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.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
145 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
146 left undeleted; the attempt to re-create it then failed - resulting in
147 the usual "SIGHUP tp have daemon reload configuration" to not work.
148 This affected any platform not supporting "abstract" Unix-domain
149 sockets (i.e. not Linux).
151 JH/42 Bug 2692: Harden against a peer which reneges on a 452 "too many
152 recipients" response to RCPT in a later response, with a 250. The
153 previous coding assumed this would not happen, and under PIPELINING
154 would result in both lost and duplicate recipients for a message.
156 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
157 Previously the weighting was incorrectly applied. Similar fix for socks
158 proxies. Found and fixed by Heiko Schlichting.
160 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
161 not handle sub-lists included using the +namedlist syntax. While
162 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
163 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
166 HS/01 Enforce absolute PID file path name.
168 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
170 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
172 PP/02 Bug 2643: Correct TLS DH constants.
173 A missing NUL termination in our code-generation tool had led to some
174 incorrect Diffie-Hellman constants in the Exim source.
175 Reported by kylon94, code-gen tool fix by Simon Arlott.
177 PP/03 Impose security length checks on various command-line options.
178 Fixes CVE-2020-SPRSS reported by Qualys.
180 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
181 better. Reported by Qualys.
183 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
184 providing a particularly obnoxious sender full name.
187 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
189 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
192 PP/08 Change default for recipients_max from unlimited to 50,000.
198 JH/01 Avoid costly startup code when not strictly needed. This reduces time
199 for some exim process initialisations. It does mean that the logging
200 of TLS configuration problems is only done for the daemon startup.
202 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
204 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
205 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
208 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
209 DKIM signing is being done. Previously a transport_filter would always
210 disable CHUNKING, falling back to traditional DATA.
212 JH/05 Regard command-line receipients as tainted.
214 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
216 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
217 PAM library frees one of the arguments given to it, despite the
218 documentation. Therefore a plain malloc must be used.
220 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
221 on-stack buffers were used, resulting in a taint trap when DSN information
222 copied from a received message was written into the buffer.
224 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
225 the ordering of its ARC headers. This caused a crash.
227 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
228 a new record was being constructed with information from the peer, a trap
231 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
232 installation would get error messages from DMARC verify, when it hit the
233 nonexistent file indicated by the default. Distros wanting DMARC enabled
234 should both provide the file and set the option.
235 Also enforce no DMARC verification for command-line sourced messages.
237 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
238 could, depending on the platform, hang at the STARTTLS response.
240 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
241 message on a connection. Previously if one message had ARC headers and
242 the following one did not, a crash could result when adding an
243 Authentication-Results: header.
245 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
246 Exim main code and Exim-related utities. The introduction of taint
247 tracking also did many adjustments to string handling. Since then, eximon
248 frequently terminated with an assert failure.
250 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
251 check for 452 responses. This slightly helps the inefficieny of doing
252 a large alias-expansion into a recipient-limited target. The max_rcpt
253 transport option still applies (and at the current default, will override
254 the new feature). The check is done for either cause of synch, and forces
255 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
256 connection. The new facility is not tunable at this time.
258 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
259 library live data was being used, so the results became garbage. Make
260 copies while it is still usable.
262 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
263 on delivery deferred (==) and failed (**) lines (if a delivery was
264 attemtped). Previously it was only on completion (=>) lines.
266 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
267 for the expansion of the server_scram_iter and server_scram_salt options.
269 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
270 are now specifically given a NO_DATA response without hitting the system
271 resolver. The library goes on to do the now-standard TXT lookup.
272 Use of dnsdb lookups is not affected.
274 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
275 only retrieve the errormessage once. Previously two calls to dlerror()
276 were used, and the second one (for mainlog/paniclog) retrieved null
279 JH/20 Taint checking: disallow use of tainted data for
280 - the appendfile transport file and directory options
281 - the pipe transport command
282 - the autoreply transport file, log and once options
283 - file names used by the redirect router (including filter files)
285 - paths used by single-key lookups
286 Previously this was permitted.
288 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
289 adjusted the size of a major service buffer; this failed because the
290 buffer was in use at the time. Change to a compile-time increase in the
291 buffer size, when this authenticator is compiled into exim.
293 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
294 previous fast-mode was untenable in the face of glibs using mmap to
295 support larger malloc requests.
297 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
298 New values supported, if defined on system where compiled:
299 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
300 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
302 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
303 running a limited number of proceses in parallel, a benefit is gained. The
304 amount varies with the platform hardware and load. The use of the option
305 queue_run_in_order means we cannot do this, as ordering becomes
308 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
309 had introduced a string-copy (for ensuring NUL-termination) which was not
310 appropriate for that case, which can include embedded NUL bytes in the
311 block of data. Investigation showed the copy to actually be needless, the
312 data being length-specified.
314 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
315 done during a receiving connection, and both used TLS, global info was
316 used rather than per-connection info for tracking the state of data
317 queued for transmission. This could result in a connection hang.
319 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
320 Previously, when delivering serveral messages down a single connection
321 only the first would provide a SIZE. This was due to the size information
322 not being properly tracked.
324 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
325 TAI (at 37 seconds currently), pretend to be in UTC for time-related
326 expansion and logging. Previously, spurious values such as a future
327 minute could be seen.
329 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
330 it could crash from a null-deref. This could also affect the
331 ${addresses: } operator and ${readsock } item.
333 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
334 message following a mime one, the variable was not reset.
336 JH/30 When an pipelined-connect fails at the first response, assume incorrect
337 cached capability (perhaps the peer reneged?) and immediately retry in
340 JH/31 Fix spurious detection of timeout while writing to transport filter.
342 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
343 an attempt to copy the string was made before checking it.
345 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
346 taint of the lookup key was maintained; we now regard the presence in the
347 filesystem as sufficient validation.
349 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
350 argument is supplied.
352 JH/35 The dsearch lookup now requires that the directory is an absolute path.
353 Previously this was not checked, and nonempty relative paths made an
354 access under Exim's current working directory.
356 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
357 Previously no event was raised.
359 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
360 parameter supplied by the sender MAIL FROM command. Previously it was
361 ignored, and only the check_spool_space option value for the required
364 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
365 the size of the signing public-key. Previously it was instead giving
366 the size of the signature hash.
368 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
369 the default. See the (new) dkim_verify_min_keysizes option.
371 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
372 and an ACL use a lookup for checking either the local_part or domain,
373 stale data could be accessed. Ensure that variable references are
374 dropped between messages.
376 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
377 by the client was not checked as pointing within response data before
378 being used. A malicious client could thus cause an out-of-bounds read and
379 possibly gain authentication. Fix by adding the check.
381 JH/42 Internationalisation: change the default for downconversion in the smtp
382 transport to be "if needed". Previously it was "as previously set" for
383 the message, which usually meant "if needed" for message-submission but
384 "no" for everything else. However, MTAs have been seen using SMTPUTF8
385 even when the envelope addresses did not need it, resulting in forwarding
386 failures to non-supporting MTAs. A downconvert in such cases will be
387 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
388 transport. The change does mean that addresses needing conversion will
389 be converted when previously a delivery failure would occur.
391 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
392 response was received it would be used unchecked in a fail-DSN, violating
393 standards on line-length limits. Truncate if needed.
395 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
396 doesn't work. (Additionally add a new main config option to configure the
403 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
404 side, to decode a TLS 1.3 packet capture.
406 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
407 Previously the default library behaviour applied, sending two, each in
410 JH/03 Debug output for ACL now gives the config file name and line number for
413 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
415 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
417 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
418 buffer overrun for (non-chunking) other transports.
420 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
421 TLS1.3, means that a server rejecting a client certificate is not visible
422 to the client until the first read of encrypted data (typically the
423 response to EHLO). Add detection for that case and treat it as a failed
424 TLS connection attempt, so that the normal retry-in-clear can work (if
425 suitably configured).
427 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
428 and/or domain. Found and fixed by Jason Betts.
430 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
431 configuration). If a CNAME target was not a wellformed name pattern, a
434 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
435 the OS reports them interleaved with other addresses.
437 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
438 used both for input and for a verify callout, both encrypted, SMTP
439 responses being sent by the server could be lost. This resulted in
440 dropped connections and sometimes bounces generated by a peer sending
443 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
444 string. Previously it was possible to cause undefined behaviour in a
445 library routine (usually a crash). Found by "zerons".
447 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
450 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
451 API was removed, so update to use the newer ones.
453 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
454 any timeout set, is taking a long time. Previously we would hang on to a
455 rotated logfile "forever" if the input was arriving with long gaps
456 (a previous attempt to fix addressed lack, for a long time, of initial
459 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
460 shared (NFS) environment. The length of the tempfile name is now
461 4 + 16 ("hdr.$message_exim_id") which might break on file
462 systems which restrict the file name length to lower values.
465 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
466 shared (NFS) environment.
468 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
469 did for all versions <4.90). Notably -M, -m, --invert, -I may be
472 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
473 on some platforms for bit 31.
475 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
476 to changes apparently associated with TLS1.3 handling some of the APIs
477 previously used were either nonfunctional or inappropriate. Strings
478 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
479 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
480 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
481 This affects log line X= elements, the $tls_{in,out}_cipher variables,
482 and the use of specific cipher names in the encrypted= ACL condition.
484 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
486 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
487 verification result was not updated unless hosts_require_ocsp applied.
489 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
490 queue_list_requires_admin set to false, non-admin users were denied the
493 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
494 directory-of-certs mode. Previously they were advertised despite the
497 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
498 A single TCP connection by a client will now hold a TLS connection open
499 for multiple message deliveries, by default. Previously the default was to
502 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
503 default. If built with the facility, DANE will be used. The facility
504 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
506 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
507 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
508 must be defined and you must still, unless you define DISABLE_TLS, manage
509 the the include-dir and library-file requirements that go with that
510 choice. Non-TLS builds are still supported.
512 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
515 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
516 default. If the platform supports and has the facility enabled, it will
517 be requested on all coneections.
519 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
520 controlled by the build-time option SUPPORT_PIPE_CONNECT.
522 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
524 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
525 success-DSN messages. Previously the From: header was always the default
526 one for these; the option was ignored.
528 JH/28 Fix the timeout on smtp response to apply to the whole response.
529 Previously it was reset for every read, so a teergrubing peer sending
530 single bytes within the time limit could extend the connection for a
531 long time. Credit to Qualsys Security Advisory Team for the discovery.
533 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
534 delivery address, which leaked information of the results of local
535 forwarding. Change to the original envelope recipient address, per
538 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
539 requested. Previously not bounce was generated and a log entry of
540 error ignored was made.
542 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
544 JH/32 Introduce a general tainting mechanism for values read from the input
545 channel, and values derived from them. Refuse to expand any tainted
546 values, to catch one form of exploit.
548 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
549 was unused and the unexpanded text used for the test. Found and
550 fixed by Ruben Jenster.
552 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
553 an attempt to use a TLS library read routine dereffed a nul pointer,
556 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
557 them in our smtp response.
559 JH/36 Have the general router option retry_use_local_part default to true when
560 any of the restrictive preconditions are set (to anything). Previously it
561 was only for check_local user. The change removes one item of manual
562 configuration which is required for proper retries when a remote router
563 handles a subset of addresses for a domain.
565 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
566 link count into consideration.
568 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
569 caused the extension of big_buffer, the following lines were ignored.
571 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
572 accordance with RFC 2308. Previously there was no expiry, so a longlived
573 receive process (eg. due to ACL delays) versus a short SOA value could
576 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
578 JH/39 Promote DMARC support to mainline.
580 JH/40 Bug 2452: Add a References: header to DSNs.
582 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
583 parameters. The relevant library call is documented as "Deprecated: This
584 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
585 3.6.0, DH parameters are negotiated following RFC7919."
587 HS/06 Change the default of dnssec_request_domains to "*"
589 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
590 carried on and emitted a BDAT command, even when PIPELINING was not
593 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
594 buffer was used for the filename, resulting in a trap when tainted
595 arguments (eg. $domain) were used.
597 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
598 recommended to avoid a possible server-load attack. The feature can be
599 re-enabled via the openssl_options main cofiguration option.
601 JH/45 local_scan API: documented the current smtp_printf() call. This changed
602 for version 4.90 - adding a "more data" boolean to the arguments.
603 Bumped the ABI version number also, this having been missed previously;
604 release versions 4.90 to 4.92.3 inclusive were effectively broken in
605 respect of usage of smtp_printf() by either local_scan code or libraries
606 accessed via the ${dlfunc } expansion item. Both will need coding
607 adjustment for any calls to smtp_printf() to match the new function
608 signature; a FALSE value for the new argument is always safe.
610 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
611 the file-offset (which the Linux syscall does, and exim expects); this
612 resulted in an indefinite loop.
614 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
615 to do ARC verification. The Authentication-Results: header line added
616 by the configuration then had no ARC item.
622 JH/01 Remove code calling the customisable local_scan function, unless a new
623 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
625 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
626 non-signal-safe functions being used.
628 JH/03 Bug 2269: When presented with a received message having a stupidly large
629 number of DKIM-Signature headers, disable DKIM verification to avoid
630 a resource-consumption attack. The limit is set at twenty.
632 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
633 report of oldest_pass in ${authres } in consequence, and separate out
634 some descriptions of reasons for verification fail.
636 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
637 files in the spool were present and unlocked. A queue-runner could spot
638 them, resulting in a duplicate delivery. Fix that by doing the unlock
639 after the unlink. Investigation by Tim Stewart. Take the opportunity to
640 add more error-checking on spoolfile handling while that code is being
643 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
644 No known attacks, no CVE, this is defensive hardening.
646 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
647 a queue-runner could start a delivery while other operations were ongoing.
648 Cutthrough delivery was a common victim, resulting in duplicate delivery.
649 Found and investigated by Tim Stewart. Fix by using the open message data
650 file handle rather than opening another, and not locally closing it (which
651 releases a lock) for that case, while creating the temporary .eml format
652 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
654 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
655 $sender_verify_failure/$recipient_verify_failure to "random".
657 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
660 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
661 Previously this would segfault.
663 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
666 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
667 like zero, since the resolver should be doing this for us, But we need one
668 as a CNAME but no MX presence gets the CNAME returned; we need to check
669 that doesn't point to an MX to declare it "no MX returned" rather than
670 "error, loop". A new main option is added so the older capability of
671 following some limited number of chain links is maintained.
673 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
675 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
676 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
677 now-deprecated earlier definitions used only specified the range up to TLS
678 1.2 (in the older-version library docs).
680 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
682 JH/15 Rework TLS client-side context management. Stop using a global, and
683 explicitly pass a context around. This enables future use of TLS for
684 connections to service-daemons (eg. malware scanning) while a client smtp
685 connection is using TLS; with cutthrough connections this is quite likely.
687 JH/16 Fix ARC verification to do AS checks in reverse order.
689 JH/17 Support a "tls" option on the ${readsocket } expansion item.
691 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
692 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
693 Previously the "utf8" would be re-prepended for every additional message.
695 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
696 Previously thery were accepted, resulting in issues when attempting to
697 forward messages to a non-supporting MTA.
699 PP/02 Let -n work with printing macros too, not just options.
701 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
702 one parent address was copied, and bogus data was used at delivery-logging
703 time. Either a crash (after delivery) or bogus log data could result.
704 Discovery and analysis by Tim Stewart.
706 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
707 Previously if the string ended mid-character, we did not insert the
708 promised '?' replacement.
710 PP/04 Documentation: current string operators work on bytes, not codepoints.
712 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
713 should pack well giving a smaller memory footprint so better caching and
714 therefore performance. Group the declarations where this can't be done so
715 that the byte-sized flag variables are not interspersed among pointer
716 variables, giving a better chance of good packing by the compiler.
718 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
719 non-null, to avoid issues with sites running BATV. Previously reports were
720 sent with an empty envelope sender so looked like bounces.
722 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
723 The ignore_error flag wasn't being returned from the filter subprocess so
724 was not set for later routers. Investigation and fix by Matthias Kurz.
726 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
727 and a msg:complete for the whole, when a message is manually removed using
728 -Mrm. Developement by Matthias Kurz, hacked on by JH.
730 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
731 a "Gnu special" function, asprintf() in the DB utility binary builds; I
732 hope that is portable enough.
734 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
735 requiring a known-CA anchor certificate; make it now rely entirely on the
736 TLSA as an anchor. Checking the name on the leaf cert against the name
737 on the A-record for the host is still done for TA (but not for EE mode).
739 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
740 included in delivery lines for non-proxied connections, when compiled with
741 SUPPORT_SOCKS and running with proxy logging enabled.
743 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
744 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
745 move the existing event to fire before the normal logging of message
746 failure so that custom logging is bracketed by normal logging.
748 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
749 msg:fail:internal event. Developement by Matthias Kurz.
751 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
752 far too small for todays use of crypto signatures stored there. Go all
753 the way to the max DNS message size of 64kB, even though this might be
754 overmuch for IOT constrained device use.
756 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
757 copy a string over itself. The library routine is documented as not
758 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
760 JH/32 For main options check_spool_space and check_inode_space, where the
761 platform supports 64b integers, support more than the previous 2^31 kB
762 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
763 the previous G, M, k.
765 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
766 $authenticated_fail_id variable on authentication failure. Previously
769 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
770 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
771 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
772 GNUTLS_SEC_PARAM_MEDIUM.
774 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
775 side. Previously we would continue as if no SNI had been received.
777 JH/36 Harden the handling of string-lists. When a list consisted of a sole
778 "<" character, which should be a list-separator specification, we walked
779 off past the nul-terimation.
781 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
782 causes) even when the retry time is not yet met. Previously they were
783 not, meaning that when (say) an account was over-quota and temp-rejecting,
784 and multiple senders' messages were queued, only one sender would get
785 notified on each configured delay_warning cycle.
787 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
789 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
790 cipher-suites, an error can be left on the stack even for a succeeding
791 accept; this results in impossible error messages when a later operation
794 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
795 return error codes indicating retry. Under TLS1.3 this becomes required.
797 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
798 it only wrote the new authenticators, resulting in a lack of tracking of
799 peer changes of ESMTP extensions until the next cache flush.
801 JH/41 Fix the loop reading a message header line to check for integer overflow,
802 and more-often against header_maxsize. Previously a crafted message could
803 induce a crash of the recive process; now the message is cleanly rejected.
805 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
806 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
812 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
813 When redis_servers is set to a list of > 1 element, and the Redis servers
814 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
815 case of MOVED into a DEFER case instead, thus moving the query onto the
816 next server in the list. For a cluster of N elements, all N servers must
817 be defined in redis_servers.
819 GF/02 Catch and remove uninitialized value warning in exiqsumm
820 Check for existence of @ARGV before looking at $ARGV[0]
822 JH/01 Replace the store_release() internal interface with store_newblock(),
823 which internalises the check required to safely use the old one, plus
824 the allocate and data copy operations duplicated in both (!) of the
825 extant use locations.
827 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
828 modifier. This matches the restriction on the commandline.
830 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
831 Previously only the last row was returned.
833 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
834 we assumed that tags in the header were well-formed, and parsed the
835 element content after inspecting only the first char of the tag.
836 Assumptions at that stage could crash the receive process on malformed
839 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
840 While running the DKIM ACL we operate on the Permanent memory pool so that
841 variables created with "set" persist to the DATA ACL. Also (at any time)
842 DNS lookups that fail create cache records using the Permanent pool. But
843 expansions release any allocations made on the current pool - so a dnsdb
844 lookup expansion done in the DKIM ACL releases the memory used for the
845 DNS negative-cache, and bad things result. Solution is to switch to the
846 Main pool for expansions.
847 While we're in that code, add checks on the DNS cache during store_reset,
848 active in the testsuite.
849 Problem spotted, and debugging aided, by Wolfgang Breyha.
851 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
852 When none of the hosts presented to a transport match an already-open
853 connection, close it and proceed with the list. Previously we would
854 queue the message. Spotted by Lena with Yahoo, probably involving
857 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
858 Previously a spurious "250 OK id=" response was appended to the proper
861 JH/08 The "support for" informational output now, which built with Content
862 Scanning support, has a line for the malware scanner interfaces compiled
863 in. Interface can be individually included or not at build time.
865 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
866 by the template makefile "src/EDITME". The "STREAM" support for an older
867 ClamAV interface method is removed.
869 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
870 rows affected is given instead).
872 JH/11 The runtime Berkeley DB library version is now additionally output by
873 "exim -d -bV". Previously only the compile-time version was shown.
875 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
876 SMTP connection. Previously, when one had more receipients than the
877 first, an abortive onward connection was made. Move to full support for
878 multiple onward connections in sequence, handling cutthrough connection
879 for all multi-message initiating connections.
881 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
882 routers. Previously, a multi-recipient message would fail to match the
883 onward-connection opened for the first recipient, and cause its closure.
885 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
886 a timeout on read on a GnuTLS initiating connection, resulting in the
887 initiating connection being dropped. This mattered most when the callout
888 was marked defer_ok. Fix to keep the two timeout-detection methods
891 JH/15 Relax results from ACL control request to enable cutthrough, in
892 unsupported situations, from error to silently (except under debug)
893 ignoring. This covers use with PRDR, frozen messages, queue-only and
896 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
898 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
899 metadata, resulting in a crash in free().
901 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
902 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
903 Broken also in d185889f4, with init system revamp.
905 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
906 Previously we abruptly closed the connection after reading a malware-
907 found indication; now we go on to read the "scan ok" response line,
910 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
911 ACL. Previously, a crash would result.
913 JH/19 Speed up macro lookups during configuration file read, by skipping non-
914 macro text after a replacement (previously it was only once per line) and
915 by skipping builtin macros when searching for an uppercase lead character.
917 JH/20 DANE support moved from Experimental to mainline. The Makefile control
918 for the build is renamed.
920 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
921 was allocated for every new TLS startup, meaning one per message. Fix
922 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
924 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
925 reported the original. Fix to report (as far as possible) the ACL
926 result replacing the original.
928 JH/23 Fix memory leak during multi-message connections using STARTTLS under
929 OpenSSL. Certificate information is loaded for every new TLS startup,
930 and the resources needed to be freed.
932 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
934 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
937 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
938 DATA response info to the (existing) per-recipient response info for
939 the "C=" log element. It can have useful tracking info from the
940 destination system. Patch from Simon Arlott.
942 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
943 length value. Previously this would segfault.
945 HS/02 Support Avast multiline protoocol, this allows passing flags to
946 newer versions of the scanner.
948 JH/28 Ensure that variables possibly set during message acceptance are marked
949 dead before release of memory in the daemon loop. This stops complaints
950 about them when the debug_store option is enabled. Discovered specifically
951 for sender_rate_period, but applies to a whole set of variables.
952 Do the same for the queue-runner and queue-list loops, for variables set
953 from spool message files. Do the same for the SMTP per-message loop, for
954 certain variables indirectly set in ACL operations.
956 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
957 as a multi-recipient message from a mailinglist manager). The coding had
958 an arbitrary cutoff number of characters while checking for more input;
959 enforced by writing a NUL into the buffer. This corrupted long / fast
960 input. The problem was exposed more widely when more pipelineing of SMTP
961 responses was introduced, and one Exim system was feeding another.
962 The symptom is log complaints of SMTP syntax error (NUL chars) on the
963 receiving system, and refused recipients seen by the sending system
964 (propating to people being dropped from mailing lists).
965 Discovered and pinpointed by David Carter.
967 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
968 replaced by the ${authresults } expansion.
970 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
972 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
973 allows proper process termination in container environments.
975 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
976 Previously the "final dot" had a newline after it; ensure it is CR,LF.
978 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
979 and "err_perm", deprecated since 4.83 when the RFC-defined words
980 "temperror" and "permerror" were introduced.
982 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
983 transport-filters or DKIM-signing. The restriction was lost in the
984 consolidation of verify-callout and delivery SMTP handling.
985 Extend the restriction to also cover ARC-signing.
987 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
988 in defer=pass mode supply a 450 to the initiator. Previously the message
991 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
992 tls_require_ciphers is used as before.
994 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
995 "pass_unscanned". Only tmpfails from the scanner are written to
996 the paniclog, as they may require admin intervention (permission
997 denied, license issues). Other scanner errors (like decompression
998 bombs) do not cause a paniclog entry.
1000 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1001 Previously it was possible to log spurious information in receive log
1004 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1005 triggered odd behaviour from Outlook Express clients.
1007 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1010 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1011 since the IETF WG has not yet settled on that versus the original
1012 "bare" representation.
1014 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1015 Previously the millisecond value corrupted the output.
1016 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1017 corrupted the output.
1023 JH/01 Rework error string handling in TLS interface so that the caller in
1024 more cases is responsible for logging. This permits library-sourced
1025 string to be attached to addresses during delivery, and collapses
1026 pairs of long lines into single ones.
1028 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1029 during configuration. Wildcards are allowed and expanded.
1031 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1032 This permits better logging.
1034 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1035 do not close it down and have the child transport start it up again on
1036 the passed-on TCP connection. Instead, proxy the child (and any
1037 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1038 affected: the continued delivery log lines do not have any DNSSEC, TLS
1039 Certificate or OCSP information. TLS cipher information is still logged.
1041 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1042 identical IP addresses on different listening ports. Will also affect
1045 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1046 add noisy ifdef guards to special-case this sillyness.
1047 Patch from Bernd Kuhls.
1049 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1050 than 255 are no longer allowed.
1052 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1053 Disable the session-cache too, which might reduce our load. Since we
1054 currrectly use a new context for every connection, both as server and
1055 client, there is no benefit for these.
1056 GnuTLS appears to not support tickets server-side by default (we don't
1057 call gnutls_session_ticket_enable_server()) but client side is enabled
1058 by default on recent versions (3.1.3 +) unless the PFS priority string
1061 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1062 <https://reproducible-builds.org/specs/source-date-epoch/>.
1064 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1065 the check for any unsuccessful recipients did not notice the limit, and
1066 erroneously found still-pending ones.
1068 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1069 MSG_MORE. Only in-clear (not on TLS connections).
1071 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1072 transport-filter is involved we can buffer the headers in memory for
1073 creating the signature, and read the spool data file once for the
1074 signature and again for transmission.
1076 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1077 4.77 as the kernel support then wasn't solid, having issues in 64bit
1078 mode. Now, it's been long enough. Add support for FreeBSD also.
1080 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1081 case where the routing stage had gathered several addresses to send to
1082 a host before calling the transport for the first, we previously failed
1083 to close down TLS in the old transport process before passing the TCP
1084 connection to the new process. The new one sent a STARTTLS command
1085 which naturally failed, giving a failed delivery and bloating the retry
1086 database. Investigation and fix prototype from Wolfgang Breyha.
1088 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1089 false-negatives in the check that the sender had not preempted a response
1090 or prompt from Exim (running as a server), due to that code's lack of
1091 awareness of the SMTP input buffering.
1093 PP/04 Add commandline_checks_require_admin option.
1094 Exim drops privileges sanely, various checks such as -be aren't a
1095 security problem, as long as you trust local users with access to their
1096 own account. When invoked by services which pass untrusted data to
1097 Exim, this might be an issue. Set this option in main configuration
1098 AND make fixes to the calling application, such as using `--` to stop
1101 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1102 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1103 responses to those, into a single TLS record each way (this usually means
1104 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1107 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1108 affects you only if you're dancing at the edge of the param size limits.
1109 If you are, and this message makes sense to you, then: raise the
1110 configured limit or use OpenSSL 1.1. Nothing we can do for older
1113 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1114 cmdline element to get the documented default one. Previously it was
1117 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1118 are unsuable for non-message-specific reasons - eg. connection timeout,
1119 banner-time rejection.
1121 JH/16 Fix logging of delivery remote port, when specified by router, under
1124 PP/06 Repair manualroute's ability to take options in any order, even if one
1125 is the name of a transport.
1128 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1130 JH/17 Change the list-building routines interface to use the expanding-string
1131 triplet model, for better allocation and copying behaviour.
1133 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1134 Previously it was constructed the first time a possibly-matching string
1135 was met in the configuration file input during startup; now it is done
1138 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1139 compatible one, to avoid the (poorly documented) possibility of a config
1140 file in the working directory redirecting the DB files, possibly correpting
1141 some existing file. CVE-2017-10140 assigned for BDB.
1143 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1144 cache-hot. Previously, although the result was properly cached, the
1145 initial verify call returned a defer.
1147 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1148 the main verify for receipient in uncached-mode.
1150 JH/22 Retire historical build files to an "unsupported" subdir. These are
1151 defined as "ones for which we have no current evidence of testing".
1153 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1154 if present. Previously it was ignored.
1156 JH/24 Start using specified-initialisers in C structure init coding. This is
1157 a C99 feature (it's 2017, so now considered safe).
1159 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1160 if was a fixed-sized field and bitmask ops via macros; it is now more
1163 PP/07 GitHub PR 56: Apply MariaDB build fix.
1164 Patch provided by Jaroslav Škarvada.
1166 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1167 during Coverity cleanups [4.87 JH/47]
1168 Diagnosis and fix provided by Michael Fischer v. Mollard.
1170 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1171 the right size to place the terminating semicolon on its own folded
1172 line, the header hash was calculated to an incorrect value thanks to
1173 the (relaxed) space the fold became.
1175 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1176 and confused the parent.
1178 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1179 which could crash as a result. This could lead to undeliverable messages.
1181 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1184 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1185 matters on fast-turnover and PID-randomising systems, which were getting
1186 out-of-order delivery.
1188 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1189 a possibly-overlapping copy. The symptom was that "Remote host closed
1190 connection in response to HELO" was logged instead of the actual 4xx
1193 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1194 Previously only that bufferd was discarded, resulting in SYMTP command
1197 JH/32 DKIM: when a message has multiple signatures matching an identity given
1198 in dkim_verify_signers, run the dkim acl once for each. Previously only
1199 one run was done. Bug 2189.
1201 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1202 "panic the current process" to "deliberately defer". The panic log is
1203 still written with the problem list name; the mail and reject logs now
1204 get a temp-reject line for the message that was being handled, saying
1205 something like "domains check lookup or other defer". The SMTP 451
1206 message is still "Temporary local problem".
1208 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1209 A crafted sequence of BDAT commands could result in in-use memory beeing
1210 freed. CVE-2017-16943.
1212 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1213 from SMTP input. Previously it was always done; now only done for DATA
1214 and not BDAT commands. CVE-2017-16944.
1216 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1217 to the message (such as an overlong header line). Previously this was
1218 not done and we did not exit BDAT mode. Followon from the previous item
1219 though a different problem.
1225 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1226 than -2003 did; needs libidn2 in addition to libidn.
1228 JH/02 The path option on a pipe transport is now expanded before use.
1230 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1231 Patch provided by "Björn", documentation fix added too.
1233 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1234 missing a wire-to-host endian conversion.
1236 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1237 close after a BDAT command line could be taken as a following command,
1238 giving a synch failure. Fix by only checking for synch immediately
1239 before acknowledging the chunk.
1241 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1242 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1243 macro. Patches provided by Josh Soref.
1245 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1246 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1247 by IANA as a service extension. However, John Klensin suggests that we
1250 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1251 the dkim code may be unix-mode line endings rather than smtp wire-format
1252 CRLF, so prepend a CR to any bare LF.
1254 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1255 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1257 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1258 the first were themselves being wrongly included in the feed into dkim
1259 processing; with most chunk sizes in use this resulted in an incorrect
1260 body hash calculated value.
1262 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1263 DKIM signature block, for verification. Although advised against by
1264 standards it is specifically not ruled illegal.
1266 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1268 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1269 missing a body hash (the bh= tag).
1271 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1272 It seems that HAProxy sends the Proxy Protocol information in clear and
1273 only then does a TLS startup, so do the same.
1275 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1276 TCP connections (such as for Spamd) unless the daemon successfully set
1277 Fast Open mode on its listening sockets. This fixes breakage seen on
1278 too-old kernels or those not configured for Fast Open, at the cost of
1279 requiring both directions being enabled for TFO, and TFO never being used
1280 by non-daemon-related Exim processes.
1282 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1283 endings, at least on the first header line. Try to canonify any that get
1284 past that check, despite the cost.
1286 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1287 now limited to an arbitrary five deep, while parsing addresses with the
1288 strip_excess_angle_brackets option enabled.
1290 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1291 instead leave the unprompted TLS handshake in socket buffer for the
1292 TLS library to consume.
1294 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1296 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1298 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1299 region in one operation, for speed, and this leaves assigned pointers
1300 dangling. Add checks run only under the testsuite which checks all
1301 variables at a store-reset and panics on a dangling pointer; add code
1302 explicitly nulling out all the variables discovered. Fixes one known
1303 bug: a transport crash, where a dangling pointer for $sending_ip_address
1304 originally assigned in a verify callout, is re-used.
1306 PP/06 Drop '.' from @INC in various Perl scripts.
1308 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1310 PP/08 Reduce a number of compilation warnings under clang; building with
1311 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1312 should be warning-free.
1314 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1316 HS/01 Fix portability problems introduced by PP/08 for platforms where
1317 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1319 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1320 chunk. This allows us to accept broken chunked messages. We need a more
1321 general solution here.
1323 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1324 already-broken messages in the queue.
1326 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1328 JH/19 Fix reference counting bug in routing-generated-address tracking.
1334 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1335 supports it and a size is available (ie. the sending peer gave us one).
1337 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1338 years of being deprecated). The replacements are the ACLs
1339 acl_smtp_mime and acl_not_smtp_mime.
1341 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1342 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1343 or even an in-clear connection were permitted. Now, if the host lookup
1344 was dnssec and dane was requested then the host is only used if the
1345 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1346 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1347 if one fails this test.
1348 This means that a poorly-configured remote DNS will make it incommunicado;
1349 but it protects against a DNS-interception attack on it.
1351 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1352 non-noisy when a race steals the message being considered.
1354 JH/05 If main configuration option tls_certificate is unset, generate a
1355 self-signed certificate for inbound TLS connections.
1357 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1358 in rewrites and routers.
1360 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1361 and logged a warning sing 4.83; now they are a configuration file error.
1363 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1364 (lacking @domain). Apply the same qualification processing as RCPT.
1366 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1368 JH/10 Support ${sha256:} applied to a string (as well as the previous
1371 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1372 a cutthrough deliver is pending, as we always want to make a connection.
1373 This also avoids re-routing the message when later placing the cutthrough
1374 connection after a verify cache hit.
1375 Do not update it with the verify result either.
1377 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1378 when routing results in more than one destination address.
1380 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1381 signing (which inhibits the cutthrough capability). Previously only
1382 the presence of an option was tested; now an expansion evaluating as
1383 empty is permissible (obviously it should depend only on data available
1384 when the cutthrough connection is made).
1386 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1387 the relevant preceding SMTP command did not note the pipelining mode.
1389 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1390 Previously they were not counted.
1392 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1393 as one having no matching records. Previously we deferred the message
1394 that needed the lookup.
1396 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1397 distinguished as "(=".
1399 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1400 for missing MX records. Previously it only worked for missing A records.
1402 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1404 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1405 after the data-go-ahead and data-ack. Patch from Jason Betts.
1407 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1408 even for a "none" policy. Patch from Tony Meyer.
1410 JH/22 Fix continued use of a connection for further deliveries. If a port was
1411 specified by a router, it must also match for the delivery to be
1414 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1415 When one of the recipients of a message was unsuitable for the connection
1416 (has no matching addresses), we lost track of needing to mark it
1417 deferred. As a result mail would be lost.
1419 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1421 JH/25 Decoding ACL controls is now done using a binary search; the source code
1422 takes up less space and should be simpler to maintain. Merge the ACL
1423 condition decode tables also, with similar effect.
1425 JH/26 Fix problem with one_time used on a redirect router which returned the
1426 parent address unchanged. A retry would see the parent address marked as
1427 delivered, so not attempt the (identical) child. As a result mail would
1430 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1431 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1432 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1435 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1436 Main config options check_{log,spool}_{inodes,space} are now
1437 100 inodes, 10MB unless set otherwise in the configuration.
1439 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1440 Previously it only applied to the main-section connection policy
1443 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1445 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1446 by me. Added RFC7919 DH primes as an alternative.
1448 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1449 are not in the system include path.
1451 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1452 GnuTLS, when a session startup failed (eg because the client disconnected)
1453 Exim did stdio operations after fclose. This was exposed by a recent
1454 change which nulled out the file handle after the fclose.
1456 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1457 signed directly by the cert-signing cert, rather than an intermediate
1458 OCSP-signing cert. This is the model used by LetsEncrypt.
1460 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1462 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1463 an incoming connection.
1465 HS/02 Bug 1802: Do not half-close the connection after sending a request
1468 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1469 fallback to "prime256v1".
1471 JH/34 SECURITY: Use proper copy of DATA command in error message.
1472 Could leak key material. Remotely exploitable. CVE-2016-9963.
1478 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1479 and 3.4.4 - once the server is enabled to respond to an OCSP request
1480 it does even when not requested, resulting in a stapling non-aware
1481 client dropping the TLS connection.
1483 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1484 support variable-length bit vectors. No functional change.
1486 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1487 The I= interface field on outgoing lines is now after the H= remote
1488 host field, same as incoming lines. There is a separate
1489 outgoing_interface log selector which allows you to disable the
1492 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1493 If not running log_selector +smtp_connection the mainlog would be held
1494 open indefinitely after a "too many connections" event, including to a
1495 deleted file after a log rotate. Leave the per net connection logging
1496 leaving it open for efficiency as that will be quickly detected by the
1497 check on the next write.
1499 HS/01 Bug 1671: Fix post transport crash.
1500 Processing the wait-<transport> messages could crash the delivery
1501 process if the message IDs didn't exist for some reason. When
1502 using 'split_spool_directory=yes' the construction of the spool
1503 file name failed already, exposing the same netto behaviour.
1505 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1506 mime_regex ACL conditions.
1508 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1509 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1510 to HELO, local diagnostic string.
1512 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1513 log line to debug. Even when configured with a tls authenticator many
1514 client connections are expected to not authenticate in this way, so
1515 an authenticate fail is not an error.
1517 HS/02 Add the Exim version string to the process info. This way exiwhat
1518 gives some more detail about the running daemon.
1520 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1521 matter for fast-change records such as DNSBLs.
1523 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1524 retry record, even if constant. There may be multiple transports with
1525 different interface settings and the retry behaviour needs to be kept
1528 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1530 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1532 JH/10 Bug 840: fix log_defer_output option of pipe transport
1534 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1535 hosts_randomize. This matters a lot when combined with mua_wrapper.
1537 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1538 ${quote_pgsql:<string>} operator.
1540 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1542 JH/14 Change header file name for freeradius-client. Relevant if compiling
1543 with Radius support; from the Gentoo tree and checked under Fedora.
1545 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1547 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1548 option transport_filter, meaning no filtering.
1550 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1552 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1553 defaults to "*" (all hosts). The variable is now available when not built
1554 with TLS, default unset, mainly to enable keeping the testsuite sane.
1555 If a server certificate is not supplied (via tls_certificate) an error is
1556 logged, and clients will find TLS connections fail on startup. Presumably
1557 they will retry in-clear.
1558 Packagers of Exim are strongly encouraged to create a server certificate
1559 at installation time.
1561 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1562 with the $config_file variable.
1564 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1565 in transport context, after the attempt, and per-recipient. The latter type
1566 is per host attempted. The event data is the error message, and the errno
1567 information encodes the lookup type (A vs. MX) used for the (first) host,
1568 and the trailing two digits of the smtp 4xx response.
1570 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1571 to write to mainlog (or rejectlog, paniclog) in the window between file
1572 creation and permissions/ownership being changed. Particularly affects
1573 installations where exicyclog is run as root, rather than exim user;
1574 result is that the running daemon panics and dies.
1576 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1578 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1579 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1580 "pri" and "weight". Note that the previous implicit priority given by the
1581 list order is no longer honoured.
1583 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1584 for DKIM processing.
1586 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1587 by defining SUPPORT_SOCKS.
1589 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1590 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1591 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1592 variables are renamed to proxy_{local,external}_{address,port}.
1594 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1595 for a build by defining SUPPORT_I18N
1597 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1598 of the query string, and make ${quote_redis:} do that quoting.
1600 JH/29 Move Events support from Experimental to mainline, enabled by default
1601 and removable for a build by defining DISABLE_EVENT.
1603 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1605 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1606 cached by the daemon.
1608 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1609 by defining LOOKUP_REDIS. The libhiredis library is required.
1611 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1612 keys are given for lookup.
1614 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1615 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1616 only supported when built with TLS support. The PolarSSL SHA routines
1617 are still used when the TLS library is too old for convenient support.
1619 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1620 openssl_options), for security. OpenSSL forces this from version 1.1.0
1621 server-side so match that on older versions.
1623 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1624 allocation for $value could be released as the expansion processing
1625 concluded, but leaving the global pointer active for it.
1627 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1628 and to use the domains and local_parts ACL conditions.
1630 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1631 incorrectly not doubled on cutthrough transmission, hence seen as a
1632 body-termination at the receiving system - resulting in truncated mails.
1633 Commonly the sender saw a TCP-level error, and retransmitted the message
1634 via the normal store-and-forward channel. This could result in duplicates
1635 received - but deduplicating mailstores were liable to retain only the
1636 initial truncated version.
1638 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1640 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1642 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1643 we're in there, support oversigning also; bug 1309.
1645 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1647 HS/04 Add support for keep_environment and add_environment options.
1649 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1650 either intentional arithmetic overflow during PRNG, or testing config-
1653 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1654 delivery resulted in actual delivery. Cancel cutthrough before DATA
1657 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1658 rejects a recipient: pass the reject to the originator.
1660 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1661 Many were false-positives and ignorable, but it's worth fixing the
1664 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1665 for the new environment-manipulation done at startup. Move the routines
1666 from being local to tls.c to being global via the os.c file.
1668 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1669 an extract embedded as result-arg for a map, the first arg for extract
1670 is unavailable so we cannot tell if this is a numbered or keyed
1671 extraction. Accept either.
1677 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1680 JH/02 The smtp transport option "multi_domain" is now expanded.
1682 JH/03 The smtp transport now requests PRDR by default, if the server offers
1685 JH/04 Certificate name checking on server certificates, when exim is a client,
1686 is now done by default. The transport option tls_verify_cert_hostnames
1687 can be used to disable this per-host. The build option
1688 EXPERIMENTAL_CERTNAMES is withdrawn.
1690 JH/05 The value of the tls_verify_certificates smtp transport and main options
1691 default to the word "system" to access the system default CA bundle.
1692 For GnuTLS, only version 3.0.20 or later.
1694 JH/06 Verification of the server certificate for a TLS connection is now tried
1695 (but not required) by default. The verification status is now logged by
1696 default, for both outbound TLS and client-certificate supplying inbound
1699 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1702 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1703 Status Notification (bounce) messages are now MIME format per RFC 3464.
1704 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1705 under the control of the dsn_advertise_hosts option, and routers may
1706 have a dsn_lasthop option.
1708 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1709 default, modifiable by a malware= option. The list separator for
1710 the options can now be changed in the usual way. Bug 68.
1712 JH/10 The smtp_receive_timeout main option is now expanded before use.
1714 JH/11 The incoming_interface log option now also enables logging of the
1715 local interface on delivery outgoing connections.
1717 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1718 if the interface and destination host and port all match.
1720 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1723 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1724 Patch from Andrew Lewis.
1726 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1727 now supports optional time-restrictions, weighting, and priority
1728 modifiers per server. Patch originally by <rommer@active.by>.
1730 JH/16 The spamd_address main option now supports a mixed list of local
1731 and remote servers. Remote servers can be IPv6 addresses, and
1732 specify a port-range.
1734 JH/17 Bug 68: The spamd_address main option now supports an optional
1735 timeout value per server.
1737 JH/18 Bug 1581: Router and transport options headers_add/remove can
1738 now have the list separator specified.
1740 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1743 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1746 JH/21 Support for the A6 type of dns record is withdrawn.
1748 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1749 rather than the verbs used.
1751 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1752 from 255 to 1024 chars.
1754 JH/24 Verification callouts now attempt to use TLS by default.
1756 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1757 are generic router options now. The defaults didn't change.
1759 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1760 Original patch from Alexander Shikoff, worked over by JH.
1762 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1763 files if ZCAT_COMMAND is not executable.
1765 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1767 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1769 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1770 Normally benign, it bites when the pair was led to by a CNAME;
1771 modern usage is to not canonicalize the domain to a CNAME target
1772 (and we were inconsistent anyway for A-only vs AAAA+A).
1774 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1776 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1777 when evaluating $sender_host_dnssec.
1779 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1780 $sender_helo_dnssec variable.
1782 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1784 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1786 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1788 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1789 documented as working, but never had. Support all but $spam_report.
1791 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1792 added for tls authenticator.
1794 HS/03 Add perl_taintmode main config option
1800 TL/01 When running the test suite, the README says that variables such as
1801 no_msglog_check are global and can be placed anywhere in a specific
1802 test's script, however it was observed that placement needed to be near
1803 the beginning for it to behave that way. Changed the runtest perl
1804 script to read through the entire script once to detect and set these
1805 variables, reset to the beginning of the script, and then run through
1806 the script parsing/test process like normal.
1808 TL/02 The BSD's have an arc4random API. One of the functions to induce
1809 adding randomness was arc4random_stir(), but it has been removed in
1810 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1811 function when detected.
1813 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1814 cause callback expansion.
1816 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1817 syntax errors in an expansion can be treated as a string instead of
1818 logging or causing an error, due to the internal use of bool_lax
1819 instead of bool when processing it.
1821 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1822 server certificates when making smtp deliveries.
1824 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1826 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1828 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1830 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1831 Merged patch from Sebastian Wiedenroth.
1833 JH/05 Fix results-pipe from transport process. Several recipients, combined
1834 with certificate use, exposed issues where response data items split
1835 over buffer boundaries were not parsed properly. This eventually
1836 resulted in duplicates being sent. This issue only became common enough
1837 to notice due to the introduction of connection certificate information,
1838 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1840 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1841 size buffer was used, resulting in syntax errors when an expansion
1844 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1845 version 3.3.6 or later.
1847 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1848 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1849 both become "event_action", the variables become $event_name, $event_data
1850 and $event_defer_errno. There is a new variable $verify_mode, usable in
1851 routers, transports and related events. The tls:cert event is now also
1852 raised for inbound connections, if the main configuration event_action
1855 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1856 early OCSP support, but no stapling (appears to be less than 1.0.0).
1858 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1859 server certificate names available under the smtp transport option
1860 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1863 JH/10 Time-related extraction expansions from certificates now use the main
1864 option "timezone" setting for output formatting, and are consistent
1865 between OpenSSL and GnuTLS compilations. Bug 1541.
1867 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1868 encoded parameter in the incoming message. Bug 1558.
1870 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1871 include certificate info, eximon was claiming there were spoolfile
1874 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1876 JH/14 Log delivery-related information more consistently, using the sequence
1877 "H=<name> [<ip>]" wherever possible.
1879 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1880 are problematic for Debian distribution, omit them from the release
1883 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1885 JH/16 Fix string representation of time values on 64bit time_t architectures.
1888 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1889 output list separator was used.
1894 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1895 checkers that were complaining about end of non-void function with no
1898 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1899 This was a regression introduced in 4.83 by another bugfix.
1901 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1903 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1904 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1910 TF/01 Correctly close the server side of TLS when forking for delivery.
1912 When a message was received over SMTP with TLS, Exim failed to clear up
1913 the incoming connection properly after forking off the child process to
1914 deliver the message. In some situations the subsequent outgoing
1915 delivery connection happened to have the same fd number as the incoming
1916 connection previously had. Exim would try to use TLS and fail, logging
1917 a "Bad file descriptor" error.
1919 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1920 utilities have not been installed.
1922 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1923 temporary space as the ACL may create new global variables.
1925 TL/01 LDAP support uses per connection or global context settings, depending
1926 upon the detected version of the libraries at build time.
1928 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1929 to extract and use the src ip:port in logging and expansions as if it
1930 were a direct connection from the outside internet. PPv2 support was
1931 updated based on HAProxy spec change in May 2014.
1933 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1935 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1936 Properly escape header and check for NULL return.
1938 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1941 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1943 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1944 characters in header names, implemented as a verify condition.
1945 Contributed by Michael Fischer v. Mollard.
1947 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1948 results permerror and temperror. Previous values are deprecated but
1949 still accepted. In a future release, err_perm and err_temp will be
1950 completely removed, which will be a backward incompatibility if the
1951 ACL tests for either of these two old results. Patch contributed by
1952 user bes-internal on the mailing list.
1954 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1956 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1957 selectors, in both main and reject logs.
1959 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1962 JH/07 Add malware type "sock" for talking to simple daemon.
1964 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1966 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1967 routers/transports under cutthrough routing.
1969 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1970 numbers. Touch up "bool" conditional to keep the same definition.
1972 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1974 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1976 JH/12 Expand items in router/transport headers_add or headers_remove lists
1977 individually rather than the list as a whole. Bug 1452.
1979 Required for reasonable handling of multiple headers_ options when
1980 they may be empty; requires that headers_remove items with embedded
1981 colons must have them doubled (or the list-separator changed).
1983 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1984 view the policy declared in the DMARC record. Currently, $dmarc_status
1985 is a combined value of both the record presence and the result of the
1988 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1990 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1991 dnslookup router and the smtp transport (applying to the forward
1994 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1995 of ldap servers used for a specific lookup. Patch provided by Heiko
1998 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1999 New variable $lookup_dnssec_authenticated for observability.
2001 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2002 Patch submitted by Lars Timman.
2004 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2006 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2007 Requires trusted mode and valid format message id, aborts otherwise.
2008 Patch contributed by Heiko Schlichting.
2010 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2011 certextract with support for various fields. Bug 1358.
2013 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2014 is requested by default, modifiable by smtp transport option
2017 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2018 operate on certificate variables to give certificate fingerprints
2019 Also new ${sha256:cert_variable}.
2021 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2023 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2024 Christian Aistleitner.
2026 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2028 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2029 file. Patch from Wolfgang Breyha.
2031 JH/25 Expand the coverage of the delivery $host and $host_address to
2032 client authenticators run in verify callout. Bug 1476.
2034 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2035 align with daemon_smtp_ports. Bug 72.
2037 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2038 support and error reporting did not work properly.
2040 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2041 and is readable. Patch from Andrew Colin Kissa.
2043 TL/14 Enhance documentation of ${run expansion and how it parses the
2044 commandline after expansion, particularly in the case when an
2045 unquoted variable expansion results in an empty value.
2047 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2049 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2050 by renaming away from it. Addresses GH issue 15, reported by
2053 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2055 TL/15 SECURITY: prevent double expansion in math comparison functions
2056 (can expand unsanitized data). Not remotely exploitable.
2063 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2065 PP/02 Make -n do something, by making it not do something.
2066 When combined with -bP, the name of an option is not output.
2068 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2071 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2072 $sender_host_name and config options to manage this, and basic check
2075 PP/05 DSCP support for outbound connections and control modifier for inbound.
2077 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2078 (Only plugin which currently uses this is kerberos4, which nobody should
2079 be using, but we should make it available and other future plugins might
2080 conceivably use it, even though it would break NAT; stuff *should* be
2081 using channel bindings instead).
2083 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2084 name; added for Sendmail compatibility; requires admin caller.
2085 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2086 just ignore it); requires trusted caller.
2087 Also parse but ignore: -Ac -Am -X<logfile>
2090 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2092 TL/02 Add +smtp_confirmation as a default logging option.
2094 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2095 Patch by Magnus Holmgren from 2007-02-20.
2097 TL/04 Bugzilla 1281 - Spec typo.
2098 Bugzilla 1283 - Spec typo.
2099 Bugzilla 1290 - Spec grammar fixes.
2101 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2103 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2105 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2106 mailing list by by Dmitry Isaikin.
2108 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2110 JH/02 Support "G" suffix to numbers in ${if comparisons.
2112 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2114 NM/01 Bugzilla 1197 - Spec typo
2115 Bugzilla 1196 - Spec examples corrections
2117 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2119 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2120 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2123 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2124 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2126 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2127 "acl {{name}{arg}...}", and optional args on acl condition
2130 JH/05 Permit multiple router/transport headers_add/remove lines.
2132 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2134 JH/07 Avoid using a waiting database for a single-message-only transport.
2135 Performance patch from Paul Fisher. Bugzilla 1262.
2137 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2140 JH/09 Add $headers_added variable, with content from use of ACL modifier
2141 add_header (but not yet added to the message). Bugzilla 199.
2143 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2144 Pulled from Bugzilla 817 by Wolfgang Breyha.
2146 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2148 (nb: this is the same fix as in Exim 4.80.1)
2150 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2153 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2156 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2158 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2159 advertise SMTP AUTH mechanism to us, instead of a generic
2160 protocol violation error. Also, make Exim more robust to bad
2161 data from the Dovecot auth socket.
2163 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2165 When a queue runner is handling a message, Exim first routes the
2166 recipient addresses, during which it prunes them based on the retry
2167 hints database. After that it attempts to deliver the message to
2168 any remaining recipients. It then updates the hints database using
2171 So if a recipient address works intermittently, it can get repeatedly
2172 deferred at routing time. The retry hints record remains fresh so the
2173 address never reaches the final cutoff time.
2175 This is a fairly common occurrence when a user is bumping up against
2176 their storage quota. Exim had some logic in its local delivery code
2177 to deal with this. However it did not apply to per-recipient defers
2178 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2180 This change adds a proper retry rule check during routing so that the
2181 final cutoff time is checked against the message's age. We only do
2182 this check if there is an address retry record and there is not a
2183 domain retry record; this implies that previous attempts to handle
2184 the address had the retry_use_local_parts option turned on. We use
2185 this as an approximation for the destination being like a local
2186 delivery, as in LMTP.
2188 I suspect this new check makes the old local delivery cutoff check
2189 redundant, but I have not verified this so I left the code in place.
2191 TF/02 Correct gecos expansion when From: is a prefix of the username.
2193 Test 0254 submits a message to Exim with the header
2197 When I ran the test suite under the user fanf2, Exim expanded
2198 the header to contain my full name, whereas it should have added
2199 a Resent-Sender: header. It erroneously treats any prefix of the
2200 username as equal to the username.
2202 This change corrects that bug.
2204 GF/01 DCC debug and logging tidyup
2205 Error conditions log to paniclog rather than rejectlog.
2206 Debug lines prefixed by "DCC: " to remove any ambiguity.
2208 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2210 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2211 Bug spotted by Jeremy Harris; was flawed since initial commit.
2212 Would have resulted in OCSP responses post-SNI triggering an Exim
2213 NULL dereference and crash.
2215 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2217 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2218 Bug detection, analysis and fix by Samuel Thibault.
2219 Bugzilla 1331, Debian bug #698092.
2221 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2223 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2224 Server implementation by Todd Lyons, client by JH.
2225 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2226 config variable "prdr_enable" controls whether the server
2227 advertises the facility. If the client requests PRDR a new
2228 acl_data_smtp_prdr ACL is called once for each recipient, after
2229 the body content is received and before the acl_smtp_data ACL.
2230 The client is controlled by both of: a hosts_try_prdr option
2231 on the smtp transport, and the server advertisement.
2232 Default client logging of deliveries and rejections involving
2233 PRDR are flagged with the string "PRDR".
2235 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2236 fclose(). Diagnosis by Todd Lyons.
2238 PP/17 Update configure.default to handle IPv6 localhost better.
2239 Patch by Alain Williams (plus minor tweaks).
2242 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2243 This is now consistent with GnuTLS, and is now documented: the
2244 previous undocumented portable approach to treating the option as
2245 unset was to force an expansion failure. That still works, and
2246 an empty string is now equivalent.
2248 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2249 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2250 not performing validation itself.
2252 PP/20 Added force_command boolean option to pipe transport.
2253 Patch from Nick Koston, of cPanel Inc.
2255 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2258 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2260 PP/21 Fix eximon continuous updating with timestamped log-files.
2261 Broken in a format-string cleanup in 4.80, missed when I repaired the
2262 other false fix of the same issue.
2263 Report and fix from Heiko Schlichting.
2266 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2267 Report from Prashanth Katuri.
2269 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2270 It's SecureTransport, so affects any MacOS clients which use the
2271 system-integrated TLS libraries, including email clients.
2273 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2274 using a MIME ACL for non-SMTP local injection.
2275 Report and assistance in diagnosis by Warren Baker.
2277 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2279 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2281 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2282 last id that failed so it may be referenced in subsequent ACL's.
2284 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2287 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2288 ldap library initialization, allowing self-signed CA's to be
2289 used. Also properly sets require_cert option later in code by
2290 using NULL (global ldap config) instead of ldap handle (per
2291 session). Bug diagnosis and testing by alxgomz.
2293 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2294 the src/util/ subdirectory.
2296 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2297 renamed to Transport Post Delivery Action by Jeremy Harris, as
2300 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2301 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2302 redis_servers = needs to be configured which will be used by the redis
2303 lookup. Patch from Warren Baker, of The Packet Hub.
2305 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2307 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2308 hostname or reverse DNS when processing a host list. Used suggestions
2309 from multiple comments on this bug.
2311 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2313 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2314 Missed a few lines, added it to make the runtest require no keyboard
2317 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2318 contains upper case chars. Make router use caseful_local_part.
2320 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2321 support when GnuTLS has been built with p11-kit.
2327 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2329 This, or similar/improved, will also be change PP/11 of 4.82.
2335 PP/01 Handle short writes when writing local log-files.
2336 In practice, only affects FreeBSD (8 onwards).
2337 Bugzilla 1053, with thanks to Dmitry Isaikin.
2339 NM/01 Bugzilla 949 - Documentation tweak
2341 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2344 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2346 PP/02 Implemented gsasl authenticator.
2348 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2350 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2351 `pkg-config foo` for cflags/libs.
2353 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2354 with rest of GSASL and with heimdal_gssapi.
2356 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2357 `pkg-config foo` for cflags/libs for the TLS implementation.
2359 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2360 properties get this fed in as external SSF. A number of robustness
2361 and debugging improvements to the cyrus_sasl authenticator.
2363 PP/08 cyrus_sasl server now expands the server_realm option.
2365 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2366 Patch by Jeremy Harris.
2368 PP/10 Added dbmjz lookup type.
2370 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2372 PP/12 MAIL args handles TAB as well as SP, for better interop with
2373 non-compliant senders.
2374 Analysis and variant patch by Todd Lyons.
2376 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2377 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2378 Patch from Dirk Mueller <dmueller@suse.com>
2380 PP/13 tls_peerdn now print-escaped for spool files.
2381 Observed some $tls_peerdn in wild which contained \n, which resulted
2382 in spool file corruption.
2384 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2385 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2386 or write after TLS renegotiation, which otherwise led to messages
2389 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2390 as a tracking header (ie: a signed header comes before the signature).
2391 Patch from Wolfgang Breyha.
2393 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2394 comma-sep list; embedded commas doubled.
2396 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2398 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2400 Report and patch from Dmitry Banschikov.
2402 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2403 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2404 failures. We appear to now support TLS1.1+ with Exim.
2406 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2407 lets Exim select keys and certificates based upon TLS SNI from client.
2408 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2409 before an outbound SMTP session. New log_selector, +tls_sni.
2411 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2412 NULL dereference. Report and patch from Alun Jones.
2414 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2415 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2416 Not seeing resolver debug output on NetBSD, but suspect this is a
2417 resolver implementation change.
2419 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2420 Left warnings. Added "eximon gdb" invocation mode.
2422 PP/21 Defaulting "accept_8bitmime" to true, not false.
2424 PP/22 Added -bw for inetd wait mode support.
2426 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2427 locate the relevant includes and libraries. Made this the default.
2429 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2430 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2432 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2433 This may cause build issues on older platforms.
2435 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2436 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2437 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2438 Added SNI support via GnuTLS too.
2439 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2441 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2443 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2444 Applied second patch from Janne, implementing suggestion to default
2445 multiple-strings-in-record handling to match SPF spec.
2447 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2449 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2450 read-only, out of scope).
2451 Patch from Wolfgang Breyha, report from Stuart Northfield.
2453 PP/29 Fix three issues highlighted by clang analyser static analysis.
2454 Only crash-plausible issue would require the Cambridge-specific
2455 iplookup router and a misconfiguration.
2456 Report from Marcin Mirosław.
2458 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2460 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2461 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2462 As part of this, removing so much warning spew let me fix some minor
2463 real issues in debug logging.
2465 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2466 assignment on my part. Fixed.
2468 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2469 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2470 Janne Snabb (who went above and beyond: thank you).
2472 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2473 string otherwise requires a connection and a bunch more work and it's
2474 relatively easy to get wrong. Should also expose TLS library linkage
2477 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2478 64-bit ${eval} (JH/03).
2480 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2481 GNU libc to support some of the 64-bit stuff, should not lead to
2482 conflicts. Defined before os.h is pulled in, so if a given platform
2483 needs to override this, it can.
2485 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2486 protection layer was required, which is not implemented.
2487 Bugzilla 1254, patch from Wolfgang Breyha.
2489 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2490 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2491 tls_dhparam take prime identifiers. Also unbreak combination of
2492 OpenSSL+DH_params+TLSSNI.
2494 PP/39 Disable SSLv2 by default in OpenSSL support.
2500 PP/01 Solaris build fix for Oracle's LDAP libraries.
2501 Bugzilla 1109, patch from Stephen Usher.
2503 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2505 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2508 TF/02 Fix a couple more cases where we did not log the error message
2509 when unlink() failed. See also change 4.74-TF/03.
2511 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2512 lock up or crash if it happened to be inside a call to libc when it
2513 got a SIGUSR1 from exiwhat.
2515 The SIGUSR1 handler appends the current process status to the process
2516 log which is later printed by exiwhat. It used to use the general
2517 purpose logging code to do this, but several functions it calls are
2518 not safe for signals.
2520 The new output code in the SIGUSR1 handler is specific to the process
2521 log, and simple enough that it's easy to inspect for signal safety.
2522 Removing some special cases also simplifies the general logging code.
2523 Removing the spurious timestamps from the process log simplifies
2526 TF/04 Improved ratelimit ACL condition.
2528 The /noupdate option has been deprecated in favour of /readonly which
2529 has clearer semantics. The /leaky, /strict, and /readonly update modes
2530 are mutually exclusive. The update mode is no longer included in the
2531 database key; it just determines when the database is updated. (This
2532 means that when you upgrade Exim will forget old rate measurements.)
2534 Exim now checks that the per_* options are used with an update mode that
2535 makes sense for the current ACL. For example, when Exim is processing a
2536 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2537 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2538 must specify per_mail/readonly. If you omit the update mode it defaults to
2539 /leaky where that makes sense (as before) or /readonly where required.
2541 The /noupdate option is now undocumented but still supported for
2542 backwards compatibility. It is equivalent to /readonly except that in
2543 ACLs where /readonly is required you may specify /leaky/noupdate or
2544 /strict/noupdate which are treated the same as /readonly.
2546 A useful new feature is the /count= option. This is a generalization
2547 of the per_byte option, so that you can measure the throughput of other
2548 aggregate values. For example, the per_byte option is now equivalent
2549 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2551 The per_rcpt option has been generalized using the /count= mechanism
2552 (though it's more complicated than the per_byte equivalence). When it is
2553 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2554 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2555 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2556 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2557 Note that using per_rcpt with a non-readonly update mode in more than
2558 one ACL will cause the recipients to be double-counted. (The per_mail
2559 and per_byte options don't have this problem.)
2561 The handling of very low rates has changed slightly. If the computed rate
2562 is less than the event's count (usually one) then this event is the first
2563 after a long gap. In this case the rate is set to the same as this event's
2564 count, so that the first message of a spam run is counted properly.
2566 The major new feature is a mechanism for counting the rate of unique
2567 events. The new per_addr option counts the number of different
2568 recipients that someone has sent messages to in the last time period. It
2569 behaves like per_rcpt if all the recipient addresses are different, but
2570 duplicate recipient addresses do not increase the measured rate. Like
2571 the /count= option this is a general mechanism, so the per_addr option
2572 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2573 example, measure the rate that a client uses different sender addresses
2574 with the options per_mail/unique=$sender_address. There are further
2575 details in the main documentation.
2577 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2579 TF/06 Removed a few PCRE remnants.
2581 TF/07 Automatically extract Exim's version number from tags in the git
2582 repository when doing development or release builds.
2584 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2585 Bugzilla 879. Patch from Paul Fisher.
2587 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2588 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2591 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2593 PP/05 Variable $av_failed, true if the AV scanner deferred.
2594 Bugzilla 1078. Patch from John Horne.
2596 PP/06 Stop make process more reliably on build failure.
2597 Bugzilla 1087. Patch from Heiko Schlittermann.
2599 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2600 Bugzilla 1089. Patch from Heiko Schlittermann.
2602 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2603 Bugzilla 1131. Patch from Holger Weiß.
2605 PP/09 Handle IPv6 addresses with SPF.
2606 Bugzilla 860. Patch from Wolfgang Breyha.
2608 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2610 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2613 PP/11 match_* no longer expand right-hand-side by default.
2614 New compile-time build option, EXPAND_LISTMATCH_RHS.
2615 New expansion conditions, "inlist", "inlisti".
2617 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2619 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2621 PP/14 fix log_write() format string regression from TF/03.
2622 Bugzilla 1152. Patch from Dmitry Isaikin.
2628 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2630 PP/02 Harmonised TLS library version reporting; only show if debugging.
2631 Layout now matches that introduced for other libraries in 4.74 PP/03.
2633 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2635 PP/04 New "dns_use_edns0" global option.
2637 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2640 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2641 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2643 TK/01 Updated PolarSSL code to 0.14.2.
2644 Bugzilla 1097. Patch from Andreas Metzler.
2646 PP/07 Catch divide-by-zero in ${eval:...}.
2647 Fixes bugzilla 1102.
2649 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2652 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2653 format-string attack -- SECURITY: remote arbitrary code execution.
2655 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2656 time unintentionally subject to list matching rules, letting the header
2657 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2658 arbitrary string expansion). This allowed for information disclosure.
2660 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2661 INT_MIN/-1 -- value coerced to INT_MAX.
2667 NM/01 Workaround for PCRE version dependency in version reporting
2670 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2671 This fixes portability to compilers other than gcc, notably
2672 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2674 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2675 makefiles for portability to HP-UX and POSIX correctness.
2677 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2678 Also via indented variable definition in the Makefile.
2679 (Debugging by Oliver Heesakkers).
2681 PP/02 Restore caching of spamd results with expanded spamd_address.
2682 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2684 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2685 Improves build reliability. Fix from: Frank Elsner
2687 NM/02 Fix wide character breakage in the rfc2047 coding
2688 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2690 NM/03 Allow underscore in dnslist lookups
2691 Fixes bug 1026. Patch from Graeme Fowler
2693 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2694 Code patches from Adam Ciarcinski of NetBSD.
2696 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2699 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2700 is logged, to avoid truncation. Patch from John Horne.
2702 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2703 Patch from Jakob Hirsch.
2705 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2706 SQL string expansion failure details.
2707 Patch from Andrey Oktyabrski.
2709 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2710 Patch from Simon Arlott.
2712 PP/09 New lookups functionality failed to compile on old gcc which rejects
2713 extern declarations in function scope.
2714 Patch from Oliver Fleischmann
2716 PP/10 Use sig_atomic_t for flags set from signal handlers.
2717 Check getgroups() return and improve debugging.
2718 Fixed developed for diagnosis in bug 927 (which turned out to be
2721 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2722 Patch from Mark Zealey.
2724 PP/12 Bugzilla 1056: Improved spamd server selection.
2725 Patch from Mark Zealey.
2727 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2728 Based on patch from Heiko Schlittermann.
2730 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2731 Patch from Uwe Doering, sign-off by Michael Haardt.
2733 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2734 variable declaration deep within a block. Bug and patch from
2737 PP/15 lookups-Makefile IRIX compatibility coercion.
2739 PP/16 Make DISABLE_DKIM build knob functional.
2741 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2742 Patch by Simon Arlott
2744 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2745 variable argument macros. Our copy now differs from upstream.
2751 TF/01 Failure to get a lock on a hints database can have serious
2752 consequences so log it to the panic log.
2754 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2755 controlled using the smtp_confirmation log selector.
2757 TF/03 Include the error message when we fail to unlink a spool file.
2759 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2760 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2761 for maintaining out-of-tree patches for some time.
2763 PP/01 Bugzilla 139: Documentation and portability issues.
2764 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2765 Handle per-OS dynamic-module compilation flags.
2767 PP/02 Let /dev/null have normal permissions.
2768 The 4.73 fixes were a little too stringent and complained about the
2769 permissions on /dev/null. Exempt it from some checks.
2770 Reported by Andreas M. Kirchwitz.
2772 PP/03 Report version information for many libraries, including
2773 Exim version information for dynamically loaded libraries. Created
2774 version.h, now support a version extension string for distributors
2775 who patch heavily. Dynamic module ABI change.
2777 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2778 privilege escalation vulnerability whereby the Exim run-time user
2779 can cause root to append content of the attacker's choosing to
2782 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2785 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2786 If dropping privileges for untrusted macros, we disabled normal logging
2787 on the basis that it would fail; for the Exim run-time user, this is not
2788 the case, and it resulted in successful deliveries going unlogged.
2789 Fixed. Reported by Andreas Metzler.
2795 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2796 only prepend for the Resent-* case. Fixes regression introduced in
2797 Exim 4.70 by NM/22 for Bugzilla 607.
2799 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2800 increasing numbers of administrators be bitten by this.
2802 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2804 PP/03 Bugzilla 994: added openssl_options main configuration option.
2806 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2808 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2810 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2812 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2813 without a peer certificate, leading to a segfault because of an
2814 assumption that peers always have certificates. Be a little more
2815 paranoid. Problem reported by Martin Tscholak.
2817 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2818 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2819 NB: ClamAV planning to remove STREAM in "middle of 2010".
2820 CL also introduces -bmalware, various -d+acl logging additions and
2821 more caution in buffer sizes.
2823 PP/09 Implemented reverse_ip expansion operator.
2825 PP/10 Bugzilla 937: provide a "debug" ACL control.
2827 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2829 PP/12 Bugzilla 973: Implement --version.
2831 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2833 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2835 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2837 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2838 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2839 ignore trailing whitespace.
2841 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2843 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2846 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2847 Notification from Dr Andrew Aitchison.
2849 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2850 ExtendedDetectionInfo response format.
2851 Notification from John Horne.
2853 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2856 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2857 XSL and documented dependency on system catalogs, with examples of how
2860 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2863 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2864 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2865 configuration file which is writeable by the Exim user or group.
2867 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2868 of configuration files to cover files specified with the -C option if
2869 they are going to be used with root privileges, not just the default
2872 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2873 option (effectively making it always true).
2875 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2876 files to be used while preserving root privileges.
2878 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2879 that rogue child processes cannot use them.
2881 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2882 run-time user, instead of root.
2884 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2885 Exim run-time user without dropping privileges.
2887 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2888 result string, instead of calling string_vformat() twice with the same
2891 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2892 for other users. Others should always drop root privileges if they use
2893 -C on the command line, even for a whitelisted configure file.
2895 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2897 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2903 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2904 $data_path, and $header_path variables; fixed documentation bugs and
2907 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2908 exipick to access non-standard spools, including the "frozen" queue
2911 NM/01 Bugzilla 965: Support mysql stored procedures.
2912 Patch from Alain Williams
2914 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2916 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2917 Patch from Andreas Metzler
2919 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2920 Patch from Kirill Miazine
2922 NM/05 Bugzilla 671: Added umask to procmail example.
2924 JJ/03 installed exipick 20100323.0, fixing doc bug
2926 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2927 directory. Notification and patch from Dan Rosenberg.
2929 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2931 TK/02 Improve log output when DKIM signing operation fails.
2933 MH/01 Treat the transport option dkim_domain as a colon separated
2934 list, not as a single string, and sign the message with each element,
2935 omitting multiple occurences of the same signer.
2937 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2938 Bugzilla 985, 986. Patch by Simon Arlott
2940 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2941 Patch by Simon Arlott
2943 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2944 MBX locking. Notification from Dan Rosenberg.
2950 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2952 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2954 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2956 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2958 NM/04 Bugzilla 915: Flag broken perl installation during build.
2964 TK/01 Added patch by Johannes Berg that expands the main option
2965 "spamd_address" if it starts with a dollar sign.
2967 TK/02 Write list of recipients to X-Envelope-Sender header when building
2968 the mbox-format spool file for content scanning (suggested by Jakob
2971 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2972 (http://www.dcc-servers.net/) support via dccifd. Activated by
2973 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2975 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2976 by Mark Daniel Reidel <mr@df.eu>.
2978 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2979 When building exim an external PCRE library is now needed -
2980 PCRE is a system library on the majority of modern systems.
2981 See entry on PCRE_LIBS in EDITME file.
2983 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2984 conversation. Added nologin parameter to request.
2985 Patch contributed by Kirill Miazine.
2987 TF/01 Do not log submission mode rewrites if they do not change the address.
2989 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2991 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2992 log files in place. Contributed by Roberto Lima.
2994 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2996 TF/03 Bugzilla 615: When checking the local_parts router precondition
2997 after a local_part_suffix or local_part_prefix option, Exim now
2998 does not use the address's named list lookup cache, since this
2999 contains cached lookups for the whole local part.
3001 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3002 Robert Millan. Documentation is in experimental-spec.txt.
3004 TF/04 Bugzilla 668: Fix parallel build (make -j).
3006 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3008 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3009 Patch provided by Jan Srzednicki.
3011 TF/05 Leading white space used to be stripped from $spam_report which
3012 wrecked the formatting. Now it is preserved.
3014 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3015 that they are available at delivery time.
3017 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3019 TF/08 TLS error reporting now respects the incoming_interface and
3020 incoming_port log selectors.
3022 TF/09 Produce a more useful error message if an SMTP transport's hosts
3023 setting expands to an empty string.
3025 NM/06 Bugzilla 744: EXPN did not work under TLS.
3026 Patch provided by Phil Pennock.
3028 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3029 Patch provided by Richard Godbee.
3031 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3032 acl_smtp_notquit, added index entry.
3034 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3035 Patch provided by Eugene Bujak.
3037 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3038 accept(). Patch provided by Maxim Dounin.
3040 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3041 Patch provided by Phil Pennock.
3043 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3045 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3046 Patch provided by Brad "anomie" Jorsch.
3048 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3049 Patch provided by Dean Brooks.
3051 TK/05 Add native DKIM support (does not depend on external libraries).
3053 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3054 Patch provided by Graeme Fowler.
3056 NM/16 Bugzilla 851: Documentation example syntax fix.
3058 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3060 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3063 NM/19 Bugzilla 745: TLS version reporting.
3064 Patch provided by Phil Pennock.
3066 NM/20 Bugzilla 167: bool: condition support.
3067 Patch provided by Phil Pennock.
3069 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3070 clients. Patch provided by Phil Pennock.
3072 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3073 Patch provided by Brad "anomie" Jorsch.
3075 NM/23 Bugzilla 687: Fix misparses in eximstats.
3076 Patch provided by Heiko Schlittermann.
3078 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3079 Patch provided by Heiko Schlittermann.
3081 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3082 plus update to original patch.
3084 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3086 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3087 Patch provided by David Brownlee.
3089 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3091 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3093 NM/30 Bugzilla 888: TLS documentation bugfixes.
3095 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3097 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3098 Unlike the original bugzilla I have changed all shell scripts in src tree.
3100 NM/33 Bugzilla 898: Transport filter timeout fix.
3101 Patch by Todd Rinaldo.
3103 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3104 Patch by Serge Demonchaux.
3106 NM/35 Bugzilla 39: Base64 decode bug fixes.
3107 Patch by Jakob Hirsch.
3109 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3111 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3113 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3115 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3121 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3122 ALT-N's libdkim that I have put here:
3123 http://duncanthrax.net/exim-experimental/
3125 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3126 were called 'true' and it seems that C99 defines that as a reserved
3127 keyword to be used with 'bool' variable types. That means you could
3128 not include C99-style headers which use bools without triggering
3129 build errors in sieve.c.
3131 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3132 as mailq or other aliases. Changed the --help handling significantly
3133 to do whats expected. exim_usage() emits usage/help information.
3135 SC/01 Added the -bylocaldomain option to eximstats.
3137 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3139 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3141 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3147 PH/01 Another patch from the Sieve maintainer.
3149 PH/02 When an IPv6 address is converted to a string for single-key lookup
3150 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3151 dots are used instead of colons so that keys in lsearch files need not
3152 contain colons. This was done some time before quoting was made available
3153 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3154 (notated using the quote facility) so as to distinguish them from IPv4
3155 keys. This meant that lookups for IP addresses in host lists did not work
3156 for iplsearch lookups.
3158 This has been fixed by arranging for IPv6 addresses to be expressed with
3159 colons if the lookup type is iplsearch. This is not incompatible, because
3160 previously such lookups could never work.
3162 The situation is now rather anomalous, since one *can* have colons in
3163 ordinary lsearch keys. However, making the change in all cases is
3164 incompatible and would probably break a number of configurations.
3166 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3169 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3170 conversion specification without a maximum field width, thereby enabling
3171 a rogue spamd server to cause a buffer overflow. While nobody in their
3172 right mind would setup Exim to query an untrusted spamd server, an
3173 attacker that gains access to a server running spamd could potentially
3174 exploit this vulnerability to run arbitrary code as the Exim user.
3176 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3177 $primary_hostname instead of what libspf2 thinks the hosts name is.
3179 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3180 a directory entry by the name of the lookup key. Previously, if a
3181 symlink pointed to a non-existing file or a file in a directory that
3182 Exim lacked permissions to read, a lookup for a key matching that
3183 symlink would fail. Now it is enough that a matching directory entry
3184 exists, symlink or not. (Bugzilla 503.)
3186 PH/03 The body_linecount and body_zerocount variables are now exported in the
3189 PH/04 Added the $dnslist_matched variable.
3191 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3192 This means they are set thereafter only if the connection becomes
3195 PH/06 Added the client_condition to authenticators so that some can be skipped
3196 by clients under certain conditions.
3198 PH/07 The error message for a badly-placed control=no_multiline_responses left
3199 "_responses" off the end of the name.
3201 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3203 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3204 (without spaces) instead of just copying the configuration text.
3206 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3208 PH/11 Added $max_received_linelength.
3210 PH/12 Added +ignore_defer and +include_defer to host lists.
3212 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3213 way in which PCRE > 7.0 is built.
3215 PH/14 Implemented queue_only_load_latch.
3217 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3218 MAIL command. The effect was to mangle the value on 64-bit systems.
3220 PH/16 Another patch from the Sieve maintainer.
3222 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3224 PH/18 If a system quota error occurred while trying to create the file for
3225 a maildir delivery, the message "Mailbox is full" was not appended to the
3226 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3227 only to a quota excession during the actual writing of the file.
3229 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3230 characters?) which causes problems in log lines. The DN values are now
3231 passed through string_printing() before being added to log lines.
3233 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3234 and InterBase are left for another time.)
3236 PH/21 Added message_body_newlines option.
3238 PH/22 Guard against possible overflow in moan_check_errorcopy().
3240 PH/23 POSIX allows open() to be a macro; guard against that.
3242 PH/24 If the recipient of an error message contained an @ in the local part
3243 (suitably quoted, of course), incorrect values were put in $domain and
3244 $local_part during the evaluation of errors_copy.
3250 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3251 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3254 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3255 issue a MAIL command.
3257 PH/02 In an ACL statement such as
3259 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3261 if a client was not listed at all, or was listed with a value other than
3262 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3263 the condition was not true (as it should be), so access was not denied.
3264 The bug was that the ! inversion was incorrectly passed on to the second
3265 item. This has been fixed.
3267 PH/03 Added additional dnslists conditions == and =& which are different from
3268 = and & when the dns lookup returns more than one IP address.
3270 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3271 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3273 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3274 FSYNC, which compiles an option called disable_fsync that allows for
3275 bypassing fsync(). The documentation is heavily laced with warnings.
3277 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3279 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3280 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3281 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3282 including adding "make clean"; (3) Added -fPIC when compiling the test
3283 dynamically loaded module, to get rid of a warning.
3285 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3286 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3287 = 0s. The bug is otherwise harmless.
3289 PH/07 There was a bug in the dovecot authenticator such that the value of
3290 $auth1 could be overwritten, and so not correctly preserved, after a
3291 successful authentication. This usually meant that the value preserved by
3292 the server_setid option was incorrect.
3294 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3296 PH/09 Installed PCRE release 7.0.
3298 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3299 run for batched SMTP input. It is now run at the start of every message
3300 in the batch. While fixing this I discovered that the process information
3301 (output by running exiwhat) was not always getting set for -bs and -bS
3302 input. This is fixed, and it now also says "batched" for BSMTP.
3304 PH/11 Added control=no_pipelining.
3306 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3307 patch, slightly modified), and move the expansion of helo_data till after
3308 the connection is made in the smtp transport (so it can use these
3311 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3313 PH/14 Added log_selector = +pid.
3315 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3317 PH/16 Add ${if forany and ${if forall.
3319 PH/17 Added dsn_from option to vary the From: line in DSNs.
3321 PH/18 Flush SMTP output before performing a callout, unless control =
3322 no_callout_flush is set.
3324 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3325 was true (the default) a successful delivery failed to delete the retry
3326 item, thus causing premature timeout of the address. The bug is now
3329 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3331 PH/21 Long custom messages for fakedefer and fakereject are now split up
3332 into multiline responses in the same way that messages for "deny" and
3333 other ACL rejections are.
3335 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3336 with slight modification.
3338 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3339 draft, changing the syntax and factoring some duplicate code".
3341 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3342 for deliveries of the second and subsequent messages over the same SMTP
3345 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3346 ${reduce, with only minor "tidies".
3348 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3350 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3351 expansion side effects.
3353 PH/27 When a message times out after an over-quota error from an Exim-imposed
3354 quota, the bounce message says "mailbox is full". This message was not
3355 being given when it was a system quota that was exceeded. It now should
3358 MH/03 Made $recipients available in local_scan(). local_scan() already has
3359 better access to the recipient list through recipients_list[], but
3360 $recipients can be useful in postmaster-provided expansion strings.
3362 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3363 in the case of a MAIL command with additional options following the
3364 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3365 were accidentally chopped off.
3367 PH/29 SMTP synchronization checks are implemented when a command is read -
3368 there is a check that no more input is waiting when there shouldn't be
3369 any. However, for some commands, a delay in an ACL can mean that it is
3370 some time before the response is written. In this time, more input might
3371 arrive, invalidly. So now there are extra checks after an ACL has run for
3372 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3373 pipelining has not been advertised.
3375 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3377 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3378 returned in the SMTP response when smtp_return_error_details was false.
3379 This has been fixed.
3381 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3382 instead of the C I/O that was originally supplied, because problems were
3383 reported on Solaris.
3385 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3386 Exim which did not show up earlier: it was assuming that a call to
3387 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3388 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3389 was a macro that became an assignment, so it seemed to work. This has
3390 changed to a proper function call with a void return, hence the compile
3391 error. Exim's code has been fixed.
3393 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3396 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3398 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3400 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3401 because it turns out to be too big or there is a timeout) there is no
3402 "Completed" line in the log. When some messages of this type were
3403 selected by exigrep, they were listed as "not completed". Others were
3404 picked up by some special patterns. I have improved the selection
3405 criteria to be more general.
3407 PH/38 The host_find_failed option in the manualroute router can now be set
3408 to "ignore", to completely ignore a host whose IP address cannot be
3409 found. If all hosts are ignored, the behaviour is controlled by the new
3410 host_all_ignored option.
3412 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3413 homing or because of multiple MX records with /mx) generated more than
3414 one IP address, and the following item turned out to be the local host,
3415 all the secondary addresses of the first item were incorrectly removed
3416 from the list, along with the local host and any following hosts (which
3417 is what is supposed to happen).
3419 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3420 whoever called Exim into the -H file. In the case of the daemon it was
3421 behaving confusingly. When first started, it used values for whoever
3422 started the daemon, but after a SIGHUP it used the Exim user (because it
3423 calls itself on a restart). I have changed the code so that it now always
3426 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3427 message are rejected with the same error (e.g. no authentication or bad
3428 sender address), and a DATA command is nevertheless sent (as can happen
3429 with PIPELINING or a stupid MUA), the error message that was given to the
3430 RCPT commands is included in the rejection of the DATA command. This is
3431 intended to be helpful for MUAs that show only the final error to their
3434 PH/42 Another patch from the Sieve maintainer.
3436 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3437 Eximstats - Fixed some broken HTML links and added missing column headers
3439 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3440 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3442 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3444 PH/43 Yet another patch from the Sieve maintainer.
3446 PH/44 I found a way to check for a TCP/IP connection going away before sending
3447 the response to the final '.' that terminates a message, but only in the
3448 case where the client has not sent further data following the '.'
3449 (unfortunately, this is allowed). However, in many cases there won't be
3450 any further data because there won't be any more messages to send. A call
3451 to select() can be used: if it shows that the input is "ready", there is
3452 either input waiting, or the socket has been closed. An attempt to read
3453 the next input character can distinguish the two cases. Previously, Exim
3454 would have sent an OK response which the client would never have see.
3455 This could lead to message repetition. This fix should cure that, at
3456 least in a lot of common cases.
3458 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3459 advertised in response to EHLO.
3465 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3466 fixed by 4.65/MH/01 (is this a record?) are fixed:
3468 (i) An empty string was always treated as zero by the numeric comparison
3469 operators. This behaviour has been restored.
3471 (ii) It is documented that the numeric comparison operators always treat
3472 their arguments as decimal numbers. This was broken in that numbers
3473 starting with 0 were being interpreted as octal.
3475 While fixing these problems I realized that there was another issue that
3476 hadn't been noticed. Values of message_size_limit (both the global option
3477 and the transport option) were treated as octal if they started with 0.
3478 The documentation was vague. These values are now always treated as
3479 decimal, and I will make that clear in the documentation.
3485 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3486 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3489 MH/01 Don't check that the operands of numeric comparison operators are
3490 integers when their expansion is in "skipping" mode (fixes bug
3491 introduced by 4.64-PH/07).
3493 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3494 child addresses, Exim now panics and dies. Previously, because the count
3495 is held in a short int, deliveries were likely to be lost. As such a
3496 large number of recipients for a single message is ridiculous
3497 (performance will be very, very poor), I have chosen to impose a limit
3498 rather than extend the field.
3504 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3505 leftover -K file (the existence of which was triggered by #402).
3506 While we were at it, introduced process PID as part of the -K
3507 filename. This should rule out race conditions when creating
3510 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3511 processing considerably. Previous code took too long for large mails,
3512 triggering a timeout which in turn triggers #401.
3514 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3515 in the DK code in transports.c. sendfile() is not really portable,
3516 hence the _LINUX specificness.
3518 TF/01 In the add_headers option to the mail command in an Exim filter,
3519 there was a bug that Exim would claim a syntax error in any
3520 header after the first one which had an odd number of characters
3523 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3524 callout verification, Exim cached a "reject" for the entire domain. This
3525 is correct for most verifications, but it is not correct for a recipient
3526 verification with use_sender or use_postmaster set, because in that case
3527 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3528 case of MAIL FROM:<> rejection from other early rejections (e.g.
3529 rejection of HELO). When verifying a recipient using a non-null MAIL
3530 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3531 Whatever the result of the callout, the value of the domain cache is
3532 left unchanged (for any other kind of callout, getting as far as trying
3533 RCPT means that the domain itself is ok).
3535 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3538 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3539 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3540 session, the daemon ignores ECONNECT errors and logs others; it now
3541 ignores EPIPE as well.
3543 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3544 (quoted-printable decoding).
3546 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3547 later the small subsequent patch to fix an introduced bug.
3549 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3551 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3553 PH/08 An error is now given if message_size_limit is specified negative.
3555 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3556 to be given (somewhat) arbitrary names.
3558 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3561 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3562 miscellaneous code fixes
3564 PH/10 Added the log_reject_target ACL modifier to specify where to log
3567 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3568 hostname. This is wrong, because it relates to the incoming message (and
3569 probably the interface on which it is arriving) and not to the outgoing
3570 callout (which could be using a different interface). This has been
3571 changed to use the value of the helo_data option from the smtp transport
3572 instead - this is what is used when a message is actually being sent. If
3573 there is no remote transport (possible with a router that sets up host
3574 addresses), $smtp_active_hostname is used.
3576 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3577 tweaks were necessary in order to get it to work (see also 21 below):
3578 (a) The code assumed that strncpy() returns a negative number on buffer
3579 overflow, which isn't the case. Replaced with Exim's string_format()
3581 (b) There were several signed/unsigned issues. I just did the minimum
3582 hacking in of casts. There is scope for a larger refactoring.
3583 (c) The code used strcasecmp() which is not a standard C function.
3584 Replaced with Exim's strcmpic() function.
3585 (d) The code set only $1; it now sets $auth1 as well.
3586 (e) A simple test gave the error "authentication client didn't specify
3587 service in request". It would seem that Dovecot has changed its
3588 interface. Fortunately there's a specification; I followed it and
3589 changed what the client sends and it appears to be working now.
3591 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3594 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3595 address A is aliased to B and C, where B exists and C does not. Without
3596 -v the output is "A verified" because verification stops after a
3597 successful redirection if more than one address is generated. However,
3598 with -v the child addresses are also verified. Exim was outputting "A
3599 failed to verify" and then showing the successful verification for C,
3600 with its parentage. It now outputs "B failed to verify", showing B's
3601 parentage before showing the successful verification of C.
3603 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3604 look up a TXT record in a specific list after matching in a combined
3607 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3608 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3609 they consult the DNS. I had assumed they would set it the way they
3610 wanted; and indeed my experiments on Linux seem to show that in some
3611 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3612 To be on the safe side, however, I have now made the interface to
3613 host_find_byname() similar to host_find_bydns(), with an argument
3614 containing the DNS resolver options. The host_find_byname() function now
3615 sets these options at its start, just as host_find_bydns() does. The smtp
3616 transport options dns_qualify_single and dns_search_parents are passed to
3617 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3618 of host_find_byname() use the default settings of RES_DEFNAMES
3619 (qualify_single) but not RES_DNSRCH (search_parents).
3621 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3622 spool_read_header() do less string testing, by means of a preliminary
3623 switch on the second character of optional "-foo" lines. (This is
3624 overdue, caused by the large number of possibilities that now exist.
3625 Originally there were few.) While I was there, I also converted the
3626 str(n)cmp tests so they don't re-test the leading "-" and the first
3627 character, in the hope this might squeeze out yet more improvement.
3629 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3630 flag allowing group syntax was set by the header_syntax check but not
3631 turned off, possible causing trouble later; (2) The flag was not being
3632 set at all for the header_verify test, causing "group"-style headers to
3633 be rejected. I have now set it in this case, and also caused header_
3634 verify to ignore an empty address taken from a group. While doing this, I
3635 came across some other cases where the code for allowing group syntax
3636 while scanning a header line wasn't quite right (mostly, not resetting
3637 the flag correctly in the right place). These bugs could have caused
3638 trouble for malformed header lines. I hope it is now all correct.
3640 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3641 with the "reply" argument non-NULL. The code, however (which originally
3642 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3643 but it didn't always do it. This confused somebody who was copying the
3644 code for some other use. I have removed all the tests.
3646 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3647 feature that was used to support insecure browsers during the U.S. crypto
3648 embargo. It requires special client support, and Exim is probably the
3649 only MTA that supported it -- and would never use it because real RSA is
3650 always available. This code has been removed, because it had the bad
3651 effect of slowing Exim down by computing (never used) parameters for the
3652 RSA_EXPORT functionality.
3654 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3655 authenticator to fail if there's a tab character in the incoming data
3656 (there should never be unless someone is messing about, as it's supposed
3657 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3658 if the connection is using TLS or if the remote IP is the same as the
3659 local IP, and the "valid-client-cert option" if a client certificate has
3662 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3663 authenticators. This can be used for authorization after authentication
3664 succeeds. (In the case of plaintext, it servers for both authentication
3667 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3668 if any retry times were supplied.
3670 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3671 connection, where there is no client IP address to check. In this
3672 situation, the verify now always succeeds.
3674 PH/25 Applied John Jetmore's -Mset patch.
3676 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3678 PH/27 In a string expansion for a processed (not raw) header when multiple
3679 headers of the same name were present, leading whitespace was being
3680 removed from all of them, but trailing whitespace was being removed only
3681 from the last one. Now trailing whitespace is removed from each header
3682 before concatenation. Completely empty headers in a concatenation (as
3683 before) are ignored.
3685 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3686 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3688 PH/29 [Removed. This was a change that I later backed out, and forgot to
3689 correct the ChangeLog entry (that I had efficiently created) before
3690 committing the later change.]
3692 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3693 address errors (4xx response to RCPT) over the same connection as other
3694 messages routed to the same hosts. Such deliveries are always "forced",
3695 so retry times are not inspected. This resulted in far too many retries
3696 for the affected addresses. The effect occurred only when there were more
3697 hosts than the hosts_max_try setting in the smtp transport when it had
3698 the 4xx errors. Those hosts that it had tried were not added to the list
3699 of hosts for which the message was waiting, so if all were tried, there
3700 was no problem. Two fixes have been applied:
3702 (i) If there are any address or message errors in an SMTP delivery, none
3703 of the hosts (tried or untried) are now added to the list of hosts
3704 for which the message is waiting, so the message should not be a
3705 candidate for sending over the same connection that was used for a
3706 successful delivery of some other message. This seems entirely
3707 reasonable: after all the message is NOT "waiting for some host".
3708 This is so "obvious" that I'm not sure why it wasn't done
3709 previously. Hope I haven't missed anything, but it can't do any
3710 harm, as the worst effect is to miss an optimization.
3712 (ii) If, despite (i), such a delivery is accidentally attempted, the
3713 routing retry time is respected, so at least it doesn't keep
3714 hammering the server.
3716 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3717 in ${readsocket because some servers need this prod.
3719 PH/32 Added some extra debug output when updating a wait-xxx database.
3721 PH/33 The hint "could be header name not terminated by colon", which has been
3722 given for certain expansion errors for a long time, was not being given
3723 for the ${if def:h_colon_omitted{... case.
3725 PH/34 The spec says: "With one important exception, whenever a domain list is
3726 being scanned, $domain contains the subject domain." There was at least
3727 one case where this was not true.
3729 PH/35 The error "getsockname() failed: connection reset by peer" was being
3730 written to the panic log as well as the main log, but it isn't really
3731 panic-worthy as it just means the connection died rather early on. I have
3732 removed the panic log writing for the ECONNRESET error when getsockname()
3735 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3736 runs only) independently of the message's sender address. This meant
3737 that, if the 4xx error was in fact related to the sender, a different
3738 message to the same recipient with a different sender could confuse
3739 things. In particular, this can happen when sending to a greylisting
3740 server, but other circumstances could also provoke similar problems.
3741 I have changed the default so that the retry time for these errors is now
3742 based a combination of the sender and recipient addresses. This change
3743 can be overridden by setting address_retry_include_sender=false in the
3746 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3747 remote server are returned as part of bounce messages. This was not
3748 happening for LMTP over a pipe (the lmtp transport), but now it is the
3749 same for both kinds of LMTP.
3751 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3752 in header lines that were in fact CNAMEs. This is no longer the case.
3754 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3755 and queue runs started by the daemon processed all messages. This has
3756 been fixed so that -R and -S can now usefully be given with -q<time>.
3758 PH/40 Import PCRE release 6.7 (fixes some bugs).
3760 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3762 PH/42 Give an error if -q is specified more than once.
3764 PH/43 Renamed the variables $interface_address and $interface_port as
3765 $received_ip_address and $received_port, to make it clear that these
3766 values apply to message reception, and not to the outgoing interface when
3767 a message is delivered. (The old names remain recognized, of course.)
3769 PH/44 There was no timeout on the connect() call when using a Unix domain
3770 socket in the ${readsocket expansion. There now is.
3772 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3773 be meaningful with "accept".
3775 SC/01 Eximstats V1.43
3776 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3778 SC/02 Eximstats V1.44
3779 Use a glob alias rather than an array ref in the generated
3780 parser. This improves both readability and performance.
3782 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3783 Collect SpamAssassin and rejection statistics.
3784 Don't display local sender or destination tables unless
3785 there is data to show.
3786 Added average volumes into the top table text output.
3788 SC/04 Eximstats V1.46
3789 Collect data on the number of addresses (recipients)
3790 as well as the number of messages.
3792 SC/05 Eximstats V1.47
3793 Added 'Message too big' to the list of mail rejection
3794 reasons (thanks to Marco Gaiarin).
3796 SC/06 Eximstats V1.48
3797 Mainlog lines which have GMT offsets and are too short to
3798 have a flag are now skipped.
3800 SC/07 Eximstats V1.49 (Alain Williams)
3801 Added the -emptyok flag.
3803 SC/08 Eximstats V1.50
3804 Fixes for obtaining the IP address from reject messages.
3806 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3807 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3808 whitespace changes from 4.64-PH/27
3810 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3813 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3814 are found, allow negative numbers in numeric criteria)
3816 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3818 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3819 to match changes made in 4.64-PH/43
3821 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3823 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3824 "domain" is actually an IP address, and force a failure. This locks out
3825 those revolvers/nameservers that support "A-for-A" lookups, in
3826 contravention of the specifications.
3828 PH/48 When a host name was looked up from an IP address, and the subsequent
3829 forward lookup of the name timed out, the host name was left in
3830 $sender_host_name, contrary to the specification.
3832 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3833 restricted to single-key lookups, Exim was not diagnosing an error if
3834 * or *@ was used with a query-style lookup.
3836 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3838 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3839 long ago, but noone interested enough thought of it. Let's just say that
3840 the "1.1" means that there are some new functions that weren't there at
3841 some point in the past.
3843 PH/51 Error processing for expansion failure of helo_data from an smtp
3844 transport during callout processing was broken.
3846 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3847 tested/used via the -bh/-bhc/-bs options.
3849 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3850 bug, fixed in subsequent PCRE releases).
3852 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3853 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3855 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3861 SC/01 Use a glob alias rather than an array ref in eximstats generated
3862 parser. This improves both readability and performance.
3864 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3865 Don't display local sender or destination tables in eximstats unless
3866 there is data to show.
3867 Added average volumes into the eximstats top table text output.
3869 SC/03 Collect data on the number of addresses (recipients) as well
3870 as the number of messages in eximstats.
3872 TF/01 Correct an error in the documentation for the redirect router. Exim
3873 does (usually) call initgroups() when daemonizing.
3875 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3876 with consistent privilege compared to when running as a daemon.
3878 TF/03 Note in the spec that $authenticated_id is not set for local
3879 submissions from trusted users.
3881 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3882 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3884 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3885 by adding some example configuration directives to the default
3886 configuration file. A little bit of work is required to uncomment the
3887 directives and define how usernames and passwords are checked, but
3888 there is now a framework to start from.
3890 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3891 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3892 without this. I don't know how relevant this is to other LDAP libraries.
3894 PH/02 Add the verb name to the "unknown ACL verb" error.
3896 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3898 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3900 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3901 directory not expanded when it should be if an expanded home directory
3902 was set for the address (which is overridden by the transport).
3904 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3907 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3908 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3909 because it is too late at that time, and has no effect.
3911 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3912 security issue with \' (bugzilla #107). I could not use the
3913 PQescapeStringConn() function, because it needs a PGconn value as one of
3916 PH/08 When testing addresses using -bt, indicate those final addresses that
3917 are duplicates that would not cause an additional delivery. At least one
3918 person was confused, thinking that -bt output corresponded to deliveries.
3919 (Suppressing duplicates isn't a good idea as you lose the information
3920 about possibly different redirections that led to the duplicates.)
3922 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3923 systems where poll() doesn't work, in particular OS X.
3925 PH/10 Added more information to debugging output for retry time not reached.
3927 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3928 operations in malware.c.
3930 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3933 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3934 syslog_duplication set false, log lines that would normally be written
3935 both the the main log and to the reject log were not written to syslog at
3938 PH/14 In the default configuration, change the use of "message" in ACL warn
3939 statements to "add_header".
3941 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3942 not followed by a command (e.g. "seen endif").
3944 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3945 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3948 PH/17 Added extra conditions to the default value of delay_warning_condition
3952 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3953 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3954 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3957 The Auto-Submitted: and various List- headers are standardised, whereas I
3958 don't think Precedence: ever was.
3960 PH/18 Refactored debugging code in route_finduser() to show more information,
3961 in particular, the error code if getpwnam() issues one.
3963 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3964 This is apparently needed in addition to the PH/07 change above to avoid
3965 any possible encoding problems.
3967 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3968 but not after initializing Perl.
3970 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3971 output them only if debugging. By default they are written stderr,
3972 apparently, which is not desirable.
3974 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3977 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3980 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3982 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3983 authenticated or an ident call has been made. Suppress the default
3984 values for $authenticated_id and $authenticated_sender (but permit -oMai
3985 and -oMas) when testing with -bh.
3987 PH/24 Re-jigged the order of the tests in the default configuration so that the
3988 tests for valid domains and recipients precede the DNS black list and CSA
3989 tests, on the grounds that those ones are more expensive.
3991 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3992 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3993 This bug exists in every version of Exim that I still have, right back to
3996 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3997 However, an attempt to turn on case-sensitivity in a regex key by
3998 including (?-i) didn't work because the subject string was already
3999 lowercased, and the effects were non-intuitive. It turns out that a
4000 one-line patch can be used to allow (?-i) to work as expected.
4006 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4007 other effects) broke the use of negated acl sub-conditions.
4009 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4012 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4013 "Deny" causes Exim to reject the incoming connection with a 554 error.
4014 Unfortunately, if there is a major crisis, such as a disk failure,
4015 tcp-wrappers gives "deny", whereas what one would like would be some
4016 kind of temporary error. A kludge has been added to help with this.
4017 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4018 554 error is used if errno is still zero or contains ENOENT (which occurs
4019 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4022 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4024 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4025 errors. Otherwise a message that provokes a temporary error (when other
4026 messages do not) can cause a whole host to time out.
4028 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4029 addresses were routed directly to files or pipes from a redirect router.
4030 File deliveries just didn't batch; pipe deliveries might have suffered
4033 PH/06 A failure to get a lock for a hints database would erroneously always say
4034 "Failed to get write lock", even when it was really a read lock.
4036 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4037 of 0600. This has been changed to use the value of the lockfile_mode
4038 option (which defaults to 0600).
4040 PH/08 Applied small patch from the Sieve maintainer.
4042 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4043 folder from quota calculations, a direct delivery into this folder messed
4044 up the contents of the maildirsize file. This was because the regex was
4045 used only to exclude .Trash (or whatever) when the size of the mailbox
4046 was calculated. There was no check that a delivery was happening into an
4047 excluded directory. This bug has been fixed by ignoring all quota
4048 processing for deliveries into excluded directories.
4050 PH/10 Added the maildirfolder_create_regex option to appendfile.
4056 PH/01 The code for finding all the local interface addresses on a FreeBSD
4057 system running IPv6 was broken. This may well have applied to all BSD
4058 systems, as well as to others that have similar system calls. The broken
4059 code found IPv4 interfaces correctly, but gave incorrect values for the
4060 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4061 that it would not match correctly against @[] and not recognize the IPv6
4064 PH/02 The ipliteral router was not recognizing addresses of the form user@
4065 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4067 PH/03 Added disable_ipv6.
4069 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4070 decoded form, because it is most often used to construct To: headers
4071 lines in autoreplies, and the decoded form may well be syntactically
4072 invalid. However, $reply_address has leading white space removed, and all
4073 newlines turned into spaces so that the autoreply transport does not
4076 PH/05 If group was specified without a user on a router, and no group or user
4077 was specified on a transport, the group from the router was ignored.
4079 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4080 for visible compile-time settings that can be used to change these
4081 numbers, for those that want even more. Backwards compatibility with old
4082 spool files has been maintained. However, going back to a previous Exim
4083 release will lost any variables that are in spool files.
4085 PH/07 Two small changes when running in the test harness: increase delay when
4086 passing a TCP/IP connection to a new process, in case the original
4087 process has to generate a bounce, and remove special handling of
4088 127.0.0.2 (sic), which is no longer necessary.
4090 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4091 be the same on different OS.
4093 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4096 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4097 whether --show-vars was specified or not
4099 JJ/02 exipick: Added support for new ACL variable spool format introduced
4102 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4103 syntactically invalid From: or Reply-to: line, and a filter used this to
4104 generate an autoreply, and therefore failed to obtain an address for the
4105 autoreply, Exim could try to deliver to a non-existent relative file
4106 name, causing unrelated and misleading errors. What now happens is that
4107 it logs this as a hard delivery error, but does not attempt to create a
4110 PH/11 The exinext utility has a -C option for testing purposes, but although
4111 the given file was scanned by exinext itself; it wasn't being passed on
4112 when Exim was called.
4114 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4115 an end-of-file indication when reading a command response.
4117 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4118 compiled. In many other places in Exim, IPv6 addresses are always
4119 recognized, so I have changed this. It also means that IPv4 domain
4120 literals of the form [IPV4:n.n.n.n] are now always recognized.
4122 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4123 used if the router is not running as root, for example, when verifying at
4124 ACL time, or when using -bh. The debugging output from this situation was
4125 non-existent - all you got was a failure to exec. I have made two
4128 (a) Failures to set uid/gid, the current directory, or a process leader
4129 in a subprocess such as that created by queryprogram now generate
4130 suitable debugging output when -d is set.
4132 (b) The queryprogram router detects when it is not running as root,
4133 outputs suitable debugging information if -d is set, and then runs
4134 the subprocess without attempting to change uid/gid.
4136 PH/15 Minor change to Makefile for building test_host (undocumented testing
4139 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4140 additional section of a DNS packet that returns MX or SRV records.
4141 Instead, it always explicitly searches for A/AAAA records. This avoids
4142 major problems that occur when a DNS server includes only records of one
4143 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4144 fixed another bug: if SRV records were looked up and the corresponding
4145 address records were *not* found in the additional section, the port
4146 values from the SRV records were lost.
4148 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4149 using the correct key (the original address) when searching the retry
4150 rules in order to find which one to use for generating the retry hint.
4152 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4153 adding the default one. Similarly, if it contains a Reply-To: header, the
4154 errors_reply_to option, if set, is not used.
4156 PH/19 When calculating a retry time, Exim used to measure the "time since
4157 failure" by looking at the "first failed" field in the retry record. Now
4158 it does not use this if it is later than than the arrival time of the
4159 message. Instead it uses the arrival time. This makes for better
4160 behaviour in cases where some deliveries succeed, thus re-setting the
4161 "first failed" field. An example is a quota failure for a huge message
4162 when small messages continue to be delivered. Without this change, the
4163 "time since failure" will always be short, possible causing more frequent
4164 delivery attempts for the huge message than are intended.
4165 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4167 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4168 $1, $2, $3) because the numerical variables can be reset during some
4169 expansion items (e.g. "match"), thereby losing the authentication data.
4171 PH/21 Make -bV show the size of off_t variables so that the test suite can
4172 decide whether to run tests for quotas > 2G.
4174 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4175 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4176 filecount value is greater than 2G or if a quota value is greater than 2G
4177 on a system where the size of off_t is not greater than 4, a panic error
4180 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4181 never match. The debug and -bh output now contains an explicit error
4182 message indicating a malformed IPv4 address or mask.
4184 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4185 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4186 PH/23 above applies.
4188 PH/25 Do not write to syslog when running in the test harness. The only
4189 occasion when this arises is a failure to open the main or panic logs
4190 (for which there is an explicit test).
4192 PH/26 Added the /no_tell option to "control=freeze".
4194 PH/27 If a host name lookup failed very early in a connection, for example, if
4195 the IP address matched host_lookup and the reverse lookup yielded a name
4196 that did not have a forward lookup, an error message of the form "no IP
4197 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4198 could be logged. Now it outputs the IP address instead of "NULL".
4200 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4201 allows the sender and the authenticated sender to be set when
4202 submitting a message from within Exim. Since child_open_exim() is
4203 documented for local_scan(), the new function should be too.
4205 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4206 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4207 results in an empty string is now treated as unset.
4209 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4211 PH/31 Added sender_verify_fail logging option.
4213 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4214 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4215 tidied the source and removed it altogether.
4217 PH/33 When a queue run was abandoned because the load average was too high, a
4218 log line was always written; now it is written only if the queue_run log
4219 selector is set. In addition, the log line for abandonment now contains
4220 information about the queue run such as the pid. This is always present
4221 in "start" and "stop" lines but was omitted from the "abandon" line.
4223 PH/34 Omit spaces between a header name and the colon in the error message that
4224 is given when verify = headers_syntax fails (if there are lots of them,
4225 the message gets confusing).
4227 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4228 names, as there are now some PTR records that contain slashes. This check
4229 is only to protect against broken name servers that fall over on strange
4230 characters, so the fact that it applies to all lookups doesn't matter.
4232 PH/36 Now that the new test suite is complete, we can remove some of the
4233 special code in Exim that was needed for the old test suite. For example,
4234 sorting DNS records because real resolvers return them in an arbitrary
4235 order. The new test suite's fake resolver always returns records in the
4238 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4239 bounces) except when queue_only is set, to avoid logging races between
4240 the different processes.
4242 PH/38 Panic-die if .include specifies a non-absolute path.
4244 PH/39 A tweak to the "H" retry rule from its user.
4246 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4247 a label. They prevented compilation on older perls.
4249 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4250 a warning to be raised on newish perls.
4252 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4253 on queue. Changes to match documented behaviour of showing count of
4254 messages matching specified criteria.
4256 PH/40 Changed the default ident timeout from 30s to 5s.
4258 PH/41 Added support for the use of login_cap features, on those BSD systems
4259 that have them, for controlling the resources used by pipe deliveries.
4261 PH/42 The content-scanning code uses fopen() to create files in which to put
4262 message data. Previously it was not paying any attention to the mode of
4263 the files. Exim runs with umask(0) because the rest of the code creates
4264 files with open(), and sets the required mode explicitly. Thus, these
4265 files were ending up world-writeable. This was not a big issue, because,
4266 being within the spool directory, they were not world-accessible. I have
4267 created a function called modefopen, which takes an additional mode
4268 argument. It sets umask(777), creates the file, chmods it to the required
4269 mode, then resets the umask. All the relevant calls to fopen() in the
4270 content scanning code have been changed to use this function.
4272 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4273 to 24 hours. This avoids potential overflow problems when processing G
4274 and H retry rules. I suspect nobody ever tinkers with this value.
4276 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4278 PH/45 When the plaintext authenticator is running as a client, the server's
4279 challenges are checked to ensure they are valid base64 strings. By
4280 default, the authentication attempt is cancelled if an invalid string is
4281 received. Setting client_ignore_invalid_base64 true ignores these errors.
4282 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4283 they are received. Thus, the responses can be made to depend on the
4284 challenges. If an invalid string is ignored, an empty string is placed in
4287 PH/46 Messages that are created by the autoreply transport now contains a
4288 References: header, in accordance with RFCs 2822 and 3834.
4290 PH/47 Added authenticated_sender_force to the smtp transport.
4292 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4294 PH/49 Installed latest patch from the Sieve maintainer.
4296 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4297 was also set, the appendfile transport was unnecessarily scanning a
4298 directory of message files (e.g. for maildir delivery) to find the count
4299 of files (along with the size), even though it did not need this
4300 information. It now does the scan only if it needs to find either the
4301 size of the count of files.
4303 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4305 PH/52 Two bugs concerned with error handling when the smtp transport is
4308 (i) Exim was not creating retry information for temporary errors given
4309 for individual recipients after the DATA command when the smtp transport
4310 was used in LMTP mode. This meant that they could be retried too
4311 frequently, and not timed out correctly.
4313 (ii) Exim was setting the flag that allows error details to be returned
4314 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4315 recipients that were returned after the DATA command.
4317 PH/53 This is related to PH/52, but is more general: for any failing address,
4318 when detailed error information was permitted to be returned to the
4319 sender, but the error was temporary, then after the final timeout, only
4320 "retry timeout exceeded" was returned. Now it returns the full error as
4321 well as "retry timeout exceeded".
4323 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4324 do this, and (what is worse) MTAs that accept it.
4326 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4327 will now be deprecated.
4329 PH/56 New os.c-cygwin from the Cygwin maintainer.
4331 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4332 formats (previously only available in exim formats via -bpr, -bpru,
4333 and -bpra. Now also available in native and exiqgrep formats)
4335 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4336 with very large, slow to parse queues
4338 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4340 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4342 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4343 responses to authentication challenges, though it was showing the
4344 challenges; (ii) I've removed the CR characters from the debug output for
4347 PH/58 Allow for the insertion of a newline as well as a space when a string
4348 is turned into more than one encoded-word during RFC 2047 encoding. The
4349 Sieve code now uses this.
4351 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4352 data_4xx, lost_connection, tls_required.
4354 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4355 message was being sent as an SMTP response.
4357 PH/61 Add -l and -k options to exicyclog.
4359 PH/62 When verifying, if an address was redirected to one new address, so that
4360 verification continued, and the new address failed or deferred after
4361 having set something in $address_data, the value of $address_data was not
4362 passed back to the ACL. This was different to the case when no
4363 redirection occurred. The value is now passed back in both cases.
4365 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4366 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4367 use login_cap.h, so on its own it isn't the distinguishing feature. The
4368 new name refers directly to the setclassresources() function.
4370 PH/65 Added configuration files for NetBSD3.
4372 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4374 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4375 is preferred over IPv4.
4377 PH/68 The bounce_return_message and bounce_return_body options were not being
4378 honoured for bounces generated during the reception of non-SMTP messages.
4379 In particular, this applied to messages rejected by the ACL. This bug has
4380 been fixed. However, if bounce_return_message is true and bounce_return_
4381 body is false, the headers that are returned for a non-SMTP message
4382 include only those that have been read before the error was detected.
4383 (In the case of an ACL rejection, they have all been read.)
4385 PH/69 The HTML version of the specification is now built in a directory called
4386 spec_html instead of spec.html, because the latter looks like a path with
4387 a MIME-type, and this confuses some software.
4389 PH/70 Catch two compiler warnings in sieve.c.
4391 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4392 function verify_get_ident() calls ip_connect() to connect a socket, but
4393 if the "connect()" function timed out, ip_connect() used to close the
4394 socket. However, verify_get_ident() also closes the socket later, and in
4395 between Exim writes to the log, which may get opened at this point. When
4396 the socket was closed in ip_connect(), the log could get the same file
4397 descriptor number as the socket. This naturally causes chaos. The fix is
4398 not to close the socket in ip_connect(); the socket should be closed by
4399 the function that creates it. There was only one place in the code where
4400 this was missing, in the iplookup router, which I don't think anybody now
4401 uses, but I've fixed it anyway.
4403 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4404 well as to direct DNS lookups. Otherwise the handling of names in host
4405 lists is inconsistent and therefore confusing.
4411 PH/01 Two changes to the default runtime configuration:
4413 (1) Move the checks for relay_from_hosts and authenticated clients from
4414 after to before the (commented out) DNS black list checks.
4416 (2) Add control=submission to the relay_from_hosts and authenticated
4417 clients checks, on the grounds that messages accepted by these
4418 statements are most likely to be submissions.
4420 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4422 (1) Generate an error if the third argument for the ${prvs expansion is
4425 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4428 (3) Reset the variables that are obtained from the first argument of
4429 ${prvscheck and used in the second argument before leaving the code,
4430 because their memory is reclaimed, so using them afterwards may do
4433 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4434 one (it's much easier than Tom thought :-).
4436 (5) Because of (4), we can now allow for the use of $prvscheck_result
4437 inside the third argument.
4439 PH/03 For some reason, the default setting of PATH when running a command from
4440 a pipe transport was just "/usr/bin". I have changed it to
4443 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4444 anything to be listed in the output from -bV.
4446 PH/05 When a filter generated an autoreply, the entire To: header line was
4447 quoted in the delivery log line, like this:
4449 => >A.N.Other <ano@some.domain> <original@ddress> ...
4451 This has been changed so that it extracts the operative address. There
4452 may be more than one such address. If so, they are comma-separated, like
4455 => >ano@some.domain,ona@other.domain <original@ddress> ...
4457 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4458 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4459 not being looked up in the rDNS to get a host name, Exim was showing the
4460 IP address twice in Received: lines, even though the IP addresses were
4461 identical. For example:
4463 Received: from [1.2.3.4] (helo=[1.2.3.4])
4465 However, if the real host name was known, it was omitting the HELO data
4466 if it matched the actual IP address. This has been tidied up so that it
4467 doesn't show the same IP address twice.
4469 PH/07 When both +timestamp and +memory debugging was on, the value given by
4470 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4471 called by the debug printing, thereby overwriting the timestamp buffer.
4472 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4474 PH/08 When the original message was included in an autoreply transport, it
4475 always said "this is a copy of the message, including all the headers",
4476 even if body_only or headers_only was set. It now gives an appropriate
4479 PH/09 Applied a patch from the Sieve maintainer which:
4481 o fixes some comments
4482 o adds the (disabled) notify extension core
4483 o adds some debug output for the result of if/elsif tests
4484 o points to the current vacation draft in the documentation
4485 and documents the missing references header update
4489 o fixes a bug in processing the envelope test (when testing
4490 multiple envelope elements, the last element determined the
4493 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4494 Electronic Mail") by including:
4496 Auto-submitted: auto-generated
4498 in the messages that it generates (bounce messages and others, such as
4499 warnings). In the case of bounce messages for non-SMTP messages, there was
4500 also a typo: it was using "Auto_submitted" (underscore instead of
4501 hyphen). Since every message generated by Exim is necessarily in response
4502 to another message, thes have all been changed to:
4504 Auto-Submitted: auto-replied
4506 in accordance with these statements in the RFC:
4508 The auto-replied keyword:
4510 - SHOULD be used on messages sent in direct response to another
4511 message by an automatic process,
4513 - MUST NOT be used on manually-generated messages,
4515 - MAY be used on Delivery Status Notifications (DSNs) and Message
4516 Disposition Notifications (MDNs),
4518 - MUST NOT be used on messages generated by automatic or periodic
4519 processes, except for messages which are automatic responses to
4522 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4523 to the default Received: header definition.
4525 PH/12 Added log selector acl_warn_skipped (default on).
4527 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4528 variables because (a) they are in the wrong storage pool and (b) even if
4529 they were copied, it wouldn't work properly because of the caching.
4531 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4532 checking when decoding. Apparently there are clients that generate
4533 overlong encoded strings. Why am I not surprised?
4535 PH/15 If the first argument of "${if match_address" was not empty, but did not
4536 contain an "@" character, Exim crashed. Now it writes a panic log message
4537 and treats the condition as false.
4539 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4541 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4542 extension "envelope-auth". The code is finished and in agreement with
4543 other implementations, but there is no documentation so far and in fact,
4544 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4545 not changing the active code.
4547 Print executed "if" and "elsif" statements when debugging is used. This
4548 helps a great deal to understand what a filter does.
4550 Document more things not specified clearly in RFC3028. I had all this
4551 sorted out, when out of a sudden new issues came to my mind. Oops."
4553 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4556 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4557 canonical form (as documented). However, after a host name lookup from
4558 the IP address, check_host() was doing a simple string comparison with
4559 addresses acquired from the DNS when checking that the found name did
4560 have the original IP as one of its addresses. Since any found IPv6
4561 addresses are likely to be in abbreviated form, the comparison could
4562 fail. Luckily, there already exists a function for doing the comparison
4563 by converting both addresses to binary, so now that is used instead of
4564 the text comparison.
4566 PH/20 There was another similar case to PH/19, when a complete host name was
4567 given in a host list; looking up its IP address could give an abbreviated
4568 form, whereas the current host's name might or might not be abbreviated.
4569 The same fix has been applied.
4575 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4576 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4579 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4580 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4582 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4584 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4585 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4586 Also, for "Auto-Submitted", treat anything other than "no" as
4587 non-personal, in accordance with RFC 3834. (Previously it treated
4588 anything starting "auto-" as non-personal.)
4590 TF/01 The control=submission/name=... option had a problem with syntax
4591 errors if the name included a slash character. The /name= option
4592 now slurps the rest of the string, so it can include any characters
4593 but it must come last in the list of options (after /sender_retain
4596 PH/05 Some modifications to the interface to the fake nameserver for the new
4604 TK/01 Added the "success_on_redirect" address verification option. See
4605 NewStuff for rationale and an example.
4607 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4609 PH/02 Patch to exigrep to allow it to work on syslog lines.
4611 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4612 fread() to skip over the body file's header line, because in Cygwin the
4613 header line is locked and is inaccessible.
4615 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4616 co-exist for some time) to make it clear that it is the Exim ID that is
4617 referenced, not the Message-ID: header line.
4619 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4620 string_format() function, because snprintf() does not exist on all
4623 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4624 ${stat: expansion item.
4626 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4627 protocol synchronization error", to keep the pedants happy.
4629 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4630 well as for IRIX systems, when gcc is being used. See the host.c source
4633 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4635 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4638 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4639 transport and to the smtp transport in LMTP mode.
4641 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4643 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4644 run a filter in a subprocess. This could lead to confusion in subsequent
4645 lookups in the parent process. There should also be a search_tidyup() at
4646 the end of the subprocess.
4648 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4649 only if the host matched helo_try_verify_hosts, which caused the
4650 verification to occur when the EHLO/HELO command was issued. The ACL just
4651 tested the remembered result. Now, if a previous verification attempt has
4652 not happened, "verify = helo" does it there and then.
4654 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4656 TK/03 Fix log output including CR from clamd.
4658 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4659 exist provoked a memory error which could cause a segfault.
4661 PH/15 Installed PCRE 6.2
4663 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4665 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4666 of the problem. Specifically, suggested +O2 rather than +O1 for the
4669 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4671 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4672 record, the port was not correctly logged when the outgoing_port log
4673 selector was set (it logged the transort's default port).
4675 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4676 fallback_hosts, and "hosts" in the smtp transport.
4678 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4679 host errors such as "Connection refused".
4681 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4682 authentication with radiusclient 0.4.9:
4684 - Error returned from rc_read_config was caught wrongly
4685 - Username/password not passed on to radius server due to wrong length.
4687 The presumption is that some radiusclient API changes for 4.51/PH/17
4688 were not taken care of correctly. The code is still untested by me (my
4689 Linux distribution still has 0.3.2 of radiusclient), but it was
4690 contributed by a Radius user.
4692 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4693 expanding the "port" option of the smtp transport.
4695 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4696 while reading a MIME header. Thanks to Tom Hughes for a patch.
4698 PH/24 Include config.h inside local_scan.h so that configuration settings are
4701 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4702 that in an ACL for RCPT (for example), you can examine exactly what was
4705 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4706 commands, but it was not correctly comparing the address with the actual
4707 client host address. Thus, it would show the EHLO address in Received:
4708 header lines when this was not necessary.
4710 PH/27 Added the % operator to ${eval:}.
4712 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4713 it should be ignoring failures (because with -C, for example, it has lost
4714 privilege). It wasn't ignoring creation failures other than "already
4717 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4720 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4721 because an input line was too long, either on its own, or by virtue of
4722 too many continuations, the temporary file was not being removed, and the
4723 return code was incorrect.
4725 PH/31 Missing "BOOL" in function definition in filtertest.c.
4727 PH/32 Applied Sieve patches from the maintainer.
4729 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4731 PH/33 Added "verify = not_blind".
4733 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4734 Local/Makefile (with some defaults set). These are used in built scripts
4735 such as exicyclog, but they have never been used in the exim_install
4736 script (though there are many overriding facilities there). I have
4737 arranged that the exim_install script now takes note of these two
4740 PH/35 Installed configuration files for Dragonfly.
4742 PH/36 When a locally submitted message by a trusted user did not contain a
4743 From: header, and the sender address was obtained from -f or from an SMTP
4744 MAIL command, and the trusted user did not use -F to supply a sender
4745 name, $originator_name was incorrectly used when constructing a From:
4746 header. Furthermore, $originator_name was used for submission mode
4747 messages from external hosts without From: headers in a similar way,
4748 which is clearly wrong.
4750 PH/37 Added control=suppress_local_fixups.
4752 PH/38 When log_selector = +received_sender was set, and the addition of the
4753 sender made the log line's construction buffer exactly full, or one byte
4754 less than full, an overflow happened when the terminating "\n" was
4757 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4758 when the result of a list match is failure because a DNS lookup failed.
4760 PH/40 RM_COMMAND is now used in the building process.
4762 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4763 the "build-* directories that it finds.
4765 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4766 address in a domain literal was a prefix of an interface address.
4768 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4769 when verifying a sender address, unless rewrite_headers is false.
4771 PH/44 Wrote a long comment about why errors_to addresses are verified as
4772 recipients, not senders.
4774 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4775 the ratelimit ACL was added.
4777 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4779 PH/46 Added extra information about PostgreSQL errors to the error string.
4781 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4782 avoiding the need to install special zones in a real server. This is
4783 backwards compatible; if it can't find the fake resolver, it drops back.
4784 Thus, both old and new test suites can be run.
4786 TF/02 Added util/ratelimit.pl
4788 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4789 clock is set back in time.
4791 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4792 Candler <B.Candler@pobox.com>.
4794 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4795 OK for addresses that are the result of redirections.
4797 PH/48 A number of further additions for the benefit of the new test suite,
4798 including a fake gethostbyname() that interfaces to the fake DNS resolver
4801 TF/06 The fix for widen_domains has also been applied to qualify_single and
4802 search_parents which are the other dnslookup options that can cause
4805 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4808 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4810 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4811 service" if it encountered an MX record with an empty target hostname.
4812 The message is now "an MX or SRV record indicated no SMTP service".
4814 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4815 if the DNS of the sending site is misconfigured. This is quite a
4816 common situation. This change restores the behaviour of treating a
4817 helo verification defer as a failure.
4819 PH/51 If self=fail was set on a router, the bounce message did not include the
4820 actual error message.
4826 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4828 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4829 command itself ended in error, the underlying message about the transport
4830 filter timeout was being overwritten with the pipe command error. Now the
4831 underlying error message should be appended to the second error message.
4833 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4835 PH/02 Reduce the amount of output that "make" produces by default. Full output
4836 can still be requested.
4838 PH/03 The warning log line about a condition test deferring for a "warn" verb
4839 was being output only once per connection, rather than after each
4840 occurrence (because it was using the same function as for successful
4841 "warn" verbs). This seems wrong, so I have changed it.
4843 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4844 it should not have, which might have caused a crash in the right
4845 circumstances, but probably never did.
4847 PH/04 Installed a modified version of Tony Finch's patch to make submission
4848 mode fix the return path as well as the Sender: header line, and to
4849 add a /name= option so that you can make the user's friendly name appear
4852 TF/03 Added the control = fakedefer ACL modifier.
4854 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4855 Mark Lowes for thorough testing.
4857 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4859 TK/03 Merged latest SRS patch from Miles Wilton.
4861 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4862 with the definition in sysexits.h (which is #included earlier).
4863 Fortunately, Exim does not actually use EX_OK. The code used to try to
4864 preserve the sysexits.h value, by assuming that macro definitions were
4865 scanned for macro replacements. I have been disabused of this notion,
4866 so now the code just undefines EX_OK before #including unistd.h.
4868 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4869 in the smtp transport. When a block could not be written in a single
4870 write() function, the timeout was being re-applied to each part-write.
4871 This seems wrong - if the receiver was accepting one byte at a time it
4872 would take for ever. The timeout is now adjusted when this happens. It
4873 doesn't have to be particularly precise.
4875 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4876 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4878 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4879 without a domain if the check to <postmaster@domain> fails.
4881 SC/01 Eximstats: added -xls and the ability to specify output files
4882 (patch written by Frank Heydlauf).
4884 SC/02 Eximstats: use FileHandles for outputting results.
4886 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4888 SC/04 Eximstats: fixed display of large numbers with -nvr option
4890 SC/05 Eximstats: fixed merging of reports with empty tables.
4892 SC/06 Eximstats: added the -include_original_destination flag
4894 SC/07 Eximstats: removed tabs and trailing whitespace.
4896 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4898 TK/06 MBOX spool code: Add real "From " MBOX separator line
4899 so the .eml file is really in mbox format (even though
4900 most programs do not really care). Patch from Alex Miller.
4902 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4903 The latter is generated from $received_to and is only set if the
4904 message has one envelope recipient. SA can use these headers,
4905 obviously out-of-the-box. Patch from Alex Miller.
4907 PH/08 The ${def test on a variable was returning false if the variable's
4908 value was "0", contrary to what the specification has always said!
4909 The result should be true unless the variable is empty.
4911 PH/09 The syntax error of a character other than { following "${if
4912 def:variable_name" (after optional whitespace) was not being diagnosed.
4913 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4914 accidental colon was present, for example, could give incorrect results.
4916 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4917 result is used (not including appendfile, where other changes are about
4920 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4921 This involved changing a lot of size variables from int to off_t. It
4922 should work with maildirs and everything.
4924 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4925 spamd dying while we are connected to it.
4927 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4930 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4931 function for BDB 4.3.
4933 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4935 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4936 the sources for PCRE have been re-arranged and more files are now
4939 PH/15 The code I had for printing potentially long long variables in PH/11
4940 above was not the best (it lost precision). The length of off_t variables
4941 is now inspected at build time, and an appropriate printing format (%ld
4942 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4943 to be "long long int" or "long int". This is needed for the internal
4944 formatting function string_vformat().
4946 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4947 the configuration file to be ":syslog", then the script "guesses" where
4948 the logs files are, rather than using the compiled in default. In our
4949 case the guess is not the same as the compiled default, so the script
4950 suddenly stopped working when I started to use syslog. The patch checks
4951 to see if log_file_path is "". If so, it attempts to read it from exim
4952 with no configuration file to get the compiled in version, before it
4953 falls back to the previous guessing code."
4955 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4956 implementing BATV in an Exim configuration. See NewStuff for the gory
4959 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4960 Makefile that are specific to HP-UX.
4962 PH/18 If the "use_postmaster" option was set for a recipient callout together
4963 with the "random" option, the postmaster address was used as the MAIL
4964 FROM address for the random test, but not for the subsequent recipient
4965 test. It is now used for both.
4967 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4968 patch removes a few documentation additions to RFC 3028, because the
4969 latest draft now contains them. It adds the new en;ascii-case comparator
4970 and a new error check for 8bit text in MIME parts. Comparator and
4971 require names are now matched exactly. I enabled the subaddress
4972 extension, but it is not well tested yet (read: it works for me)."
4974 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4975 rework some of the code of TK/09 above to avoid the hardwired use of
4976 "%lld" and "long long". Replaced the call to snprintf() with a call to
4979 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4980 records point to non-existent hosts", "retry timeout exceeded", and
4981 "retry time not reached for any host after a long failure period".
4983 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4984 experimental DomainKeys support:
4986 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4987 (2) On an error such as an illegally used "control", the wrong name for
4988 the control was given.
4990 These problems did NOT occur unless DomainKeys support was compiled.
4992 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4994 PH/24 Added ${if match_ip condition.
4996 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4997 databases so that it will be absolutely obvious if a crash occurs in the
4998 DB library. This is a regular occurrence (often caused by mis-matched
5001 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5002 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5003 picked up on a user's system that detects such things. There doesn't seem
5004 to be a gcc warning option for this - only an attribute that has to be
5005 put on the function's prototype. It seems that in Fedora Core 4 they have
5006 set this on a number of new functions. No doubt there will be more in due
5009 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5010 specify a transport. However, if an address that was verified by such a
5011 router was the subject of a callout, Exim crashed because it tried to
5012 read the rcpt_include_affixes from the non-existent transport. Now it
5013 just assumes that the setting of that option is false. This bug was
5014 introduced by 4.51/PH/31.
5016 PH/28 Changed -d+all to exclude +memory, because that information is very
5017 rarely of interest, but it makes the output a lot bigger. People tend to
5018 do -d+all out of habit.
5020 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5021 code in os-type was giving problems when libc.so lives in lib64, like on
5024 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5025 aren't the modern standard, and it seems that some systems' include files
5026 don't always have them. Exim was already checking for some of the newer
5027 ones like T_AAAA, and defining it itself. I've added checks for all the
5028 record types that Exim uses.
5030 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5031 not automatically generating a new one, as it is supposed to. This
5032 prevented TLS from working. If the file did exist, but contained invalid
5033 data, a new version was generated, as expected. It was only the case of a
5034 non-existent file that was broken.
5036 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5037 with a change in libdomainkeys > 0.64.
5039 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5040 from DNS. If the selector record carries the flag, it now has
5041 precedence over the domain-wide flag.
5043 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5045 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5046 the use of a port name that isn't defined in /etc/services) occurred, the
5047 message was deferred as in a normal delivery, and thus remained on the
5048 spool, instead of being failed because of the mua_wrapper setting. This
5049 is now fixed, and I tidied up some of the mua_wrapper messages at the
5052 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5053 the messages in a hash of arrays rather than using individual hashes.
5054 This is a bit cleaner and results in dramatic memory savings, albeit
5055 at a slight CPU cost.
5057 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5058 as requested by Marc Sherman.
5060 SC/10 Eximstats: added histograms for user specified patterns as requested
5063 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5065 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5066 fopen() in the content-scanning modules that did not already have it.
5072 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5073 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5075 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5077 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5079 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5080 to test Sieve filters that use "vacation".
5082 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5083 that changes the way the GnuTLS parameters are stored in the cache file.
5084 The new format can be generated externally. For backward compatibility,
5085 if the data in the cache doesn't make sense, Exim assumes it has read an
5086 old-format file, and it generates new data and writes a new file. This
5087 means that you can't go back to an older release without removing the
5090 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5091 work if there are any delivery delays because "one_time" forces the
5092 parent to be marked "delivered", so its unseen clone is never tried
5093 again. For this reason, Exim now forbids the simultaneous setting of
5096 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5097 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5098 entry if you want to know the details. The fix, however, neglected to
5099 consider the case when local delivery batching is involved. The test for
5100 "previously delivered" was not happening when checking to see if an
5101 address could be batched with a previous (undelivered) one; under
5102 certain circumstances this could lead to multiple deliveries to the same
5105 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5106 in its include files, and this causes problems building Exim.
5108 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5109 header_syntax) but Exim was just ignoring anything given after a slash.
5110 In particular, this caused confusion with an attempt to use "verify =
5111 reverse_host_lookup/defer_ok". An error is now given when options are
5112 supplied for verify items that do not have them. (Maybe reverse_host_
5113 lookup should have a defer_ok option, but that's a different point.)
5115 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5116 defined by RFC 821) to 2048, because there were problems with some AUTH
5117 commands, and RFC 1869 says the size should be increased for extended
5118 SMTP commands that take arguments.
5120 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5123 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5124 "unknown" error; now it says that the functionality isn't in the binary.
5126 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5127 an address' error message when a string expansion fails (syntax or
5128 whatever). Otherwise the password may appear in the log. Following change
5129 PH/42 below, there is no longer a chance of it appearing in a bounce
5132 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5134 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5135 of its addresses was ever tried. (Bugzilla bug #2.)
5137 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5138 the result incorrectly in the debug output. (It correctly added a newline
5139 to what was transported.)
5141 TF/01 Added $received_time.
5143 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5144 commented out examples of how to interface to a virus scanner and to
5145 SpamAssassin. Also added commented examples of av_scanner and
5146 spamd_address settings.
5148 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5149 and controls are allowed in which ACLs. There were a couple of minor
5150 errors. Some of the entries in the conditions table (which is a table of
5151 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5152 negation of where the condition IS allowed.
5154 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5156 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5157 header file does not have a version number, so I've had to invent a new
5158 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5159 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5160 radiusclient), but it was contributed by a Radius user.
5162 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5163 files or directories, for OpenSSL.
5165 PH/19 When an Exim process that is running as root has to create an Exim log
5166 file, it does so in a subprocess that runs as exim:exim so as to get the
5167 ownership right at creation (otherwise, other Exim processes might see
5168 the file with the wrong ownership). There was no test for failure of this
5169 fork() call, which would lead to the process getting stuck as it waited
5170 for a non-existent subprocess. Forks do occasionally fail when resources
5171 run out. I reviewed all the other calls to fork(); they all seem to check
5174 PH/20 When checking for unexpected SMTP input at connect time (before writing
5175 the banner), Exim was not dealing correctly with a non-positive return
5176 from the read() function. If the client had disconnected by this time,
5177 the result was a log entry for a synchronization error with an empty
5178 string after "input=" when read() returned zero. If read() returned -1
5179 (an event I could not check), uninitialized data bytes were printed.
5180 There were reports of junk text (parts of files, etc) appearing after
5183 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5185 PH/22 Added support for macro redefinition, and (re)definition in between
5186 driver and ACL definitions.
5188 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5189 forgetting to use the resulting value; it was using the unexpanded value.
5191 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5192 hadn't been configured. The fix is from Juergen Kreileder, who
5193 understands it better than I do:
5195 "Here's what I see happening with three configured cyrus_sasl
5196 authenticators configured (plain, login, cram-md5):
5198 On startup auth_cyrus_sasl_init() gets called for each of these.
5199 This means three calls to sasl_listmech() without a specified mech_list.
5200 => SASL tests which mechs of all available mechs actually work
5201 => three warnings about OTP not working
5202 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5204 With the patch, sasl_listmech() also gets called three times. But now
5205 SASL's mech_list option is set to the server_mech specified in the the
5206 authenticator. Or in other words, the answer from sasl_listmech()
5207 gets limited to just the mech you're testing for (which is different
5209 => the return list contains just 'plain' or 'login', 'cram-md5' or
5210 nothing depending on the value of ob->server_mech.
5212 I've just tested the patch: Authentication still works fine,
5213 unavailable mechs specified in the exim configuration are still
5214 caught, and the auth.log warnings about OTP are gone."
5216 PH/25 When debugging is enabled, the contents of the command line are added
5217 to the debugging output, even when log_selector=+arguments is not
5220 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5221 answer is "GNU", and only if the return is "GNU/something" is the answer
5224 PH/27 $acl_verify_message is now set immediately after the failure of a
5225 verification in an ACL, and so is available in subsequent modifiers. In
5226 particular, the message can be preserved by coding like this:
5228 warn !verify = sender
5229 set acl_m0 = $acl_verify_message
5231 Previously, $acl_verify_message was set only while expanding "message"
5232 and "log_message" when a very denied access.
5234 PH/28 Modified OS/os.c-Linux with
5236 -#ifndef OS_LOAD_AVERAGE
5237 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5239 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5240 nomenclature these days.)
5242 PH/29 Installed patch from the Sieve maintainer that adds the options
5243 sieve_useraddress and sieve_subaddress to the redirect router.
5245 PH/30 In these circumstances:
5246 . Two addresses routed to the same list of hosts;
5247 . First host does not offer TLS;
5248 . First host accepts first address;
5249 . First host gives temporary error to second address;
5250 . Second host offers TLS and a TLS session is established;
5251 . Second host accepts second address.
5252 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5253 and peerdn, if requested) that were in fact used only for the second
5256 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5257 attention to any local part prefix or suffix that was matched by the
5258 router that accepted the address. It now behaves in the same way as it
5259 does for delivery: the affixes are removed from the local part unless
5260 rcpt_include_affixes is set on the transport.
5262 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5263 timeout during the DATA phase of an incoming message.
5265 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5266 applied a patch sanctioned by the Sieve maintainer.
5268 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5269 the uid or gid is negative. A case of a negative gid caused this to be
5270 noticed. The fix allows for either to be negative.
5272 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5273 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5276 PH/36 The change PH/12 above was broken. Fixed it.
5278 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5279 the grounds that routing the same address twice would always produce the
5280 same answer. This might have been true once, but it is certainly no
5281 longer true now. Routing a child address may depend on the previous
5282 routing that produced that child. Some complicated redirection strategies
5283 went wrong when messages had multiple recipients, and made Exim's
5284 behaviour dependent on the order in which the addresses were given.
5286 I have moved the duplicate checking until after the routing is complete.
5287 Exim scans the addresses that are assigned to local and remote
5288 transports, and removes any duplicates. This means that more work will be
5289 done, as duplicates will always all be routed, but duplicates are
5290 presumably rare, so I don't expect this is of any significance.
5292 For deliveries to pipes, files, and autoreplies, the duplicate checking
5293 still happens during the routing process, since they are not going to be
5296 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5297 It corrects a timeout issue with spamd. This is Ian's comment: "The
5298 background is that sometimes spamd either never reads data from a
5299 connection it has accepted, or it never writes response data. The exiscan
5300 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5301 blindly assumes that writes won't block so it may never time out."
5303 PH/39 Allow G after quota size as well as K and M.
5305 PH/40 The value set for $authenticated_id in an authenticator may not contain
5306 binary zeroes or newlines because the value is written to log lines and
5307 to spool files. There was no check on this. Now the value is run through
5308 the string_printing() function so that such characters are converted to
5309 printable escape sequences.
5311 PH/41 $message_linecount is a new variable that contains the total number of
5312 lines in the message. Compare $body_linecount, which is the count for the
5315 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5316 bounce and delay warning messages, except in certain special cases, which
5319 (a) An SMTP error message from a remote host;
5320 (b) A message specified in a :fail: redirection;
5321 (c) A message specified in a "fail" command in a system filter;
5322 (d) A message specified in a FAIL return from the queryprogram router;
5323 (e) A message specified by the cannot_route_message router option.
5325 In these cases only, Exim does include the error details in bounce and
5326 warning messages. There are also a few cases where bland messages such
5327 as "unrouteable address" or "local delivery error" are given.
5329 PH/43 $value is now also set for the "else" part of a ${run expansion.
5331 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5332 being worked on, but at least Exim now implements the latest version to
5335 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5336 process to complete was treated as a delivery failure, a timeout while
5337 writing the message to the pipe was logged, but erroneously treated as a
5338 successful delivery. Such timeouts include transport filter timeouts. For
5339 consistency with the overall process timeout, these timeouts are now
5340 treated as errors, giving rise to delivery failures by default. However,
5341 there is now a new Boolean option for the pipe transport called
5342 timeout_defer, which, if set TRUE, converts the failures into defers for
5343 both kinds of timeout. A transport filter timeout is now identified in
5346 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5347 systems where "make" and "gmake" are different, calling "gmake" at top
5348 level broke things. I've arranged for the value of $(MAKE) to be passed
5349 from the Makefile to this script so that it can call the same version of
5353 A note about Exim versions 4.44 and 4.50
5354 ----------------------------------------
5356 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5357 changes of various kinds. As a consequence, a big documentation update was
5358 needed. This delayed the release for rather longer than seemed good, especially
5359 in the light of a couple of (minor) security issues. Therefore, the changes
5360 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5361 release. So 4.44 and 4.50 are in effect two different branches that both start
5364 I have left the 4.50 change log unchanged; it contains all the changes since
5365 4.43. The change log for 4.44 is below; many of its items are identical to
5366 those for 4.50. This seems to be the most sensible way to preserve the
5367 historical information.
5373 1. Minor wording change to the doc/README.SIEVE file.
5375 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5376 computation of the current number of files was incorrect.
5378 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5379 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5382 4. Give more explanation in the error message when the command for a transport
5383 filter fails to execute.
5385 5. There are several places where Exim runs a non-Exim command in a
5386 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5387 was being done only for the command run by the queryprogram router. It is
5388 now done for all such subprocesses. The other cases are: ${run, transport
5389 filters, and the commands run by the lmtp and pipe transports.
5391 6. Added CONFIGURE_GROUP build-time option.
5393 7. Some older OS have a limit of 256 on the maximum number of file
5394 descriptors. Exim was using setrlimit() to set 1000 as a large value
5395 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5396 systems. I've change it so that if it can't get 1000, it tries for 256.
5398 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5399 was an oversight, and furthermore, ever since the addition of extra
5400 controls (e.g. 4.43/32), the checks on when to allow different forms of
5401 "control" were broken. There should now be diagnostics for all cases when a
5402 control that does not make sense is encountered.
5404 9. Added the /retain_sender option to "control=submission".
5406 10. $recipients is now available in the predata ACL (oversight).
5408 11. Tidy the search cache before the fork to do a delivery from a message
5409 received from the command line. Otherwise the child will trigger a lookup
5410 failure and thereby defer the delivery if it tries to use (for example) a
5411 cached ldap connection that the parent has called unbind on.
5413 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5414 of $address_data from the recipient verification was clobbered by the
5415 sender verification.
5417 13. The value of address_data from a sender verification is now available in
5418 $sender_address_data in subsequent conditions in the ACL statement.
5420 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5422 15. Added a new option "connect=<time>" to callout options, to set a different
5425 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5426 was its contents. (It was OK if the option was not defined at all.)
5428 17. A "Completed" log line is now written for messages that are removed from
5429 the spool by the -Mrm option.
5431 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5432 information about exactly what failed.
5434 19. Added -dd to debug only the daemon process.
5436 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5437 handles timeouts, both on the server side and network timeouts. Renamed the
5438 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5440 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5441 It is now set to "smtps".
5443 22. $host_address is now set to the target address during the checking of
5444 ignore_target_hosts.
5446 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5447 being passed; this would have caused $sender_host_name to have been used if
5448 matching the list had actually called for a host name (not very likely,
5449 since this list is usually IP addresses). A host name is now passed as
5452 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5453 code that specifies a non-restarting handler (typically sigaction() in
5454 modern systems) in an attempt to fix a rare and obscure crash bug.
5456 25. Narrowed the window for a race in the daemon that could cause it to ignore
5457 SIGCHLD signals. This is not a major problem, because they are used only to
5458 wake it up if nothing else does.
5460 26. A malformed maildirsize file could cause Exim to calculate negative values
5461 for the mailbox size or file count. Odd effects could occur as a result.
5462 The maildirsize information is now recalculated if the size or filecount
5465 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5466 support for a long time. Removed HAVE_SYS_VFS_H.
5468 28. Installed the latest version of exipick from John Jetmore.
5470 29. In an address list, if the pattern was not a regular expression, an empty
5471 subject address (from a bounce message) matched only if the pattern was an
5472 empty string. Non-empty patterns were not even tested. This was the wrong
5473 because it is perfectly reasonable to use an empty address as part of a
5474 database query. An empty address is now tested by patterns that are
5475 lookups. However, all the other forms of pattern expect the subject to
5476 contain a local part and a domain, and therefore, for them, an empty
5477 address still always fails if the pattern is not itself empty.
5479 30. Exim went into a mad DNS loop when attempting to do a callout where the
5480 host was specified on an smtp transport, and looking it up yielded more
5481 than one IP address.
5483 31. Re-factored the code for checking spool and log partition space into a
5484 function that finds that data and another that does the check. The former
5485 is then used to implement four new variables: $spool_space, $log_space,
5486 $spool_inodes, and $log_inodes.
5488 32. The RFC2047 encoding function was originally intended for short strings
5489 such as real names; it was not keeping to the 75-character limit for
5490 encoded words that the RFC imposes. It now respects the limit, and
5491 generates multiple encoded words if necessary. To be on the safe side, I
5492 have increased the buffer size for the ${rfc2047: expansion operator from
5495 33. It is now permitted to omit both strings after an "if" condition; if the
5496 condition is true, the result is "true". As before, when the second string
5497 is omitted, a false condition yields an empty string. This makes it less
5498 cumbersome to write custom ACL and router conditions.
5500 34. Failure to deliver a bounce message always caused it to be frozen, even if
5501 there was an errors_to setting on the router. The errors_to setting is now
5504 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5505 canonical form (fully expanded) before being placed in
5506 $sender_host_address.
5508 36. The table in the code that translates DNS record types into text (T_A to
5509 "A" for instance) was missing entries for NS and CNAME. It is just possible
5510 that this could have caused confusion if both these types were looked up
5511 for the same domain, because the text type is used as part of Exim's
5512 per-process caching. But the chance of anyone hitting this buglet seems
5515 37. The dnsdb lookup has been extended in a number of ways.
5517 (1) There is a new type, "zns", which walks up the domain tree until it
5518 finds some nameserver records. It should be used with care.
5520 (2) There is a new type, "mxh", which is like "mx" except that it returns
5521 just the host names, not the priorities.
5523 (3) It is now possible to give a list of domains (or IP addresses) to be
5524 looked up. The behaviour when one of the lookups defers can be
5525 controlled by a keyword.
5527 (4) It is now possible to specify the separator character for use when
5528 multiple records are returned.
5530 38. The dnslists ACL condition has been extended: it is now possible to supply
5531 a list of IP addresses and/or domains to be looked up in a particular DNS
5534 39. Added log_selector=+queue_time_overall.
5536 40. When running the queue in the test harness, wait just a tad after forking a
5537 delivery process, to get repeatability of debugging output.
5539 41. Include certificate and key file names in error message when GnuTLS fails
5540 to set them up, because the GnuTLS error message doesn't include the name
5541 of the failing file when there is a problem reading it.
5543 42. Allow both -bf and -bF in the same test run.
5545 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5547 44. The "Exiscan patch" is now merged into the mainline Exim source.
5549 45. Sometimes the final signoff response after QUIT could fail to get
5550 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5551 < 0 before doing a fflush(). This bug looks as though it goes back to the
5552 introduction of TLS in release 3.20, but "sometimes" must have been rare
5553 because the tests only now provoked it.
5555 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5556 (this can affect the format of dates).
5558 47. exim_tidydb, when checking for the continued existence of a message for
5559 which it has found a message-specific retry record, was not finding
5560 messages that were in split spool directories. Consequently, it was
5561 deleting retry records that should have stayed in existence.
5563 48. Steve fixed some bugs in eximstats.
5565 49. The SPA authentication driver was not abandoning authentication and moving
5566 on to the next authenticator when an expansion was forced to fail,
5567 contradicting the general specification for all authenticators. Instead it
5568 was generating a temporary error. It now behaves as specified.
5570 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5571 (the order specifies the preference for clients). The order is now AES256,
5572 AES128, 3DES, ARCFOUR128.
5574 51. Small patch to Sieve code - explicitly set From: when generating an
5577 52. Exim crashed if a remote delivery caused a very long error message to be
5578 recorded - for instance if somebody sent an entire SpamAssassin report back
5579 as a large number of 550 error lines. This bug was coincidentally fixed by
5580 increasing the size of one of Exim's internal buffers (big_buffer) that
5581 happened as part of the Exiscan merge. However, to be on the safe side, I
5582 have made the code more robust (and fixed the comments that describe what
5585 53. Now that there can be additional text after "Completed" in log lines (if
5586 the queue_time_overall log selector is set), a one-byte patch to exigrep
5587 was needed to allow it to recognize "Completed" as not the last thing in
5590 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5591 patch that reportedly fixes this has been added. I am not expert enough to
5592 create a test for it. This is what the patch creator wrote:
5594 "I found a little strange behaviour of ldap code when working with
5595 Windows 2003 AD Domain, where users was placed in more than one
5596 Organization Units. When I tried to give exim partial DN, the exit code
5597 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5598 But simultaneously result of request was absolutely normal ldap result,
5599 so I produce this patch..."
5601 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5602 REFERENCE, so I have modified the code to exclude the patch when that macro
5605 55. Some experimental protocols are using DNS PTR records for new purposes. The
5606 keys for these records are domain names, not reversed IP addresses. The
5607 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5608 leaves it alone. Component reversal etc. now happens only for IP addresses.
5611 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5613 57. Double the size of the debug message buffer (to 2048) so that more of very
5614 long debug lines gets shown.
5616 58. The exicyclog utility now does better if the number of log files to keep
5617 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5619 59. Two changes related to the smtp_active_hostname option:
5621 (1) $smtp_active_hostname is now available as a variable.
5622 (2) The default for smtp_banner uses $smtp_active_hostname instead
5623 of $primary_hostname.
5625 60. The host_aton() function is supposed to be passed a string that is known
5626 to be a valid IP address. However, in the case of IPv6 addresses, it was
5627 not checking this. This is a hostage to fortune. Exim now panics and dies
5628 if the condition is not met. A case was found where this could be provoked
5629 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5630 components; fortuitously, this particular loophole had already been fixed
5631 by change 4.50/55 above.
5633 If there are any other similar loopholes, the new check in host_aton()
5634 itself should stop them being exploited. The report I received stated that
5635 data on the command line could provoke the exploit when Exim was running as
5636 exim, but did not say which command line option was involved. All I could
5637 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5638 running as the user.
5641 61. There was a buffer overflow vulnerability in the SPA authentication code
5642 (which came originally from the Samba project). I have added a test to the
5643 spa_base64_to_bits() function which I hope fixes it.
5646 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5647 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5649 63. The daemon start-up calls getloadavg() while still root for those OS that
5650 need the first call to be done as root, but it missed one case: when
5651 deliver_queue_load_max is set with deliver_drop_privilege. This is
5652 necessary for the benefit of the queue runner, because there is no re-exec
5653 when deliver_drop_privilege is set.
5655 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5656 This has been fixed.
5658 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5659 was in use, was not putting the data itself into the right store pool;
5660 consequently, it could be overwritten for a subsequent message in the same
5661 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5664 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5666 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5667 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5668 the function were treating the return as a boolean value, which happened to
5669 work because 0=false and not-0=true, but is not correct code.
5671 68. The host_aton() function was not handling scoped IPv6 addresses (those
5672 with, for example, "%eth0" on the end) correctly.
5674 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5675 negated items (that is, ~something) in unsigned ints. Some compilers
5676 apparently mutter when there is no cast.
5678 70. If an address verification called from an ACL failed, and did not produce a
5679 user-specific message (i.e. there was only a "system" message), nothing was
5680 put in $acl_verify_message. In this situation, it now puts the system
5683 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5684 session; change 4.31/43 added the unwanted input to the log line - except
5685 that it did not do this in the start of session case. It now does.
5687 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5688 This is wrong and can cause the other end to generate a synchronization
5689 error if it is another Exim or anything else that does the synchronization
5690 check. A QUIT command is no longer sent after a timeout.
5692 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5693 during host lookups.
5695 74. The defer_ok option of callout verification was not working if it was used
5696 when verifying addresses in header lines, that is, for this case:
5698 verify = header_sender/callout=defer_ok
5700 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5701 those file descriptors could be used for SMTP connections. If anything
5702 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5703 could be sent to the SMTP client, causing chaos. The daemon now opens
5704 stdin, stdout, and stderr to /dev/null when it puts itself into the
5707 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5708 log by default. The user can override this with suitable Perl magic.
5710 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5711 the log message when discard triggers, was not working for the DATA ACL or
5712 for the non-SMTP ACL.
5714 78. Error message wording change in sieve.c.
5716 79. If smtp_accept_max_per_host was set, the number of connections could be
5717 restricted to fewer than expected, because the daemon was trying to set up
5718 a new connection before checking whether the processes handling previous
5719 connections had finished. The check for completed processes is now done
5720 earlier. On busy systems, this bug wouldn't be noticed because something
5721 else would have woken the daemon, and it would have reaped the completed
5724 80. If a message was submitted locally by a user whose login name contained one
5725 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5726 It caused a spool format error. I have fixed the spool reading code. A
5727 related problem was that the "from" clause in the Received: line became
5728 illegal because of the space(s). It is now covered by ${quote_local_part.
5730 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5733 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5735 83. Patch from Sieve maintainer to fix unterminated string problem in
5736 "vacation" handling.
5738 84. Some minor changes to the Linux configuration files to help with other
5739 OS variants using glibc.
5741 85. One more patch for Sieve to update vacation handling to latest spec.
5744 ----------------------------------------------------
5745 See the note above about the 4.44 and 4.50 releases.
5746 ----------------------------------------------------
5752 1. Change 4.43/35 introduced a bug that caused file counts to be
5753 incorrectly computed when quota_filecount was set in an appendfile
5756 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5757 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5760 3. Give more explanation in the error message when the command for a transport
5761 filter fails to execute.
5763 4. There are several places where Exim runs a non-Exim command in a
5764 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5765 was being done only for the command run by the queryprogram router. It is
5766 now done for all such subprocesses. The other cases are: ${run, transport
5767 filters, and the commands run by the lmtp and pipe transports.
5769 5. Some older OS have a limit of 256 on the maximum number of file
5770 descriptors. Exim was using setrlimit() to set 1000 as a large value
5771 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5772 systems. I've change it so that if it can't get 1000, it tries for 256.
5774 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5775 was an oversight, and furthermore, ever since the addition of extra
5776 controls (e.g. 4.43/32), the checks on when to allow different forms of
5777 "control" were broken. There should now be diagnostics for all cases when a
5778 control that does not make sense is encountered.
5780 7. $recipients is now available in the predata ACL (oversight).
5782 8. Tidy the search cache before the fork to do a delivery from a message
5783 received from the command line. Otherwise the child will trigger a lookup
5784 failure and thereby defer the delivery if it tries to use (for example) a
5785 cached ldap connection that the parent has called unbind on.
5787 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5788 of $address_data from the recipient verification was clobbered by the
5789 sender verification.
5791 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5792 was its contents. (It was OK if the option was not defined at all.)
5794 11. A "Completed" log line is now written for messages that are removed from
5795 the spool by the -Mrm option.
5797 12. $host_address is now set to the target address during the checking of
5798 ignore_target_hosts.
5800 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5801 being passed; this would have caused $sender_host_name to have been used if
5802 matching the list had actually called for a host name (not very likely,
5803 since this list is usually IP addresses). A host name is now passed as
5806 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5807 code that specifies a non-restarting handler (typically sigaction() in
5808 modern systems) in an attempt to fix a rare and obscure crash bug.
5810 15. Narrowed the window for a race in the daemon that could cause it to ignore
5811 SIGCHLD signals. This is not a major problem, because they are used only to
5812 wake it up if nothing else does.
5814 16. A malformed maildirsize file could cause Exim to calculate negative values
5815 for the mailbox size or file count. Odd effects could occur as a result.
5816 The maildirsize information is now recalculated if the size or filecount
5819 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5820 support for a long time. Removed HAVE_SYS_VFS_H.
5822 18. Updated exipick to current release from John Jetmore.
5824 19. Allow an empty sender to be matched against a lookup in an address list.
5825 Previously the only cases considered were a regular expression, or an
5828 20. Exim went into a mad DNS lookup loop when doing a callout where the
5829 host was specified on the transport, if the DNS lookup yielded more than
5832 21. The RFC2047 encoding function was originally intended for short strings
5833 such as real names; it was not keeping to the 75-character limit for
5834 encoded words that the RFC imposes. It now respects the limit, and
5835 generates multiple encoded words if necessary. To be on the safe side, I
5836 have increased the buffer size for the ${rfc2047: expansion operator from
5839 22. Failure to deliver a bounce message always caused it to be frozen, even if
5840 there was an errors_to setting on the router. The errors_to setting is now
5843 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5844 canonical form (fully expanded) before being placed in
5845 $sender_host_address.
5847 24. Updated eximstats to version 1.33
5849 25. Include certificate and key file names in error message when GnuTLS fails
5850 to set them up, because the GnuTLS error message doesn't include the name
5851 of the failing file when there is a problem reading it.
5853 26. Expand error message when OpenSSL has problems setting up cert/key files.
5856 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5857 (this can affect the format of dates).
5859 28. exim_tidydb, when checking for the continued existence of a message for
5860 which it has found a message-specific retry record, was not finding
5861 messages that were in split spool directories. Consequently, it was
5862 deleting retry records that should have stayed in existence.
5864 29. eximstats updated to version 1.35
5865 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5866 1.35 - bugfix such that pie charts by volume are generated correctly
5868 30. The SPA authentication driver was not abandoning authentication and moving
5869 on to the next authenticator when an expansion was forced to fail,
5870 contradicting the general specification for all authenticators. Instead it
5871 was generating a temporary error. It now behaves as specified.
5873 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5874 (the order specifies the preference for clients). The order is now AES256,
5875 AES128, 3DES, ARCFOUR128.
5877 31. Small patch to Sieve code - explicitly set From: when generating an
5880 32. Exim crashed if a remote delivery caused a very long error message to be
5881 recorded - for instance if somebody sent an entire SpamAssassin report back
5882 as a large number of 550 error lines. This bug was coincidentally fixed by
5883 increasing the size of one of Exim's internal buffers (big_buffer) that
5884 happened as part of the Exiscan merge. However, to be on the safe side, I
5885 have made the code more robust (and fixed the comments that describe what
5888 33. Some experimental protocols are using DNS PTR records for new purposes. The
5889 keys for these records are domain names, not reversed IP addresses. The
5890 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5891 leaves it alone. Component reversal etc. now happens only for IP addresses.
5894 34. The host_aton() function is supposed to be passed a string that is known
5895 to be a valid IP address. However, in the case of IPv6 addresses, it was
5896 not checking this. This is a hostage to fortune. Exim now panics and dies
5897 if the condition is not met. A case was found where this could be provoked
5898 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5899 components; fortuitously, this particular loophole had already been fixed
5900 by change 4.50/55 or 4.44/33 above.
5902 If there are any other similar loopholes, the new check in host_aton()
5903 itself should stop them being exploited. The report I received stated that
5904 data on the command line could provoke the exploit when Exim was running as
5905 exim, but did not say which command line option was involved. All I could
5906 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5907 running as the user.
5910 35. There was a buffer overflow vulnerability in the SPA authentication code
5911 (which came originally from the Samba project). I have added a test to the
5912 spa_base64_to_bits() function which I hope fixes it.
5915 36. The daemon start-up calls getloadavg() while still root for those OS that
5916 need the first call to be done as root, but it missed one case: when
5917 deliver_queue_load_max is set with deliver_drop_privilege. This is
5918 necessary for the benefit of the queue runner, because there is no re-exec
5919 when deliver_drop_privilege is set.
5921 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5922 was in use, was not putting the data itself into the right store pool;
5923 consequently, it could be overwritten for a subsequent message in the same
5924 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5927 38. Sometimes the final signoff response after QUIT could fail to get
5928 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5929 < 0 before doing a fflush(). This bug looks as though it goes back to the
5930 introduction of TLS in release 3.20, but "sometimes" must have been rare
5931 because the tests only now provoked it.
5937 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5938 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5939 Now it returns an integer. A number of calls were still expecting a T/F
5940 return. Fortuitously, in all cases, the tests worked in OK situations,
5941 which is the norm. However, things would have gone wrong on any write
5942 failures on the smtp file descriptor. This function is used when sending
5943 messages over SMTP and also when doing verify callouts.
5945 2. When Exim is called to do synchronous delivery of a locally submitted
5946 message (the -odf or -odi options), it no longer closes stderr before doing
5949 3. Implemented the mua_wrapper option.
5951 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5953 5. Implemented the functions header_remove(), header_testname(),
5954 header_add_at_position(), and receive_remove_recipient(), and exported them
5957 6. If an ACL "warn" statement specified the addition of headers, Exim already
5958 inserted X-ACL-Warn: at the start if there was no header name. However, it
5959 was not making this test for the second and subsequent header lines if
5960 there were newlines in the string. This meant that an invalid header could
5961 be inserted if Exim was badly configured.
5963 7. Allow an ACL "warn" statement to add header lines at the start or after all
5964 the Received: headers, as well as at the end.
5966 8. Added the rcpt_4xx retry error code.
5968 9. Added postmaster_mailfrom=xxx to callout verification option.
5970 10. Added mailfrom=xxxx to the callout verification option, for verify=
5973 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5974 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5976 12. Inserted some casts to stop certain compilers warning when using pointer
5977 differences as field lengths or precisions in printf-type calls (mostly
5978 affecting debugging statements).
5980 13. Added optional readline() support for -be (dynamically loaded).
5982 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5983 same clock tick as a message's arrival, so that its received time was the
5984 same as the "first fail" time on the retry record, and that message
5985 remained on the queue past the ultimate address timeout, every queue runner
5986 would try a delivery (because it was past the ultimate address timeout) but
5987 after another failure, the ultimate address timeout, which should have then
5988 bounced the address, did not kick in. This was a "< instead of <=" error;
5989 in most cases the first failure would have been in the next clock tick
5990 after the received time, and all would be well.
5992 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5993 being recognized when the domain list was tested by the match_domain
5994 condition in an expansion string.
5996 16. Added the ${str2b64: operator.
5998 17. Exim was always calling setrlimit() to set a large limit for the number of
5999 processes, without checking whether the existing limit was already
6000 adequate. (It did check for the limit on file descriptors.) Furthermore,
6001 errors from getrlimit() and setrlimit() were being ignored. Now they are
6002 logged to the main and panic logs, but Exim does carry on, to try to do its
6003 job under whatever limits there are.
6005 18. Imported PCRE 5.0.
6007 19. Trivial typo in log message " temporarily refused connection" (the leading
6010 20. If the log selector return_path_on_delivery was set and an address was
6011 redirected to /dev/null, the delivery process crashed because it assumed
6012 that a return path would always be set for a "successful" delivery. In this
6013 case, the whole delivery is bypassed as an optimization, and therefore no
6016 21. Internal re-arrangement: the function for sending a challenge and reading
6017 a response while authentication was assuming a zero-terminated challenge
6018 string. It's now changed to take a pointer and a length, to allow for
6019 binary data in such strings.
6021 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6023 23. Exim was not respecting finduser_retries when seeking the login of the
6024 uid under which it was called; it was always trying 10 times. (The default
6025 setting of finduser_retries is zero.) Also, it was sleeping after the final
6026 failure, which is pointless.
6028 24. Implemented tls_on_connect_ports.
6030 25. Implemented acl_smtp_predata.
6032 26. If the domain in control=submission is set empty, Exim assumes that the
6033 authenticated id is a complete email address when it generates From: or
6034 Sender: header lines.
6036 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6037 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6038 chown and chgrp in /bin and hostname in /usr/bin.
6040 28. Exim was keeping the "process log" file open after each use, just as it
6041 does for the main log. This opens the possibility of it remaining open for
6042 long periods when the USR1 signal hits a daemon. Occasional processlog
6043 errors were reported, that could have been caused by this. Anyway, it seems
6044 much more sensible not to leave this file open at all, so that is what now
6047 29. The long-running daemon process does not normally write to the log once it
6048 has entered its main loop, and it closes the log before doing so. This is
6049 so that log files can straightforwardly be renamed and moved. However,
6050 there are a couple of unusual error situations where the daemon does write
6051 log entries, and I had neglected to close the log afterwards.
6053 30. The text of an SMTP error response that was received during a remote
6054 delivery was being truncated at 512 bytes. This is too short for some of
6055 the long messages that one sometimes sees. I've increased the limit to
6058 31. It is now possible to make retry rules that apply only when a message has a
6059 specific sender, in particular, an empty sender.
6061 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6062 it possible to be selective about when SMTP synchronization is enforced.
6064 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6066 32. Implemented hosts_connection_nolog.
6068 33. Added an ACL for QUIT.
6070 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6073 35. Added mailbox_size and mailbox_filecount to appendfile.
6075 36. Added control = no_multiline_responses to ACLs.
6077 37. There was a bug in the logic of the code that waits for the clock to tick
6078 in the case where the clock went backwards by a substantial amount such
6079 that the microsecond fraction of "now" was more than the microsecond
6080 fraction of "then" (but the whole seconds number was less).
6082 38. Added support for the libradius Radius client library this is found on
6083 FreeBSD (previously only the radiusclient library was supported).
6089 1. When certain lookups returned multiple values in the form name=value, the
6090 quoting of the values was not always being done properly. Specifically:
6091 (a) If the value started with a double quote, but contained no whitespace,
6093 (b) If the value contained whitespace other than a space character (i.e.
6094 tabs or newlines or carriage returns) it was not quoted.
6095 This fix has been applied to the mysql and pgsql lookups by writing a
6096 separate quoting function and calling it from the lookup code. The fix
6097 should probably also be applied to nisplus, ibase and oracle lookups, but
6098 since I cannot test any of those, I have not disturbed their existing code.
6100 2. A hit in the callout cache for a specific address caused a log line with no
6101 reason for rejecting RCPT. Now it says "Previous (cached) callout
6102 verification failure".
6104 3. There was an off-by-one bug in the queryprogram router. An over-long
6105 return line was truncated at 256 instead of 255 characters, thereby
6106 overflowing its buffer with the terminating zero. As well as fixing this, I
6107 have increased the buffer size to 1024 (and made a note to document this).
6109 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6110 when Exim is waiting for an SMTP response from a remote server, Exim
6111 restarts its select() call on the socket, thereby resetting its timeout.
6112 This is not a problem when such interrupts are rare. Somebody set up a cron
6113 job to run exiwhat every 2 minutes, which is less than the normal select()
6114 timeout (5 or 10 minutes). This meant that the select() timeout never
6115 kicked in because it was always reset. I have fixed this by comparing the
6116 time when an interrupt arrives with the time at the start of the first call
6117 to select(). If more time than the timeout has elapsed, the interrupt is
6118 treated as a timeout.
6120 5. Some internal re-factoring in preparation for the addition of Sieve
6121 extensions (by MH). In particular, the "personal" test is moved to a
6122 separate function, and given an option for scanning Cc: and Bcc: (which is
6123 not set for Exim filters).
6125 6. When Exim created an email address using the login of the caller as the
6126 local part (e.g. when creating a From: or Sender: header line), it was not
6127 quoting the local part when it contained special characters such as @.
6129 7. Installed new OpenBSD configuration files.
6131 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6132 try to make them clearer.
6134 9. Callout options, other than the timeout value, were being ignored when
6135 verifying sender addresses in header lines. For example, when using
6137 verify = header_sender/callout=no_cache
6139 the cache was (incorrectly) being used.
6141 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6142 only the Cygwin environment.
6144 11. When return_path_on_delivery was set as a log selector, if different remote
6145 addresses in the same message used different return paths and parallel
6146 remote delivery occurred, the wrong values would sometimes be logged.
6147 (Whenever a remote delivery process finished, the return path value from
6148 the most recently started remote delivery process was logged.)
6150 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6151 lines when AUTH and/or TLS are in use. This is the "received protocol"
6152 field. Exim used to use "asmtp" for authenticated SMTP, without any
6153 indication (in the protocol name) for TLS use. Now it follows the RFC and
6154 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6155 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6156 names appear in log lines as well as in Received: header lines.
6158 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6159 extensions, and comparison tests, and to fix some bugs.
6161 14. Changes to the "personal" filter test:
6163 (1) The test was buggy in that it was just doing the equivalent of
6164 "contains" tests on header lines. For example, if a user's address was
6165 anne@some.where, the "personal" test would incorrectly be true for
6167 To: susanne@some.where
6169 This test is now done by extracting each address from the header in turn,
6170 and checking the entire address. Other tests that are part of "personal"
6171 are now done using regular expressions (for example, to check local parts
6172 of addresses in From: header lines).
6174 (2) The list of non-personal local parts in From: addresses has been
6175 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6176 taken from the Sieve specification recommendations.
6178 (3) If the message contains any header line starting with "List-" it is
6179 treated as non-personal.
6181 (4) The test for "circular" in the Subject: header line has been removed
6182 because it now seems ill-conceived.
6184 15. Minor typos in src/EDITME comments corrected.
6186 16. Installed latest exipick from John Jetmore.
6188 17. If headers_add on a router specified a text string that was too long for
6189 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6190 of string_sprintf() is now avoided.
6192 18. $message_body_size was not set (it was always zero) when running the DATA
6193 ACL and the local_scan() function.
6195 19. For the "mail" command in an Exim filter, no default was being set for
6196 the once_repeat time, causing a random time value to be used if "once" was
6197 specified. (If the value happened to be <= 0, no repeat happened.) The
6198 default is now 0s, meaning "never repeat". The "vacation" command was OK
6199 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6200 (I found it when inspecting the code).
6202 20. There is now an overall timeout for performing a callout verification. It
6203 defaults to 4 times the callout timeout, which applies to individual SMTP
6204 commands during the callout. The overall timeout applies when there is more
6205 than one host that can be tried. The timeout is checked before trying the
6206 next host. This prevents very long delays if there are a large number of
6207 hosts and all are timing out (e.g. when the network connections are timing
6208 out). The value of the overall timeout can be changed by specifying an
6209 additional sub-option for "callout", called "maxwait". For example:
6211 verify = sender/callout=5s,maxwait=20s
6213 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6214 to the end before writing, but this should make it even safer).
6216 22. Exim was forgetting that it had advertised PIPELINING for the second and
6217 subsequent messages on an SMTP connection. It was also not resetting its
6218 memory on STARTTLS and an internal HELO.
6220 23. When Exim logs an SMTP synchronization error within a session, it now
6221 records whether PIPELINING has been advertised or not.
6223 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6224 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6225 rather than long int.
6227 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6229 26. Added the never_mail option to autoreply.
6235 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6236 crash if the getsockname() call failed; this can happen if a connection is
6237 closed very soon after it is established. The problem was simply in the
6238 order in which certain operations were done, causing Exim to try to write
6239 to the SMTP stream before it had set up the file descriptor. The bug has
6240 been fixed by making things happen in the correct order.
6246 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6247 before the connection was closed, thus losing the rejection response.
6249 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6250 some early Solaris releases, but causes trouble in current releases where
6251 socklen_t is defined.
6253 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6256 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6259 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6260 that releasing the top of it at the end releases what was used for sub-
6261 expansions (unless the block got too big). However, discard this block if
6262 the first thing is a variable or header, so that we can use its block when
6263 it is dynamic (useful for very large $message_headers, for example).
6265 6. Lookups now cache *every* query, not just the most recent. A new, separate
6266 store pool is used for this. It can be recovered when all lookup caches are
6267 flushed. Lookups now release memory at the end of their result strings.
6268 This has involved some general refactoring of the lookup sources.
6270 7. Some code has been added to the store_xxx() functions to reduce the amount
6271 of flapping under certain conditions.
6273 8. log_incoming_interface used to affect only the <= reception log lines. Now
6274 it causes the local interface and port to be added to several more SMTP log
6275 lines, for example "SMTP connection from", and rejection lines.
6277 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6279 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6281 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6282 could be overwritten at the end of the current message (or the start of a
6283 new message if it was set in a HELO ACL). The value is now preserved for
6284 the duration of the SMTP connection.
6286 12. If a transport had a headers_rewrite setting, and a matching header line
6287 contained an unqualified address, that address was qualified, even if it
6288 did not match any rewriting rules. The underlying bug was that the values
6289 of the flags that permit the existence of unqualified sender and recipient
6290 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6291 non-local messages, and by -bnq for local messages) were not being
6292 preserved with the message after it was received.
6294 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6295 "next input=" as part of the text comprising the host identity instead of
6296 the correct text. The code was using the same buffer for two different
6297 strings. However, depending on which order the printing function evaluated
6298 its arguments, the bug did not always show up. Under Linux, for example, my
6299 test suite worked just fine.
6301 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6302 doesn't work with some older versions of Perl. It has been changed to "my",
6303 which in any case is probably the better facility to use.
6305 15. A really picky compiler found some instances of statements for creating
6306 error messages that either had too many or two few arguments for the format
6309 16. The size of the buffer for calls to the DNS resolver has been increased
6310 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6311 for addresses that have a lot of PTR records. This alleviates a problem; it
6312 does not fully solve it.
6314 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6315 buffer now truncates the list and logs the incident, which is the same
6316 action as happens when Exim is looking up a host name and its aliases.
6317 Previously in this situation something unpredictable would happen;
6318 sometimes it was "internal error: store_reset failed".
6320 18. If a server dropped the connection unexpectedly when an Exim client was
6321 using GnuTLS and trying to read a response, the client delivery process
6322 crashed while trying to generate an error log message.
6324 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6325 string, for example:
6327 warn message = H1: something\nH2: something
6329 the text was added as a single header line from Exim's point of view
6330 though it ended up OK in the delivered message. However, searching for the
6331 second and subsequent header lines using $h_h2: did not work. This has been
6332 fixed. Similarly, if a system filter added multiple headers in this way,
6333 the routers could not see them.
6335 20. Expanded the error message when iplsearch is called with an invalid key to
6336 suggest using net-iplsearch in a host list.
6338 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6339 ACLs are no longer actually imposed (and a message to that effect is
6342 22. If a "gecos" field in a passwd entry contained escaped characters, in
6343 particular, if it contained a \" sequence, Exim got it wrong when building
6344 a From: or a Sender: header from that name. A second bug also caused
6345 incorrect handling when an unquoted " was present following a character
6346 that needed quoting.
6348 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6349 was not being matched caselessly.
6351 24. Arranged for all hyphens in the exim.8 source to be escaped with
6354 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6355 properly. Recipient callout cache records were still being keyed to include
6356 the sender, even when use_sender was set false. This led to far more
6357 callouts that were necessary. The sender is no longer included in the key
6358 when use_sender is false.
6360 26. Added "control = submission" modifier to ACLs.
6362 27. Added the ${base62d: operator to decode base 62 numbers.
6364 28. dnsdb lookups can now access SRV records.
6366 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6367 the configuration file.
6369 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6370 (-v) mode. This makes the output for a verbose queue run more intelligible.
6372 31. Added a use_postmaster feature to recipient callouts.
6374 32. Added the $body_zerocount variable, containing the number of binary zero
6375 bytes in the message body.
6377 33. The time of last modification of the "new" subdirectory is now used as the
6378 "mailbox time last read" when there is a quota error for a maildir
6381 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6383 35. Added +ignore_unknown as a special item in host lists.
6385 36. Code for decoding IPv6 addresses in host lists is now included, even if
6386 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6387 address was recognized as an IP address, but was then not correctly decoded
6388 into binary, causing unexpected and incorrect effects when compared with
6395 1. Very minor rewording of debugging text in manualroute to say "list of
6396 hosts" instead of "hostlist".
6398 2. If verify=header_syntax was set, and a header line with an unqualified
6399 address (no domain) and a large number of spaces between the end of the
6400 name and the colon was received, the reception process suffered a buffer
6401 overflow, and (when I tested it) crashed. This was caused by some obsolete
6402 code that should have been removed. The fix is to remove it!
6404 3. When running in the test harness, delay a bit after writing a bounce
6405 message to get a bit more predictability in the log output.
6407 4. Added a call to search_tidyup() just before forking a reception process. In
6408 theory, someone could use a lookup in the expansion of smtp_accept_max_
6409 per_host which, without the tidyup, could leave open a database connection.
6411 5. Added the variables $recipient_data and $sender_data which get set from a
6412 lookup success in an ACL "recipients" or "senders" condition, or a router
6413 "senders" option, similar to $domain_data and $local_part_data.
6415 6. Moved the writing of debug_print from before to after the "senders" test
6418 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6419 problems for message scanning, either using a data ACL, or using
6420 local_scan() because the Received: header was not generated till after they
6421 were called (in order to set the time as the time of reception completion).
6422 I have revised the way this works. The header is now generated after the
6423 body is received, but before the ACL or local_scan() are called. After they
6424 are run, the timestamp in the header is updated.
6430 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6431 before starting a queue runner without re-exec. This happened only when
6432 deliver_drop_privilege was set or when the Exim user was set to root. The
6433 effect of the bug was that timeouts during subsequent deliveries caused
6434 crashes instead of being properly handled. The handler is now left at its
6435 default (and expected) setting.
6437 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6438 message, again when deliver_drop_privilege is set or Exim is run as root.
6439 The bug described in (1) was not present in this case, but the tidying up
6440 of the other signals was missing. I have made the two cases consistent.
6442 3. The ignore_target_hosts setting on a manualroute router was being ignored
6443 for hosts that were looked up using the /MX notation.
6445 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6448 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6449 operated on the sender address. After changing the $sender_address to <>
6450 for the sender address verify, Exim was re-instated it as the original
6451 (before rewriting) address, but remembering that it had rewritten it, so it
6452 wasn't rewriting it again. This bug also had the effect of breaking the
6453 sender address verification caching when the sender address was rewritten.
6455 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6456 This has been changed so that if the ip literal address matches
6457 ignore_target_hosts, the router declines.
6459 7. Added expansion conditions match_domain, match_address, and match_local_
6460 part (NOT match_host).
6462 8. The placeholder for the Received: header didn't have a length field set.
6464 9. Added code to Exim itself and to exim_lock to test for a specific race
6465 condition that could lead to file corruption when using MBX delivery. The
6466 issue is with the lockfile that is created in /tmp. If this file is removed
6467 after a process has opened it but before that process has acquired a lock,
6468 there is the potential for a second process to recreate the file and also
6469 acquire a lock. This could lead to two Exim processes writing to the file
6470 at the same time. The added code performs the same test as UW imapd; it
6471 checks after acquiring the lock that its file descriptor still refers to
6472 the same named file.
6474 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6475 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6478 11. Added the smtp_active_hostname option. If used, this will typically be made
6479 to depend on the incoming interface address. Because $interface_address is
6480 not set up until the daemon has forked a reception process, error responses
6481 that can happen earlier (such as "too many connections") no longer contain
6484 12. If an expansion in a condition on a "warn" statement fails because a lookup
6485 defers, the "warn" statement is abandoned, and the next ACL statement is
6486 processed. Previously this caused the whole ACL to be aborted.
6488 13. Added the iplsearch lookup type.
6490 14. Added ident_timeout as a log selector.
6492 15. Added tls_certificate_verified as a log selector.
6494 16. Added a global option tls_require_ciphers (compare the smtp transport
6495 option of the same name). This controls incoming TLS connections.
6497 17. I finally figured out how to make tls_require_ciphers do a similar thing
6498 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6499 before starting the TLS session.
6501 18. Tabs are now shown as \t in -bP output.
6503 19. If the log selector return_path_on_delivery was set, Exim crashed when
6504 bouncing a message because it had too many Received: header lines.
6506 20. If two routers both had headers_remove settings, and the first one included
6507 a superfluous trailing colon, the final name in the first list and the
6508 first name in the second list were incorrectly joined into one item (with a
6509 colon in the middle).
6515 1. Added -C and -D options to the exinext utility, mainly to make it easier
6516 to include in the automated testing, but these could be helpful when
6517 multiple configurations are in use.
6519 2. The exinext utility was not formatting the output nicely when there was
6520 an alternate port involved in the retry record key, nor when there was a
6521 message id as well (for retries that were specific to a specific message
6522 and a specific host). It was also confused by IPv6 addresses, because of
6523 the additional colons they contain. I have fixed the IPv4 problem, and
6524 patched it up to do a reasonable job for IPv6.
6526 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6527 delivery, the log line now contains "pipelined" if PIPELINING was used.
6529 4. An SMTP transport process used to panic and die if the bind() call to set
6530 an explicit outgoing interface failed. This has been changed; it is now
6531 treated in the same way as a connect() failure.
6533 5. A reference to $sender_host_name in the part of a conditional expansion
6534 that was being skipped was still causing a DNS lookup. This no longer
6537 6. The def: expansion condition was not recognizing references to header lines
6538 that used bh_ and bheader_.
6540 7. Added the _cache feature to named lists.
6542 8. The code for checking quota_filecount in the appendfile transport was
6543 allowing one more file than it should have been.
6545 9. For compatibility with Sendmail, the command line option
6553 and sets the incoming protocol and host name (for trusted callers). The
6554 host name and its colon can be omitted when only the protocol is to be set.
6555 Note the Exim already has two private options, -pd and -ps, that refer to
6556 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6557 "s", but I don't think that's a major issue.
6559 10. A number of refactoring changes to the code, none of which should affect
6562 (a) The number of logging options was getting close to filling up the
6563 32-bit word that was used as a bit map. I have split them into two classes:
6564 those that are passed in the argument to log_write(), and those that are
6565 only ever tested independently outside of that function. These are now in
6566 separate 32-bit words, so there is plenty of room for expansion again.
6567 There is no change in the user interface or the logging behaviour.
6569 (b) When building, for example, log lines, the code previously used a
6570 macro that called string_cat() twice, in order to add two strings. This is
6571 not really sufficiently general. Furthermore, there was one instance where
6572 it was actually wrong because one of the argument was used twice, and in
6573 one call a function was used. (As it happened, calling the function twice
6574 did not affect the overall behaviour.) The macro has been replaced by a
6575 function that can join an arbitrary number of extra strings onto a growing
6578 (c) The code for expansion conditions now uses a table and a binary chop
6579 instead of a serial search (which was left over from when there were very
6580 few conditions). Also, it now recognizes conditions like "pam" even when
6581 the relevant support is not compiled in: a suitably worded error message is
6582 given if an attempt is made to use such a condition.
6584 11. Added ${time_interval:xxxxx}.
6586 12. A bug was causing one of the ddress fields not to be passed back correctly
6587 from remote delivery subprocesses. The field in question was not being
6588 subsequently used, so this caused to problems in practice.
6590 13. Added new log selectors queue_time and deliver_time.
6592 14. Might have fixed a bug in maildirsizefile handling that threw up
6593 "unexpected character" debug warnings, and recalculated the data
6594 unnecessarily. In any case, I expanded the warning message to give more
6597 15. Added the message "Restricted characters in address" to the statements in
6598 the default ACL that block characters like @ and % in local parts.
6600 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6601 Three changes have been made:
6603 (a) There was a serious bug; a negative response to MAIL caused the whole
6604 recipient domain to be cached as invalid, thereby blocking all messages
6605 to all local parts at the same domain, from all senders. This bug has
6606 been fixed. The domain is no longer cached after a negative response to
6607 MAIL if the sender used is not empty.
6609 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6612 (c) A new callout option, "use_sender" has been added for people who want
6613 the modified behaviour.
6619 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6622 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6623 indeed breaks things for older releases.
6625 3. Added additional logging to the case where there is a problem reading data
6626 from a filter that is running in a subprocess using a pipe, in order to
6627 try to track down a specific problem.
6629 4. Testing facility fudge: when running in the test harness and attempting
6630 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6631 getting "No route to host". Convert this to a timeout.
6633 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6636 6. Some OS don't have socklen_t but use size_t instead. This affects the
6637 fifth argument of getsockopt() amongst other things. This is now
6638 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6639 can be set for individual OS. I have set it for SunOS5, OSF1, and
6640 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6641 some earlier ones do not.
6643 7. Change 4.30/15 was not doing the test caselessly.
6645 8. The standard form for an IPv6 address literal was being rejected by address
6646 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6647 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6648 this, as well as the form without the "IPv6" on the front (but only when
6649 address literals are enabled, of course).
6651 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6653 10. Exim crashed if a message with an empty sender address specified by -f
6654 encountered a router with an errors_to setting. This could be provoked only
6655 by a command such as
6659 where an empty string was supplied; "<>" did not hit this bug.
6661 11. Installed PCRE release 4.5.
6663 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6664 remained set. It is now erased.
6666 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6667 times from message ids (which are base 36 rather than the normal 62).
6669 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6670 were being counted as actual protocol errors, and logged if the log
6671 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6672 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6673 and DATA following a set of rejected RCPTs do not count as protocol errors.
6674 In other words, Exim assumes they were pipelined, though this may not
6675 actually be the case. Of course, in all cases the client gets an
6676 appropriate error code.
6678 15. If a lookup fails in an ACL condition, a message about the failure may
6679 be available; it is used if testing the ACL cannot continue, because most
6680 such messages specify what the cause of the deferral is. However, some
6681 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6682 that caused an old message to be retained and used if a later statement
6683 caused a defer, replacing the real cause of the deferral.
6685 16. If an IP address had so many PTR records that the DNS lookup buffer
6686 was not large enough to hold them, Exim could crash while trying to process
6687 the truncated data. It now detects and logs this case.
6689 17. Further to 4.21/58, another change has been made: if (and only if) the
6690 first line of a message (the first header line) ends with CRLF, a bare LF
6691 in a subsequent header line has a space inserted after it, so as not to
6692 terminate the header.
6694 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6695 unnecessarily, used atoi() instead of strtol(), and didn't check the
6696 termination when getting file sizes from file names by regex.
6698 19. Completely re-implemented the support for maildirsize files, in the light
6699 of a number of problems with the previous contributed implementation
6700 (4.30/29). In particular:
6702 . If the quota is zero, the maildirsize file is maintained, but no quota is
6705 . If the maildir directory does not exist, it is created before any attempt
6706 to write a maildirsize file.
6708 . The quota value in the file is just a cache; if the quota is changed in
6709 the transport, the new value overrides.
6711 . A regular expression is available for excluding directories from the
6714 20. The autoreply transport checks the characters in options that define the
6715 message's headers; it allows continued headers, but it was checking with
6716 isspace() after an embedded newline instead of explicitly looking for a
6719 21. If all the "regular" hosts to which an address was routed had passed their
6720 expiry times, and had not reached their retry times, the address was
6721 bounced, even if fallback hosts were defined. Now Exim should go on to try
6724 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6725 equivalent code in the SMTP transport. Some hosts send humungous responses
6726 to HELO/EHLO, more than 1024 it seems.
6728 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6729 gives compiler warnings in some environments. I've now done it "properly",
6732 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6733 (because of problems with the built-in one) was declared to return uschar *
6734 instead of char *, causing compiler failure.
6736 25. Fixed a file descriptor leak when processing alias/forward files.
6738 26. Fixed a minor format string issue in dbfn.c.
6740 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6742 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6743 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6746 29. When a sender address is verified, it is cached, to save repeating the test
6747 when there is more than one recipient in a message. However, when the
6748 verification involves a callout, it is possible for different callout
6749 options to be set for different recipients. It is too complicated to keep
6750 track of this in the cache, so now Exim always runs a verification when a
6751 callout is required, relying on the callout cache for the optimization.
6752 The overhead is duplication of the address routing, but this should not be
6755 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6756 to be verified with callout=postmaster, and the main callout worked but the
6757 postmaster check failed, the verification correctly failed. However, if a
6758 subsequent RCPT command asked for sender verification *without* the
6759 postmaster check, incorrect caching caused this verification also to fail,
6762 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6763 it was not caching the DNS options (qualify_single, search_parents) that
6764 were used when the lookup failed. A subsequent lookup with different
6765 options therefore always gave the same answer, though there were cases
6766 where it should not have. (Example: a "domains = !$mx_any" option on a
6767 dnslookup router: the "domains" option is always processed without any
6768 widening, but the router might have qualify_single set.) Now Exim uses the
6769 cached value only when the same options are set.
6771 32. Added John Jetmore's "exipick" utility to the distribution.
6773 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6774 than a timeout (e.g. a certificate is required, and is not provided), an
6775 Exim server now closes the connection immediately. Previously it waited for
6776 the client to close - but if the client is SSL, it seems that they each
6777 wait for each other, leading to a delay before one of them times out.
6779 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6780 maintained 0.8.x compatibility because I don't think many are using it, and
6781 it is clearly obsolete.
6783 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6786 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6787 client certificate was expired. A simple patch fixes this, though I don't
6788 understand the full logic of why the verify callback is called multiple
6791 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6792 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6793 which causes problems with some clients (such as the Certicom SSL Plus
6794 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6795 disables the coutermeasure allowing Eudora to connect."
6797 38. Exim was not checking that a write() to a log file succeeded. This could
6798 lead to Bad Things if a log got too big, in particular if it hit a file
6799 size limit. Exim now panics and dies if it cannot write to a log file, just
6800 as it does if it cannot open a log file.
6802 39. Modified OS/Makefile-Linux so that it now contains
6804 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6806 The two -D definitions ensure that Exim is compiled with large file
6807 support, which makes it possible to handle log files that are bigger than
6810 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6811 instance) a domain was checked against a named list that involved a lookup,
6812 causing $domain_data to be set, then another domain was checked against the
6813 same list, then the first domain was re-checked, the value of $domain_data
6814 after the final check could be wrong. In particular, if the second check
6815 failed, it could be set empty. This bug probably also applied to
6818 41. The strip_trailing_dot option was not being applied to the address given
6819 with the -f command-line option.
6821 42. The code for reading a message's header from the spool was incrementing
6822 $received_count, but never initializing it. This meant that the value was
6823 incorrect (doubled) while delivering a message in the same process in which
6824 it was received. In the most common configuration of Exim, this never
6825 happens - a fresh exec is done - but it can happen when
6826 deliver_drop_privilege is set.
6828 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6829 it now includes up to 150 characters of the unexpected data in the log
6832 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6833 and building data strings. The size of both of these buffers was 10 000
6834 bytes - far larger than anybody would *ever* want, thought I. Needless to
6835 say, somebody hit the limit. I have increased the maximum line length to
6836 20 000 and the maximum data length of concatenated lines to 100 000. I have
6837 also fixed two bugs, because there was no checking on these buffers. Tsk,
6838 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6839 buffer is too small.
6841 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6842 lsearch lookups. Now it does.
6844 46. When parsing a route_list item in a manualroute router, a fixed-length
6845 buffer was used for the list of hosts. I made this 1024 bytes long,
6846 thinking that nobody would ever have a list of hosts that long. Wrong.
6847 Somebody had a whole pile of complicated expansion conditions, and the
6848 string was silently truncated, leading to an expansion error. It turns out
6849 that it is easier to change to an unlimited length (owing to other changes
6850 that have happened since this code was originally written) than to build
6851 structure for giving a limitation error. The length of the item that
6852 expands into the list of hosts is now unlimited.
6854 47. The lsearch lookup could not handle data where the length of text line was
6855 more than 4095 characters. Such lines were truncated, leading to shortened
6856 data being returned. It should now handle lines of any length.
6858 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6859 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6862 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6863 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6864 no longer happens in comment lines. A list of replacements is now placed
6865 at the head of all of the source files, except those whose only change is
6866 to replace PERL_COMMAND in the very first #! line.
6868 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6869 messages on the queue, with a bottom-up merge sort, using code contributed
6870 by Michael Haardt. This should make operations like -bp somewhat faster on
6871 large queues. It won't affect queue runners, except when queue_run_in_order
6874 51. Installed eximstats 1.31 in the distribution.
6876 52. Added support for SRV lookups to the dnslookup router.
6878 53. If an ACL referred to $message_body or $message_body_end, the value was not
6879 reset for any messages that followed in the same SMTP session.
6881 54. The store-handling optimization for building very long strings was not
6882 differentiating between the different store pools. I don't think this
6883 actually made any difference in practice, but I've tidied it.
6885 55. While running the routers to verify a sender address, $sender_address
6886 was still set to the sender address. This is wrong, because when routing to
6887 send a bounce to the sender, it would be empty. Therefore, I have changed
6888 it so that, while verifying a sender address, $sender_address is set to <>.
6889 (There is no change to what happens when verifying a recipient address.)
6891 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6892 target A or AAAA records (if not already returned) without resetting the
6893 qualify_single or search_parents options of the DNS resolver. These are
6894 inappropriate in this case because the targets of MX and SRV records must
6895 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6896 target that, when qualified, matched something in the local domain. These
6897 two options are now turned off when doing these lookups.
6899 57. It seems that at least some releases of Reiserfs (which does not have the
6900 concept of a fixed number of inodes) returns zero and not -1 for the
6901 number of available inodes. This interacted badly with check_spool_inodes,
6902 which assumed that -1 was the "no such thing" setting. What I have done is
6903 to check that the total number of inodes is greater than zero before doing
6904 the test of how many are available.
6906 58. When a "warn" ACL statement has a log_message modifier, the message is
6907 remembered, and not repeated. This is to avoid a lot of repetition when a
6908 message has many recipients that cause the same warning to be written.
6909 However, Exim was preserving the list of already written lines for an
6910 entire SMTP session, which doesn't seem right. The memory is now reset if a
6911 new message is started.
6913 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6914 debugging output unless log_rewrite was also set.
6916 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6917 of (int)(handle) when we know that handle contains (void *)(-1).
6919 61. The Exim daemon panic-logs an error return when it closes the incoming
6920 connection. However "connection reset by peer" seems to be common, and
6921 isn't really an error worthy of noting specially, so that particular error
6924 62. When Exim is trying to find all the local interfaces, it used to panic and
6925 die if the ioctl to get the interface flags failed. However, it seems that
6926 on at least one OS (Solaris 9) it is possible to have an interface that is
6927 included in the list of interfaces, but for which you get a failure error
6928 for this call. This happens when the interface is not "plumbed" into a
6929 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6930 failure of the "get flags" call assumes that the interface is down.
6932 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6933 makes life easier for people who are doing arithmetic on fields extracted
6934 from dates, where you often get leading zeros that should not be
6935 interpreted as octal.
6937 64. Added qualify_domain to the redirect router, to override the global
6940 65. If a pathologically long header line contained very many addresses (the
6941 report of this problem mentioned 10 000) and each of them was rewritten,
6942 Exim could use up a very large amount of memory. (It kept on making new
6943 copies of the header line as it rewrote, and never released the old ones.)
6944 At the expense of a bit more processing, the header rewriting function has
6945 been changed so that it no longer eats memory in this way.
6947 66. The generation of the Received: header has been moved from the time that a
6948 message starts to be received, to the time that it finishes. The timestamp
6949 in the Received: header should now be very close to that of the <= log
6950 line. There are two side-effects of this change:
6952 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6953 logged header lines no longer include the local Received: line, because
6954 it has not yet been created. The same applies to a copy of the message
6955 that is returned to a non-SMTP sender when a message is rejected.
6957 (b) When a filter file is tested using -bf, no additional Received: header
6958 is added to the test message. After some thought, I decided that this
6961 This change does not affect the value of $received_for. It is still set
6962 after address rewriting, but before local_scan() is called.
6964 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6966 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6967 gave an unhelpful panic error message, and a defer error. I have managed to
6968 change this behaviour so that it now rejects any supplied certificate,
6969 which seems right, as the list of acceptable certificates is empty.
6971 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6972 gave an unhelpful defer error. I have not managed to make this reject any
6973 supplied certificates, but the error message it gives is "no certificate
6974 supplied", which is not helpful.
6976 70. exigrep's output now also includes lines that are not associated with any
6977 message, but which match the given pattern. Implemented by a patch from
6978 Martin Sluka, which also tidied up the Perl a bit.
6980 71. Recipient callout verification, like sender verification, was using <> in
6981 the MAIL FROM command. This isn't really the right thing, since the actual
6982 sender may affect whether the remote host accepts the recipient or not. I
6983 have changed it to use the actual sender in the callout; this means that
6984 the cache record is now keyed on a recipient/sender pair, not just the
6985 recipient address. There doesn't seem to be a real danger of callout loops,
6986 since a callout by the remote host to check the sender would use <>.
6987 [SEE ABOVE: changed after hitting problems.]
6989 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6990 temporary errors. However, in the case of such a code being given after
6991 the end of a data transmission (i.e. after ".") Exim was failing to write
6992 a retry record for the message. (Yes, there was some broken host that was
6993 actually sending 8xx at this point.)
6995 73. An unknown lookup type in a host list could cause Exim to panic-die when
6996 the list was checked. (An example that provoked this was putting <; in the
6997 middle of a list instead of at the start.) If this happened during a DATA
6998 ACL check, a -D file could be left lying around. This kind of configuration
6999 error no longer causes Exim to die; instead it causes a defer error. The
7000 incident is still logged to the main and panic logs.
7002 74. Buglet left over from Exim 3 conversion. The message "too many messages
7003 in one connection" was written to the rejectlog but not the mainlog, except
7004 when address rewriting (yes!) was being logged.
7006 75. Added write_rejectlog option.
7008 76. When a system filter was run not as root (that is, when system_filter_user
7009 was set), the values of the $n variables were not being returned to the
7010 main process; thus, they were not subsequently available in the $sn
7013 77. Added +return_path_on_delivery log selector.
7015 78. A connection timeout was being treated differently from recipients deferred
7016 when testing hosts_max_try with a message that was older than the host's
7017 retry timeout. (The host should not be counted, thus allowing all hosts to
7018 be tried at least once before bouncing.) This may have been the cause of an
7019 occasionally reported bug whereby a message would remain on the queue
7020 longer than the retry timeout, but would be bounced if a delivery was
7021 forced. I say "may" because I never totally pinned down the problem;
7022 setting up timeout/retry tests is difficult. See also the next item.
7024 79. The ultimate address timeout was not being applied to errors that involved
7025 a combination of host plus message (for example, a timeout on a MAIL
7026 command). When an address resolved to a number of possible hosts, and they
7027 were not all tried for each delivery (e.g. because of hosts_max_try), a
7028 message could remain on the queue longer than the retry timeout.
7030 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7033 81. Fixed an obscure SMTP outgoing bug which required at least the following
7034 conditions: (a) there was another message waiting for the same server;
7035 (b) the server returned 5xx to all RCPT commands in the first message so
7036 that the message was not completed; (c) the server dropped the connection
7037 or gave a negative response to the RSET that Exim sends to abort the
7038 transaction. The observed case was a dropped connection after DATA that had
7039 been sent in pipelining mode. That is, the server had advertised PIPELINING
7040 but was not implementing it correctly. The effect of the bug was incorrect
7041 behaviour, such as trying another host, and this could lead to a crash.
7047 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7048 and daemon.c were passed as pointers to ints; they should have been
7049 pointers to socklen_t variables (which are typically unsigned ints).
7051 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7054 3. Fixed a really odd bug that affected only the testing scheme; patching a
7055 certain fixed string in the binary changed the value of another string that
7056 happened to be identical to the end of the original first string.
7058 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7059 name", it returns that address as the IP address. On some operating
7060 systems (e.g. Solaris), it also passes back the IP address string as the
7061 "host name". However, on others (e.g. Linux), it passes back an empty
7062 string. Exim wasn't checking for this, and was changing the host name to an
7063 empty string, assuming it had been canonicalized.
7065 5. Although rare, it is permitted to have more than one PTR record for a given
7066 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7067 all the names associated with an address, because they do in Solaris.
7068 However, it seems that they do not in Linux for data that comes from the
7069 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7070 I found this out when I moved to a new Linux workstation and tried to run
7071 the Exim test suite.
7073 To get round this problem I have changed the code so that it now does its
7074 own call to the DNS to look up PTR records when searching for a host name.
7075 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7076 addresses that are only in /etc/hosts are still found.
7078 This behaviour is, however, controlled by an option called host_lookup_
7079 order, which defaults to "bydns:byaddr". If people want to use the other
7080 order, or indeed, just use one or the other means of lookup, they can
7081 specify it in this variable.
7083 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7084 some operating systems, this comes back from gethostbyaddr() as an empty
7085 string, and this is what Exim used to test for. However, it seems that in
7086 other systems, "." is yielded. Exim now tests for this case too.
7088 7. The values of check_spool_space and check_log_space are now held internally
7089 as a number of kilobytes instead of an absolute number of bytes. If a
7090 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7091 kilobyte. This means that much larger values can be stored.
7093 8. Exim monitor: an attempt to get the action menu when not actually pointing
7094 at a message produces an empty menu entitled "No message selected". This
7095 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7096 no entries in it ("Shell widget menu has zero width and/or height"). So I
7097 have added a single, blank menu entry in this case.
7099 9. Added ${quote_local_part.
7101 10. MIME decoding is now applied to the contents of Subject: header lines when
7104 11. Now that a reference to $sender_host_address automatically causes a reverse
7105 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7106 host lookup before query-style lookups in lists that might use this
7107 variable. This has therefore been abolished, and the "net-" prefix is no
7108 longer necessary for query-style lookups.
7110 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7111 have been a typo for LFLAGS, so it has been changed.
7113 13. The install script calls Exim with "-C /dev/null" in order to find the
7114 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7115 to be output. However, since Exim outputs its version number before the
7116 error, it didn't break the script. It just looked ugly. I fixed this by
7117 always allowing "-C /dev/null" if the caller is root.
7119 14. Ignore overlarge ACL variable number when reading spool file - insurance
7120 against a later release with more variables having written the file.
7122 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7123 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7124 this, as well as the form without the "IPv6" on the front.
7126 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7127 OS/Makefile-Darwin file.
7129 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7130 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7132 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7133 variable. However, some OS use ints for inodes. I've added cast to long int
7134 to get rid of the compiler warning.
7136 19. I had forgotten to lock out "/../" in configuration file names when
7137 ALT_CONFIG_PREFIX was set.
7139 20. Routers used for verification do not need to specify transports. However,
7140 if such a router generated a host list, and callout was configured, Exim
7141 crashed, because it could not find a port number from the (non-existent)
7142 transport. It now assumes port 25 in this circumstance.
7144 21. Added the -t option to exigrep.
7146 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7147 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7148 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7149 them is set, LOOKUP_LSEARCH is forced.
7151 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7153 24. Added sender and host information to the "rejected by local_scan()" log
7154 line; previously there was no indication of these.
7156 25. Added .include_if_exists.
7158 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7159 Linux. It turns out that not all file systems support this. Apparently some
7160 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7161 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7162 not supported on the file descriptor, is now ignored when Exim is trying to
7163 sync a directory. This applies only to Linux.
7165 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7167 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7168 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7171 29. Added support for maildirsize files from supplied patch (modified a bit).
7173 30. The use of :fail: followed by an empty string could lead Exim to respond to
7174 sender verification failures with (e.g.):
7176 550 Verification failed for <xxx>
7177 550 Sender verify failed
7179 where the first response line was missing the '-' that indicates it is not
7180 the final line of the response.
7182 31. The loop for finding the name of the user that called Exim had a hardwired
7183 limit of 10; it now uses the value of finduser_retries, which is used for
7184 all other user lookups.
7186 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7189 33. Exim was neglecting to zero errno before one call of strtol() when
7190 expanding a string and expecting an integer value. On some systems this
7191 resulted in spurious "integer overflow" errors. Also, it was casting the
7192 result into an int without checking.
7194 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7195 did not work. The code looks as if it has *never* worked, though it appears
7196 to have been documented since at least release 1.62. I have made it work.
7198 35. The "timeout_DNS" error in retry rules, also documented since at least
7199 1.62, also never worked. As it isn't clear exactly what this means, and
7200 clearly it isn't a major issue, I have abolished the feature by treating it
7201 as "timeout", and writing a warning to the main and panic logs.
7203 36. The display of retry rules for -brt wasn't always showing the error code
7206 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7207 timeout_connect_A, timeout_connect_MX.
7209 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7210 to the empty sender.
7212 39. The daemon was not analysing the content of -oX till after it had closed
7213 stderr and disconnected from the controlling terminal. This meant that any
7214 syntax errors were only noted on the panic log, and the return code from
7215 the command was 0. By re-arranging the code a little, I've made the
7216 decoding happen first, so such errors now appear on stderr, and the return
7217 code is 1. However, the actual setting up of the sockets still happens in
7218 the disconnected process, so errors there are still only recorded on the
7221 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7222 connections (as happens on some IP stacks) was logged at start up time as
7223 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7224 it from "IPv6 and IPv4", which means that two separate sockets are being
7227 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7228 says whether AF_INET or AF_INET6 was passed as an argument.
7230 42. Exiwhat output was messed up when time zones were included in log
7233 43. Exiwhat now gives more information about the daemon's listening ports,
7234 and whether -tls-on-connect was used.
7236 44. The "port" option of the smtp transport is now expanded.
7238 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7239 silently ignored. Now an error message is written to the main and panic
7242 46. There's a new ACL modifier called "logwrite" which writes to a log file
7243 as soon as it is encountered.
7245 47. Added $local_user_uid and $local_user_gid at routing time.
7247 48. Exim crashed when trying to verify a sender address that was being
7250 49. Exim was recognizing only a space character after ".include". It now also
7251 recognizes a tab character.
7253 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7254 extracting the relevant information from the specification. The man page no
7255 longer contains scrambled data for the -d option, and I've added a section
7256 at the front about calling Exim under different names.
7258 51. Added "extra_headers" argument to the "mail" command in filter files.
7260 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7263 53. Installed eximstats 1.29.
7265 54. Added transport_filter_timeout as a generic transport option.
7267 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7268 Cc: header lines. This was required by RFC 822, but it not required by RFC
7271 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7272 incoming messages that did not have them. Now it does so only if the
7273 message originates locally, that is, if there is no associated remote host
7274 address. When Resent- header lines are present, this applies to the Resent-
7275 lines rather than the non-Resent- lines.
7277 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7278 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7280 58. Messages for configuration errors now include the name of the main
7281 configuration file - useful now that there may be more than one file in a
7282 list (.included file names were always shown).
7284 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7285 for those rare installations that do not start the daemon as root or run it
7286 setuid root. I've cut out the call to initgroups() if the daemon is not
7289 60. The Exim user and group can now be bound into the binary as text strings
7290 that are looked up at the start of Exim's processing.
7292 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7294 62. Added $mailstore_basename variable.
7296 63. Installed patch to sieve.c from Michael Haardt.
7298 64. When Exim failed to open the panic log after failing to open the main log,
7299 the original message it was trying to log was written to stderr and debug
7300 output, but if they were not available (the usual case in production), it
7301 was lost. Now it is written to syslog before the two lines that record the
7302 failures to open the logs.
7304 65. Users' Exim filters run in subprocesses under the user's uid. It is
7305 possible for a "deliver" command or an alias in a "personal" command to
7306 provoke an address rewrite. If logging of address rewriting is configured,
7307 this fails because the process is not running as root or exim. There may be
7308 a better way of dealing with this, but for the moment (because 4.30 needs
7309 to be released), I have disabled address rewrite logging when running a
7310 filter in a non-root, non-exim process.
7316 1. The buildconfig auxiliary program wasn't quoting the value set for
7317 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7318 not defined. This bug was masked in 4.22 by the effect that was fixed in
7321 2. Some messages that were rejected after a message id was allocated were
7322 shown as "incomplete" by exigrep. It no longer does this for messages that
7323 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7325 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7326 have allow_domain_literals set, the ID did not get logged in the <= line.
7327 Domain literals are now always recognized in Message-ID: header lines.
7329 4. The first argument for a ${extract expansion item is the key name or field
7330 number. Leading and trailing spaces in this item were not being ignored,
7331 causing some misleading effects.
7333 5. When deliver_drop_privilege was set, single queue runner processes started
7334 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7335 same command in the process it spins off) were not dropping privilege.
7337 6. When the daemon running as "exim" started a queue runner, it always
7338 re-executed Exim in the spun-off process. This is a waste of effort when
7339 deliver_drop_privilege is set. The new process now just calls the
7340 queue-runner function directly.
7346 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7349 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7350 ignored. Though the use of -oP was forcing the writing of a pid file, it
7351 was always written to the default place.
7353 3. If the message "no IP address found for host xxxx" is generated during
7354 incoming verification, it is now followed by identification of the incoming
7355 connection (so you can more easily find what provoked it).
7357 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7359 5. Added some features to "harden" Exim a bit more against certain attacks:
7361 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7362 be put in Local/Makefile. This is like the never_users runtime option,
7363 but it cannot be overridden. The default setting is "root".
7365 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7366 prefix string with which any file named in a -C command line option
7369 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7370 is retained for -C and -D only if the caller of Exim is root. Without
7371 it, the exim user may also use -C and -D and retain privilege.
7373 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7374 command line option is disabled.
7376 6. Macro names set by the -D option must start with an upper case letter, just
7377 like macro names defined in the configuration file.
7379 7. Added "dereference=" facility to LDAP.
7381 8. Two instances of the typo "uknown" in the source files are fixed.
7383 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7384 the Configure-Makefile script screwed up while processing it.
7386 10. Incorporated PCRE 4.4.
7388 11. The SMTP synchronization check was not operating right at the start of an
7389 SMTP session. For example, it could not catch a HELO sent before the client
7390 waited for the greeting. There is now a check for outstanding input at the
7391 point when the greeting is written. Because of the duplex, asynchronous
7392 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7393 way, but not yet received, when the check is performed.
7395 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7396 on TCP/IP sockets, because this apparently causes some broken clients to
7399 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7400 unchanged) from the Cygwin maintainer.
7402 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7403 was supported instead of testing for mbx. Effectively a typo.
7405 15. The spa authenticator server code was not checking that the input it
7406 received was valid base64.
7408 16. The debug output line for the "set" modifier in ACLs was not showing the
7409 name of the variable that was being set.
7411 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7413 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7414 lookup on its own. Something else was needed to trigger the lookup. For
7415 example, a match in host_lookup or the need for a host name in a host list.
7416 Now, if $sender_host_name is referenced and the host name has not yet been
7417 looked up, a lookup is performed. If the lookup fails, the variable remains
7418 empty, and $host_lookup_failed is set to "1".
7420 19. Added "eqi" as a case-independent comparison operator.
7422 20. The saslauthd authentication condition could segfault if neither service
7423 nor realm was specified.
7425 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7426 error message that was logged was misleading, and incoming SMTP
7427 connections were dropped. The message is now more accurate, and temporary
7428 errors are given to SMTP connections.
7430 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7431 (except RSET) with a 421 error, until QUIT is received. However, it was
7432 failing to send a response to QUIT.
7434 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7435 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7436 ACL is now run after the helo_try_verify_hosts code.
7438 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7439 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7440 case-independent, so other case variants are also recognized). Apparently
7441 some systems use these upper case variants.
7443 25. If more than two messages were waiting for the same host, and a transport
7444 filter was specified for the transport, Exim sent two messages over the
7445 same TCP/IP connection, and then failed with "socket operation on non-
7446 socket" when it tried to send the third.
7448 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7450 27. The extern definition of crypt16() in expand.c was not being excluded when
7451 the OS had its own crypt16() function.
7453 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7454 as a preferred synonym for return_size_limit, both as an option and as an
7457 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7459 30. Changed the default configuration ACL to relax the local part checking rule
7460 for addresses that are not in any local domains. For these addresses,
7461 slashes and pipe symbols are allowed within local parts, but the sequence
7462 /../ is explicitly forbidden.
7464 31. SPA server authentication was not clearing the challenge buffer before
7467 32. log_message in a "warn" ACL statement was writing to the reject log as
7468 well as to the main log, which contradicts the documentation and doesn't
7469 seem right (because no rejection is happening). So I have stopped it.
7471 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7472 However, I am unable to do any testing of this.
7474 34. Fixed an infelicity in the appendfile transport. When checking directories
7475 for a mailbox, to see if any needed to be created, it was accidentally
7476 using path names with one or more superfluous leading slashes; tracing
7477 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7479 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7480 contents are added to the log line that is written for every discarded
7481 recipient. (Previously a log_message setting was ignored.)
7483 36. The ${quote: operator now quotes the string if it is empty.
7485 37. The install script runs exim in order to find its version number. If for
7486 some reason other than non-existence or emptiness, which it checks, it
7487 could not run './exim', it was installing it with an empty version number,
7488 i.e. as "exim-". This error state is now caught, and the installation is
7491 38. An argument was missing from the function that creates an error message
7492 when Exim fails to connect to the socket for saslauthd authentication.
7493 This could cause Exim to crash, or give a corrupted message.
7495 39. Added isip, isip4, and isip6 to ${if conditions.
7497 40. The ACL variables $acl_xx are now saved with the message, and can be
7498 accessed later in routers, transports, and filters.
7500 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7501 strings in the file are not string-expanded.
7503 42. If a MAIL command specified a SIZE value that was too large to fit into an
7504 int variable, the check against message_size_limit failed. Such values are
7505 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7506 day this will have to be increased, but I don't think I want to be around
7507 when emails are that large.
7514 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7515 iconv() is not standard in FreeBSD.
7517 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7518 IPv6 enabled. The observed symptom was a segmentation fault on return from
7519 the function os_common_find_running_interfaces() in src/os.c.
7521 3. In the check_special_case() function in daemon.c I had used "errno" as an
7522 argument name, which causes warnings on some systems. This was basically a
7523 typo, since it was named "eno" in the comments!
7525 4. The code that waits for the clock to tick (at a resolution of some fraction
7526 of a second) so as to ensure message-id uniqueness was always waiting for
7527 at least one whole tick, when it could have waited for less. [This is
7528 almost certainly not relevant at current processor speeds, where it is
7529 unlikely to ever wait at all. But we try to future-proof.]
7531 5. The function that sleeps for a time interval that includes fractions of a
7532 second contained a race. It did not block SIGALRM between setting the
7533 timer, and suspending (a couple of lines later). If the interval was short
7534 and the sigsuspend() was delayed until after it had expired, the suspension
7535 never ended. On busy systems this could lead to processes getting stuck for
7538 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7539 process, before it forks any delivery processes. The open lookup caching
7540 mechanism meant that the open file or database connection was passed into
7541 the delivery process. The problem was that delivery processes always tidy
7542 up cached lookup data. This could cause a problem for the next delivery
7543 process started by the queue runner, because the external queue runner
7544 process does not know about the closure. So the next delivery process
7545 still has data in the lookup cache. In the case of a file lookup, there was
7546 no problem because closing a file descriptor in a subprocess doesn't affect
7547 the parent. However, if the lookup was caching a connection to a database,
7548 the connection was closed, and the second delivery process was likely to
7549 see errors such as "PGSQL: query failed: server closed the connection
7550 unexpectedly". The problem has been fixed by closing all cached lookups
7551 in a queue runner before running a delivery process.
7553 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7554 seem to have the "const" qualifier which it has on other OS. I've
7557 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7558 *of the same type* (client or server) with the same public name that an
7559 error should be diagnosed.
7561 9. When Exim looked up a host name for an IP address, but failed to find the
7562 original IP address when looking up the host name (a safety check), it
7563 output the message "<ip address> does not match any IP for NULL", which was
7564 confusing, to say the least. The bug was that the host name should have
7565 appeared instead of "NULL".
7567 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7568 user that is built into the binary, and the -C or -D options is used, root
7569 privilege is dropped before the configuration file is read. In addition,
7570 logging is switched to stderr instead of the normal log files. If the
7571 configuration then re-defines the Exim user, the unprivileged environment
7572 is probably not what is expected, so Exim logs a panic warning message (but
7575 However, if deliver_drop_privilege is set, the unprivileged state may well
7576 be exactly what is intended, so the warning has been cut out in that case,
7577 and Exim is allowed to try to write to its normal log files.
7583 1. smtp_return_error_details was not giving details for temporary sender
7584 or receiver verification errors.
7586 2. Diagnose a configuration error if two authenticators have the same public
7589 3. Exim used not to create the message log file for a message until the first
7590 delivery attempt. This could be confusing when incoming messages were held
7591 for policy or load reasons. The message log file is now created at the time
7592 the message is received, and an initial "Received" line is written to it.
7594 4. The automatically generated man page for command line options had a minor
7595 bug that caused no ill effects; however, a more serious problem was that
7596 the procedure for building the man page automatically didn't always
7597 operate. Consequently, release 4.20 contains an out-of-date version. This
7598 shouldn't happen again.
7600 5. When building Exim with embedded Perl support, the script that builds the
7601 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7602 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7604 6. The freeze_tell option was not being used for messages that were frozen on
7605 arrival, either by an ACL or by local_scan().
7607 7. Added the smtp_incomplete_transaction log selector.
7609 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7610 was accepting AUTH without a new EHLO.
7612 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7613 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7616 10. Logging of TCP/IP connections (when configured) now happens in the main
7617 daemon process instead of the child process, so that the TCP/IP connection
7618 count is more accurate (but it can never be perfect).
7620 11. The use of "drop" in a nested ACL was not being handled correctly in the
7621 outer ACL. Now, if condition failure induced by the nested "drop" causes
7622 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7623 or "require"), the connection is dropped.
7625 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7626 that yield "discard" can now be used with an "accept" or a "discard" verb,
7627 but an error is generated for any others (because I can't see a useful way
7628 to define what should happen).
7630 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7631 are now processed in the same way as lines in the Exim configuration file.
7632 In particular, continuation lines are supported.
7634 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7636 15. Added -ti meaning -t -i.
7638 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7639 domains, and warn by logging if others are found.
7641 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7642 returned from GIFCONF when Exim is trying to find the list of interfaces on
7643 a host. The code in os.c has been modified to copy each ifreq to an aligned
7644 structure in all cases.
7646 Also, in some cases, the returned ifreq's were being copied to a 'struct
7647 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7648 means the last couple of bytes of an IPv6 address could be chopped if the
7649 ifreq contained only a normal sockaddr (14 bytes storage).
7651 18. Named domain lists were not supported in the hosts_treat_as_local option.
7652 An entry such as +xxxx was not recognized, and was treated as a literal
7655 19. Ensure that header lines added by a DATA ACL are included in the reject log
7656 if the ACL subsequently rejects the message.
7658 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7659 MD5 (which is deprecated).
7661 21. When testing a filter file using -bf, Exim was writing a message when it
7662 took the sender from a "From " line in the message, but it was not doing so
7663 when it took $return_path from a Return-Path: header line. It now does.
7665 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7666 with a valid header line field name (a series of printing characters
7667 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7669 23. Changed "disc" in the source to "disk" to conform to the documentation and
7670 the book and for uniformity.
7672 24. Ignore Sendmail's -Ooption=value command line item.
7674 25. When execve() failed while trying to run a command in a pipe transport,
7675 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7676 could be confused with a return value of 69 from the command itself. This
7677 has been changed to 127, the value the shell returns if it is asked to run
7678 a non-existent command. The wording for the related log line suggests a
7679 non-existent command as the problem.
7681 26. If received_header_text expands to an empty string, do not add a Received:
7682 header line to the message. (Well, it adds a token one on the spool, but
7683 marks it "old" so that it doesn't get used or transmitted.)
7685 27. Installed eximstats 1.28 (addition of -nt option).
7687 28. There was no check for failure on the call to getsockname() in the daemon
7688 code. This can fail if there is a shortage of resources on the system, with
7689 ENOMEM, for example. A temporary error is now given on failure.
7691 29. Contrary to the C standard, it seems that in some environments, the
7692 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7693 program. Exim now does this explicitly; it affects the formatting of
7694 timestamps using strftime().
7696 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7697 complaints. I've now initialized all the variables, to avoid this.
7699 32. Header lines added by a system filter were not being "seen" during
7700 transport-time rewrites.
7702 33. The info_callback() function passed to OpenSSL is set up with type void
7703 (*)(SSL *, int, int), as described somewhere. However, when calling the
7704 function (actually a macro) that sets it up, the type void(*)() is
7705 expected. I've put in a cast to prevent warnings from picky compilers.
7707 34. If a DNS black list lookup found a CNAME record, but there were no A
7708 records associated with the domain it pointed at, Exim crashed.
7710 35. If a DNS black list lookup returned more than one A record, Exim ignored
7711 all but the first. It now scans all returned addresses if a particular IP
7712 value is being sought. In this situation, the contents of the
7713 $dnslist_value variable are a list of all the addresses, separated by a
7716 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7717 to accept a host name and all its aliases if the forward lookup for any of
7718 them yielded the IP address of the incoming connection. Now it accepts only
7719 those names whose forward lookup yields the correct IP address. Any other
7720 names are discarded. This closes a loophole whereby a rogue DNS
7721 administrator could create reverse DNS records to break through a
7722 wildcarded host restriction in an ACL.
7724 37. If a user filter or a system filter that ran in a subprocess used any of
7725 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7726 the wrong values were passed to the pipe command ($thisaddress had the
7727 value of $0, $0 had the value of $1, etc). This bug was introduced by
7728 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7730 38. Improved the line breaking for long SMTP error messages from ACLs.
7731 Previously, if there was no break point between 40 and 75 characters, Exim
7732 left the rest of the message alone. Two changes have been made: (a) I've
7733 reduced the minimum length to 35 characters; (b) if it can't find a break
7734 point between 35 and 75 characters, it looks ahead and uses the first one
7735 that it finds. This may give the occasional overlong line, but at least the
7736 remaining text gets split now.
7738 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7739 file descriptors might be low, and that setting 1000 would always raise it.
7740 It turns out that in some environments, the limit is already over 1000 and
7741 that lowering it causes trouble. So now Exim takes care not to decrease it.
7743 40. When delivering a message, the value of $return_path is set to $sender_
7744 address at the start of routing (routers may change the value). By an
7745 oversight, this default was not being set up when an address was tested by
7746 -bt or -bv, which affected the outcome if any router or filter referred to
7749 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7750 log only when "message" or "log_message" are set. However, if one of the
7751 conditions was an address verification, or a call to a nested ACL, the
7752 messages generated by the underlying test were being passed through. This
7753 no longer happens. The underlying message is available in $acl_verify_
7754 message for both "message" and "log_message" expansions, so it can be
7755 passed through if needed.
7757 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7758 new expansion $bh_ to give the encoded byte string without charset
7759 translation. Translation happens only if iconv() is available; HAVE_ICONV
7760 indicates this at build time. HEADERS_CHARSET gives the charset to
7761 translate to; headers_charset can change it in the configuration, and
7762 "headers charset" can change it in an individual filter file.
7764 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7765 that creates RFC 2047 encoded "words" labels them as that charset instead
7766 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7767 expansion operator; (ii) when Exim creates a From: line for a local
7768 message; (iii) when a header line is rewritten to include a "phrase" part.
7770 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7771 buggy, causing it to skip the first lines of messages whose message ID
7772 ended in 'D'. This would not have bitten before Exim release 4.14, because
7773 message IDs were unlikely to end in 'D' before then. The effect was to have
7774 incorrect size information for certain domains.
7776 45. #include "config.h" was missing at the start of the crypt16.c module. This
7777 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7780 46. If there was a timeout during a "random" callout check, Exim treated it as
7781 a failure of the random address, and carried on sending RSET and the real
7782 address. If the delay was just some slowness somewhere, the response to the
7783 original RCPT would be taken as a response to RSET and so on, causing
7784 mayhem of various kinds.
7786 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7787 when I implemented it. It didn't allow for the fact that some option values
7788 may legitimately be negative (e.g. size_addition), and it didn't even do
7789 the right test for positive values.
7791 48. Domain names in DNS records are case-independent. Exim always looks them up
7792 in lower case. Some resolvers return domain names in exactly the case they
7793 appear in the zone file, that is, they may contain uppercase letters. Not
7794 all resolvers do this - some return always lower case. Exim was treating a
7795 change of case by a resolver as a change of domain, similar to a widening
7796 of a domain abbreviation. This triggered its re-routing code and so it was
7797 trying to route what was effectively the same domain again. This normally
7798 caused routing to fail (because the router wouldn't handle the domain
7799 twice). Now Exim checks for this case specially, and just changes the
7800 casing of the domain that it ultimately uses when it transmits the message
7803 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7806 50. If a filter generated a file delivery with a non-absolute name (possible if
7807 no home directory exists for the router), the forbid_file option was not
7810 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7811 the existing equality matching.
7813 52. Exim was using ints instead of ino_t variables in some places where it was
7814 dealing with inode numbers.
7816 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7817 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7818 called TMPDIR, and if it finds it is different, it changes its value.
7820 54. The smtp_printf() function is now made available to local_scan() so
7821 additional output lines can be written before returning. There is also an
7822 smtp_fflush() function to enable the detection of a dropped connection.
7823 The variables smtp_input and smtp_batched_input are exported to
7826 55. Changed the default runtime configuration: the message "Unknown user"
7827 has been removed from the ACL, and instead placed on the localuser router,
7828 using the cannot_route_message feature. This means that any verification
7829 failures that generate their own messages won't get overridden. Similarly,
7830 the "Unrouteable address" message that was in the ACL for unverifiable
7831 relay addresses has also been removed.
7833 56. Added hosts_avoid_esmtp to the smtp transport.
7835 57. The exicyclog script was not checking for the esoteric option
7836 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7837 will work only if exicyclog is run under the appropriate euid.
7839 58. Following a discussion on the list, the rules by which Exim recognises line
7840 endings on incoming messages have been changed. The -dropcr and drop_cr
7841 options are now no-ops, retained only for backwards compatibility. The
7842 following line terminators are recognized: LF CRLF CR. However, special
7843 processing applies to CR:
7845 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7846 nor a local message in the state where . is a terminator.
7848 (ii) If a bare CR is encountered in a header line, an extra space is added
7849 after the line terminator so as not to end the header. The reasoning
7850 behind this is that bare CRs in header lines are most likely either
7851 to be mistakes, or people trying to play silly games.
7853 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7854 was being incorrectly given as 18 bytes larger than it should have been.
7855 This is a VOB (very old bug).
7857 60. This may never have affected anything current, but just in case it has:
7858 When the local host is found other than at the start of a list of hosts,
7859 the local host, those with the same MX, and any that follow, are discarded.
7860 When the list in question was part of a longer list of hosts, the following
7861 hosts (not currently being processed) were also being discarded. This no
7862 longer happens. I'm not sure if this situation could ever has previously
7865 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7868 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7869 header. This is something that is recommended in a new Internet Draft, and
7870 is something that is documented as being done by Sendmail. There are two
7871 possible values. For messages generated by the autoreply transport, Exim
7874 Auto-Submitted: auto-replied
7876 whereas for all other generated messages (e.g. bounces) it adds
7878 Auto-Submitted: auto-generated
7880 63. The "personal" condition in filters now includes a test for the
7881 Auto-Submitted: header. If it contains the string "auto-" the message it
7882 not considered personal.
7884 64. Added rcpt_include_affixes as a generic transport option.
7886 65. Added queue_only_override (default true).
7888 66. Added the syslog_duplication option.
7890 67. If what should have been the first header line of a message consisted of
7891 a space followed by a colon, Exim was mis-interpreting it as a header line.
7892 It isn't of course - it is syntactically invalid and should therefore be
7893 treated as the start of the message body. The misbehaviour could have
7894 caused a number of strange effects, including loss of data in subsequent
7895 header lines, and spool format errors.
7897 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7898 client host had authenticated. This control can now be exercised by an ACL
7899 for more flexibility.
7901 69. By default, callouts do not happen when testing with -bh. There is now a
7902 variant, -bhc, which does actually run the callout code, including
7903 consulting and updating the callout cache.
7905 70. Added support for saslauthd authentication, courtesy of Alexander
7908 71. If statvfs() failed on the spool or log directories while checking their
7909 size for availability, Exim confusingly gave the error "space shortage".
7910 Furthermore, in debugging mode it crashed with a floating point exception.
7911 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7912 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7913 serious problem, Exim now writes to the main and panic logs when this
7914 happens, with details of the failure. It then refuses to accept the
7915 incoming message, giving the message "spool directory problem" or "log
7916 directory problem" with a 421 code for SMTP messages.
7918 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7919 0, 1, and 2 exist, because some OS complain for execs without them (see
7920 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7921 descriptors. However, the code omitted to check that the open succeeded,
7922 causing mysterious errors if for some reason the permissions on /dev/null
7923 got screwed. Now Exim writes a message to the main and panic logs, and
7924 bombs out if it can't open /dev/null.
7926 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7927 interact so that it is all more flexible. It is supposed to remain
7928 backwards compatible. Also added extra_local_interfaces.
7930 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7931 to bomb out with an assertion failure - to the client this appears as a
7932 connection drop. This problem occurs in the part of the code that was taken
7933 from the Samba project. Fortunately, the assertion is in a very simple
7934 function, so I have fixed this by reproducing the function inline in the
7935 one place where it is called, and arranging for authentication to fail
7936 instead of killing the process with assert().
7938 75. The SPA client code was not working when the server requested OEM rather
7939 than Unicode encoding.
7941 76. Added code to make require_files with a specific uid setting more usable in
7942 the case where statting the file as root fails - usually a non-root-mounted
7943 NFS file system. When this happens and the failure is EACCES, Exim now
7944 forks a subprocess and does the per-uid checking as the relevant uid.
7946 77. Added process_log_path.
7948 78. If log_file_path was not explicitly set, a setting of check_log_space or
7949 check_log_inodes was ignored.
7951 79. If a space check for the spool or log partitions fails, the incident is now
7952 logged. Of course, in the latter case the data may get lost...
7954 80. Added the %p formatting code to string_format() so that it can be used to
7955 print addresses in debug_print(). Adjusted all the address printing in the
7956 debugging in store.c to use %p rather than %d.
7958 81. There was a concern that a line of code in smtp_in.c could overflow a
7959 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7960 initially expressed, the concern was not well-founded, because trailing
7961 spaces are removed early. However, if the trailing spaces were followed by
7962 a NULL, they did not get removed, so the overflow was possible. Two fixes
7965 (a) I re-wrote the offending code in a cleaner fashion.
7966 (b) If an incoming SMTP command contains a NULL character, it is rejected
7969 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7970 runs initgroups(), so that if the Exim user is in any additional groups,
7971 they will be used during message reception.
7977 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob