaddress had the
7515 value of $0, $0 had the value of $1, etc). This bug was introduced by
7516 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7518 38. Improved the line breaking for long SMTP error messages from ACLs.
7519 Previously, if there was no break point between 40 and 75 characters, Exim
7520 left the rest of the message alone. Two changes have been made: (a) I've
7521 reduced the minimum length to 35 characters; (b) if it can't find a break
7522 point between 35 and 75 characters, it looks ahead and uses the first one
7523 that it finds. This may give the occasional overlong line, but at least the
7524 remaining text gets split now.
7526 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7527 file descriptors might be low, and that setting 1000 would always raise it.
7528 It turns out that in some environments, the limit is already over 1000 and
7529 that lowering it causes trouble. So now Exim takes care not to decrease it.
7531 40. When delivering a message, the value of $return_path is set to $sender_
7532 address at the start of routing (routers may change the value). By an
7533 oversight, this default was not being set up when an address was tested by
7534 -bt or -bv, which affected the outcome if any router or filter referred to
7537 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7538 log only when "message" or "log_message" are set. However, if one of the
7539 conditions was an address verification, or a call to a nested ACL, the
7540 messages generated by the underlying test were being passed through. This
7541 no longer happens. The underlying message is available in $acl_verify_
7542 message for both "message" and "log_message" expansions, so it can be
7543 passed through if needed.
7545 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7546 new expansion $bh_ to give the encoded byte string without charset
7547 translation. Translation happens only if iconv() is available; HAVE_ICONV
7548 indicates this at build time. HEADERS_CHARSET gives the charset to
7549 translate to; headers_charset can change it in the configuration, and
7550 "headers charset" can change it in an individual filter file.
7552 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7553 that creates RFC 2047 encoded "words" labels them as that charset instead
7554 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7555 expansion operator; (ii) when Exim creates a From: line for a local
7556 message; (iii) when a header line is rewritten to include a "phrase" part.
7558 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7559 buggy, causing it to skip the first lines of messages whose message ID
7560 ended in 'D'. This would not have bitten before Exim release 4.14, because
7561 message IDs were unlikely to end in 'D' before then. The effect was to have
7562 incorrect size information for certain domains.
7564 45. #include "config.h" was missing at the start of the crypt16.c module. This
7565 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7568 46. If there was a timeout during a "random" callout check, Exim treated it as
7569 a failure of the random address, and carried on sending RSET and the real
7570 address. If the delay was just some slowness somewhere, the response to the
7571 original RCPT would be taken as a response to RSET and so on, causing
7572 mayhem of various kinds.
7574 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7575 when I implemented it. It didn't allow for the fact that some option values
7576 may legitimately be negative (e.g. size_addition), and it didn't even do
7577 the right test for positive values.
7579 48. Domain names in DNS records are case-independent. Exim always looks them up
7580 in lower case. Some resolvers return domain names in exactly the case they
7581 appear in the zone file, that is, they may contain uppercase letters. Not
7582 all resolvers do this - some return always lower case. Exim was treating a
7583 change of case by a resolver as a change of domain, similar to a widening
7584 of a domain abbreviation. This triggered its re-routing code and so it was
7585 trying to route what was effectively the same domain again. This normally
7586 caused routing to fail (because the router wouldn't handle the domain
7587 twice). Now Exim checks for this case specially, and just changes the
7588 casing of the domain that it ultimately uses when it transmits the message
7591 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7594 50. If a filter generated a file delivery with a non-absolute name (possible if
7595 no home directory exists for the router), the forbid_file option was not
7598 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7599 the existing equality matching.
7601 52. Exim was using ints instead of ino_t variables in some places where it was
7602 dealing with inode numbers.
7604 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7605 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7606 called TMPDIR, and if it finds it is different, it changes its value.
7608 54. The smtp_printf() function is now made available to local_scan() so
7609 additional output lines can be written before returning. There is also an
7610 smtp_fflush() function to enable the detection of a dropped connection.
7611 The variables smtp_input and smtp_batched_input are exported to
7614 55. Changed the default runtime configuration: the message "Unknown user"
7615 has been removed from the ACL, and instead placed on the localuser router,
7616 using the cannot_route_message feature. This means that any verification
7617 failures that generate their own messages won't get overridden. Similarly,
7618 the "Unrouteable address" message that was in the ACL for unverifiable
7619 relay addresses has also been removed.
7621 56. Added hosts_avoid_esmtp to the smtp transport.
7623 57. The exicyclog script was not checking for the esoteric option
7624 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7625 will work only if exicyclog is run under the appropriate euid.
7627 58. Following a discussion on the list, the rules by which Exim recognises line
7628 endings on incoming messages have been changed. The -dropcr and drop_cr
7629 options are now no-ops, retained only for backwards compatibility. The
7630 following line terminators are recognized: LF CRLF CR. However, special
7631 processing applies to CR:
7633 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7634 nor a local message in the state where . is a terminator.
7636 (ii) If a bare CR is encountered in a header line, an extra space is added
7637 after the line terminator so as not to end the header. The reasoning
7638 behind this is that bare CRs in header lines are most likely either
7639 to be mistakes, or people trying to play silly games.
7641 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7642 was being incorrectly given as 18 bytes larger than it should have been.
7643 This is a VOB (very old bug).
7645 60. This may never have affected anything current, but just in case it has:
7646 When the local host is found other than at the start of a list of hosts,
7647 the local host, those with the same MX, and any that follow, are discarded.
7648 When the list in question was part of a longer list of hosts, the following
7649 hosts (not currently being processed) were also being discarded. This no
7650 longer happens. I'm not sure if this situation could ever has previously
7653 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7656 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7657 header. This is something that is recommended in a new Internet Draft, and
7658 is something that is documented as being done by Sendmail. There are two
7659 possible values. For messages generated by the autoreply transport, Exim
7662 Auto-Submitted: auto-replied
7664 whereas for all other generated messages (e.g. bounces) it adds
7666 Auto-Submitted: auto-generated
7668 63. The "personal" condition in filters now includes a test for the
7669 Auto-Submitted: header. If it contains the string "auto-" the message it
7670 not considered personal.
7672 64. Added rcpt_include_affixes as a generic transport option.
7674 65. Added queue_only_override (default true).
7676 66. Added the syslog_duplication option.
7678 67. If what should have been the first header line of a message consisted of
7679 a space followed by a colon, Exim was mis-interpreting it as a header line.
7680 It isn't of course - it is syntactically invalid and should therefore be
7681 treated as the start of the message body. The misbehaviour could have
7682 caused a number of strange effects, including loss of data in subsequent
7683 header lines, and spool format errors.
7685 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7686 client host had authenticated. This control can now be exercised by an ACL
7687 for more flexibility.
7689 69. By default, callouts do not happen when testing with -bh. There is now a
7690 variant, -bhc, which does actually run the callout code, including
7691 consulting and updating the callout cache.
7693 70. Added support for saslauthd authentication, courtesy of Alexander
7696 71. If statvfs() failed on the spool or log directories while checking their
7697 size for availability, Exim confusingly gave the error "space shortage".
7698 Furthermore, in debugging mode it crashed with a floating point exception.
7699 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7700 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7701 serious problem, Exim now writes to the main and panic logs when this
7702 happens, with details of the failure. It then refuses to accept the
7703 incoming message, giving the message "spool directory problem" or "log
7704 directory problem" with a 421 code for SMTP messages.
7706 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7707 0, 1, and 2 exist, because some OS complain for execs without them (see
7708 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7709 descriptors. However, the code omitted to check that the open succeeded,
7710 causing mysterious errors if for some reason the permissions on /dev/null
7711 got screwed. Now Exim writes a message to the main and panic logs, and
7712 bombs out if it can't open /dev/null.
7714 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7715 interact so that it is all more flexible. It is supposed to remain
7716 backwards compatible. Also added extra_local_interfaces.
7718 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7719 to bomb out with an assertion failure - to the client this appears as a
7720 connection drop. This problem occurs in the part of the code that was taken
7721 from the Samba project. Fortunately, the assertion is in a very simple
7722 function, so I have fixed this by reproducing the function inline in the
7723 one place where it is called, and arranging for authentication to fail
7724 instead of killing the process with assert().
7726 75. The SPA client code was not working when the server requested OEM rather
7727 than Unicode encoding.
7729 76. Added code to make require_files with a specific uid setting more usable in
7730 the case where statting the file as root fails - usually a non-root-mounted
7731 NFS file system. When this happens and the failure is EACCES, Exim now
7732 forks a subprocess and does the per-uid checking as the relevant uid.
7734 77. Added process_log_path.
7736 78. If log_file_path was not explicitly set, a setting of check_log_space or
7737 check_log_inodes was ignored.
7739 79. If a space check for the spool or log partitions fails, the incident is now
7740 logged. Of course, in the latter case the data may get lost...
7742 80. Added the %p formatting code to string_format() so that it can be used to
7743 print addresses in debug_print(). Adjusted all the address printing in the
7744 debugging in store.c to use %p rather than %d.
7746 81. There was a concern that a line of code in smtp_in.c could overflow a
7747 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7748 initially expressed, the concern was not well-founded, because trailing
7749 spaces are removed early. However, if the trailing spaces were followed by
7750 a NULL, they did not get removed, so the overflow was possible. Two fixes
7753 (a) I re-wrote the offending code in a cleaner fashion.
7754 (b) If an incoming SMTP command contains a NULL character, it is rejected
7757 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7758 runs initgroups(), so that if the Exim user is in any additional groups,
7759 they will be used during message reception.
7765 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob