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.
194 PP/09 Fix security issue with too many recipients on a message (to remove a
195 known security problem if someone does set recipients_max to unlimited,
196 or if local additions add to the recipient list).
197 Fixes CVE-2020-RCPTL reported by Qualys.
199 PP/10 Fix security issue in SMTP verb option parsing
200 Fixes CVE-2020-EXOPT reported by Qualys.
202 PP/11 Fix security issue in BDAT state confusion.
203 Ensure we reset known-good where we know we need to not be reading BDAT
204 data, as a general case fix, and move the places where we switch to BDAT
205 mode until after various protocol state checks.
206 Fixes CVE-2020-BDATA reported by Qualys.
208 HS/03 Die on "/../" in msglog file names
210 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
211 the Exim runtime user are allowed to create files.
213 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
216 QS/03 When reading the output from interpreted forward files we do not
217 pass the pipe between the parent and the interpreting process to
218 executed child processes (if any).
223 JH/01 Avoid costly startup code when not strictly needed. This reduces time
224 for some exim process initialisations. It does mean that the logging
225 of TLS configuration problems is only done for the daemon startup.
227 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
229 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
230 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
233 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
234 DKIM signing is being done. Previously a transport_filter would always
235 disable CHUNKING, falling back to traditional DATA.
237 JH/05 Regard command-line receipients as tainted.
239 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
241 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
242 PAM library frees one of the arguments given to it, despite the
243 documentation. Therefore a plain malloc must be used.
245 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
246 on-stack buffers were used, resulting in a taint trap when DSN information
247 copied from a received message was written into the buffer.
249 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
250 the ordering of its ARC headers. This caused a crash.
252 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
253 a new record was being constructed with information from the peer, a trap
256 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
257 installation would get error messages from DMARC verify, when it hit the
258 nonexistent file indicated by the default. Distros wanting DMARC enabled
259 should both provide the file and set the option.
260 Also enforce no DMARC verification for command-line sourced messages.
262 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
263 could, depending on the platform, hang at the STARTTLS response.
265 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
266 message on a connection. Previously if one message had ARC headers and
267 the following one did not, a crash could result when adding an
268 Authentication-Results: header.
270 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
271 Exim main code and Exim-related utities. The introduction of taint
272 tracking also did many adjustments to string handling. Since then, eximon
273 frequently terminated with an assert failure.
275 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
276 check for 452 responses. This slightly helps the inefficieny of doing
277 a large alias-expansion into a recipient-limited target. The max_rcpt
278 transport option still applies (and at the current default, will override
279 the new feature). The check is done for either cause of synch, and forces
280 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
281 connection. The new facility is not tunable at this time.
283 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
284 library live data was being used, so the results became garbage. Make
285 copies while it is still usable.
287 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
288 on delivery deferred (==) and failed (**) lines (if a delivery was
289 attemtped). Previously it was only on completion (=>) lines.
291 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
292 for the expansion of the server_scram_iter and server_scram_salt options.
294 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
295 are now specifically given a NO_DATA response without hitting the system
296 resolver. The library goes on to do the now-standard TXT lookup.
297 Use of dnsdb lookups is not affected.
299 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
300 only retrieve the errormessage once. Previously two calls to dlerror()
301 were used, and the second one (for mainlog/paniclog) retrieved null
304 JH/20 Taint checking: disallow use of tainted data for
305 - the appendfile transport file and directory options
306 - the pipe transport command
307 - the autoreply transport file, log and once options
308 - file names used by the redirect router (including filter files)
310 - paths used by single-key lookups
311 Previously this was permitted.
313 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
314 adjusted the size of a major service buffer; this failed because the
315 buffer was in use at the time. Change to a compile-time increase in the
316 buffer size, when this authenticator is compiled into exim.
318 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
319 previous fast-mode was untenable in the face of glibs using mmap to
320 support larger malloc requests.
322 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
323 New values supported, if defined on system where compiled:
324 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
325 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
327 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
328 running a limited number of proceses in parallel, a benefit is gained. The
329 amount varies with the platform hardware and load. The use of the option
330 queue_run_in_order means we cannot do this, as ordering becomes
333 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
334 had introduced a string-copy (for ensuring NUL-termination) which was not
335 appropriate for that case, which can include embedded NUL bytes in the
336 block of data. Investigation showed the copy to actually be needless, the
337 data being length-specified.
339 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
340 done during a receiving connection, and both used TLS, global info was
341 used rather than per-connection info for tracking the state of data
342 queued for transmission. This could result in a connection hang.
344 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
345 Previously, when delivering serveral messages down a single connection
346 only the first would provide a SIZE. This was due to the size information
347 not being properly tracked.
349 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
350 TAI (at 37 seconds currently), pretend to be in UTC for time-related
351 expansion and logging. Previously, spurious values such as a future
352 minute could be seen.
354 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
355 it could crash from a null-deref. This could also affect the
356 ${addresses: } operator and ${readsock } item.
358 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
359 message following a mime one, the variable was not reset.
361 JH/30 When an pipelined-connect fails at the first response, assume incorrect
362 cached capability (perhaps the peer reneged?) and immediately retry in
365 JH/31 Fix spurious detection of timeout while writing to transport filter.
367 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
368 an attempt to copy the string was made before checking it.
370 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
371 taint of the lookup key was maintained; we now regard the presence in the
372 filesystem as sufficient validation.
374 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
375 argument is supplied.
377 JH/35 The dsearch lookup now requires that the directory is an absolute path.
378 Previously this was not checked, and nonempty relative paths made an
379 access under Exim's current working directory.
381 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
382 Previously no event was raised.
384 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
385 parameter supplied by the sender MAIL FROM command. Previously it was
386 ignored, and only the check_spool_space option value for the required
389 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
390 the size of the signing public-key. Previously it was instead giving
391 the size of the signature hash.
393 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
394 the default. See the (new) dkim_verify_min_keysizes option.
396 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
397 and an ACL use a lookup for checking either the local_part or domain,
398 stale data could be accessed. Ensure that variable references are
399 dropped between messages.
401 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
402 by the client was not checked as pointing within response data before
403 being used. A malicious client could thus cause an out-of-bounds read and
404 possibly gain authentication. Fix by adding the check.
406 JH/42 Internationalisation: change the default for downconversion in the smtp
407 transport to be "if needed". Previously it was "as previously set" for
408 the message, which usually meant "if needed" for message-submission but
409 "no" for everything else. However, MTAs have been seen using SMTPUTF8
410 even when the envelope addresses did not need it, resulting in forwarding
411 failures to non-supporting MTAs. A downconvert in such cases will be
412 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
413 transport. The change does mean that addresses needing conversion will
414 be converted when previously a delivery failure would occur.
416 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
417 response was received it would be used unchecked in a fail-DSN, violating
418 standards on line-length limits. Truncate if needed.
420 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
421 doesn't work. (Additionally add a new main config option to configure the
428 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
429 side, to decode a TLS 1.3 packet capture.
431 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
432 Previously the default library behaviour applied, sending two, each in
435 JH/03 Debug output for ACL now gives the config file name and line number for
438 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
440 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
442 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
443 buffer overrun for (non-chunking) other transports.
445 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
446 TLS1.3, means that a server rejecting a client certificate is not visible
447 to the client until the first read of encrypted data (typically the
448 response to EHLO). Add detection for that case and treat it as a failed
449 TLS connection attempt, so that the normal retry-in-clear can work (if
450 suitably configured).
452 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
453 and/or domain. Found and fixed by Jason Betts.
455 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
456 configuration). If a CNAME target was not a wellformed name pattern, a
459 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
460 the OS reports them interleaved with other addresses.
462 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
463 used both for input and for a verify callout, both encrypted, SMTP
464 responses being sent by the server could be lost. This resulted in
465 dropped connections and sometimes bounces generated by a peer sending
468 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
469 string. Previously it was possible to cause undefined behaviour in a
470 library routine (usually a crash). Found by "zerons".
472 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
475 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
476 API was removed, so update to use the newer ones.
478 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
479 any timeout set, is taking a long time. Previously we would hang on to a
480 rotated logfile "forever" if the input was arriving with long gaps
481 (a previous attempt to fix addressed lack, for a long time, of initial
484 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
485 shared (NFS) environment. The length of the tempfile name is now
486 4 + 16 ("hdr.$message_exim_id") which might break on file
487 systems which restrict the file name length to lower values.
490 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
491 shared (NFS) environment.
493 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
494 did for all versions <4.90). Notably -M, -m, --invert, -I may be
497 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
498 on some platforms for bit 31.
500 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
501 to changes apparently associated with TLS1.3 handling some of the APIs
502 previously used were either nonfunctional or inappropriate. Strings
503 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
504 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
505 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
506 This affects log line X= elements, the $tls_{in,out}_cipher variables,
507 and the use of specific cipher names in the encrypted= ACL condition.
509 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
511 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
512 verification result was not updated unless hosts_require_ocsp applied.
514 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
515 queue_list_requires_admin set to false, non-admin users were denied the
518 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
519 directory-of-certs mode. Previously they were advertised despite the
522 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
523 A single TCP connection by a client will now hold a TLS connection open
524 for multiple message deliveries, by default. Previously the default was to
527 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
528 default. If built with the facility, DANE will be used. The facility
529 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
531 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
532 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
533 must be defined and you must still, unless you define DISABLE_TLS, manage
534 the the include-dir and library-file requirements that go with that
535 choice. Non-TLS builds are still supported.
537 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
540 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
541 default. If the platform supports and has the facility enabled, it will
542 be requested on all coneections.
544 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
545 controlled by the build-time option SUPPORT_PIPE_CONNECT.
547 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
549 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
550 success-DSN messages. Previously the From: header was always the default
551 one for these; the option was ignored.
553 JH/28 Fix the timeout on smtp response to apply to the whole response.
554 Previously it was reset for every read, so a teergrubing peer sending
555 single bytes within the time limit could extend the connection for a
556 long time. Credit to Qualsys Security Advisory Team for the discovery.
558 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
559 delivery address, which leaked information of the results of local
560 forwarding. Change to the original envelope recipient address, per
563 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
564 requested. Previously not bounce was generated and a log entry of
565 error ignored was made.
567 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
569 JH/32 Introduce a general tainting mechanism for values read from the input
570 channel, and values derived from them. Refuse to expand any tainted
571 values, to catch one form of exploit.
573 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
574 was unused and the unexpanded text used for the test. Found and
575 fixed by Ruben Jenster.
577 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
578 an attempt to use a TLS library read routine dereffed a nul pointer,
581 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
582 them in our smtp response.
584 JH/36 Have the general router option retry_use_local_part default to true when
585 any of the restrictive preconditions are set (to anything). Previously it
586 was only for check_local user. The change removes one item of manual
587 configuration which is required for proper retries when a remote router
588 handles a subset of addresses for a domain.
590 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
591 link count into consideration.
593 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
594 caused the extension of big_buffer, the following lines were ignored.
596 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
597 accordance with RFC 2308. Previously there was no expiry, so a longlived
598 receive process (eg. due to ACL delays) versus a short SOA value could
601 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
603 JH/39 Promote DMARC support to mainline.
605 JH/40 Bug 2452: Add a References: header to DSNs.
607 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
608 parameters. The relevant library call is documented as "Deprecated: This
609 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
610 3.6.0, DH parameters are negotiated following RFC7919."
612 HS/06 Change the default of dnssec_request_domains to "*"
614 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
615 carried on and emitted a BDAT command, even when PIPELINING was not
618 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
619 buffer was used for the filename, resulting in a trap when tainted
620 arguments (eg. $domain) were used.
622 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
623 recommended to avoid a possible server-load attack. The feature can be
624 re-enabled via the openssl_options main cofiguration option.
626 JH/45 local_scan API: documented the current smtp_printf() call. This changed
627 for version 4.90 - adding a "more data" boolean to the arguments.
628 Bumped the ABI version number also, this having been missed previously;
629 release versions 4.90 to 4.92.3 inclusive were effectively broken in
630 respect of usage of smtp_printf() by either local_scan code or libraries
631 accessed via the ${dlfunc } expansion item. Both will need coding
632 adjustment for any calls to smtp_printf() to match the new function
633 signature; a FALSE value for the new argument is always safe.
635 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
636 the file-offset (which the Linux syscall does, and exim expects); this
637 resulted in an indefinite loop.
639 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
640 to do ARC verification. The Authentication-Results: header line added
641 by the configuration then had no ARC item.
647 JH/01 Remove code calling the customisable local_scan function, unless a new
648 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
650 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
651 non-signal-safe functions being used.
653 JH/03 Bug 2269: When presented with a received message having a stupidly large
654 number of DKIM-Signature headers, disable DKIM verification to avoid
655 a resource-consumption attack. The limit is set at twenty.
657 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
658 report of oldest_pass in ${authres } in consequence, and separate out
659 some descriptions of reasons for verification fail.
661 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
662 files in the spool were present and unlocked. A queue-runner could spot
663 them, resulting in a duplicate delivery. Fix that by doing the unlock
664 after the unlink. Investigation by Tim Stewart. Take the opportunity to
665 add more error-checking on spoolfile handling while that code is being
668 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
669 No known attacks, no CVE, this is defensive hardening.
671 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
672 a queue-runner could start a delivery while other operations were ongoing.
673 Cutthrough delivery was a common victim, resulting in duplicate delivery.
674 Found and investigated by Tim Stewart. Fix by using the open message data
675 file handle rather than opening another, and not locally closing it (which
676 releases a lock) for that case, while creating the temporary .eml format
677 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
679 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
680 $sender_verify_failure/$recipient_verify_failure to "random".
682 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
685 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
686 Previously this would segfault.
688 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
691 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
692 like zero, since the resolver should be doing this for us, But we need one
693 as a CNAME but no MX presence gets the CNAME returned; we need to check
694 that doesn't point to an MX to declare it "no MX returned" rather than
695 "error, loop". A new main option is added so the older capability of
696 following some limited number of chain links is maintained.
698 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
700 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
701 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
702 now-deprecated earlier definitions used only specified the range up to TLS
703 1.2 (in the older-version library docs).
705 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
707 JH/15 Rework TLS client-side context management. Stop using a global, and
708 explicitly pass a context around. This enables future use of TLS for
709 connections to service-daemons (eg. malware scanning) while a client smtp
710 connection is using TLS; with cutthrough connections this is quite likely.
712 JH/16 Fix ARC verification to do AS checks in reverse order.
714 JH/17 Support a "tls" option on the ${readsocket } expansion item.
716 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
717 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
718 Previously the "utf8" would be re-prepended for every additional message.
720 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
721 Previously thery were accepted, resulting in issues when attempting to
722 forward messages to a non-supporting MTA.
724 PP/02 Let -n work with printing macros too, not just options.
726 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
727 one parent address was copied, and bogus data was used at delivery-logging
728 time. Either a crash (after delivery) or bogus log data could result.
729 Discovery and analysis by Tim Stewart.
731 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
732 Previously if the string ended mid-character, we did not insert the
733 promised '?' replacement.
735 PP/04 Documentation: current string operators work on bytes, not codepoints.
737 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
738 should pack well giving a smaller memory footprint so better caching and
739 therefore performance. Group the declarations where this can't be done so
740 that the byte-sized flag variables are not interspersed among pointer
741 variables, giving a better chance of good packing by the compiler.
743 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
744 non-null, to avoid issues with sites running BATV. Previously reports were
745 sent with an empty envelope sender so looked like bounces.
747 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
748 The ignore_error flag wasn't being returned from the filter subprocess so
749 was not set for later routers. Investigation and fix by Matthias Kurz.
751 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
752 and a msg:complete for the whole, when a message is manually removed using
753 -Mrm. Developement by Matthias Kurz, hacked on by JH.
755 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
756 a "Gnu special" function, asprintf() in the DB utility binary builds; I
757 hope that is portable enough.
759 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
760 requiring a known-CA anchor certificate; make it now rely entirely on the
761 TLSA as an anchor. Checking the name on the leaf cert against the name
762 on the A-record for the host is still done for TA (but not for EE mode).
764 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
765 included in delivery lines for non-proxied connections, when compiled with
766 SUPPORT_SOCKS and running with proxy logging enabled.
768 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
769 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
770 move the existing event to fire before the normal logging of message
771 failure so that custom logging is bracketed by normal logging.
773 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
774 msg:fail:internal event. Developement by Matthias Kurz.
776 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
777 far too small for todays use of crypto signatures stored there. Go all
778 the way to the max DNS message size of 64kB, even though this might be
779 overmuch for IOT constrained device use.
781 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
782 copy a string over itself. The library routine is documented as not
783 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
785 JH/32 For main options check_spool_space and check_inode_space, where the
786 platform supports 64b integers, support more than the previous 2^31 kB
787 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
788 the previous G, M, k.
790 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
791 $authenticated_fail_id variable on authentication failure. Previously
794 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
795 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
796 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
797 GNUTLS_SEC_PARAM_MEDIUM.
799 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
800 side. Previously we would continue as if no SNI had been received.
802 JH/36 Harden the handling of string-lists. When a list consisted of a sole
803 "<" character, which should be a list-separator specification, we walked
804 off past the nul-terimation.
806 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
807 causes) even when the retry time is not yet met. Previously they were
808 not, meaning that when (say) an account was over-quota and temp-rejecting,
809 and multiple senders' messages were queued, only one sender would get
810 notified on each configured delay_warning cycle.
812 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
814 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
815 cipher-suites, an error can be left on the stack even for a succeeding
816 accept; this results in impossible error messages when a later operation
819 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
820 return error codes indicating retry. Under TLS1.3 this becomes required.
822 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
823 it only wrote the new authenticators, resulting in a lack of tracking of
824 peer changes of ESMTP extensions until the next cache flush.
826 JH/41 Fix the loop reading a message header line to check for integer overflow,
827 and more-often against header_maxsize. Previously a crafted message could
828 induce a crash of the recive process; now the message is cleanly rejected.
830 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
831 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
837 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
838 When redis_servers is set to a list of > 1 element, and the Redis servers
839 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
840 case of MOVED into a DEFER case instead, thus moving the query onto the
841 next server in the list. For a cluster of N elements, all N servers must
842 be defined in redis_servers.
844 GF/02 Catch and remove uninitialized value warning in exiqsumm
845 Check for existence of @ARGV before looking at $ARGV[0]
847 JH/01 Replace the store_release() internal interface with store_newblock(),
848 which internalises the check required to safely use the old one, plus
849 the allocate and data copy operations duplicated in both (!) of the
850 extant use locations.
852 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
853 modifier. This matches the restriction on the commandline.
855 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
856 Previously only the last row was returned.
858 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
859 we assumed that tags in the header were well-formed, and parsed the
860 element content after inspecting only the first char of the tag.
861 Assumptions at that stage could crash the receive process on malformed
864 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
865 While running the DKIM ACL we operate on the Permanent memory pool so that
866 variables created with "set" persist to the DATA ACL. Also (at any time)
867 DNS lookups that fail create cache records using the Permanent pool. But
868 expansions release any allocations made on the current pool - so a dnsdb
869 lookup expansion done in the DKIM ACL releases the memory used for the
870 DNS negative-cache, and bad things result. Solution is to switch to the
871 Main pool for expansions.
872 While we're in that code, add checks on the DNS cache during store_reset,
873 active in the testsuite.
874 Problem spotted, and debugging aided, by Wolfgang Breyha.
876 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
877 When none of the hosts presented to a transport match an already-open
878 connection, close it and proceed with the list. Previously we would
879 queue the message. Spotted by Lena with Yahoo, probably involving
882 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
883 Previously a spurious "250 OK id=" response was appended to the proper
886 JH/08 The "support for" informational output now, which built with Content
887 Scanning support, has a line for the malware scanner interfaces compiled
888 in. Interface can be individually included or not at build time.
890 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
891 by the template makefile "src/EDITME". The "STREAM" support for an older
892 ClamAV interface method is removed.
894 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
895 rows affected is given instead).
897 JH/11 The runtime Berkeley DB library version is now additionally output by
898 "exim -d -bV". Previously only the compile-time version was shown.
900 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
901 SMTP connection. Previously, when one had more receipients than the
902 first, an abortive onward connection was made. Move to full support for
903 multiple onward connections in sequence, handling cutthrough connection
904 for all multi-message initiating connections.
906 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
907 routers. Previously, a multi-recipient message would fail to match the
908 onward-connection opened for the first recipient, and cause its closure.
910 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
911 a timeout on read on a GnuTLS initiating connection, resulting in the
912 initiating connection being dropped. This mattered most when the callout
913 was marked defer_ok. Fix to keep the two timeout-detection methods
916 JH/15 Relax results from ACL control request to enable cutthrough, in
917 unsupported situations, from error to silently (except under debug)
918 ignoring. This covers use with PRDR, frozen messages, queue-only and
921 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
923 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
924 metadata, resulting in a crash in free().
926 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
927 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
928 Broken also in d185889f4, with init system revamp.
930 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
931 Previously we abruptly closed the connection after reading a malware-
932 found indication; now we go on to read the "scan ok" response line,
935 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
936 ACL. Previously, a crash would result.
938 JH/19 Speed up macro lookups during configuration file read, by skipping non-
939 macro text after a replacement (previously it was only once per line) and
940 by skipping builtin macros when searching for an uppercase lead character.
942 JH/20 DANE support moved from Experimental to mainline. The Makefile control
943 for the build is renamed.
945 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
946 was allocated for every new TLS startup, meaning one per message. Fix
947 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
949 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
950 reported the original. Fix to report (as far as possible) the ACL
951 result replacing the original.
953 JH/23 Fix memory leak during multi-message connections using STARTTLS under
954 OpenSSL. Certificate information is loaded for every new TLS startup,
955 and the resources needed to be freed.
957 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
959 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
962 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
963 DATA response info to the (existing) per-recipient response info for
964 the "C=" log element. It can have useful tracking info from the
965 destination system. Patch from Simon Arlott.
967 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
968 length value. Previously this would segfault.
970 HS/02 Support Avast multiline protoocol, this allows passing flags to
971 newer versions of the scanner.
973 JH/28 Ensure that variables possibly set during message acceptance are marked
974 dead before release of memory in the daemon loop. This stops complaints
975 about them when the debug_store option is enabled. Discovered specifically
976 for sender_rate_period, but applies to a whole set of variables.
977 Do the same for the queue-runner and queue-list loops, for variables set
978 from spool message files. Do the same for the SMTP per-message loop, for
979 certain variables indirectly set in ACL operations.
981 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
982 as a multi-recipient message from a mailinglist manager). The coding had
983 an arbitrary cutoff number of characters while checking for more input;
984 enforced by writing a NUL into the buffer. This corrupted long / fast
985 input. The problem was exposed more widely when more pipelineing of SMTP
986 responses was introduced, and one Exim system was feeding another.
987 The symptom is log complaints of SMTP syntax error (NUL chars) on the
988 receiving system, and refused recipients seen by the sending system
989 (propating to people being dropped from mailing lists).
990 Discovered and pinpointed by David Carter.
992 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
993 replaced by the ${authresults } expansion.
995 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
997 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
998 allows proper process termination in container environments.
1000 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1001 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1003 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1004 and "err_perm", deprecated since 4.83 when the RFC-defined words
1005 "temperror" and "permerror" were introduced.
1007 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1008 transport-filters or DKIM-signing. The restriction was lost in the
1009 consolidation of verify-callout and delivery SMTP handling.
1010 Extend the restriction to also cover ARC-signing.
1012 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1013 in defer=pass mode supply a 450 to the initiator. Previously the message
1016 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1017 tls_require_ciphers is used as before.
1019 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1020 "pass_unscanned". Only tmpfails from the scanner are written to
1021 the paniclog, as they may require admin intervention (permission
1022 denied, license issues). Other scanner errors (like decompression
1023 bombs) do not cause a paniclog entry.
1025 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1026 Previously it was possible to log spurious information in receive log
1029 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1030 triggered odd behaviour from Outlook Express clients.
1032 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1035 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1036 since the IETF WG has not yet settled on that versus the original
1037 "bare" representation.
1039 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1040 Previously the millisecond value corrupted the output.
1041 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1042 corrupted the output.
1048 JH/01 Rework error string handling in TLS interface so that the caller in
1049 more cases is responsible for logging. This permits library-sourced
1050 string to be attached to addresses during delivery, and collapses
1051 pairs of long lines into single ones.
1053 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1054 during configuration. Wildcards are allowed and expanded.
1056 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1057 This permits better logging.
1059 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1060 do not close it down and have the child transport start it up again on
1061 the passed-on TCP connection. Instead, proxy the child (and any
1062 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1063 affected: the continued delivery log lines do not have any DNSSEC, TLS
1064 Certificate or OCSP information. TLS cipher information is still logged.
1066 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1067 identical IP addresses on different listening ports. Will also affect
1070 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1071 add noisy ifdef guards to special-case this sillyness.
1072 Patch from Bernd Kuhls.
1074 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1075 than 255 are no longer allowed.
1077 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1078 Disable the session-cache too, which might reduce our load. Since we
1079 currrectly use a new context for every connection, both as server and
1080 client, there is no benefit for these.
1081 GnuTLS appears to not support tickets server-side by default (we don't
1082 call gnutls_session_ticket_enable_server()) but client side is enabled
1083 by default on recent versions (3.1.3 +) unless the PFS priority string
1086 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1087 <https://reproducible-builds.org/specs/source-date-epoch/>.
1089 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1090 the check for any unsuccessful recipients did not notice the limit, and
1091 erroneously found still-pending ones.
1093 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1094 MSG_MORE. Only in-clear (not on TLS connections).
1096 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1097 transport-filter is involved we can buffer the headers in memory for
1098 creating the signature, and read the spool data file once for the
1099 signature and again for transmission.
1101 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1102 4.77 as the kernel support then wasn't solid, having issues in 64bit
1103 mode. Now, it's been long enough. Add support for FreeBSD also.
1105 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1106 case where the routing stage had gathered several addresses to send to
1107 a host before calling the transport for the first, we previously failed
1108 to close down TLS in the old transport process before passing the TCP
1109 connection to the new process. The new one sent a STARTTLS command
1110 which naturally failed, giving a failed delivery and bloating the retry
1111 database. Investigation and fix prototype from Wolfgang Breyha.
1113 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1114 false-negatives in the check that the sender had not preempted a response
1115 or prompt from Exim (running as a server), due to that code's lack of
1116 awareness of the SMTP input buffering.
1118 PP/04 Add commandline_checks_require_admin option.
1119 Exim drops privileges sanely, various checks such as -be aren't a
1120 security problem, as long as you trust local users with access to their
1121 own account. When invoked by services which pass untrusted data to
1122 Exim, this might be an issue. Set this option in main configuration
1123 AND make fixes to the calling application, such as using `--` to stop
1126 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1127 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1128 responses to those, into a single TLS record each way (this usually means
1129 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1132 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1133 affects you only if you're dancing at the edge of the param size limits.
1134 If you are, and this message makes sense to you, then: raise the
1135 configured limit or use OpenSSL 1.1. Nothing we can do for older
1138 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1139 cmdline element to get the documented default one. Previously it was
1142 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1143 are unsuable for non-message-specific reasons - eg. connection timeout,
1144 banner-time rejection.
1146 JH/16 Fix logging of delivery remote port, when specified by router, under
1149 PP/06 Repair manualroute's ability to take options in any order, even if one
1150 is the name of a transport.
1153 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1155 JH/17 Change the list-building routines interface to use the expanding-string
1156 triplet model, for better allocation and copying behaviour.
1158 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1159 Previously it was constructed the first time a possibly-matching string
1160 was met in the configuration file input during startup; now it is done
1163 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1164 compatible one, to avoid the (poorly documented) possibility of a config
1165 file in the working directory redirecting the DB files, possibly correpting
1166 some existing file. CVE-2017-10140 assigned for BDB.
1168 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1169 cache-hot. Previously, although the result was properly cached, the
1170 initial verify call returned a defer.
1172 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1173 the main verify for receipient in uncached-mode.
1175 JH/22 Retire historical build files to an "unsupported" subdir. These are
1176 defined as "ones for which we have no current evidence of testing".
1178 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1179 if present. Previously it was ignored.
1181 JH/24 Start using specified-initialisers in C structure init coding. This is
1182 a C99 feature (it's 2017, so now considered safe).
1184 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1185 if was a fixed-sized field and bitmask ops via macros; it is now more
1188 PP/07 GitHub PR 56: Apply MariaDB build fix.
1189 Patch provided by Jaroslav Škarvada.
1191 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1192 during Coverity cleanups [4.87 JH/47]
1193 Diagnosis and fix provided by Michael Fischer v. Mollard.
1195 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1196 the right size to place the terminating semicolon on its own folded
1197 line, the header hash was calculated to an incorrect value thanks to
1198 the (relaxed) space the fold became.
1200 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1201 and confused the parent.
1203 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1204 which could crash as a result. This could lead to undeliverable messages.
1206 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1209 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1210 matters on fast-turnover and PID-randomising systems, which were getting
1211 out-of-order delivery.
1213 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1214 a possibly-overlapping copy. The symptom was that "Remote host closed
1215 connection in response to HELO" was logged instead of the actual 4xx
1218 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1219 Previously only that bufferd was discarded, resulting in SYMTP command
1222 JH/32 DKIM: when a message has multiple signatures matching an identity given
1223 in dkim_verify_signers, run the dkim acl once for each. Previously only
1224 one run was done. Bug 2189.
1226 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1227 "panic the current process" to "deliberately defer". The panic log is
1228 still written with the problem list name; the mail and reject logs now
1229 get a temp-reject line for the message that was being handled, saying
1230 something like "domains check lookup or other defer". The SMTP 451
1231 message is still "Temporary local problem".
1233 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1234 A crafted sequence of BDAT commands could result in in-use memory beeing
1235 freed. CVE-2017-16943.
1237 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1238 from SMTP input. Previously it was always done; now only done for DATA
1239 and not BDAT commands. CVE-2017-16944.
1241 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1242 to the message (such as an overlong header line). Previously this was
1243 not done and we did not exit BDAT mode. Followon from the previous item
1244 though a different problem.
1250 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1251 than -2003 did; needs libidn2 in addition to libidn.
1253 JH/02 The path option on a pipe transport is now expanded before use.
1255 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1256 Patch provided by "Björn", documentation fix added too.
1258 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1259 missing a wire-to-host endian conversion.
1261 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1262 close after a BDAT command line could be taken as a following command,
1263 giving a synch failure. Fix by only checking for synch immediately
1264 before acknowledging the chunk.
1266 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1267 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1268 macro. Patches provided by Josh Soref.
1270 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1271 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1272 by IANA as a service extension. However, John Klensin suggests that we
1275 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1276 the dkim code may be unix-mode line endings rather than smtp wire-format
1277 CRLF, so prepend a CR to any bare LF.
1279 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1280 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1282 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1283 the first were themselves being wrongly included in the feed into dkim
1284 processing; with most chunk sizes in use this resulted in an incorrect
1285 body hash calculated value.
1287 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1288 DKIM signature block, for verification. Although advised against by
1289 standards it is specifically not ruled illegal.
1291 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1293 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1294 missing a body hash (the bh= tag).
1296 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1297 It seems that HAProxy sends the Proxy Protocol information in clear and
1298 only then does a TLS startup, so do the same.
1300 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1301 TCP connections (such as for Spamd) unless the daemon successfully set
1302 Fast Open mode on its listening sockets. This fixes breakage seen on
1303 too-old kernels or those not configured for Fast Open, at the cost of
1304 requiring both directions being enabled for TFO, and TFO never being used
1305 by non-daemon-related Exim processes.
1307 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1308 endings, at least on the first header line. Try to canonify any that get
1309 past that check, despite the cost.
1311 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1312 now limited to an arbitrary five deep, while parsing addresses with the
1313 strip_excess_angle_brackets option enabled.
1315 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1316 instead leave the unprompted TLS handshake in socket buffer for the
1317 TLS library to consume.
1319 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1321 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1323 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1324 region in one operation, for speed, and this leaves assigned pointers
1325 dangling. Add checks run only under the testsuite which checks all
1326 variables at a store-reset and panics on a dangling pointer; add code
1327 explicitly nulling out all the variables discovered. Fixes one known
1328 bug: a transport crash, where a dangling pointer for $sending_ip_address
1329 originally assigned in a verify callout, is re-used.
1331 PP/06 Drop '.' from @INC in various Perl scripts.
1333 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1335 PP/08 Reduce a number of compilation warnings under clang; building with
1336 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1337 should be warning-free.
1339 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1341 HS/01 Fix portability problems introduced by PP/08 for platforms where
1342 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1344 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1345 chunk. This allows us to accept broken chunked messages. We need a more
1346 general solution here.
1348 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1349 already-broken messages in the queue.
1351 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1353 JH/19 Fix reference counting bug in routing-generated-address tracking.
1359 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1360 supports it and a size is available (ie. the sending peer gave us one).
1362 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1363 years of being deprecated). The replacements are the ACLs
1364 acl_smtp_mime and acl_not_smtp_mime.
1366 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1367 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1368 or even an in-clear connection were permitted. Now, if the host lookup
1369 was dnssec and dane was requested then the host is only used if the
1370 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1371 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1372 if one fails this test.
1373 This means that a poorly-configured remote DNS will make it incommunicado;
1374 but it protects against a DNS-interception attack on it.
1376 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1377 non-noisy when a race steals the message being considered.
1379 JH/05 If main configuration option tls_certificate is unset, generate a
1380 self-signed certificate for inbound TLS connections.
1382 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1383 in rewrites and routers.
1385 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1386 and logged a warning sing 4.83; now they are a configuration file error.
1388 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1389 (lacking @domain). Apply the same qualification processing as RCPT.
1391 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1393 JH/10 Support ${sha256:} applied to a string (as well as the previous
1396 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1397 a cutthrough deliver is pending, as we always want to make a connection.
1398 This also avoids re-routing the message when later placing the cutthrough
1399 connection after a verify cache hit.
1400 Do not update it with the verify result either.
1402 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1403 when routing results in more than one destination address.
1405 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1406 signing (which inhibits the cutthrough capability). Previously only
1407 the presence of an option was tested; now an expansion evaluating as
1408 empty is permissible (obviously it should depend only on data available
1409 when the cutthrough connection is made).
1411 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1412 the relevant preceding SMTP command did not note the pipelining mode.
1414 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1415 Previously they were not counted.
1417 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1418 as one having no matching records. Previously we deferred the message
1419 that needed the lookup.
1421 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1422 distinguished as "(=".
1424 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1425 for missing MX records. Previously it only worked for missing A records.
1427 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1429 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1430 after the data-go-ahead and data-ack. Patch from Jason Betts.
1432 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1433 even for a "none" policy. Patch from Tony Meyer.
1435 JH/22 Fix continued use of a connection for further deliveries. If a port was
1436 specified by a router, it must also match for the delivery to be
1439 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1440 When one of the recipients of a message was unsuitable for the connection
1441 (has no matching addresses), we lost track of needing to mark it
1442 deferred. As a result mail would be lost.
1444 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1446 JH/25 Decoding ACL controls is now done using a binary search; the source code
1447 takes up less space and should be simpler to maintain. Merge the ACL
1448 condition decode tables also, with similar effect.
1450 JH/26 Fix problem with one_time used on a redirect router which returned the
1451 parent address unchanged. A retry would see the parent address marked as
1452 delivered, so not attempt the (identical) child. As a result mail would
1455 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1456 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1457 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1460 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1461 Main config options check_{log,spool}_{inodes,space} are now
1462 100 inodes, 10MB unless set otherwise in the configuration.
1464 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1465 Previously it only applied to the main-section connection policy
1468 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1470 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1471 by me. Added RFC7919 DH primes as an alternative.
1473 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1474 are not in the system include path.
1476 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1477 GnuTLS, when a session startup failed (eg because the client disconnected)
1478 Exim did stdio operations after fclose. This was exposed by a recent
1479 change which nulled out the file handle after the fclose.
1481 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1482 signed directly by the cert-signing cert, rather than an intermediate
1483 OCSP-signing cert. This is the model used by LetsEncrypt.
1485 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1487 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1488 an incoming connection.
1490 HS/02 Bug 1802: Do not half-close the connection after sending a request
1493 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1494 fallback to "prime256v1".
1496 JH/34 SECURITY: Use proper copy of DATA command in error message.
1497 Could leak key material. Remotely exploitable. CVE-2016-9963.
1503 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1504 and 3.4.4 - once the server is enabled to respond to an OCSP request
1505 it does even when not requested, resulting in a stapling non-aware
1506 client dropping the TLS connection.
1508 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1509 support variable-length bit vectors. No functional change.
1511 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1512 The I= interface field on outgoing lines is now after the H= remote
1513 host field, same as incoming lines. There is a separate
1514 outgoing_interface log selector which allows you to disable the
1517 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1518 If not running log_selector +smtp_connection the mainlog would be held
1519 open indefinitely after a "too many connections" event, including to a
1520 deleted file after a log rotate. Leave the per net connection logging
1521 leaving it open for efficiency as that will be quickly detected by the
1522 check on the next write.
1524 HS/01 Bug 1671: Fix post transport crash.
1525 Processing the wait-<transport> messages could crash the delivery
1526 process if the message IDs didn't exist for some reason. When
1527 using 'split_spool_directory=yes' the construction of the spool
1528 file name failed already, exposing the same netto behaviour.
1530 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1531 mime_regex ACL conditions.
1533 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1534 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1535 to HELO, local diagnostic string.
1537 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1538 log line to debug. Even when configured with a tls authenticator many
1539 client connections are expected to not authenticate in this way, so
1540 an authenticate fail is not an error.
1542 HS/02 Add the Exim version string to the process info. This way exiwhat
1543 gives some more detail about the running daemon.
1545 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1546 matter for fast-change records such as DNSBLs.
1548 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1549 retry record, even if constant. There may be multiple transports with
1550 different interface settings and the retry behaviour needs to be kept
1553 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1555 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1557 JH/10 Bug 840: fix log_defer_output option of pipe transport
1559 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1560 hosts_randomize. This matters a lot when combined with mua_wrapper.
1562 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1563 ${quote_pgsql:<string>} operator.
1565 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1567 JH/14 Change header file name for freeradius-client. Relevant if compiling
1568 with Radius support; from the Gentoo tree and checked under Fedora.
1570 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1572 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1573 option transport_filter, meaning no filtering.
1575 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1577 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1578 defaults to "*" (all hosts). The variable is now available when not built
1579 with TLS, default unset, mainly to enable keeping the testsuite sane.
1580 If a server certificate is not supplied (via tls_certificate) an error is
1581 logged, and clients will find TLS connections fail on startup. Presumably
1582 they will retry in-clear.
1583 Packagers of Exim are strongly encouraged to create a server certificate
1584 at installation time.
1586 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1587 with the $config_file variable.
1589 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1590 in transport context, after the attempt, and per-recipient. The latter type
1591 is per host attempted. The event data is the error message, and the errno
1592 information encodes the lookup type (A vs. MX) used for the (first) host,
1593 and the trailing two digits of the smtp 4xx response.
1595 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1596 to write to mainlog (or rejectlog, paniclog) in the window between file
1597 creation and permissions/ownership being changed. Particularly affects
1598 installations where exicyclog is run as root, rather than exim user;
1599 result is that the running daemon panics and dies.
1601 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1603 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1604 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1605 "pri" and "weight". Note that the previous implicit priority given by the
1606 list order is no longer honoured.
1608 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1609 for DKIM processing.
1611 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1612 by defining SUPPORT_SOCKS.
1614 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1615 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1616 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1617 variables are renamed to proxy_{local,external}_{address,port}.
1619 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1620 for a build by defining SUPPORT_I18N
1622 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1623 of the query string, and make ${quote_redis:} do that quoting.
1625 JH/29 Move Events support from Experimental to mainline, enabled by default
1626 and removable for a build by defining DISABLE_EVENT.
1628 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1630 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1631 cached by the daemon.
1633 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1634 by defining LOOKUP_REDIS. The libhiredis library is required.
1636 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1637 keys are given for lookup.
1639 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1640 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1641 only supported when built with TLS support. The PolarSSL SHA routines
1642 are still used when the TLS library is too old for convenient support.
1644 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1645 openssl_options), for security. OpenSSL forces this from version 1.1.0
1646 server-side so match that on older versions.
1648 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1649 allocation for $value could be released as the expansion processing
1650 concluded, but leaving the global pointer active for it.
1652 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1653 and to use the domains and local_parts ACL conditions.
1655 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1656 incorrectly not doubled on cutthrough transmission, hence seen as a
1657 body-termination at the receiving system - resulting in truncated mails.
1658 Commonly the sender saw a TCP-level error, and retransmitted the message
1659 via the normal store-and-forward channel. This could result in duplicates
1660 received - but deduplicating mailstores were liable to retain only the
1661 initial truncated version.
1663 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1665 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1667 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1668 we're in there, support oversigning also; bug 1309.
1670 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1672 HS/04 Add support for keep_environment and add_environment options.
1674 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1675 either intentional arithmetic overflow during PRNG, or testing config-
1678 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1679 delivery resulted in actual delivery. Cancel cutthrough before DATA
1682 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1683 rejects a recipient: pass the reject to the originator.
1685 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1686 Many were false-positives and ignorable, but it's worth fixing the
1689 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1690 for the new environment-manipulation done at startup. Move the routines
1691 from being local to tls.c to being global via the os.c file.
1693 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1694 an extract embedded as result-arg for a map, the first arg for extract
1695 is unavailable so we cannot tell if this is a numbered or keyed
1696 extraction. Accept either.
1702 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1705 JH/02 The smtp transport option "multi_domain" is now expanded.
1707 JH/03 The smtp transport now requests PRDR by default, if the server offers
1710 JH/04 Certificate name checking on server certificates, when exim is a client,
1711 is now done by default. The transport option tls_verify_cert_hostnames
1712 can be used to disable this per-host. The build option
1713 EXPERIMENTAL_CERTNAMES is withdrawn.
1715 JH/05 The value of the tls_verify_certificates smtp transport and main options
1716 default to the word "system" to access the system default CA bundle.
1717 For GnuTLS, only version 3.0.20 or later.
1719 JH/06 Verification of the server certificate for a TLS connection is now tried
1720 (but not required) by default. The verification status is now logged by
1721 default, for both outbound TLS and client-certificate supplying inbound
1724 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1727 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1728 Status Notification (bounce) messages are now MIME format per RFC 3464.
1729 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1730 under the control of the dsn_advertise_hosts option, and routers may
1731 have a dsn_lasthop option.
1733 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1734 default, modifiable by a malware= option. The list separator for
1735 the options can now be changed in the usual way. Bug 68.
1737 JH/10 The smtp_receive_timeout main option is now expanded before use.
1739 JH/11 The incoming_interface log option now also enables logging of the
1740 local interface on delivery outgoing connections.
1742 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1743 if the interface and destination host and port all match.
1745 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1748 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1749 Patch from Andrew Lewis.
1751 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1752 now supports optional time-restrictions, weighting, and priority
1753 modifiers per server. Patch originally by <rommer@active.by>.
1755 JH/16 The spamd_address main option now supports a mixed list of local
1756 and remote servers. Remote servers can be IPv6 addresses, and
1757 specify a port-range.
1759 JH/17 Bug 68: The spamd_address main option now supports an optional
1760 timeout value per server.
1762 JH/18 Bug 1581: Router and transport options headers_add/remove can
1763 now have the list separator specified.
1765 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1768 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1771 JH/21 Support for the A6 type of dns record is withdrawn.
1773 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1774 rather than the verbs used.
1776 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1777 from 255 to 1024 chars.
1779 JH/24 Verification callouts now attempt to use TLS by default.
1781 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1782 are generic router options now. The defaults didn't change.
1784 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1785 Original patch from Alexander Shikoff, worked over by JH.
1787 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1788 files if ZCAT_COMMAND is not executable.
1790 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1792 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1794 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1795 Normally benign, it bites when the pair was led to by a CNAME;
1796 modern usage is to not canonicalize the domain to a CNAME target
1797 (and we were inconsistent anyway for A-only vs AAAA+A).
1799 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1801 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1802 when evaluating $sender_host_dnssec.
1804 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1805 $sender_helo_dnssec variable.
1807 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1809 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1811 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1813 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1814 documented as working, but never had. Support all but $spam_report.
1816 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1817 added for tls authenticator.
1819 HS/03 Add perl_taintmode main config option
1825 TL/01 When running the test suite, the README says that variables such as
1826 no_msglog_check are global and can be placed anywhere in a specific
1827 test's script, however it was observed that placement needed to be near
1828 the beginning for it to behave that way. Changed the runtest perl
1829 script to read through the entire script once to detect and set these
1830 variables, reset to the beginning of the script, and then run through
1831 the script parsing/test process like normal.
1833 TL/02 The BSD's have an arc4random API. One of the functions to induce
1834 adding randomness was arc4random_stir(), but it has been removed in
1835 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1836 function when detected.
1838 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1839 cause callback expansion.
1841 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1842 syntax errors in an expansion can be treated as a string instead of
1843 logging or causing an error, due to the internal use of bool_lax
1844 instead of bool when processing it.
1846 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1847 server certificates when making smtp deliveries.
1849 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1851 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1853 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1855 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1856 Merged patch from Sebastian Wiedenroth.
1858 JH/05 Fix results-pipe from transport process. Several recipients, combined
1859 with certificate use, exposed issues where response data items split
1860 over buffer boundaries were not parsed properly. This eventually
1861 resulted in duplicates being sent. This issue only became common enough
1862 to notice due to the introduction of connection certificate information,
1863 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1865 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1866 size buffer was used, resulting in syntax errors when an expansion
1869 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1870 version 3.3.6 or later.
1872 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1873 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1874 both become "event_action", the variables become $event_name, $event_data
1875 and $event_defer_errno. There is a new variable $verify_mode, usable in
1876 routers, transports and related events. The tls:cert event is now also
1877 raised for inbound connections, if the main configuration event_action
1880 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1881 early OCSP support, but no stapling (appears to be less than 1.0.0).
1883 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1884 server certificate names available under the smtp transport option
1885 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1888 JH/10 Time-related extraction expansions from certificates now use the main
1889 option "timezone" setting for output formatting, and are consistent
1890 between OpenSSL and GnuTLS compilations. Bug 1541.
1892 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1893 encoded parameter in the incoming message. Bug 1558.
1895 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1896 include certificate info, eximon was claiming there were spoolfile
1899 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1901 JH/14 Log delivery-related information more consistently, using the sequence
1902 "H=<name> [<ip>]" wherever possible.
1904 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1905 are problematic for Debian distribution, omit them from the release
1908 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1910 JH/16 Fix string representation of time values on 64bit time_t architectures.
1913 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1914 output list separator was used.
1919 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1920 checkers that were complaining about end of non-void function with no
1923 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1924 This was a regression introduced in 4.83 by another bugfix.
1926 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1928 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1929 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1935 TF/01 Correctly close the server side of TLS when forking for delivery.
1937 When a message was received over SMTP with TLS, Exim failed to clear up
1938 the incoming connection properly after forking off the child process to
1939 deliver the message. In some situations the subsequent outgoing
1940 delivery connection happened to have the same fd number as the incoming
1941 connection previously had. Exim would try to use TLS and fail, logging
1942 a "Bad file descriptor" error.
1944 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1945 utilities have not been installed.
1947 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1948 temporary space as the ACL may create new global variables.
1950 TL/01 LDAP support uses per connection or global context settings, depending
1951 upon the detected version of the libraries at build time.
1953 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1954 to extract and use the src ip:port in logging and expansions as if it
1955 were a direct connection from the outside internet. PPv2 support was
1956 updated based on HAProxy spec change in May 2014.
1958 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1960 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1961 Properly escape header and check for NULL return.
1963 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1966 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1968 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1969 characters in header names, implemented as a verify condition.
1970 Contributed by Michael Fischer v. Mollard.
1972 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1973 results permerror and temperror. Previous values are deprecated but
1974 still accepted. In a future release, err_perm and err_temp will be
1975 completely removed, which will be a backward incompatibility if the
1976 ACL tests for either of these two old results. Patch contributed by
1977 user bes-internal on the mailing list.
1979 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1981 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1982 selectors, in both main and reject logs.
1984 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1987 JH/07 Add malware type "sock" for talking to simple daemon.
1989 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1991 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1992 routers/transports under cutthrough routing.
1994 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1995 numbers. Touch up "bool" conditional to keep the same definition.
1997 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1999 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2001 JH/12 Expand items in router/transport headers_add or headers_remove lists
2002 individually rather than the list as a whole. Bug 1452.
2004 Required for reasonable handling of multiple headers_ options when
2005 they may be empty; requires that headers_remove items with embedded
2006 colons must have them doubled (or the list-separator changed).
2008 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2009 view the policy declared in the DMARC record. Currently, $dmarc_status
2010 is a combined value of both the record presence and the result of the
2013 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2015 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2016 dnslookup router and the smtp transport (applying to the forward
2019 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2020 of ldap servers used for a specific lookup. Patch provided by Heiko
2023 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2024 New variable $lookup_dnssec_authenticated for observability.
2026 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2027 Patch submitted by Lars Timman.
2029 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2031 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2032 Requires trusted mode and valid format message id, aborts otherwise.
2033 Patch contributed by Heiko Schlichting.
2035 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2036 certextract with support for various fields. Bug 1358.
2038 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2039 is requested by default, modifiable by smtp transport option
2042 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2043 operate on certificate variables to give certificate fingerprints
2044 Also new ${sha256:cert_variable}.
2046 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2048 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2049 Christian Aistleitner.
2051 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2053 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2054 file. Patch from Wolfgang Breyha.
2056 JH/25 Expand the coverage of the delivery $host and $host_address to
2057 client authenticators run in verify callout. Bug 1476.
2059 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2060 align with daemon_smtp_ports. Bug 72.
2062 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2063 support and error reporting did not work properly.
2065 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2066 and is readable. Patch from Andrew Colin Kissa.
2068 TL/14 Enhance documentation of ${run expansion and how it parses the
2069 commandline after expansion, particularly in the case when an
2070 unquoted variable expansion results in an empty value.
2072 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2074 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2075 by renaming away from it. Addresses GH issue 15, reported by
2078 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2080 TL/15 SECURITY: prevent double expansion in math comparison functions
2081 (can expand unsanitized data). Not remotely exploitable.
2088 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2090 PP/02 Make -n do something, by making it not do something.
2091 When combined with -bP, the name of an option is not output.
2093 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2096 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2097 $sender_host_name and config options to manage this, and basic check
2100 PP/05 DSCP support for outbound connections and control modifier for inbound.
2102 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2103 (Only plugin which currently uses this is kerberos4, which nobody should
2104 be using, but we should make it available and other future plugins might
2105 conceivably use it, even though it would break NAT; stuff *should* be
2106 using channel bindings instead).
2108 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2109 name; added for Sendmail compatibility; requires admin caller.
2110 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2111 just ignore it); requires trusted caller.
2112 Also parse but ignore: -Ac -Am -X<logfile>
2115 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2117 TL/02 Add +smtp_confirmation as a default logging option.
2119 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2120 Patch by Magnus Holmgren from 2007-02-20.
2122 TL/04 Bugzilla 1281 - Spec typo.
2123 Bugzilla 1283 - Spec typo.
2124 Bugzilla 1290 - Spec grammar fixes.
2126 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2128 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2130 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2131 mailing list by by Dmitry Isaikin.
2133 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2135 JH/02 Support "G" suffix to numbers in ${if comparisons.
2137 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2139 NM/01 Bugzilla 1197 - Spec typo
2140 Bugzilla 1196 - Spec examples corrections
2142 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2144 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2145 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2148 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2149 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2151 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2152 "acl {{name}{arg}...}", and optional args on acl condition
2155 JH/05 Permit multiple router/transport headers_add/remove lines.
2157 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2159 JH/07 Avoid using a waiting database for a single-message-only transport.
2160 Performance patch from Paul Fisher. Bugzilla 1262.
2162 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2165 JH/09 Add $headers_added variable, with content from use of ACL modifier
2166 add_header (but not yet added to the message). Bugzilla 199.
2168 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2169 Pulled from Bugzilla 817 by Wolfgang Breyha.
2171 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2173 (nb: this is the same fix as in Exim 4.80.1)
2175 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2178 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2181 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2183 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2184 advertise SMTP AUTH mechanism to us, instead of a generic
2185 protocol violation error. Also, make Exim more robust to bad
2186 data from the Dovecot auth socket.
2188 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2190 When a queue runner is handling a message, Exim first routes the
2191 recipient addresses, during which it prunes them based on the retry
2192 hints database. After that it attempts to deliver the message to
2193 any remaining recipients. It then updates the hints database using
2196 So if a recipient address works intermittently, it can get repeatedly
2197 deferred at routing time. The retry hints record remains fresh so the
2198 address never reaches the final cutoff time.
2200 This is a fairly common occurrence when a user is bumping up against
2201 their storage quota. Exim had some logic in its local delivery code
2202 to deal with this. However it did not apply to per-recipient defers
2203 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2205 This change adds a proper retry rule check during routing so that the
2206 final cutoff time is checked against the message's age. We only do
2207 this check if there is an address retry record and there is not a
2208 domain retry record; this implies that previous attempts to handle
2209 the address had the retry_use_local_parts option turned on. We use
2210 this as an approximation for the destination being like a local
2211 delivery, as in LMTP.
2213 I suspect this new check makes the old local delivery cutoff check
2214 redundant, but I have not verified this so I left the code in place.
2216 TF/02 Correct gecos expansion when From: is a prefix of the username.
2218 Test 0254 submits a message to Exim with the header
2222 When I ran the test suite under the user fanf2, Exim expanded
2223 the header to contain my full name, whereas it should have added
2224 a Resent-Sender: header. It erroneously treats any prefix of the
2225 username as equal to the username.
2227 This change corrects that bug.
2229 GF/01 DCC debug and logging tidyup
2230 Error conditions log to paniclog rather than rejectlog.
2231 Debug lines prefixed by "DCC: " to remove any ambiguity.
2233 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2235 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2236 Bug spotted by Jeremy Harris; was flawed since initial commit.
2237 Would have resulted in OCSP responses post-SNI triggering an Exim
2238 NULL dereference and crash.
2240 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2242 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2243 Bug detection, analysis and fix by Samuel Thibault.
2244 Bugzilla 1331, Debian bug #698092.
2246 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2248 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2249 Server implementation by Todd Lyons, client by JH.
2250 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2251 config variable "prdr_enable" controls whether the server
2252 advertises the facility. If the client requests PRDR a new
2253 acl_data_smtp_prdr ACL is called once for each recipient, after
2254 the body content is received and before the acl_smtp_data ACL.
2255 The client is controlled by both of: a hosts_try_prdr option
2256 on the smtp transport, and the server advertisement.
2257 Default client logging of deliveries and rejections involving
2258 PRDR are flagged with the string "PRDR".
2260 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2261 fclose(). Diagnosis by Todd Lyons.
2263 PP/17 Update configure.default to handle IPv6 localhost better.
2264 Patch by Alain Williams (plus minor tweaks).
2267 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2268 This is now consistent with GnuTLS, and is now documented: the
2269 previous undocumented portable approach to treating the option as
2270 unset was to force an expansion failure. That still works, and
2271 an empty string is now equivalent.
2273 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2274 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2275 not performing validation itself.
2277 PP/20 Added force_command boolean option to pipe transport.
2278 Patch from Nick Koston, of cPanel Inc.
2280 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2283 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2285 PP/21 Fix eximon continuous updating with timestamped log-files.
2286 Broken in a format-string cleanup in 4.80, missed when I repaired the
2287 other false fix of the same issue.
2288 Report and fix from Heiko Schlichting.
2291 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2292 Report from Prashanth Katuri.
2294 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2295 It's SecureTransport, so affects any MacOS clients which use the
2296 system-integrated TLS libraries, including email clients.
2298 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2299 using a MIME ACL for non-SMTP local injection.
2300 Report and assistance in diagnosis by Warren Baker.
2302 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2304 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2306 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2307 last id that failed so it may be referenced in subsequent ACL's.
2309 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2312 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2313 ldap library initialization, allowing self-signed CA's to be
2314 used. Also properly sets require_cert option later in code by
2315 using NULL (global ldap config) instead of ldap handle (per
2316 session). Bug diagnosis and testing by alxgomz.
2318 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2319 the src/util/ subdirectory.
2321 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2322 renamed to Transport Post Delivery Action by Jeremy Harris, as
2325 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2326 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2327 redis_servers = needs to be configured which will be used by the redis
2328 lookup. Patch from Warren Baker, of The Packet Hub.
2330 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2332 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2333 hostname or reverse DNS when processing a host list. Used suggestions
2334 from multiple comments on this bug.
2336 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2338 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2339 Missed a few lines, added it to make the runtest require no keyboard
2342 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2343 contains upper case chars. Make router use caseful_local_part.
2345 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2346 support when GnuTLS has been built with p11-kit.
2352 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2354 This, or similar/improved, will also be change PP/11 of 4.82.
2360 PP/01 Handle short writes when writing local log-files.
2361 In practice, only affects FreeBSD (8 onwards).
2362 Bugzilla 1053, with thanks to Dmitry Isaikin.
2364 NM/01 Bugzilla 949 - Documentation tweak
2366 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2369 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2371 PP/02 Implemented gsasl authenticator.
2373 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2375 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2376 `pkg-config foo` for cflags/libs.
2378 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2379 with rest of GSASL and with heimdal_gssapi.
2381 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2382 `pkg-config foo` for cflags/libs for the TLS implementation.
2384 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2385 properties get this fed in as external SSF. A number of robustness
2386 and debugging improvements to the cyrus_sasl authenticator.
2388 PP/08 cyrus_sasl server now expands the server_realm option.
2390 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2391 Patch by Jeremy Harris.
2393 PP/10 Added dbmjz lookup type.
2395 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2397 PP/12 MAIL args handles TAB as well as SP, for better interop with
2398 non-compliant senders.
2399 Analysis and variant patch by Todd Lyons.
2401 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2402 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2403 Patch from Dirk Mueller <dmueller@suse.com>
2405 PP/13 tls_peerdn now print-escaped for spool files.
2406 Observed some $tls_peerdn in wild which contained \n, which resulted
2407 in spool file corruption.
2409 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2410 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2411 or write after TLS renegotiation, which otherwise led to messages
2414 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2415 as a tracking header (ie: a signed header comes before the signature).
2416 Patch from Wolfgang Breyha.
2418 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2419 comma-sep list; embedded commas doubled.
2421 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2423 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2425 Report and patch from Dmitry Banschikov.
2427 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2428 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2429 failures. We appear to now support TLS1.1+ with Exim.
2431 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2432 lets Exim select keys and certificates based upon TLS SNI from client.
2433 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2434 before an outbound SMTP session. New log_selector, +tls_sni.
2436 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2437 NULL dereference. Report and patch from Alun Jones.
2439 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2440 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2441 Not seeing resolver debug output on NetBSD, but suspect this is a
2442 resolver implementation change.
2444 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2445 Left warnings. Added "eximon gdb" invocation mode.
2447 PP/21 Defaulting "accept_8bitmime" to true, not false.
2449 PP/22 Added -bw for inetd wait mode support.
2451 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2452 locate the relevant includes and libraries. Made this the default.
2454 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2455 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2457 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2458 This may cause build issues on older platforms.
2460 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2461 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2462 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2463 Added SNI support via GnuTLS too.
2464 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2466 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2468 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2469 Applied second patch from Janne, implementing suggestion to default
2470 multiple-strings-in-record handling to match SPF spec.
2472 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2474 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2475 read-only, out of scope).
2476 Patch from Wolfgang Breyha, report from Stuart Northfield.
2478 PP/29 Fix three issues highlighted by clang analyser static analysis.
2479 Only crash-plausible issue would require the Cambridge-specific
2480 iplookup router and a misconfiguration.
2481 Report from Marcin Mirosław.
2483 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2485 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2486 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2487 As part of this, removing so much warning spew let me fix some minor
2488 real issues in debug logging.
2490 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2491 assignment on my part. Fixed.
2493 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2494 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2495 Janne Snabb (who went above and beyond: thank you).
2497 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2498 string otherwise requires a connection and a bunch more work and it's
2499 relatively easy to get wrong. Should also expose TLS library linkage
2502 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2503 64-bit ${eval} (JH/03).
2505 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2506 GNU libc to support some of the 64-bit stuff, should not lead to
2507 conflicts. Defined before os.h is pulled in, so if a given platform
2508 needs to override this, it can.
2510 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2511 protection layer was required, which is not implemented.
2512 Bugzilla 1254, patch from Wolfgang Breyha.
2514 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2515 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2516 tls_dhparam take prime identifiers. Also unbreak combination of
2517 OpenSSL+DH_params+TLSSNI.
2519 PP/39 Disable SSLv2 by default in OpenSSL support.
2525 PP/01 Solaris build fix for Oracle's LDAP libraries.
2526 Bugzilla 1109, patch from Stephen Usher.
2528 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2530 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2533 TF/02 Fix a couple more cases where we did not log the error message
2534 when unlink() failed. See also change 4.74-TF/03.
2536 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2537 lock up or crash if it happened to be inside a call to libc when it
2538 got a SIGUSR1 from exiwhat.
2540 The SIGUSR1 handler appends the current process status to the process
2541 log which is later printed by exiwhat. It used to use the general
2542 purpose logging code to do this, but several functions it calls are
2543 not safe for signals.
2545 The new output code in the SIGUSR1 handler is specific to the process
2546 log, and simple enough that it's easy to inspect for signal safety.
2547 Removing some special cases also simplifies the general logging code.
2548 Removing the spurious timestamps from the process log simplifies
2551 TF/04 Improved ratelimit ACL condition.
2553 The /noupdate option has been deprecated in favour of /readonly which
2554 has clearer semantics. The /leaky, /strict, and /readonly update modes
2555 are mutually exclusive. The update mode is no longer included in the
2556 database key; it just determines when the database is updated. (This
2557 means that when you upgrade Exim will forget old rate measurements.)
2559 Exim now checks that the per_* options are used with an update mode that
2560 makes sense for the current ACL. For example, when Exim is processing a
2561 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2562 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2563 must specify per_mail/readonly. If you omit the update mode it defaults to
2564 /leaky where that makes sense (as before) or /readonly where required.
2566 The /noupdate option is now undocumented but still supported for
2567 backwards compatibility. It is equivalent to /readonly except that in
2568 ACLs where /readonly is required you may specify /leaky/noupdate or
2569 /strict/noupdate which are treated the same as /readonly.
2571 A useful new feature is the /count= option. This is a generalization
2572 of the per_byte option, so that you can measure the throughput of other
2573 aggregate values. For example, the per_byte option is now equivalent
2574 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2576 The per_rcpt option has been generalized using the /count= mechanism
2577 (though it's more complicated than the per_byte equivalence). When it is
2578 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2579 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2580 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2581 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2582 Note that using per_rcpt with a non-readonly update mode in more than
2583 one ACL will cause the recipients to be double-counted. (The per_mail
2584 and per_byte options don't have this problem.)
2586 The handling of very low rates has changed slightly. If the computed rate
2587 is less than the event's count (usually one) then this event is the first
2588 after a long gap. In this case the rate is set to the same as this event's
2589 count, so that the first message of a spam run is counted properly.
2591 The major new feature is a mechanism for counting the rate of unique
2592 events. The new per_addr option counts the number of different
2593 recipients that someone has sent messages to in the last time period. It
2594 behaves like per_rcpt if all the recipient addresses are different, but
2595 duplicate recipient addresses do not increase the measured rate. Like
2596 the /count= option this is a general mechanism, so the per_addr option
2597 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2598 example, measure the rate that a client uses different sender addresses
2599 with the options per_mail/unique=$sender_address. There are further
2600 details in the main documentation.
2602 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2604 TF/06 Removed a few PCRE remnants.
2606 TF/07 Automatically extract Exim's version number from tags in the git
2607 repository when doing development or release builds.
2609 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2610 Bugzilla 879. Patch from Paul Fisher.
2612 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2613 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2616 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2618 PP/05 Variable $av_failed, true if the AV scanner deferred.
2619 Bugzilla 1078. Patch from John Horne.
2621 PP/06 Stop make process more reliably on build failure.
2622 Bugzilla 1087. Patch from Heiko Schlittermann.
2624 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2625 Bugzilla 1089. Patch from Heiko Schlittermann.
2627 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2628 Bugzilla 1131. Patch from Holger Weiß.
2630 PP/09 Handle IPv6 addresses with SPF.
2631 Bugzilla 860. Patch from Wolfgang Breyha.
2633 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2635 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2638 PP/11 match_* no longer expand right-hand-side by default.
2639 New compile-time build option, EXPAND_LISTMATCH_RHS.
2640 New expansion conditions, "inlist", "inlisti".
2642 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2644 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2646 PP/14 fix log_write() format string regression from TF/03.
2647 Bugzilla 1152. Patch from Dmitry Isaikin.
2653 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2655 PP/02 Harmonised TLS library version reporting; only show if debugging.
2656 Layout now matches that introduced for other libraries in 4.74 PP/03.
2658 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2660 PP/04 New "dns_use_edns0" global option.
2662 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2665 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2666 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2668 TK/01 Updated PolarSSL code to 0.14.2.
2669 Bugzilla 1097. Patch from Andreas Metzler.
2671 PP/07 Catch divide-by-zero in ${eval:...}.
2672 Fixes bugzilla 1102.
2674 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2677 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2678 format-string attack -- SECURITY: remote arbitrary code execution.
2680 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2681 time unintentionally subject to list matching rules, letting the header
2682 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2683 arbitrary string expansion). This allowed for information disclosure.
2685 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2686 INT_MIN/-1 -- value coerced to INT_MAX.
2692 NM/01 Workaround for PCRE version dependency in version reporting
2695 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2696 This fixes portability to compilers other than gcc, notably
2697 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2699 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2700 makefiles for portability to HP-UX and POSIX correctness.
2702 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2703 Also via indented variable definition in the Makefile.
2704 (Debugging by Oliver Heesakkers).
2706 PP/02 Restore caching of spamd results with expanded spamd_address.
2707 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2709 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2710 Improves build reliability. Fix from: Frank Elsner
2712 NM/02 Fix wide character breakage in the rfc2047 coding
2713 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2715 NM/03 Allow underscore in dnslist lookups
2716 Fixes bug 1026. Patch from Graeme Fowler
2718 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2719 Code patches from Adam Ciarcinski of NetBSD.
2721 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2724 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2725 is logged, to avoid truncation. Patch from John Horne.
2727 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2728 Patch from Jakob Hirsch.
2730 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2731 SQL string expansion failure details.
2732 Patch from Andrey Oktyabrski.
2734 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2735 Patch from Simon Arlott.
2737 PP/09 New lookups functionality failed to compile on old gcc which rejects
2738 extern declarations in function scope.
2739 Patch from Oliver Fleischmann
2741 PP/10 Use sig_atomic_t for flags set from signal handlers.
2742 Check getgroups() return and improve debugging.
2743 Fixed developed for diagnosis in bug 927 (which turned out to be
2746 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2747 Patch from Mark Zealey.
2749 PP/12 Bugzilla 1056: Improved spamd server selection.
2750 Patch from Mark Zealey.
2752 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2753 Based on patch from Heiko Schlittermann.
2755 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2756 Patch from Uwe Doering, sign-off by Michael Haardt.
2758 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2759 variable declaration deep within a block. Bug and patch from
2762 PP/15 lookups-Makefile IRIX compatibility coercion.
2764 PP/16 Make DISABLE_DKIM build knob functional.
2766 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2767 Patch by Simon Arlott
2769 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2770 variable argument macros. Our copy now differs from upstream.
2776 TF/01 Failure to get a lock on a hints database can have serious
2777 consequences so log it to the panic log.
2779 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2780 controlled using the smtp_confirmation log selector.
2782 TF/03 Include the error message when we fail to unlink a spool file.
2784 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2785 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2786 for maintaining out-of-tree patches for some time.
2788 PP/01 Bugzilla 139: Documentation and portability issues.
2789 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2790 Handle per-OS dynamic-module compilation flags.
2792 PP/02 Let /dev/null have normal permissions.
2793 The 4.73 fixes were a little too stringent and complained about the
2794 permissions on /dev/null. Exempt it from some checks.
2795 Reported by Andreas M. Kirchwitz.
2797 PP/03 Report version information for many libraries, including
2798 Exim version information for dynamically loaded libraries. Created
2799 version.h, now support a version extension string for distributors
2800 who patch heavily. Dynamic module ABI change.
2802 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2803 privilege escalation vulnerability whereby the Exim run-time user
2804 can cause root to append content of the attacker's choosing to
2807 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2810 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2811 If dropping privileges for untrusted macros, we disabled normal logging
2812 on the basis that it would fail; for the Exim run-time user, this is not
2813 the case, and it resulted in successful deliveries going unlogged.
2814 Fixed. Reported by Andreas Metzler.
2820 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2821 only prepend for the Resent-* case. Fixes regression introduced in
2822 Exim 4.70 by NM/22 for Bugzilla 607.
2824 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2825 increasing numbers of administrators be bitten by this.
2827 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2829 PP/03 Bugzilla 994: added openssl_options main configuration option.
2831 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2833 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2835 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2837 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2838 without a peer certificate, leading to a segfault because of an
2839 assumption that peers always have certificates. Be a little more
2840 paranoid. Problem reported by Martin Tscholak.
2842 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2843 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2844 NB: ClamAV planning to remove STREAM in "middle of 2010".
2845 CL also introduces -bmalware, various -d+acl logging additions and
2846 more caution in buffer sizes.
2848 PP/09 Implemented reverse_ip expansion operator.
2850 PP/10 Bugzilla 937: provide a "debug" ACL control.
2852 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2854 PP/12 Bugzilla 973: Implement --version.
2856 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2858 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2860 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2862 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2863 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2864 ignore trailing whitespace.
2866 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2868 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2871 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2872 Notification from Dr Andrew Aitchison.
2874 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2875 ExtendedDetectionInfo response format.
2876 Notification from John Horne.
2878 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2881 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2882 XSL and documented dependency on system catalogs, with examples of how
2885 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2888 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2889 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2890 configuration file which is writeable by the Exim user or group.
2892 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2893 of configuration files to cover files specified with the -C option if
2894 they are going to be used with root privileges, not just the default
2897 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2898 option (effectively making it always true).
2900 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2901 files to be used while preserving root privileges.
2903 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2904 that rogue child processes cannot use them.
2906 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2907 run-time user, instead of root.
2909 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2910 Exim run-time user without dropping privileges.
2912 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2913 result string, instead of calling string_vformat() twice with the same
2916 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2917 for other users. Others should always drop root privileges if they use
2918 -C on the command line, even for a whitelisted configure file.
2920 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2922 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2928 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2929 $data_path, and $header_path variables; fixed documentation bugs and
2932 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2933 exipick to access non-standard spools, including the "frozen" queue
2936 NM/01 Bugzilla 965: Support mysql stored procedures.
2937 Patch from Alain Williams
2939 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2941 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2942 Patch from Andreas Metzler
2944 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2945 Patch from Kirill Miazine
2947 NM/05 Bugzilla 671: Added umask to procmail example.
2949 JJ/03 installed exipick 20100323.0, fixing doc bug
2951 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2952 directory. Notification and patch from Dan Rosenberg.
2954 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2956 TK/02 Improve log output when DKIM signing operation fails.
2958 MH/01 Treat the transport option dkim_domain as a colon separated
2959 list, not as a single string, and sign the message with each element,
2960 omitting multiple occurences of the same signer.
2962 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2963 Bugzilla 985, 986. Patch by Simon Arlott
2965 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2966 Patch by Simon Arlott
2968 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2969 MBX locking. Notification from Dan Rosenberg.
2975 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2977 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2979 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2981 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2983 NM/04 Bugzilla 915: Flag broken perl installation during build.
2989 TK/01 Added patch by Johannes Berg that expands the main option
2990 "spamd_address" if it starts with a dollar sign.
2992 TK/02 Write list of recipients to X-Envelope-Sender header when building
2993 the mbox-format spool file for content scanning (suggested by Jakob
2996 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2997 (http://www.dcc-servers.net/) support via dccifd. Activated by
2998 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3000 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3001 by Mark Daniel Reidel <mr@df.eu>.
3003 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3004 When building exim an external PCRE library is now needed -
3005 PCRE is a system library on the majority of modern systems.
3006 See entry on PCRE_LIBS in EDITME file.
3008 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3009 conversation. Added nologin parameter to request.
3010 Patch contributed by Kirill Miazine.
3012 TF/01 Do not log submission mode rewrites if they do not change the address.
3014 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3016 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3017 log files in place. Contributed by Roberto Lima.
3019 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3021 TF/03 Bugzilla 615: When checking the local_parts router precondition
3022 after a local_part_suffix or local_part_prefix option, Exim now
3023 does not use the address's named list lookup cache, since this
3024 contains cached lookups for the whole local part.
3026 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3027 Robert Millan. Documentation is in experimental-spec.txt.
3029 TF/04 Bugzilla 668: Fix parallel build (make -j).
3031 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3033 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3034 Patch provided by Jan Srzednicki.
3036 TF/05 Leading white space used to be stripped from $spam_report which
3037 wrecked the formatting. Now it is preserved.
3039 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3040 that they are available at delivery time.
3042 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3044 TF/08 TLS error reporting now respects the incoming_interface and
3045 incoming_port log selectors.
3047 TF/09 Produce a more useful error message if an SMTP transport's hosts
3048 setting expands to an empty string.
3050 NM/06 Bugzilla 744: EXPN did not work under TLS.
3051 Patch provided by Phil Pennock.
3053 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3054 Patch provided by Richard Godbee.
3056 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3057 acl_smtp_notquit, added index entry.
3059 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3060 Patch provided by Eugene Bujak.
3062 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3063 accept(). Patch provided by Maxim Dounin.
3065 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3066 Patch provided by Phil Pennock.
3068 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3070 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3071 Patch provided by Brad "anomie" Jorsch.
3073 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3074 Patch provided by Dean Brooks.
3076 TK/05 Add native DKIM support (does not depend on external libraries).
3078 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3079 Patch provided by Graeme Fowler.
3081 NM/16 Bugzilla 851: Documentation example syntax fix.
3083 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3085 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3088 NM/19 Bugzilla 745: TLS version reporting.
3089 Patch provided by Phil Pennock.
3091 NM/20 Bugzilla 167: bool: condition support.
3092 Patch provided by Phil Pennock.
3094 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3095 clients. Patch provided by Phil Pennock.
3097 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3098 Patch provided by Brad "anomie" Jorsch.
3100 NM/23 Bugzilla 687: Fix misparses in eximstats.
3101 Patch provided by Heiko Schlittermann.
3103 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3104 Patch provided by Heiko Schlittermann.
3106 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3107 plus update to original patch.
3109 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3111 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3112 Patch provided by David Brownlee.
3114 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3116 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3118 NM/30 Bugzilla 888: TLS documentation bugfixes.
3120 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3122 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3123 Unlike the original bugzilla I have changed all shell scripts in src tree.
3125 NM/33 Bugzilla 898: Transport filter timeout fix.
3126 Patch by Todd Rinaldo.
3128 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3129 Patch by Serge Demonchaux.
3131 NM/35 Bugzilla 39: Base64 decode bug fixes.
3132 Patch by Jakob Hirsch.
3134 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3136 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3138 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3140 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3146 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3147 ALT-N's libdkim that I have put here:
3148 http://duncanthrax.net/exim-experimental/
3150 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3151 were called 'true' and it seems that C99 defines that as a reserved
3152 keyword to be used with 'bool' variable types. That means you could
3153 not include C99-style headers which use bools without triggering
3154 build errors in sieve.c.
3156 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3157 as mailq or other aliases. Changed the --help handling significantly
3158 to do whats expected. exim_usage() emits usage/help information.
3160 SC/01 Added the -bylocaldomain option to eximstats.
3162 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3164 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3166 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3172 PH/01 Another patch from the Sieve maintainer.
3174 PH/02 When an IPv6 address is converted to a string for single-key lookup
3175 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3176 dots are used instead of colons so that keys in lsearch files need not
3177 contain colons. This was done some time before quoting was made available
3178 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3179 (notated using the quote facility) so as to distinguish them from IPv4
3180 keys. This meant that lookups for IP addresses in host lists did not work
3181 for iplsearch lookups.
3183 This has been fixed by arranging for IPv6 addresses to be expressed with
3184 colons if the lookup type is iplsearch. This is not incompatible, because
3185 previously such lookups could never work.
3187 The situation is now rather anomalous, since one *can* have colons in
3188 ordinary lsearch keys. However, making the change in all cases is
3189 incompatible and would probably break a number of configurations.
3191 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3194 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3195 conversion specification without a maximum field width, thereby enabling
3196 a rogue spamd server to cause a buffer overflow. While nobody in their
3197 right mind would setup Exim to query an untrusted spamd server, an
3198 attacker that gains access to a server running spamd could potentially
3199 exploit this vulnerability to run arbitrary code as the Exim user.
3201 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3202 $primary_hostname instead of what libspf2 thinks the hosts name is.
3204 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3205 a directory entry by the name of the lookup key. Previously, if a
3206 symlink pointed to a non-existing file or a file in a directory that
3207 Exim lacked permissions to read, a lookup for a key matching that
3208 symlink would fail. Now it is enough that a matching directory entry
3209 exists, symlink or not. (Bugzilla 503.)
3211 PH/03 The body_linecount and body_zerocount variables are now exported in the
3214 PH/04 Added the $dnslist_matched variable.
3216 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3217 This means they are set thereafter only if the connection becomes
3220 PH/06 Added the client_condition to authenticators so that some can be skipped
3221 by clients under certain conditions.
3223 PH/07 The error message for a badly-placed control=no_multiline_responses left
3224 "_responses" off the end of the name.
3226 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3228 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3229 (without spaces) instead of just copying the configuration text.
3231 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3233 PH/11 Added $max_received_linelength.
3235 PH/12 Added +ignore_defer and +include_defer to host lists.
3237 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3238 way in which PCRE > 7.0 is built.
3240 PH/14 Implemented queue_only_load_latch.
3242 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3243 MAIL command. The effect was to mangle the value on 64-bit systems.
3245 PH/16 Another patch from the Sieve maintainer.
3247 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3249 PH/18 If a system quota error occurred while trying to create the file for
3250 a maildir delivery, the message "Mailbox is full" was not appended to the
3251 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3252 only to a quota excession during the actual writing of the file.
3254 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3255 characters?) which causes problems in log lines. The DN values are now
3256 passed through string_printing() before being added to log lines.
3258 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3259 and InterBase are left for another time.)
3261 PH/21 Added message_body_newlines option.
3263 PH/22 Guard against possible overflow in moan_check_errorcopy().
3265 PH/23 POSIX allows open() to be a macro; guard against that.
3267 PH/24 If the recipient of an error message contained an @ in the local part
3268 (suitably quoted, of course), incorrect values were put in $domain and
3269 $local_part during the evaluation of errors_copy.
3275 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3276 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3279 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3280 issue a MAIL command.
3282 PH/02 In an ACL statement such as
3284 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3286 if a client was not listed at all, or was listed with a value other than
3287 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3288 the condition was not true (as it should be), so access was not denied.
3289 The bug was that the ! inversion was incorrectly passed on to the second
3290 item. This has been fixed.
3292 PH/03 Added additional dnslists conditions == and =& which are different from
3293 = and & when the dns lookup returns more than one IP address.
3295 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3296 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3298 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3299 FSYNC, which compiles an option called disable_fsync that allows for
3300 bypassing fsync(). The documentation is heavily laced with warnings.
3302 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3304 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3305 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3306 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3307 including adding "make clean"; (3) Added -fPIC when compiling the test
3308 dynamically loaded module, to get rid of a warning.
3310 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3311 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3312 = 0s. The bug is otherwise harmless.
3314 PH/07 There was a bug in the dovecot authenticator such that the value of
3315 $auth1 could be overwritten, and so not correctly preserved, after a
3316 successful authentication. This usually meant that the value preserved by
3317 the server_setid option was incorrect.
3319 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3321 PH/09 Installed PCRE release 7.0.
3323 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3324 run for batched SMTP input. It is now run at the start of every message
3325 in the batch. While fixing this I discovered that the process information
3326 (output by running exiwhat) was not always getting set for -bs and -bS
3327 input. This is fixed, and it now also says "batched" for BSMTP.
3329 PH/11 Added control=no_pipelining.
3331 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3332 patch, slightly modified), and move the expansion of helo_data till after
3333 the connection is made in the smtp transport (so it can use these
3336 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3338 PH/14 Added log_selector = +pid.
3340 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3342 PH/16 Add ${if forany and ${if forall.
3344 PH/17 Added dsn_from option to vary the From: line in DSNs.
3346 PH/18 Flush SMTP output before performing a callout, unless control =
3347 no_callout_flush is set.
3349 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3350 was true (the default) a successful delivery failed to delete the retry
3351 item, thus causing premature timeout of the address. The bug is now
3354 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3356 PH/21 Long custom messages for fakedefer and fakereject are now split up
3357 into multiline responses in the same way that messages for "deny" and
3358 other ACL rejections are.
3360 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3361 with slight modification.
3363 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3364 draft, changing the syntax and factoring some duplicate code".
3366 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3367 for deliveries of the second and subsequent messages over the same SMTP
3370 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3371 ${reduce, with only minor "tidies".
3373 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3375 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3376 expansion side effects.
3378 PH/27 When a message times out after an over-quota error from an Exim-imposed
3379 quota, the bounce message says "mailbox is full". This message was not
3380 being given when it was a system quota that was exceeded. It now should
3383 MH/03 Made $recipients available in local_scan(). local_scan() already has
3384 better access to the recipient list through recipients_list[], but
3385 $recipients can be useful in postmaster-provided expansion strings.
3387 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3388 in the case of a MAIL command with additional options following the
3389 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3390 were accidentally chopped off.
3392 PH/29 SMTP synchronization checks are implemented when a command is read -
3393 there is a check that no more input is waiting when there shouldn't be
3394 any. However, for some commands, a delay in an ACL can mean that it is
3395 some time before the response is written. In this time, more input might
3396 arrive, invalidly. So now there are extra checks after an ACL has run for
3397 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3398 pipelining has not been advertised.
3400 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3402 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3403 returned in the SMTP response when smtp_return_error_details was false.
3404 This has been fixed.
3406 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3407 instead of the C I/O that was originally supplied, because problems were
3408 reported on Solaris.
3410 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3411 Exim which did not show up earlier: it was assuming that a call to
3412 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3413 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3414 was a macro that became an assignment, so it seemed to work. This has
3415 changed to a proper function call with a void return, hence the compile
3416 error. Exim's code has been fixed.
3418 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3421 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3423 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3425 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3426 because it turns out to be too big or there is a timeout) there is no
3427 "Completed" line in the log. When some messages of this type were
3428 selected by exigrep, they were listed as "not completed". Others were
3429 picked up by some special patterns. I have improved the selection
3430 criteria to be more general.
3432 PH/38 The host_find_failed option in the manualroute router can now be set
3433 to "ignore", to completely ignore a host whose IP address cannot be
3434 found. If all hosts are ignored, the behaviour is controlled by the new
3435 host_all_ignored option.
3437 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3438 homing or because of multiple MX records with /mx) generated more than
3439 one IP address, and the following item turned out to be the local host,
3440 all the secondary addresses of the first item were incorrectly removed
3441 from the list, along with the local host and any following hosts (which
3442 is what is supposed to happen).
3444 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3445 whoever called Exim into the -H file. In the case of the daemon it was
3446 behaving confusingly. When first started, it used values for whoever
3447 started the daemon, but after a SIGHUP it used the Exim user (because it
3448 calls itself on a restart). I have changed the code so that it now always
3451 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3452 message are rejected with the same error (e.g. no authentication or bad
3453 sender address), and a DATA command is nevertheless sent (as can happen
3454 with PIPELINING or a stupid MUA), the error message that was given to the
3455 RCPT commands is included in the rejection of the DATA command. This is
3456 intended to be helpful for MUAs that show only the final error to their
3459 PH/42 Another patch from the Sieve maintainer.
3461 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3462 Eximstats - Fixed some broken HTML links and added missing column headers
3464 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3465 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3467 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3469 PH/43 Yet another patch from the Sieve maintainer.
3471 PH/44 I found a way to check for a TCP/IP connection going away before sending
3472 the response to the final '.' that terminates a message, but only in the
3473 case where the client has not sent further data following the '.'
3474 (unfortunately, this is allowed). However, in many cases there won't be
3475 any further data because there won't be any more messages to send. A call
3476 to select() can be used: if it shows that the input is "ready", there is
3477 either input waiting, or the socket has been closed. An attempt to read
3478 the next input character can distinguish the two cases. Previously, Exim
3479 would have sent an OK response which the client would never have see.
3480 This could lead to message repetition. This fix should cure that, at
3481 least in a lot of common cases.
3483 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3484 advertised in response to EHLO.
3490 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3491 fixed by 4.65/MH/01 (is this a record?) are fixed:
3493 (i) An empty string was always treated as zero by the numeric comparison
3494 operators. This behaviour has been restored.
3496 (ii) It is documented that the numeric comparison operators always treat
3497 their arguments as decimal numbers. This was broken in that numbers
3498 starting with 0 were being interpreted as octal.
3500 While fixing these problems I realized that there was another issue that
3501 hadn't been noticed. Values of message_size_limit (both the global option
3502 and the transport option) were treated as octal if they started with 0.
3503 The documentation was vague. These values are now always treated as
3504 decimal, and I will make that clear in the documentation.
3510 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3511 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3514 MH/01 Don't check that the operands of numeric comparison operators are
3515 integers when their expansion is in "skipping" mode (fixes bug
3516 introduced by 4.64-PH/07).
3518 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3519 child addresses, Exim now panics and dies. Previously, because the count
3520 is held in a short int, deliveries were likely to be lost. As such a
3521 large number of recipients for a single message is ridiculous
3522 (performance will be very, very poor), I have chosen to impose a limit
3523 rather than extend the field.
3529 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3530 leftover -K file (the existence of which was triggered by #402).
3531 While we were at it, introduced process PID as part of the -K
3532 filename. This should rule out race conditions when creating
3535 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3536 processing considerably. Previous code took too long for large mails,
3537 triggering a timeout which in turn triggers #401.
3539 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3540 in the DK code in transports.c. sendfile() is not really portable,
3541 hence the _LINUX specificness.
3543 TF/01 In the add_headers option to the mail command in an Exim filter,
3544 there was a bug that Exim would claim a syntax error in any
3545 header after the first one which had an odd number of characters
3548 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3549 callout verification, Exim cached a "reject" for the entire domain. This
3550 is correct for most verifications, but it is not correct for a recipient
3551 verification with use_sender or use_postmaster set, because in that case
3552 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3553 case of MAIL FROM:<> rejection from other early rejections (e.g.
3554 rejection of HELO). When verifying a recipient using a non-null MAIL
3555 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3556 Whatever the result of the callout, the value of the domain cache is
3557 left unchanged (for any other kind of callout, getting as far as trying
3558 RCPT means that the domain itself is ok).
3560 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3563 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3564 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3565 session, the daemon ignores ECONNECT errors and logs others; it now
3566 ignores EPIPE as well.
3568 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3569 (quoted-printable decoding).
3571 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3572 later the small subsequent patch to fix an introduced bug.
3574 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3576 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3578 PH/08 An error is now given if message_size_limit is specified negative.
3580 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3581 to be given (somewhat) arbitrary names.
3583 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3586 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3587 miscellaneous code fixes
3589 PH/10 Added the log_reject_target ACL modifier to specify where to log
3592 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3593 hostname. This is wrong, because it relates to the incoming message (and
3594 probably the interface on which it is arriving) and not to the outgoing
3595 callout (which could be using a different interface). This has been
3596 changed to use the value of the helo_data option from the smtp transport
3597 instead - this is what is used when a message is actually being sent. If
3598 there is no remote transport (possible with a router that sets up host
3599 addresses), $smtp_active_hostname is used.
3601 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3602 tweaks were necessary in order to get it to work (see also 21 below):
3603 (a) The code assumed that strncpy() returns a negative number on buffer
3604 overflow, which isn't the case. Replaced with Exim's string_format()
3606 (b) There were several signed/unsigned issues. I just did the minimum
3607 hacking in of casts. There is scope for a larger refactoring.
3608 (c) The code used strcasecmp() which is not a standard C function.
3609 Replaced with Exim's strcmpic() function.
3610 (d) The code set only $1; it now sets $auth1 as well.
3611 (e) A simple test gave the error "authentication client didn't specify
3612 service in request". It would seem that Dovecot has changed its
3613 interface. Fortunately there's a specification; I followed it and
3614 changed what the client sends and it appears to be working now.
3616 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3619 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3620 address A is aliased to B and C, where B exists and C does not. Without
3621 -v the output is "A verified" because verification stops after a
3622 successful redirection if more than one address is generated. However,
3623 with -v the child addresses are also verified. Exim was outputting "A
3624 failed to verify" and then showing the successful verification for C,
3625 with its parentage. It now outputs "B failed to verify", showing B's
3626 parentage before showing the successful verification of C.
3628 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3629 look up a TXT record in a specific list after matching in a combined
3632 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3633 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3634 they consult the DNS. I had assumed they would set it the way they
3635 wanted; and indeed my experiments on Linux seem to show that in some
3636 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3637 To be on the safe side, however, I have now made the interface to
3638 host_find_byname() similar to host_find_bydns(), with an argument
3639 containing the DNS resolver options. The host_find_byname() function now
3640 sets these options at its start, just as host_find_bydns() does. The smtp
3641 transport options dns_qualify_single and dns_search_parents are passed to
3642 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3643 of host_find_byname() use the default settings of RES_DEFNAMES
3644 (qualify_single) but not RES_DNSRCH (search_parents).
3646 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3647 spool_read_header() do less string testing, by means of a preliminary
3648 switch on the second character of optional "-foo" lines. (This is
3649 overdue, caused by the large number of possibilities that now exist.
3650 Originally there were few.) While I was there, I also converted the
3651 str(n)cmp tests so they don't re-test the leading "-" and the first
3652 character, in the hope this might squeeze out yet more improvement.
3654 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3655 flag allowing group syntax was set by the header_syntax check but not
3656 turned off, possible causing trouble later; (2) The flag was not being
3657 set at all for the header_verify test, causing "group"-style headers to
3658 be rejected. I have now set it in this case, and also caused header_
3659 verify to ignore an empty address taken from a group. While doing this, I
3660 came across some other cases where the code for allowing group syntax
3661 while scanning a header line wasn't quite right (mostly, not resetting
3662 the flag correctly in the right place). These bugs could have caused
3663 trouble for malformed header lines. I hope it is now all correct.
3665 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3666 with the "reply" argument non-NULL. The code, however (which originally
3667 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3668 but it didn't always do it. This confused somebody who was copying the
3669 code for some other use. I have removed all the tests.
3671 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3672 feature that was used to support insecure browsers during the U.S. crypto
3673 embargo. It requires special client support, and Exim is probably the
3674 only MTA that supported it -- and would never use it because real RSA is
3675 always available. This code has been removed, because it had the bad
3676 effect of slowing Exim down by computing (never used) parameters for the
3677 RSA_EXPORT functionality.
3679 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3680 authenticator to fail if there's a tab character in the incoming data
3681 (there should never be unless someone is messing about, as it's supposed
3682 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3683 if the connection is using TLS or if the remote IP is the same as the
3684 local IP, and the "valid-client-cert option" if a client certificate has
3687 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3688 authenticators. This can be used for authorization after authentication
3689 succeeds. (In the case of plaintext, it servers for both authentication
3692 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3693 if any retry times were supplied.
3695 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3696 connection, where there is no client IP address to check. In this
3697 situation, the verify now always succeeds.
3699 PH/25 Applied John Jetmore's -Mset patch.
3701 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3703 PH/27 In a string expansion for a processed (not raw) header when multiple
3704 headers of the same name were present, leading whitespace was being
3705 removed from all of them, but trailing whitespace was being removed only
3706 from the last one. Now trailing whitespace is removed from each header
3707 before concatenation. Completely empty headers in a concatenation (as
3708 before) are ignored.
3710 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3711 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3713 PH/29 [Removed. This was a change that I later backed out, and forgot to
3714 correct the ChangeLog entry (that I had efficiently created) before
3715 committing the later change.]
3717 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3718 address errors (4xx response to RCPT) over the same connection as other
3719 messages routed to the same hosts. Such deliveries are always "forced",
3720 so retry times are not inspected. This resulted in far too many retries
3721 for the affected addresses. The effect occurred only when there were more
3722 hosts than the hosts_max_try setting in the smtp transport when it had
3723 the 4xx errors. Those hosts that it had tried were not added to the list
3724 of hosts for which the message was waiting, so if all were tried, there
3725 was no problem. Two fixes have been applied:
3727 (i) If there are any address or message errors in an SMTP delivery, none
3728 of the hosts (tried or untried) are now added to the list of hosts
3729 for which the message is waiting, so the message should not be a
3730 candidate for sending over the same connection that was used for a
3731 successful delivery of some other message. This seems entirely
3732 reasonable: after all the message is NOT "waiting for some host".
3733 This is so "obvious" that I'm not sure why it wasn't done
3734 previously. Hope I haven't missed anything, but it can't do any
3735 harm, as the worst effect is to miss an optimization.
3737 (ii) If, despite (i), such a delivery is accidentally attempted, the
3738 routing retry time is respected, so at least it doesn't keep
3739 hammering the server.
3741 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3742 in ${readsocket because some servers need this prod.
3744 PH/32 Added some extra debug output when updating a wait-xxx database.
3746 PH/33 The hint "could be header name not terminated by colon", which has been
3747 given for certain expansion errors for a long time, was not being given
3748 for the ${if def:h_colon_omitted{... case.
3750 PH/34 The spec says: "With one important exception, whenever a domain list is
3751 being scanned, $domain contains the subject domain." There was at least
3752 one case where this was not true.
3754 PH/35 The error "getsockname() failed: connection reset by peer" was being
3755 written to the panic log as well as the main log, but it isn't really
3756 panic-worthy as it just means the connection died rather early on. I have
3757 removed the panic log writing for the ECONNRESET error when getsockname()
3760 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3761 runs only) independently of the message's sender address. This meant
3762 that, if the 4xx error was in fact related to the sender, a different
3763 message to the same recipient with a different sender could confuse
3764 things. In particular, this can happen when sending to a greylisting
3765 server, but other circumstances could also provoke similar problems.
3766 I have changed the default so that the retry time for these errors is now
3767 based a combination of the sender and recipient addresses. This change
3768 can be overridden by setting address_retry_include_sender=false in the
3771 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3772 remote server are returned as part of bounce messages. This was not
3773 happening for LMTP over a pipe (the lmtp transport), but now it is the
3774 same for both kinds of LMTP.
3776 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3777 in header lines that were in fact CNAMEs. This is no longer the case.
3779 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3780 and queue runs started by the daemon processed all messages. This has
3781 been fixed so that -R and -S can now usefully be given with -q<time>.
3783 PH/40 Import PCRE release 6.7 (fixes some bugs).
3785 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3787 PH/42 Give an error if -q is specified more than once.
3789 PH/43 Renamed the variables $interface_address and $interface_port as
3790 $received_ip_address and $received_port, to make it clear that these
3791 values apply to message reception, and not to the outgoing interface when
3792 a message is delivered. (The old names remain recognized, of course.)
3794 PH/44 There was no timeout on the connect() call when using a Unix domain
3795 socket in the ${readsocket expansion. There now is.
3797 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3798 be meaningful with "accept".
3800 SC/01 Eximstats V1.43
3801 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3803 SC/02 Eximstats V1.44
3804 Use a glob alias rather than an array ref in the generated
3805 parser. This improves both readability and performance.
3807 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3808 Collect SpamAssassin and rejection statistics.
3809 Don't display local sender or destination tables unless
3810 there is data to show.
3811 Added average volumes into the top table text output.
3813 SC/04 Eximstats V1.46
3814 Collect data on the number of addresses (recipients)
3815 as well as the number of messages.
3817 SC/05 Eximstats V1.47
3818 Added 'Message too big' to the list of mail rejection
3819 reasons (thanks to Marco Gaiarin).
3821 SC/06 Eximstats V1.48
3822 Mainlog lines which have GMT offsets and are too short to
3823 have a flag are now skipped.
3825 SC/07 Eximstats V1.49 (Alain Williams)
3826 Added the -emptyok flag.
3828 SC/08 Eximstats V1.50
3829 Fixes for obtaining the IP address from reject messages.
3831 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3832 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3833 whitespace changes from 4.64-PH/27
3835 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3838 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3839 are found, allow negative numbers in numeric criteria)
3841 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3843 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3844 to match changes made in 4.64-PH/43
3846 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3848 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3849 "domain" is actually an IP address, and force a failure. This locks out
3850 those revolvers/nameservers that support "A-for-A" lookups, in
3851 contravention of the specifications.
3853 PH/48 When a host name was looked up from an IP address, and the subsequent
3854 forward lookup of the name timed out, the host name was left in
3855 $sender_host_name, contrary to the specification.
3857 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3858 restricted to single-key lookups, Exim was not diagnosing an error if
3859 * or *@ was used with a query-style lookup.
3861 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3863 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3864 long ago, but noone interested enough thought of it. Let's just say that
3865 the "1.1" means that there are some new functions that weren't there at
3866 some point in the past.
3868 PH/51 Error processing for expansion failure of helo_data from an smtp
3869 transport during callout processing was broken.
3871 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3872 tested/used via the -bh/-bhc/-bs options.
3874 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3875 bug, fixed in subsequent PCRE releases).
3877 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3878 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3880 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3886 SC/01 Use a glob alias rather than an array ref in eximstats generated
3887 parser. This improves both readability and performance.
3889 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3890 Don't display local sender or destination tables in eximstats unless
3891 there is data to show.
3892 Added average volumes into the eximstats top table text output.
3894 SC/03 Collect data on the number of addresses (recipients) as well
3895 as the number of messages in eximstats.
3897 TF/01 Correct an error in the documentation for the redirect router. Exim
3898 does (usually) call initgroups() when daemonizing.
3900 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3901 with consistent privilege compared to when running as a daemon.
3903 TF/03 Note in the spec that $authenticated_id is not set for local
3904 submissions from trusted users.
3906 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3907 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3909 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3910 by adding some example configuration directives to the default
3911 configuration file. A little bit of work is required to uncomment the
3912 directives and define how usernames and passwords are checked, but
3913 there is now a framework to start from.
3915 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3916 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3917 without this. I don't know how relevant this is to other LDAP libraries.
3919 PH/02 Add the verb name to the "unknown ACL verb" error.
3921 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3923 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3925 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3926 directory not expanded when it should be if an expanded home directory
3927 was set for the address (which is overridden by the transport).
3929 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3932 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3933 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3934 because it is too late at that time, and has no effect.
3936 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3937 security issue with \' (bugzilla #107). I could not use the
3938 PQescapeStringConn() function, because it needs a PGconn value as one of
3941 PH/08 When testing addresses using -bt, indicate those final addresses that
3942 are duplicates that would not cause an additional delivery. At least one
3943 person was confused, thinking that -bt output corresponded to deliveries.
3944 (Suppressing duplicates isn't a good idea as you lose the information
3945 about possibly different redirections that led to the duplicates.)
3947 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3948 systems where poll() doesn't work, in particular OS X.
3950 PH/10 Added more information to debugging output for retry time not reached.
3952 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3953 operations in malware.c.
3955 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3958 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3959 syslog_duplication set false, log lines that would normally be written
3960 both the the main log and to the reject log were not written to syslog at
3963 PH/14 In the default configuration, change the use of "message" in ACL warn
3964 statements to "add_header".
3966 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3967 not followed by a command (e.g. "seen endif").
3969 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3970 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3973 PH/17 Added extra conditions to the default value of delay_warning_condition
3977 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3978 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3979 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3982 The Auto-Submitted: and various List- headers are standardised, whereas I
3983 don't think Precedence: ever was.
3985 PH/18 Refactored debugging code in route_finduser() to show more information,
3986 in particular, the error code if getpwnam() issues one.
3988 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3989 This is apparently needed in addition to the PH/07 change above to avoid
3990 any possible encoding problems.
3992 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3993 but not after initializing Perl.
3995 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3996 output them only if debugging. By default they are written stderr,
3997 apparently, which is not desirable.
3999 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4002 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4005 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4007 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4008 authenticated or an ident call has been made. Suppress the default
4009 values for $authenticated_id and $authenticated_sender (but permit -oMai
4010 and -oMas) when testing with -bh.
4012 PH/24 Re-jigged the order of the tests in the default configuration so that the
4013 tests for valid domains and recipients precede the DNS black list and CSA
4014 tests, on the grounds that those ones are more expensive.
4016 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4017 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4018 This bug exists in every version of Exim that I still have, right back to
4021 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4022 However, an attempt to turn on case-sensitivity in a regex key by
4023 including (?-i) didn't work because the subject string was already
4024 lowercased, and the effects were non-intuitive. It turns out that a
4025 one-line patch can be used to allow (?-i) to work as expected.
4031 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4032 other effects) broke the use of negated acl sub-conditions.
4034 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4037 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4038 "Deny" causes Exim to reject the incoming connection with a 554 error.
4039 Unfortunately, if there is a major crisis, such as a disk failure,
4040 tcp-wrappers gives "deny", whereas what one would like would be some
4041 kind of temporary error. A kludge has been added to help with this.
4042 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4043 554 error is used if errno is still zero or contains ENOENT (which occurs
4044 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4047 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4049 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4050 errors. Otherwise a message that provokes a temporary error (when other
4051 messages do not) can cause a whole host to time out.
4053 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4054 addresses were routed directly to files or pipes from a redirect router.
4055 File deliveries just didn't batch; pipe deliveries might have suffered
4058 PH/06 A failure to get a lock for a hints database would erroneously always say
4059 "Failed to get write lock", even when it was really a read lock.
4061 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4062 of 0600. This has been changed to use the value of the lockfile_mode
4063 option (which defaults to 0600).
4065 PH/08 Applied small patch from the Sieve maintainer.
4067 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4068 folder from quota calculations, a direct delivery into this folder messed
4069 up the contents of the maildirsize file. This was because the regex was
4070 used only to exclude .Trash (or whatever) when the size of the mailbox
4071 was calculated. There was no check that a delivery was happening into an
4072 excluded directory. This bug has been fixed by ignoring all quota
4073 processing for deliveries into excluded directories.
4075 PH/10 Added the maildirfolder_create_regex option to appendfile.
4081 PH/01 The code for finding all the local interface addresses on a FreeBSD
4082 system running IPv6 was broken. This may well have applied to all BSD
4083 systems, as well as to others that have similar system calls. The broken
4084 code found IPv4 interfaces correctly, but gave incorrect values for the
4085 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4086 that it would not match correctly against @[] and not recognize the IPv6
4089 PH/02 The ipliteral router was not recognizing addresses of the form user@
4090 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4092 PH/03 Added disable_ipv6.
4094 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4095 decoded form, because it is most often used to construct To: headers
4096 lines in autoreplies, and the decoded form may well be syntactically
4097 invalid. However, $reply_address has leading white space removed, and all
4098 newlines turned into spaces so that the autoreply transport does not
4101 PH/05 If group was specified without a user on a router, and no group or user
4102 was specified on a transport, the group from the router was ignored.
4104 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4105 for visible compile-time settings that can be used to change these
4106 numbers, for those that want even more. Backwards compatibility with old
4107 spool files has been maintained. However, going back to a previous Exim
4108 release will lost any variables that are in spool files.
4110 PH/07 Two small changes when running in the test harness: increase delay when
4111 passing a TCP/IP connection to a new process, in case the original
4112 process has to generate a bounce, and remove special handling of
4113 127.0.0.2 (sic), which is no longer necessary.
4115 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4116 be the same on different OS.
4118 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4121 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4122 whether --show-vars was specified or not
4124 JJ/02 exipick: Added support for new ACL variable spool format introduced
4127 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4128 syntactically invalid From: or Reply-to: line, and a filter used this to
4129 generate an autoreply, and therefore failed to obtain an address for the
4130 autoreply, Exim could try to deliver to a non-existent relative file
4131 name, causing unrelated and misleading errors. What now happens is that
4132 it logs this as a hard delivery error, but does not attempt to create a
4135 PH/11 The exinext utility has a -C option for testing purposes, but although
4136 the given file was scanned by exinext itself; it wasn't being passed on
4137 when Exim was called.
4139 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4140 an end-of-file indication when reading a command response.
4142 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4143 compiled. In many other places in Exim, IPv6 addresses are always
4144 recognized, so I have changed this. It also means that IPv4 domain
4145 literals of the form [IPV4:n.n.n.n] are now always recognized.
4147 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4148 used if the router is not running as root, for example, when verifying at
4149 ACL time, or when using -bh. The debugging output from this situation was
4150 non-existent - all you got was a failure to exec. I have made two
4153 (a) Failures to set uid/gid, the current directory, or a process leader
4154 in a subprocess such as that created by queryprogram now generate
4155 suitable debugging output when -d is set.
4157 (b) The queryprogram router detects when it is not running as root,
4158 outputs suitable debugging information if -d is set, and then runs
4159 the subprocess without attempting to change uid/gid.
4161 PH/15 Minor change to Makefile for building test_host (undocumented testing
4164 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4165 additional section of a DNS packet that returns MX or SRV records.
4166 Instead, it always explicitly searches for A/AAAA records. This avoids
4167 major problems that occur when a DNS server includes only records of one
4168 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4169 fixed another bug: if SRV records were looked up and the corresponding
4170 address records were *not* found in the additional section, the port
4171 values from the SRV records were lost.
4173 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4174 using the correct key (the original address) when searching the retry
4175 rules in order to find which one to use for generating the retry hint.
4177 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4178 adding the default one. Similarly, if it contains a Reply-To: header, the
4179 errors_reply_to option, if set, is not used.
4181 PH/19 When calculating a retry time, Exim used to measure the "time since
4182 failure" by looking at the "first failed" field in the retry record. Now
4183 it does not use this if it is later than than the arrival time of the
4184 message. Instead it uses the arrival time. This makes for better
4185 behaviour in cases where some deliveries succeed, thus re-setting the
4186 "first failed" field. An example is a quota failure for a huge message
4187 when small messages continue to be delivered. Without this change, the
4188 "time since failure" will always be short, possible causing more frequent
4189 delivery attempts for the huge message than are intended.
4190 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4192 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4193 $1, $2, $3) because the numerical variables can be reset during some
4194 expansion items (e.g. "match"), thereby losing the authentication data.
4196 PH/21 Make -bV show the size of off_t variables so that the test suite can
4197 decide whether to run tests for quotas > 2G.
4199 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4200 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4201 filecount value is greater than 2G or if a quota value is greater than 2G
4202 on a system where the size of off_t is not greater than 4, a panic error
4205 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4206 never match. The debug and -bh output now contains an explicit error
4207 message indicating a malformed IPv4 address or mask.
4209 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4210 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4211 PH/23 above applies.
4213 PH/25 Do not write to syslog when running in the test harness. The only
4214 occasion when this arises is a failure to open the main or panic logs
4215 (for which there is an explicit test).
4217 PH/26 Added the /no_tell option to "control=freeze".
4219 PH/27 If a host name lookup failed very early in a connection, for example, if
4220 the IP address matched host_lookup and the reverse lookup yielded a name
4221 that did not have a forward lookup, an error message of the form "no IP
4222 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4223 could be logged. Now it outputs the IP address instead of "NULL".
4225 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4226 allows the sender and the authenticated sender to be set when
4227 submitting a message from within Exim. Since child_open_exim() is
4228 documented for local_scan(), the new function should be too.
4230 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4231 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4232 results in an empty string is now treated as unset.
4234 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4236 PH/31 Added sender_verify_fail logging option.
4238 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4239 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4240 tidied the source and removed it altogether.
4242 PH/33 When a queue run was abandoned because the load average was too high, a
4243 log line was always written; now it is written only if the queue_run log
4244 selector is set. In addition, the log line for abandonment now contains
4245 information about the queue run such as the pid. This is always present
4246 in "start" and "stop" lines but was omitted from the "abandon" line.
4248 PH/34 Omit spaces between a header name and the colon in the error message that
4249 is given when verify = headers_syntax fails (if there are lots of them,
4250 the message gets confusing).
4252 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4253 names, as there are now some PTR records that contain slashes. This check
4254 is only to protect against broken name servers that fall over on strange
4255 characters, so the fact that it applies to all lookups doesn't matter.
4257 PH/36 Now that the new test suite is complete, we can remove some of the
4258 special code in Exim that was needed for the old test suite. For example,
4259 sorting DNS records because real resolvers return them in an arbitrary
4260 order. The new test suite's fake resolver always returns records in the
4263 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4264 bounces) except when queue_only is set, to avoid logging races between
4265 the different processes.
4267 PH/38 Panic-die if .include specifies a non-absolute path.
4269 PH/39 A tweak to the "H" retry rule from its user.
4271 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4272 a label. They prevented compilation on older perls.
4274 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4275 a warning to be raised on newish perls.
4277 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4278 on queue. Changes to match documented behaviour of showing count of
4279 messages matching specified criteria.
4281 PH/40 Changed the default ident timeout from 30s to 5s.
4283 PH/41 Added support for the use of login_cap features, on those BSD systems
4284 that have them, for controlling the resources used by pipe deliveries.
4286 PH/42 The content-scanning code uses fopen() to create files in which to put
4287 message data. Previously it was not paying any attention to the mode of
4288 the files. Exim runs with umask(0) because the rest of the code creates
4289 files with open(), and sets the required mode explicitly. Thus, these
4290 files were ending up world-writeable. This was not a big issue, because,
4291 being within the spool directory, they were not world-accessible. I have
4292 created a function called modefopen, which takes an additional mode
4293 argument. It sets umask(777), creates the file, chmods it to the required
4294 mode, then resets the umask. All the relevant calls to fopen() in the
4295 content scanning code have been changed to use this function.
4297 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4298 to 24 hours. This avoids potential overflow problems when processing G
4299 and H retry rules. I suspect nobody ever tinkers with this value.
4301 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4303 PH/45 When the plaintext authenticator is running as a client, the server's
4304 challenges are checked to ensure they are valid base64 strings. By
4305 default, the authentication attempt is cancelled if an invalid string is
4306 received. Setting client_ignore_invalid_base64 true ignores these errors.
4307 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4308 they are received. Thus, the responses can be made to depend on the
4309 challenges. If an invalid string is ignored, an empty string is placed in
4312 PH/46 Messages that are created by the autoreply transport now contains a
4313 References: header, in accordance with RFCs 2822 and 3834.
4315 PH/47 Added authenticated_sender_force to the smtp transport.
4317 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4319 PH/49 Installed latest patch from the Sieve maintainer.
4321 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4322 was also set, the appendfile transport was unnecessarily scanning a
4323 directory of message files (e.g. for maildir delivery) to find the count
4324 of files (along with the size), even though it did not need this
4325 information. It now does the scan only if it needs to find either the
4326 size of the count of files.
4328 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4330 PH/52 Two bugs concerned with error handling when the smtp transport is
4333 (i) Exim was not creating retry information for temporary errors given
4334 for individual recipients after the DATA command when the smtp transport
4335 was used in LMTP mode. This meant that they could be retried too
4336 frequently, and not timed out correctly.
4338 (ii) Exim was setting the flag that allows error details to be returned
4339 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4340 recipients that were returned after the DATA command.
4342 PH/53 This is related to PH/52, but is more general: for any failing address,
4343 when detailed error information was permitted to be returned to the
4344 sender, but the error was temporary, then after the final timeout, only
4345 "retry timeout exceeded" was returned. Now it returns the full error as
4346 well as "retry timeout exceeded".
4348 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4349 do this, and (what is worse) MTAs that accept it.
4351 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4352 will now be deprecated.
4354 PH/56 New os.c-cygwin from the Cygwin maintainer.
4356 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4357 formats (previously only available in exim formats via -bpr, -bpru,
4358 and -bpra. Now also available in native and exiqgrep formats)
4360 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4361 with very large, slow to parse queues
4363 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4365 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4367 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4368 responses to authentication challenges, though it was showing the
4369 challenges; (ii) I've removed the CR characters from the debug output for
4372 PH/58 Allow for the insertion of a newline as well as a space when a string
4373 is turned into more than one encoded-word during RFC 2047 encoding. The
4374 Sieve code now uses this.
4376 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4377 data_4xx, lost_connection, tls_required.
4379 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4380 message was being sent as an SMTP response.
4382 PH/61 Add -l and -k options to exicyclog.
4384 PH/62 When verifying, if an address was redirected to one new address, so that
4385 verification continued, and the new address failed or deferred after
4386 having set something in $address_data, the value of $address_data was not
4387 passed back to the ACL. This was different to the case when no
4388 redirection occurred. The value is now passed back in both cases.
4390 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4391 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4392 use login_cap.h, so on its own it isn't the distinguishing feature. The
4393 new name refers directly to the setclassresources() function.
4395 PH/65 Added configuration files for NetBSD3.
4397 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4399 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4400 is preferred over IPv4.
4402 PH/68 The bounce_return_message and bounce_return_body options were not being
4403 honoured for bounces generated during the reception of non-SMTP messages.
4404 In particular, this applied to messages rejected by the ACL. This bug has
4405 been fixed. However, if bounce_return_message is true and bounce_return_
4406 body is false, the headers that are returned for a non-SMTP message
4407 include only those that have been read before the error was detected.
4408 (In the case of an ACL rejection, they have all been read.)
4410 PH/69 The HTML version of the specification is now built in a directory called
4411 spec_html instead of spec.html, because the latter looks like a path with
4412 a MIME-type, and this confuses some software.
4414 PH/70 Catch two compiler warnings in sieve.c.
4416 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4417 function verify_get_ident() calls ip_connect() to connect a socket, but
4418 if the "connect()" function timed out, ip_connect() used to close the
4419 socket. However, verify_get_ident() also closes the socket later, and in
4420 between Exim writes to the log, which may get opened at this point. When
4421 the socket was closed in ip_connect(), the log could get the same file
4422 descriptor number as the socket. This naturally causes chaos. The fix is
4423 not to close the socket in ip_connect(); the socket should be closed by
4424 the function that creates it. There was only one place in the code where
4425 this was missing, in the iplookup router, which I don't think anybody now
4426 uses, but I've fixed it anyway.
4428 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4429 well as to direct DNS lookups. Otherwise the handling of names in host
4430 lists is inconsistent and therefore confusing.
4436 PH/01 Two changes to the default runtime configuration:
4438 (1) Move the checks for relay_from_hosts and authenticated clients from
4439 after to before the (commented out) DNS black list checks.
4441 (2) Add control=submission to the relay_from_hosts and authenticated
4442 clients checks, on the grounds that messages accepted by these
4443 statements are most likely to be submissions.
4445 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4447 (1) Generate an error if the third argument for the ${prvs expansion is
4450 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4453 (3) Reset the variables that are obtained from the first argument of
4454 ${prvscheck and used in the second argument before leaving the code,
4455 because their memory is reclaimed, so using them afterwards may do
4458 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4459 one (it's much easier than Tom thought :-).
4461 (5) Because of (4), we can now allow for the use of $prvscheck_result
4462 inside the third argument.
4464 PH/03 For some reason, the default setting of PATH when running a command from
4465 a pipe transport was just "/usr/bin". I have changed it to
4468 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4469 anything to be listed in the output from -bV.
4471 PH/05 When a filter generated an autoreply, the entire To: header line was
4472 quoted in the delivery log line, like this:
4474 => >A.N.Other <ano@some.domain> <original@ddress> ...
4476 This has been changed so that it extracts the operative address. There
4477 may be more than one such address. If so, they are comma-separated, like
4480 => >ano@some.domain,ona@other.domain <original@ddress> ...
4482 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4483 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4484 not being looked up in the rDNS to get a host name, Exim was showing the
4485 IP address twice in Received: lines, even though the IP addresses were
4486 identical. For example:
4488 Received: from [1.2.3.4] (helo=[1.2.3.4])
4490 However, if the real host name was known, it was omitting the HELO data
4491 if it matched the actual IP address. This has been tidied up so that it
4492 doesn't show the same IP address twice.
4494 PH/07 When both +timestamp and +memory debugging was on, the value given by
4495 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4496 called by the debug printing, thereby overwriting the timestamp buffer.
4497 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4499 PH/08 When the original message was included in an autoreply transport, it
4500 always said "this is a copy of the message, including all the headers",
4501 even if body_only or headers_only was set. It now gives an appropriate
4504 PH/09 Applied a patch from the Sieve maintainer which:
4506 o fixes some comments
4507 o adds the (disabled) notify extension core
4508 o adds some debug output for the result of if/elsif tests
4509 o points to the current vacation draft in the documentation
4510 and documents the missing references header update
4514 o fixes a bug in processing the envelope test (when testing
4515 multiple envelope elements, the last element determined the
4518 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4519 Electronic Mail") by including:
4521 Auto-submitted: auto-generated
4523 in the messages that it generates (bounce messages and others, such as
4524 warnings). In the case of bounce messages for non-SMTP messages, there was
4525 also a typo: it was using "Auto_submitted" (underscore instead of
4526 hyphen). Since every message generated by Exim is necessarily in response
4527 to another message, thes have all been changed to:
4529 Auto-Submitted: auto-replied
4531 in accordance with these statements in the RFC:
4533 The auto-replied keyword:
4535 - SHOULD be used on messages sent in direct response to another
4536 message by an automatic process,
4538 - MUST NOT be used on manually-generated messages,
4540 - MAY be used on Delivery Status Notifications (DSNs) and Message
4541 Disposition Notifications (MDNs),
4543 - MUST NOT be used on messages generated by automatic or periodic
4544 processes, except for messages which are automatic responses to
4547 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4548 to the default Received: header definition.
4550 PH/12 Added log selector acl_warn_skipped (default on).
4552 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4553 variables because (a) they are in the wrong storage pool and (b) even if
4554 they were copied, it wouldn't work properly because of the caching.
4556 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4557 checking when decoding. Apparently there are clients that generate
4558 overlong encoded strings. Why am I not surprised?
4560 PH/15 If the first argument of "${if match_address" was not empty, but did not
4561 contain an "@" character, Exim crashed. Now it writes a panic log message
4562 and treats the condition as false.
4564 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4566 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4567 extension "envelope-auth". The code is finished and in agreement with
4568 other implementations, but there is no documentation so far and in fact,
4569 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4570 not changing the active code.
4572 Print executed "if" and "elsif" statements when debugging is used. This
4573 helps a great deal to understand what a filter does.
4575 Document more things not specified clearly in RFC3028. I had all this
4576 sorted out, when out of a sudden new issues came to my mind. Oops."
4578 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4581 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4582 canonical form (as documented). However, after a host name lookup from
4583 the IP address, check_host() was doing a simple string comparison with
4584 addresses acquired from the DNS when checking that the found name did
4585 have the original IP as one of its addresses. Since any found IPv6
4586 addresses are likely to be in abbreviated form, the comparison could
4587 fail. Luckily, there already exists a function for doing the comparison
4588 by converting both addresses to binary, so now that is used instead of
4589 the text comparison.
4591 PH/20 There was another similar case to PH/19, when a complete host name was
4592 given in a host list; looking up its IP address could give an abbreviated
4593 form, whereas the current host's name might or might not be abbreviated.
4594 The same fix has been applied.
4600 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4601 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4604 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4605 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4607 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4609 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4610 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4611 Also, for "Auto-Submitted", treat anything other than "no" as
4612 non-personal, in accordance with RFC 3834. (Previously it treated
4613 anything starting "auto-" as non-personal.)
4615 TF/01 The control=submission/name=... option had a problem with syntax
4616 errors if the name included a slash character. The /name= option
4617 now slurps the rest of the string, so it can include any characters
4618 but it must come last in the list of options (after /sender_retain
4621 PH/05 Some modifications to the interface to the fake nameserver for the new
4629 TK/01 Added the "success_on_redirect" address verification option. See
4630 NewStuff for rationale and an example.
4632 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4634 PH/02 Patch to exigrep to allow it to work on syslog lines.
4636 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4637 fread() to skip over the body file's header line, because in Cygwin the
4638 header line is locked and is inaccessible.
4640 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4641 co-exist for some time) to make it clear that it is the Exim ID that is
4642 referenced, not the Message-ID: header line.
4644 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4645 string_format() function, because snprintf() does not exist on all
4648 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4649 ${stat: expansion item.
4651 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4652 protocol synchronization error", to keep the pedants happy.
4654 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4655 well as for IRIX systems, when gcc is being used. See the host.c source
4658 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4660 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4663 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4664 transport and to the smtp transport in LMTP mode.
4666 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4668 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4669 run a filter in a subprocess. This could lead to confusion in subsequent
4670 lookups in the parent process. There should also be a search_tidyup() at
4671 the end of the subprocess.
4673 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4674 only if the host matched helo_try_verify_hosts, which caused the
4675 verification to occur when the EHLO/HELO command was issued. The ACL just
4676 tested the remembered result. Now, if a previous verification attempt has
4677 not happened, "verify = helo" does it there and then.
4679 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4681 TK/03 Fix log output including CR from clamd.
4683 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4684 exist provoked a memory error which could cause a segfault.
4686 PH/15 Installed PCRE 6.2
4688 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4690 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4691 of the problem. Specifically, suggested +O2 rather than +O1 for the
4694 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4696 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4697 record, the port was not correctly logged when the outgoing_port log
4698 selector was set (it logged the transort's default port).
4700 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4701 fallback_hosts, and "hosts" in the smtp transport.
4703 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4704 host errors such as "Connection refused".
4706 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4707 authentication with radiusclient 0.4.9:
4709 - Error returned from rc_read_config was caught wrongly
4710 - Username/password not passed on to radius server due to wrong length.
4712 The presumption is that some radiusclient API changes for 4.51/PH/17
4713 were not taken care of correctly. The code is still untested by me (my
4714 Linux distribution still has 0.3.2 of radiusclient), but it was
4715 contributed by a Radius user.
4717 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4718 expanding the "port" option of the smtp transport.
4720 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4721 while reading a MIME header. Thanks to Tom Hughes for a patch.
4723 PH/24 Include config.h inside local_scan.h so that configuration settings are
4726 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4727 that in an ACL for RCPT (for example), you can examine exactly what was
4730 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4731 commands, but it was not correctly comparing the address with the actual
4732 client host address. Thus, it would show the EHLO address in Received:
4733 header lines when this was not necessary.
4735 PH/27 Added the % operator to ${eval:}.
4737 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4738 it should be ignoring failures (because with -C, for example, it has lost
4739 privilege). It wasn't ignoring creation failures other than "already
4742 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4745 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4746 because an input line was too long, either on its own, or by virtue of
4747 too many continuations, the temporary file was not being removed, and the
4748 return code was incorrect.
4750 PH/31 Missing "BOOL" in function definition in filtertest.c.
4752 PH/32 Applied Sieve patches from the maintainer.
4754 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4756 PH/33 Added "verify = not_blind".
4758 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4759 Local/Makefile (with some defaults set). These are used in built scripts
4760 such as exicyclog, but they have never been used in the exim_install
4761 script (though there are many overriding facilities there). I have
4762 arranged that the exim_install script now takes note of these two
4765 PH/35 Installed configuration files for Dragonfly.
4767 PH/36 When a locally submitted message by a trusted user did not contain a
4768 From: header, and the sender address was obtained from -f or from an SMTP
4769 MAIL command, and the trusted user did not use -F to supply a sender
4770 name, $originator_name was incorrectly used when constructing a From:
4771 header. Furthermore, $originator_name was used for submission mode
4772 messages from external hosts without From: headers in a similar way,
4773 which is clearly wrong.
4775 PH/37 Added control=suppress_local_fixups.
4777 PH/38 When log_selector = +received_sender was set, and the addition of the
4778 sender made the log line's construction buffer exactly full, or one byte
4779 less than full, an overflow happened when the terminating "\n" was
4782 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4783 when the result of a list match is failure because a DNS lookup failed.
4785 PH/40 RM_COMMAND is now used in the building process.
4787 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4788 the "build-* directories that it finds.
4790 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4791 address in a domain literal was a prefix of an interface address.
4793 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4794 when verifying a sender address, unless rewrite_headers is false.
4796 PH/44 Wrote a long comment about why errors_to addresses are verified as
4797 recipients, not senders.
4799 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4800 the ratelimit ACL was added.
4802 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4804 PH/46 Added extra information about PostgreSQL errors to the error string.
4806 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4807 avoiding the need to install special zones in a real server. This is
4808 backwards compatible; if it can't find the fake resolver, it drops back.
4809 Thus, both old and new test suites can be run.
4811 TF/02 Added util/ratelimit.pl
4813 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4814 clock is set back in time.
4816 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4817 Candler <B.Candler@pobox.com>.
4819 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4820 OK for addresses that are the result of redirections.
4822 PH/48 A number of further additions for the benefit of the new test suite,
4823 including a fake gethostbyname() that interfaces to the fake DNS resolver
4826 TF/06 The fix for widen_domains has also been applied to qualify_single and
4827 search_parents which are the other dnslookup options that can cause
4830 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4833 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4835 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4836 service" if it encountered an MX record with an empty target hostname.
4837 The message is now "an MX or SRV record indicated no SMTP service".
4839 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4840 if the DNS of the sending site is misconfigured. This is quite a
4841 common situation. This change restores the behaviour of treating a
4842 helo verification defer as a failure.
4844 PH/51 If self=fail was set on a router, the bounce message did not include the
4845 actual error message.
4851 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4853 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4854 command itself ended in error, the underlying message about the transport
4855 filter timeout was being overwritten with the pipe command error. Now the
4856 underlying error message should be appended to the second error message.
4858 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4860 PH/02 Reduce the amount of output that "make" produces by default. Full output
4861 can still be requested.
4863 PH/03 The warning log line about a condition test deferring for a "warn" verb
4864 was being output only once per connection, rather than after each
4865 occurrence (because it was using the same function as for successful
4866 "warn" verbs). This seems wrong, so I have changed it.
4868 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4869 it should not have, which might have caused a crash in the right
4870 circumstances, but probably never did.
4872 PH/04 Installed a modified version of Tony Finch's patch to make submission
4873 mode fix the return path as well as the Sender: header line, and to
4874 add a /name= option so that you can make the user's friendly name appear
4877 TF/03 Added the control = fakedefer ACL modifier.
4879 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4880 Mark Lowes for thorough testing.
4882 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4884 TK/03 Merged latest SRS patch from Miles Wilton.
4886 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4887 with the definition in sysexits.h (which is #included earlier).
4888 Fortunately, Exim does not actually use EX_OK. The code used to try to
4889 preserve the sysexits.h value, by assuming that macro definitions were
4890 scanned for macro replacements. I have been disabused of this notion,
4891 so now the code just undefines EX_OK before #including unistd.h.
4893 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4894 in the smtp transport. When a block could not be written in a single
4895 write() function, the timeout was being re-applied to each part-write.
4896 This seems wrong - if the receiver was accepting one byte at a time it
4897 would take for ever. The timeout is now adjusted when this happens. It
4898 doesn't have to be particularly precise.
4900 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4901 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4903 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4904 without a domain if the check to <postmaster@domain> fails.
4906 SC/01 Eximstats: added -xls and the ability to specify output files
4907 (patch written by Frank Heydlauf).
4909 SC/02 Eximstats: use FileHandles for outputting results.
4911 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4913 SC/04 Eximstats: fixed display of large numbers with -nvr option
4915 SC/05 Eximstats: fixed merging of reports with empty tables.
4917 SC/06 Eximstats: added the -include_original_destination flag
4919 SC/07 Eximstats: removed tabs and trailing whitespace.
4921 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4923 TK/06 MBOX spool code: Add real "From " MBOX separator line
4924 so the .eml file is really in mbox format (even though
4925 most programs do not really care). Patch from Alex Miller.
4927 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4928 The latter is generated from $received_to and is only set if the
4929 message has one envelope recipient. SA can use these headers,
4930 obviously out-of-the-box. Patch from Alex Miller.
4932 PH/08 The ${def test on a variable was returning false if the variable's
4933 value was "0", contrary to what the specification has always said!
4934 The result should be true unless the variable is empty.
4936 PH/09 The syntax error of a character other than { following "${if
4937 def:variable_name" (after optional whitespace) was not being diagnosed.
4938 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4939 accidental colon was present, for example, could give incorrect results.
4941 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4942 result is used (not including appendfile, where other changes are about
4945 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4946 This involved changing a lot of size variables from int to off_t. It
4947 should work with maildirs and everything.
4949 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4950 spamd dying while we are connected to it.
4952 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4955 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4956 function for BDB 4.3.
4958 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4960 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4961 the sources for PCRE have been re-arranged and more files are now
4964 PH/15 The code I had for printing potentially long long variables in PH/11
4965 above was not the best (it lost precision). The length of off_t variables
4966 is now inspected at build time, and an appropriate printing format (%ld
4967 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4968 to be "long long int" or "long int". This is needed for the internal
4969 formatting function string_vformat().
4971 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4972 the configuration file to be ":syslog", then the script "guesses" where
4973 the logs files are, rather than using the compiled in default. In our
4974 case the guess is not the same as the compiled default, so the script
4975 suddenly stopped working when I started to use syslog. The patch checks
4976 to see if log_file_path is "". If so, it attempts to read it from exim
4977 with no configuration file to get the compiled in version, before it
4978 falls back to the previous guessing code."
4980 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4981 implementing BATV in an Exim configuration. See NewStuff for the gory
4984 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4985 Makefile that are specific to HP-UX.
4987 PH/18 If the "use_postmaster" option was set for a recipient callout together
4988 with the "random" option, the postmaster address was used as the MAIL
4989 FROM address for the random test, but not for the subsequent recipient
4990 test. It is now used for both.
4992 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4993 patch removes a few documentation additions to RFC 3028, because the
4994 latest draft now contains them. It adds the new en;ascii-case comparator
4995 and a new error check for 8bit text in MIME parts. Comparator and
4996 require names are now matched exactly. I enabled the subaddress
4997 extension, but it is not well tested yet (read: it works for me)."
4999 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5000 rework some of the code of TK/09 above to avoid the hardwired use of
5001 "%lld" and "long long". Replaced the call to snprintf() with a call to
5004 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5005 records point to non-existent hosts", "retry timeout exceeded", and
5006 "retry time not reached for any host after a long failure period".
5008 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5009 experimental DomainKeys support:
5011 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5012 (2) On an error such as an illegally used "control", the wrong name for
5013 the control was given.
5015 These problems did NOT occur unless DomainKeys support was compiled.
5017 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5019 PH/24 Added ${if match_ip condition.
5021 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5022 databases so that it will be absolutely obvious if a crash occurs in the
5023 DB library. This is a regular occurrence (often caused by mis-matched
5026 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5027 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5028 picked up on a user's system that detects such things. There doesn't seem
5029 to be a gcc warning option for this - only an attribute that has to be
5030 put on the function's prototype. It seems that in Fedora Core 4 they have
5031 set this on a number of new functions. No doubt there will be more in due
5034 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5035 specify a transport. However, if an address that was verified by such a
5036 router was the subject of a callout, Exim crashed because it tried to
5037 read the rcpt_include_affixes from the non-existent transport. Now it
5038 just assumes that the setting of that option is false. This bug was
5039 introduced by 4.51/PH/31.
5041 PH/28 Changed -d+all to exclude +memory, because that information is very
5042 rarely of interest, but it makes the output a lot bigger. People tend to
5043 do -d+all out of habit.
5045 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5046 code in os-type was giving problems when libc.so lives in lib64, like on
5049 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5050 aren't the modern standard, and it seems that some systems' include files
5051 don't always have them. Exim was already checking for some of the newer
5052 ones like T_AAAA, and defining it itself. I've added checks for all the
5053 record types that Exim uses.
5055 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5056 not automatically generating a new one, as it is supposed to. This
5057 prevented TLS from working. If the file did exist, but contained invalid
5058 data, a new version was generated, as expected. It was only the case of a
5059 non-existent file that was broken.
5061 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5062 with a change in libdomainkeys > 0.64.
5064 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5065 from DNS. If the selector record carries the flag, it now has
5066 precedence over the domain-wide flag.
5068 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5070 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5071 the use of a port name that isn't defined in /etc/services) occurred, the
5072 message was deferred as in a normal delivery, and thus remained on the
5073 spool, instead of being failed because of the mua_wrapper setting. This
5074 is now fixed, and I tidied up some of the mua_wrapper messages at the
5077 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5078 the messages in a hash of arrays rather than using individual hashes.
5079 This is a bit cleaner and results in dramatic memory savings, albeit
5080 at a slight CPU cost.
5082 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5083 as requested by Marc Sherman.
5085 SC/10 Eximstats: added histograms for user specified patterns as requested
5088 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5090 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5091 fopen() in the content-scanning modules that did not already have it.
5097 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5098 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5100 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5102 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5104 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5105 to test Sieve filters that use "vacation".
5107 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5108 that changes the way the GnuTLS parameters are stored in the cache file.
5109 The new format can be generated externally. For backward compatibility,
5110 if the data in the cache doesn't make sense, Exim assumes it has read an
5111 old-format file, and it generates new data and writes a new file. This
5112 means that you can't go back to an older release without removing the
5115 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5116 work if there are any delivery delays because "one_time" forces the
5117 parent to be marked "delivered", so its unseen clone is never tried
5118 again. For this reason, Exim now forbids the simultaneous setting of
5121 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5122 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5123 entry if you want to know the details. The fix, however, neglected to
5124 consider the case when local delivery batching is involved. The test for
5125 "previously delivered" was not happening when checking to see if an
5126 address could be batched with a previous (undelivered) one; under
5127 certain circumstances this could lead to multiple deliveries to the same
5130 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5131 in its include files, and this causes problems building Exim.
5133 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5134 header_syntax) but Exim was just ignoring anything given after a slash.
5135 In particular, this caused confusion with an attempt to use "verify =
5136 reverse_host_lookup/defer_ok". An error is now given when options are
5137 supplied for verify items that do not have them. (Maybe reverse_host_
5138 lookup should have a defer_ok option, but that's a different point.)
5140 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5141 defined by RFC 821) to 2048, because there were problems with some AUTH
5142 commands, and RFC 1869 says the size should be increased for extended
5143 SMTP commands that take arguments.
5145 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5148 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5149 "unknown" error; now it says that the functionality isn't in the binary.
5151 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5152 an address' error message when a string expansion fails (syntax or
5153 whatever). Otherwise the password may appear in the log. Following change
5154 PH/42 below, there is no longer a chance of it appearing in a bounce
5157 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5159 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5160 of its addresses was ever tried. (Bugzilla bug #2.)
5162 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5163 the result incorrectly in the debug output. (It correctly added a newline
5164 to what was transported.)
5166 TF/01 Added $received_time.
5168 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5169 commented out examples of how to interface to a virus scanner and to
5170 SpamAssassin. Also added commented examples of av_scanner and
5171 spamd_address settings.
5173 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5174 and controls are allowed in which ACLs. There were a couple of minor
5175 errors. Some of the entries in the conditions table (which is a table of
5176 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5177 negation of where the condition IS allowed.
5179 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5181 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5182 header file does not have a version number, so I've had to invent a new
5183 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5184 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5185 radiusclient), but it was contributed by a Radius user.
5187 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5188 files or directories, for OpenSSL.
5190 PH/19 When an Exim process that is running as root has to create an Exim log
5191 file, it does so in a subprocess that runs as exim:exim so as to get the
5192 ownership right at creation (otherwise, other Exim processes might see
5193 the file with the wrong ownership). There was no test for failure of this
5194 fork() call, which would lead to the process getting stuck as it waited
5195 for a non-existent subprocess. Forks do occasionally fail when resources
5196 run out. I reviewed all the other calls to fork(); they all seem to check
5199 PH/20 When checking for unexpected SMTP input at connect time (before writing
5200 the banner), Exim was not dealing correctly with a non-positive return
5201 from the read() function. If the client had disconnected by this time,
5202 the result was a log entry for a synchronization error with an empty
5203 string after "input=" when read() returned zero. If read() returned -1
5204 (an event I could not check), uninitialized data bytes were printed.
5205 There were reports of junk text (parts of files, etc) appearing after
5208 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5210 PH/22 Added support for macro redefinition, and (re)definition in between
5211 driver and ACL definitions.
5213 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5214 forgetting to use the resulting value; it was using the unexpanded value.
5216 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5217 hadn't been configured. The fix is from Juergen Kreileder, who
5218 understands it better than I do:
5220 "Here's what I see happening with three configured cyrus_sasl
5221 authenticators configured (plain, login, cram-md5):
5223 On startup auth_cyrus_sasl_init() gets called for each of these.
5224 This means three calls to sasl_listmech() without a specified mech_list.
5225 => SASL tests which mechs of all available mechs actually work
5226 => three warnings about OTP not working
5227 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5229 With the patch, sasl_listmech() also gets called three times. But now
5230 SASL's mech_list option is set to the server_mech specified in the the
5231 authenticator. Or in other words, the answer from sasl_listmech()
5232 gets limited to just the mech you're testing for (which is different
5234 => the return list contains just 'plain' or 'login', 'cram-md5' or
5235 nothing depending on the value of ob->server_mech.
5237 I've just tested the patch: Authentication still works fine,
5238 unavailable mechs specified in the exim configuration are still
5239 caught, and the auth.log warnings about OTP are gone."
5241 PH/25 When debugging is enabled, the contents of the command line are added
5242 to the debugging output, even when log_selector=+arguments is not
5245 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5246 answer is "GNU", and only if the return is "GNU/something" is the answer
5249 PH/27 $acl_verify_message is now set immediately after the failure of a
5250 verification in an ACL, and so is available in subsequent modifiers. In
5251 particular, the message can be preserved by coding like this:
5253 warn !verify = sender
5254 set acl_m0 = $acl_verify_message
5256 Previously, $acl_verify_message was set only while expanding "message"
5257 and "log_message" when a very denied access.
5259 PH/28 Modified OS/os.c-Linux with
5261 -#ifndef OS_LOAD_AVERAGE
5262 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5264 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5265 nomenclature these days.)
5267 PH/29 Installed patch from the Sieve maintainer that adds the options
5268 sieve_useraddress and sieve_subaddress to the redirect router.
5270 PH/30 In these circumstances:
5271 . Two addresses routed to the same list of hosts;
5272 . First host does not offer TLS;
5273 . First host accepts first address;
5274 . First host gives temporary error to second address;
5275 . Second host offers TLS and a TLS session is established;
5276 . Second host accepts second address.
5277 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5278 and peerdn, if requested) that were in fact used only for the second
5281 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5282 attention to any local part prefix or suffix that was matched by the
5283 router that accepted the address. It now behaves in the same way as it
5284 does for delivery: the affixes are removed from the local part unless
5285 rcpt_include_affixes is set on the transport.
5287 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5288 timeout during the DATA phase of an incoming message.
5290 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5291 applied a patch sanctioned by the Sieve maintainer.
5293 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5294 the uid or gid is negative. A case of a negative gid caused this to be
5295 noticed. The fix allows for either to be negative.
5297 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5298 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5301 PH/36 The change PH/12 above was broken. Fixed it.
5303 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5304 the grounds that routing the same address twice would always produce the
5305 same answer. This might have been true once, but it is certainly no
5306 longer true now. Routing a child address may depend on the previous
5307 routing that produced that child. Some complicated redirection strategies
5308 went wrong when messages had multiple recipients, and made Exim's
5309 behaviour dependent on the order in which the addresses were given.
5311 I have moved the duplicate checking until after the routing is complete.
5312 Exim scans the addresses that are assigned to local and remote
5313 transports, and removes any duplicates. This means that more work will be
5314 done, as duplicates will always all be routed, but duplicates are
5315 presumably rare, so I don't expect this is of any significance.
5317 For deliveries to pipes, files, and autoreplies, the duplicate checking
5318 still happens during the routing process, since they are not going to be
5321 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5322 It corrects a timeout issue with spamd. This is Ian's comment: "The
5323 background is that sometimes spamd either never reads data from a
5324 connection it has accepted, or it never writes response data. The exiscan
5325 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5326 blindly assumes that writes won't block so it may never time out."
5328 PH/39 Allow G after quota size as well as K and M.
5330 PH/40 The value set for $authenticated_id in an authenticator may not contain
5331 binary zeroes or newlines because the value is written to log lines and
5332 to spool files. There was no check on this. Now the value is run through
5333 the string_printing() function so that such characters are converted to
5334 printable escape sequences.
5336 PH/41 $message_linecount is a new variable that contains the total number of
5337 lines in the message. Compare $body_linecount, which is the count for the
5340 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5341 bounce and delay warning messages, except in certain special cases, which
5344 (a) An SMTP error message from a remote host;
5345 (b) A message specified in a :fail: redirection;
5346 (c) A message specified in a "fail" command in a system filter;
5347 (d) A message specified in a FAIL return from the queryprogram router;
5348 (e) A message specified by the cannot_route_message router option.
5350 In these cases only, Exim does include the error details in bounce and
5351 warning messages. There are also a few cases where bland messages such
5352 as "unrouteable address" or "local delivery error" are given.
5354 PH/43 $value is now also set for the "else" part of a ${run expansion.
5356 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5357 being worked on, but at least Exim now implements the latest version to
5360 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5361 process to complete was treated as a delivery failure, a timeout while
5362 writing the message to the pipe was logged, but erroneously treated as a
5363 successful delivery. Such timeouts include transport filter timeouts. For
5364 consistency with the overall process timeout, these timeouts are now
5365 treated as errors, giving rise to delivery failures by default. However,
5366 there is now a new Boolean option for the pipe transport called
5367 timeout_defer, which, if set TRUE, converts the failures into defers for
5368 both kinds of timeout. A transport filter timeout is now identified in
5371 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5372 systems where "make" and "gmake" are different, calling "gmake" at top
5373 level broke things. I've arranged for the value of $(MAKE) to be passed
5374 from the Makefile to this script so that it can call the same version of
5378 A note about Exim versions 4.44 and 4.50
5379 ----------------------------------------
5381 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5382 changes of various kinds. As a consequence, a big documentation update was
5383 needed. This delayed the release for rather longer than seemed good, especially
5384 in the light of a couple of (minor) security issues. Therefore, the changes
5385 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5386 release. So 4.44 and 4.50 are in effect two different branches that both start
5389 I have left the 4.50 change log unchanged; it contains all the changes since
5390 4.43. The change log for 4.44 is below; many of its items are identical to
5391 those for 4.50. This seems to be the most sensible way to preserve the
5392 historical information.
5398 1. Minor wording change to the doc/README.SIEVE file.
5400 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5401 computation of the current number of files was incorrect.
5403 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5404 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5407 4. Give more explanation in the error message when the command for a transport
5408 filter fails to execute.
5410 5. There are several places where Exim runs a non-Exim command in a
5411 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5412 was being done only for the command run by the queryprogram router. It is
5413 now done for all such subprocesses. The other cases are: ${run, transport
5414 filters, and the commands run by the lmtp and pipe transports.
5416 6. Added CONFIGURE_GROUP build-time option.
5418 7. Some older OS have a limit of 256 on the maximum number of file
5419 descriptors. Exim was using setrlimit() to set 1000 as a large value
5420 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5421 systems. I've change it so that if it can't get 1000, it tries for 256.
5423 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5424 was an oversight, and furthermore, ever since the addition of extra
5425 controls (e.g. 4.43/32), the checks on when to allow different forms of
5426 "control" were broken. There should now be diagnostics for all cases when a
5427 control that does not make sense is encountered.
5429 9. Added the /retain_sender option to "control=submission".
5431 10. $recipients is now available in the predata ACL (oversight).
5433 11. Tidy the search cache before the fork to do a delivery from a message
5434 received from the command line. Otherwise the child will trigger a lookup
5435 failure and thereby defer the delivery if it tries to use (for example) a
5436 cached ldap connection that the parent has called unbind on.
5438 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5439 of $address_data from the recipient verification was clobbered by the
5440 sender verification.
5442 13. The value of address_data from a sender verification is now available in
5443 $sender_address_data in subsequent conditions in the ACL statement.
5445 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5447 15. Added a new option "connect=<time>" to callout options, to set a different
5450 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5451 was its contents. (It was OK if the option was not defined at all.)
5453 17. A "Completed" log line is now written for messages that are removed from
5454 the spool by the -Mrm option.
5456 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5457 information about exactly what failed.
5459 19. Added -dd to debug only the daemon process.
5461 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5462 handles timeouts, both on the server side and network timeouts. Renamed the
5463 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5465 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5466 It is now set to "smtps".
5468 22. $host_address is now set to the target address during the checking of
5469 ignore_target_hosts.
5471 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5472 being passed; this would have caused $sender_host_name to have been used if
5473 matching the list had actually called for a host name (not very likely,
5474 since this list is usually IP addresses). A host name is now passed as
5477 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5478 code that specifies a non-restarting handler (typically sigaction() in
5479 modern systems) in an attempt to fix a rare and obscure crash bug.
5481 25. Narrowed the window for a race in the daemon that could cause it to ignore
5482 SIGCHLD signals. This is not a major problem, because they are used only to
5483 wake it up if nothing else does.
5485 26. A malformed maildirsize file could cause Exim to calculate negative values
5486 for the mailbox size or file count. Odd effects could occur as a result.
5487 The maildirsize information is now recalculated if the size or filecount
5490 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5491 support for a long time. Removed HAVE_SYS_VFS_H.
5493 28. Installed the latest version of exipick from John Jetmore.
5495 29. In an address list, if the pattern was not a regular expression, an empty
5496 subject address (from a bounce message) matched only if the pattern was an
5497 empty string. Non-empty patterns were not even tested. This was the wrong
5498 because it is perfectly reasonable to use an empty address as part of a
5499 database query. An empty address is now tested by patterns that are
5500 lookups. However, all the other forms of pattern expect the subject to
5501 contain a local part and a domain, and therefore, for them, an empty
5502 address still always fails if the pattern is not itself empty.
5504 30. Exim went into a mad DNS loop when attempting to do a callout where the
5505 host was specified on an smtp transport, and looking it up yielded more
5506 than one IP address.
5508 31. Re-factored the code for checking spool and log partition space into a
5509 function that finds that data and another that does the check. The former
5510 is then used to implement four new variables: $spool_space, $log_space,
5511 $spool_inodes, and $log_inodes.
5513 32. The RFC2047 encoding function was originally intended for short strings
5514 such as real names; it was not keeping to the 75-character limit for
5515 encoded words that the RFC imposes. It now respects the limit, and
5516 generates multiple encoded words if necessary. To be on the safe side, I
5517 have increased the buffer size for the ${rfc2047: expansion operator from
5520 33. It is now permitted to omit both strings after an "if" condition; if the
5521 condition is true, the result is "true". As before, when the second string
5522 is omitted, a false condition yields an empty string. This makes it less
5523 cumbersome to write custom ACL and router conditions.
5525 34. Failure to deliver a bounce message always caused it to be frozen, even if
5526 there was an errors_to setting on the router. The errors_to setting is now
5529 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5530 canonical form (fully expanded) before being placed in
5531 $sender_host_address.
5533 36. The table in the code that translates DNS record types into text (T_A to
5534 "A" for instance) was missing entries for NS and CNAME. It is just possible
5535 that this could have caused confusion if both these types were looked up
5536 for the same domain, because the text type is used as part of Exim's
5537 per-process caching. But the chance of anyone hitting this buglet seems
5540 37. The dnsdb lookup has been extended in a number of ways.
5542 (1) There is a new type, "zns", which walks up the domain tree until it
5543 finds some nameserver records. It should be used with care.
5545 (2) There is a new type, "mxh", which is like "mx" except that it returns
5546 just the host names, not the priorities.
5548 (3) It is now possible to give a list of domains (or IP addresses) to be
5549 looked up. The behaviour when one of the lookups defers can be
5550 controlled by a keyword.
5552 (4) It is now possible to specify the separator character for use when
5553 multiple records are returned.
5555 38. The dnslists ACL condition has been extended: it is now possible to supply
5556 a list of IP addresses and/or domains to be looked up in a particular DNS
5559 39. Added log_selector=+queue_time_overall.
5561 40. When running the queue in the test harness, wait just a tad after forking a
5562 delivery process, to get repeatability of debugging output.
5564 41. Include certificate and key file names in error message when GnuTLS fails
5565 to set them up, because the GnuTLS error message doesn't include the name
5566 of the failing file when there is a problem reading it.
5568 42. Allow both -bf and -bF in the same test run.
5570 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5572 44. The "Exiscan patch" is now merged into the mainline Exim source.
5574 45. Sometimes the final signoff response after QUIT could fail to get
5575 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5576 < 0 before doing a fflush(). This bug looks as though it goes back to the
5577 introduction of TLS in release 3.20, but "sometimes" must have been rare
5578 because the tests only now provoked it.
5580 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5581 (this can affect the format of dates).
5583 47. exim_tidydb, when checking for the continued existence of a message for
5584 which it has found a message-specific retry record, was not finding
5585 messages that were in split spool directories. Consequently, it was
5586 deleting retry records that should have stayed in existence.
5588 48. Steve fixed some bugs in eximstats.
5590 49. The SPA authentication driver was not abandoning authentication and moving
5591 on to the next authenticator when an expansion was forced to fail,
5592 contradicting the general specification for all authenticators. Instead it
5593 was generating a temporary error. It now behaves as specified.
5595 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5596 (the order specifies the preference for clients). The order is now AES256,
5597 AES128, 3DES, ARCFOUR128.
5599 51. Small patch to Sieve code - explicitly set From: when generating an
5602 52. Exim crashed if a remote delivery caused a very long error message to be
5603 recorded - for instance if somebody sent an entire SpamAssassin report back
5604 as a large number of 550 error lines. This bug was coincidentally fixed by
5605 increasing the size of one of Exim's internal buffers (big_buffer) that
5606 happened as part of the Exiscan merge. However, to be on the safe side, I
5607 have made the code more robust (and fixed the comments that describe what
5610 53. Now that there can be additional text after "Completed" in log lines (if
5611 the queue_time_overall log selector is set), a one-byte patch to exigrep
5612 was needed to allow it to recognize "Completed" as not the last thing in
5615 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5616 patch that reportedly fixes this has been added. I am not expert enough to
5617 create a test for it. This is what the patch creator wrote:
5619 "I found a little strange behaviour of ldap code when working with
5620 Windows 2003 AD Domain, where users was placed in more than one
5621 Organization Units. When I tried to give exim partial DN, the exit code
5622 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5623 But simultaneously result of request was absolutely normal ldap result,
5624 so I produce this patch..."
5626 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5627 REFERENCE, so I have modified the code to exclude the patch when that macro
5630 55. Some experimental protocols are using DNS PTR records for new purposes. The
5631 keys for these records are domain names, not reversed IP addresses. The
5632 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5633 leaves it alone. Component reversal etc. now happens only for IP addresses.
5636 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5638 57. Double the size of the debug message buffer (to 2048) so that more of very
5639 long debug lines gets shown.
5641 58. The exicyclog utility now does better if the number of log files to keep
5642 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5644 59. Two changes related to the smtp_active_hostname option:
5646 (1) $smtp_active_hostname is now available as a variable.
5647 (2) The default for smtp_banner uses $smtp_active_hostname instead
5648 of $primary_hostname.
5650 60. The host_aton() function is supposed to be passed a string that is known
5651 to be a valid IP address. However, in the case of IPv6 addresses, it was
5652 not checking this. This is a hostage to fortune. Exim now panics and dies
5653 if the condition is not met. A case was found where this could be provoked
5654 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5655 components; fortuitously, this particular loophole had already been fixed
5656 by change 4.50/55 above.
5658 If there are any other similar loopholes, the new check in host_aton()
5659 itself should stop them being exploited. The report I received stated that
5660 data on the command line could provoke the exploit when Exim was running as
5661 exim, but did not say which command line option was involved. All I could
5662 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5663 running as the user.
5666 61. There was a buffer overflow vulnerability in the SPA authentication code
5667 (which came originally from the Samba project). I have added a test to the
5668 spa_base64_to_bits() function which I hope fixes it.
5671 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5672 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5674 63. The daemon start-up calls getloadavg() while still root for those OS that
5675 need the first call to be done as root, but it missed one case: when
5676 deliver_queue_load_max is set with deliver_drop_privilege. This is
5677 necessary for the benefit of the queue runner, because there is no re-exec
5678 when deliver_drop_privilege is set.
5680 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5681 This has been fixed.
5683 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5684 was in use, was not putting the data itself into the right store pool;
5685 consequently, it could be overwritten for a subsequent message in the same
5686 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5689 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5691 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5692 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5693 the function were treating the return as a boolean value, which happened to
5694 work because 0=false and not-0=true, but is not correct code.
5696 68. The host_aton() function was not handling scoped IPv6 addresses (those
5697 with, for example, "%eth0" on the end) correctly.
5699 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5700 negated items (that is, ~something) in unsigned ints. Some compilers
5701 apparently mutter when there is no cast.
5703 70. If an address verification called from an ACL failed, and did not produce a
5704 user-specific message (i.e. there was only a "system" message), nothing was
5705 put in $acl_verify_message. In this situation, it now puts the system
5708 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5709 session; change 4.31/43 added the unwanted input to the log line - except
5710 that it did not do this in the start of session case. It now does.
5712 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5713 This is wrong and can cause the other end to generate a synchronization
5714 error if it is another Exim or anything else that does the synchronization
5715 check. A QUIT command is no longer sent after a timeout.
5717 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5718 during host lookups.
5720 74. The defer_ok option of callout verification was not working if it was used
5721 when verifying addresses in header lines, that is, for this case:
5723 verify = header_sender/callout=defer_ok
5725 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5726 those file descriptors could be used for SMTP connections. If anything
5727 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5728 could be sent to the SMTP client, causing chaos. The daemon now opens
5729 stdin, stdout, and stderr to /dev/null when it puts itself into the
5732 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5733 log by default. The user can override this with suitable Perl magic.
5735 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5736 the log message when discard triggers, was not working for the DATA ACL or
5737 for the non-SMTP ACL.
5739 78. Error message wording change in sieve.c.
5741 79. If smtp_accept_max_per_host was set, the number of connections could be
5742 restricted to fewer than expected, because the daemon was trying to set up
5743 a new connection before checking whether the processes handling previous
5744 connections had finished. The check for completed processes is now done
5745 earlier. On busy systems, this bug wouldn't be noticed because something
5746 else would have woken the daemon, and it would have reaped the completed
5749 80. If a message was submitted locally by a user whose login name contained one
5750 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5751 It caused a spool format error. I have fixed the spool reading code. A
5752 related problem was that the "from" clause in the Received: line became
5753 illegal because of the space(s). It is now covered by ${quote_local_part.
5755 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5758 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5760 83. Patch from Sieve maintainer to fix unterminated string problem in
5761 "vacation" handling.
5763 84. Some minor changes to the Linux configuration files to help with other
5764 OS variants using glibc.
5766 85. One more patch for Sieve to update vacation handling to latest spec.
5769 ----------------------------------------------------
5770 See the note above about the 4.44 and 4.50 releases.
5771 ----------------------------------------------------
5777 1. Change 4.43/35 introduced a bug that caused file counts to be
5778 incorrectly computed when quota_filecount was set in an appendfile
5781 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5782 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5785 3. Give more explanation in the error message when the command for a transport
5786 filter fails to execute.
5788 4. There are several places where Exim runs a non-Exim command in a
5789 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5790 was being done only for the command run by the queryprogram router. It is
5791 now done for all such subprocesses. The other cases are: ${run, transport
5792 filters, and the commands run by the lmtp and pipe transports.
5794 5. Some older OS have a limit of 256 on the maximum number of file
5795 descriptors. Exim was using setrlimit() to set 1000 as a large value
5796 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5797 systems. I've change it so that if it can't get 1000, it tries for 256.
5799 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5800 was an oversight, and furthermore, ever since the addition of extra
5801 controls (e.g. 4.43/32), the checks on when to allow different forms of
5802 "control" were broken. There should now be diagnostics for all cases when a
5803 control that does not make sense is encountered.
5805 7. $recipients is now available in the predata ACL (oversight).
5807 8. Tidy the search cache before the fork to do a delivery from a message
5808 received from the command line. Otherwise the child will trigger a lookup
5809 failure and thereby defer the delivery if it tries to use (for example) a
5810 cached ldap connection that the parent has called unbind on.
5812 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5813 of $address_data from the recipient verification was clobbered by the
5814 sender verification.
5816 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5817 was its contents. (It was OK if the option was not defined at all.)
5819 11. A "Completed" log line is now written for messages that are removed from
5820 the spool by the -Mrm option.
5822 12. $host_address is now set to the target address during the checking of
5823 ignore_target_hosts.
5825 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5826 being passed; this would have caused $sender_host_name to have been used if
5827 matching the list had actually called for a host name (not very likely,
5828 since this list is usually IP addresses). A host name is now passed as
5831 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5832 code that specifies a non-restarting handler (typically sigaction() in
5833 modern systems) in an attempt to fix a rare and obscure crash bug.
5835 15. Narrowed the window for a race in the daemon that could cause it to ignore
5836 SIGCHLD signals. This is not a major problem, because they are used only to
5837 wake it up if nothing else does.
5839 16. A malformed maildirsize file could cause Exim to calculate negative values
5840 for the mailbox size or file count. Odd effects could occur as a result.
5841 The maildirsize information is now recalculated if the size or filecount
5844 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5845 support for a long time. Removed HAVE_SYS_VFS_H.
5847 18. Updated exipick to current release from John Jetmore.
5849 19. Allow an empty sender to be matched against a lookup in an address list.
5850 Previously the only cases considered were a regular expression, or an
5853 20. Exim went into a mad DNS lookup loop when doing a callout where the
5854 host was specified on the transport, if the DNS lookup yielded more than
5857 21. The RFC2047 encoding function was originally intended for short strings
5858 such as real names; it was not keeping to the 75-character limit for
5859 encoded words that the RFC imposes. It now respects the limit, and
5860 generates multiple encoded words if necessary. To be on the safe side, I
5861 have increased the buffer size for the ${rfc2047: expansion operator from
5864 22. Failure to deliver a bounce message always caused it to be frozen, even if
5865 there was an errors_to setting on the router. The errors_to setting is now
5868 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5869 canonical form (fully expanded) before being placed in
5870 $sender_host_address.
5872 24. Updated eximstats to version 1.33
5874 25. Include certificate and key file names in error message when GnuTLS fails
5875 to set them up, because the GnuTLS error message doesn't include the name
5876 of the failing file when there is a problem reading it.
5878 26. Expand error message when OpenSSL has problems setting up cert/key files.
5881 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5882 (this can affect the format of dates).
5884 28. exim_tidydb, when checking for the continued existence of a message for
5885 which it has found a message-specific retry record, was not finding
5886 messages that were in split spool directories. Consequently, it was
5887 deleting retry records that should have stayed in existence.
5889 29. eximstats updated to version 1.35
5890 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5891 1.35 - bugfix such that pie charts by volume are generated correctly
5893 30. The SPA authentication driver was not abandoning authentication and moving
5894 on to the next authenticator when an expansion was forced to fail,
5895 contradicting the general specification for all authenticators. Instead it
5896 was generating a temporary error. It now behaves as specified.
5898 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5899 (the order specifies the preference for clients). The order is now AES256,
5900 AES128, 3DES, ARCFOUR128.
5902 31. Small patch to Sieve code - explicitly set From: when generating an
5905 32. Exim crashed if a remote delivery caused a very long error message to be
5906 recorded - for instance if somebody sent an entire SpamAssassin report back
5907 as a large number of 550 error lines. This bug was coincidentally fixed by
5908 increasing the size of one of Exim's internal buffers (big_buffer) that
5909 happened as part of the Exiscan merge. However, to be on the safe side, I
5910 have made the code more robust (and fixed the comments that describe what
5913 33. Some experimental protocols are using DNS PTR records for new purposes. The
5914 keys for these records are domain names, not reversed IP addresses. The
5915 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5916 leaves it alone. Component reversal etc. now happens only for IP addresses.
5919 34. The host_aton() function is supposed to be passed a string that is known
5920 to be a valid IP address. However, in the case of IPv6 addresses, it was
5921 not checking this. This is a hostage to fortune. Exim now panics and dies
5922 if the condition is not met. A case was found where this could be provoked
5923 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5924 components; fortuitously, this particular loophole had already been fixed
5925 by change 4.50/55 or 4.44/33 above.
5927 If there are any other similar loopholes, the new check in host_aton()
5928 itself should stop them being exploited. The report I received stated that
5929 data on the command line could provoke the exploit when Exim was running as
5930 exim, but did not say which command line option was involved. All I could
5931 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5932 running as the user.
5935 35. There was a buffer overflow vulnerability in the SPA authentication code
5936 (which came originally from the Samba project). I have added a test to the
5937 spa_base64_to_bits() function which I hope fixes it.
5940 36. The daemon start-up calls getloadavg() while still root for those OS that
5941 need the first call to be done as root, but it missed one case: when
5942 deliver_queue_load_max is set with deliver_drop_privilege. This is
5943 necessary for the benefit of the queue runner, because there is no re-exec
5944 when deliver_drop_privilege is set.
5946 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5947 was in use, was not putting the data itself into the right store pool;
5948 consequently, it could be overwritten for a subsequent message in the same
5949 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5952 38. Sometimes the final signoff response after QUIT could fail to get
5953 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5954 < 0 before doing a fflush(). This bug looks as though it goes back to the
5955 introduction of TLS in release 3.20, but "sometimes" must have been rare
5956 because the tests only now provoked it.
5962 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5963 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5964 Now it returns an integer. A number of calls were still expecting a T/F
5965 return. Fortuitously, in all cases, the tests worked in OK situations,
5966 which is the norm. However, things would have gone wrong on any write
5967 failures on the smtp file descriptor. This function is used when sending
5968 messages over SMTP and also when doing verify callouts.
5970 2. When Exim is called to do synchronous delivery of a locally submitted
5971 message (the -odf or -odi options), it no longer closes stderr before doing
5974 3. Implemented the mua_wrapper option.
5976 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5978 5. Implemented the functions header_remove(), header_testname(),
5979 header_add_at_position(), and receive_remove_recipient(), and exported them
5982 6. If an ACL "warn" statement specified the addition of headers, Exim already
5983 inserted X-ACL-Warn: at the start if there was no header name. However, it
5984 was not making this test for the second and subsequent header lines if
5985 there were newlines in the string. This meant that an invalid header could
5986 be inserted if Exim was badly configured.
5988 7. Allow an ACL "warn" statement to add header lines at the start or after all
5989 the Received: headers, as well as at the end.
5991 8. Added the rcpt_4xx retry error code.
5993 9. Added postmaster_mailfrom=xxx to callout verification option.
5995 10. Added mailfrom=xxxx to the callout verification option, for verify=
5998 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5999 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6001 12. Inserted some casts to stop certain compilers warning when using pointer
6002 differences as field lengths or precisions in printf-type calls (mostly
6003 affecting debugging statements).
6005 13. Added optional readline() support for -be (dynamically loaded).
6007 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6008 same clock tick as a message's arrival, so that its received time was the
6009 same as the "first fail" time on the retry record, and that message
6010 remained on the queue past the ultimate address timeout, every queue runner
6011 would try a delivery (because it was past the ultimate address timeout) but
6012 after another failure, the ultimate address timeout, which should have then
6013 bounced the address, did not kick in. This was a "< instead of <=" error;
6014 in most cases the first failure would have been in the next clock tick
6015 after the received time, and all would be well.
6017 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6018 being recognized when the domain list was tested by the match_domain
6019 condition in an expansion string.
6021 16. Added the ${str2b64: operator.
6023 17. Exim was always calling setrlimit() to set a large limit for the number of
6024 processes, without checking whether the existing limit was already
6025 adequate. (It did check for the limit on file descriptors.) Furthermore,
6026 errors from getrlimit() and setrlimit() were being ignored. Now they are
6027 logged to the main and panic logs, but Exim does carry on, to try to do its
6028 job under whatever limits there are.
6030 18. Imported PCRE 5.0.
6032 19. Trivial typo in log message " temporarily refused connection" (the leading
6035 20. If the log selector return_path_on_delivery was set and an address was
6036 redirected to /dev/null, the delivery process crashed because it assumed
6037 that a return path would always be set for a "successful" delivery. In this
6038 case, the whole delivery is bypassed as an optimization, and therefore no
6041 21. Internal re-arrangement: the function for sending a challenge and reading
6042 a response while authentication was assuming a zero-terminated challenge
6043 string. It's now changed to take a pointer and a length, to allow for
6044 binary data in such strings.
6046 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6048 23. Exim was not respecting finduser_retries when seeking the login of the
6049 uid under which it was called; it was always trying 10 times. (The default
6050 setting of finduser_retries is zero.) Also, it was sleeping after the final
6051 failure, which is pointless.
6053 24. Implemented tls_on_connect_ports.
6055 25. Implemented acl_smtp_predata.
6057 26. If the domain in control=submission is set empty, Exim assumes that the
6058 authenticated id is a complete email address when it generates From: or
6059 Sender: header lines.
6061 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6062 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6063 chown and chgrp in /bin and hostname in /usr/bin.
6065 28. Exim was keeping the "process log" file open after each use, just as it
6066 does for the main log. This opens the possibility of it remaining open for
6067 long periods when the USR1 signal hits a daemon. Occasional processlog
6068 errors were reported, that could have been caused by this. Anyway, it seems
6069 much more sensible not to leave this file open at all, so that is what now
6072 29. The long-running daemon process does not normally write to the log once it
6073 has entered its main loop, and it closes the log before doing so. This is
6074 so that log files can straightforwardly be renamed and moved. However,
6075 there are a couple of unusual error situations where the daemon does write
6076 log entries, and I had neglected to close the log afterwards.
6078 30. The text of an SMTP error response that was received during a remote
6079 delivery was being truncated at 512 bytes. This is too short for some of
6080 the long messages that one sometimes sees. I've increased the limit to
6083 31. It is now possible to make retry rules that apply only when a message has a
6084 specific sender, in particular, an empty sender.
6086 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6087 it possible to be selective about when SMTP synchronization is enforced.
6089 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6091 32. Implemented hosts_connection_nolog.
6093 33. Added an ACL for QUIT.
6095 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6098 35. Added mailbox_size and mailbox_filecount to appendfile.
6100 36. Added control = no_multiline_responses to ACLs.
6102 37. There was a bug in the logic of the code that waits for the clock to tick
6103 in the case where the clock went backwards by a substantial amount such
6104 that the microsecond fraction of "now" was more than the microsecond
6105 fraction of "then" (but the whole seconds number was less).
6107 38. Added support for the libradius Radius client library this is found on
6108 FreeBSD (previously only the radiusclient library was supported).
6114 1. When certain lookups returned multiple values in the form name=value, the
6115 quoting of the values was not always being done properly. Specifically:
6116 (a) If the value started with a double quote, but contained no whitespace,
6118 (b) If the value contained whitespace other than a space character (i.e.
6119 tabs or newlines or carriage returns) it was not quoted.
6120 This fix has been applied to the mysql and pgsql lookups by writing a
6121 separate quoting function and calling it from the lookup code. The fix
6122 should probably also be applied to nisplus, ibase and oracle lookups, but
6123 since I cannot test any of those, I have not disturbed their existing code.
6125 2. A hit in the callout cache for a specific address caused a log line with no
6126 reason for rejecting RCPT. Now it says "Previous (cached) callout
6127 verification failure".
6129 3. There was an off-by-one bug in the queryprogram router. An over-long
6130 return line was truncated at 256 instead of 255 characters, thereby
6131 overflowing its buffer with the terminating zero. As well as fixing this, I
6132 have increased the buffer size to 1024 (and made a note to document this).
6134 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6135 when Exim is waiting for an SMTP response from a remote server, Exim
6136 restarts its select() call on the socket, thereby resetting its timeout.
6137 This is not a problem when such interrupts are rare. Somebody set up a cron
6138 job to run exiwhat every 2 minutes, which is less than the normal select()
6139 timeout (5 or 10 minutes). This meant that the select() timeout never
6140 kicked in because it was always reset. I have fixed this by comparing the
6141 time when an interrupt arrives with the time at the start of the first call
6142 to select(). If more time than the timeout has elapsed, the interrupt is
6143 treated as a timeout.
6145 5. Some internal re-factoring in preparation for the addition of Sieve
6146 extensions (by MH). In particular, the "personal" test is moved to a
6147 separate function, and given an option for scanning Cc: and Bcc: (which is
6148 not set for Exim filters).
6150 6. When Exim created an email address using the login of the caller as the
6151 local part (e.g. when creating a From: or Sender: header line), it was not
6152 quoting the local part when it contained special characters such as @.
6154 7. Installed new OpenBSD configuration files.
6156 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6157 try to make them clearer.
6159 9. Callout options, other than the timeout value, were being ignored when
6160 verifying sender addresses in header lines. For example, when using
6162 verify = header_sender/callout=no_cache
6164 the cache was (incorrectly) being used.
6166 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6167 only the Cygwin environment.
6169 11. When return_path_on_delivery was set as a log selector, if different remote
6170 addresses in the same message used different return paths and parallel
6171 remote delivery occurred, the wrong values would sometimes be logged.
6172 (Whenever a remote delivery process finished, the return path value from
6173 the most recently started remote delivery process was logged.)
6175 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6176 lines when AUTH and/or TLS are in use. This is the "received protocol"
6177 field. Exim used to use "asmtp" for authenticated SMTP, without any
6178 indication (in the protocol name) for TLS use. Now it follows the RFC and
6179 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6180 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6181 names appear in log lines as well as in Received: header lines.
6183 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6184 extensions, and comparison tests, and to fix some bugs.
6186 14. Changes to the "personal" filter test:
6188 (1) The test was buggy in that it was just doing the equivalent of
6189 "contains" tests on header lines. For example, if a user's address was
6190 anne@some.where, the "personal" test would incorrectly be true for
6192 To: susanne@some.where
6194 This test is now done by extracting each address from the header in turn,
6195 and checking the entire address. Other tests that are part of "personal"
6196 are now done using regular expressions (for example, to check local parts
6197 of addresses in From: header lines).
6199 (2) The list of non-personal local parts in From: addresses has been
6200 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6201 taken from the Sieve specification recommendations.
6203 (3) If the message contains any header line starting with "List-" it is
6204 treated as non-personal.
6206 (4) The test for "circular" in the Subject: header line has been removed
6207 because it now seems ill-conceived.
6209 15. Minor typos in src/EDITME comments corrected.
6211 16. Installed latest exipick from John Jetmore.
6213 17. If headers_add on a router specified a text string that was too long for
6214 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6215 of string_sprintf() is now avoided.
6217 18. $message_body_size was not set (it was always zero) when running the DATA
6218 ACL and the local_scan() function.
6220 19. For the "mail" command in an Exim filter, no default was being set for
6221 the once_repeat time, causing a random time value to be used if "once" was
6222 specified. (If the value happened to be <= 0, no repeat happened.) The
6223 default is now 0s, meaning "never repeat". The "vacation" command was OK
6224 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6225 (I found it when inspecting the code).
6227 20. There is now an overall timeout for performing a callout verification. It
6228 defaults to 4 times the callout timeout, which applies to individual SMTP
6229 commands during the callout. The overall timeout applies when there is more
6230 than one host that can be tried. The timeout is checked before trying the
6231 next host. This prevents very long delays if there are a large number of
6232 hosts and all are timing out (e.g. when the network connections are timing
6233 out). The value of the overall timeout can be changed by specifying an
6234 additional sub-option for "callout", called "maxwait". For example:
6236 verify = sender/callout=5s,maxwait=20s
6238 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6239 to the end before writing, but this should make it even safer).
6241 22. Exim was forgetting that it had advertised PIPELINING for the second and
6242 subsequent messages on an SMTP connection. It was also not resetting its
6243 memory on STARTTLS and an internal HELO.
6245 23. When Exim logs an SMTP synchronization error within a session, it now
6246 records whether PIPELINING has been advertised or not.
6248 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6249 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6250 rather than long int.
6252 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6254 26. Added the never_mail option to autoreply.
6260 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6261 crash if the getsockname() call failed; this can happen if a connection is
6262 closed very soon after it is established. The problem was simply in the
6263 order in which certain operations were done, causing Exim to try to write
6264 to the SMTP stream before it had set up the file descriptor. The bug has
6265 been fixed by making things happen in the correct order.
6271 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6272 before the connection was closed, thus losing the rejection response.
6274 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6275 some early Solaris releases, but causes trouble in current releases where
6276 socklen_t is defined.
6278 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6281 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6284 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6285 that releasing the top of it at the end releases what was used for sub-
6286 expansions (unless the block got too big). However, discard this block if
6287 the first thing is a variable or header, so that we can use its block when
6288 it is dynamic (useful for very large $message_headers, for example).
6290 6. Lookups now cache *every* query, not just the most recent. A new, separate
6291 store pool is used for this. It can be recovered when all lookup caches are
6292 flushed. Lookups now release memory at the end of their result strings.
6293 This has involved some general refactoring of the lookup sources.
6295 7. Some code has been added to the store_xxx() functions to reduce the amount
6296 of flapping under certain conditions.
6298 8. log_incoming_interface used to affect only the <= reception log lines. Now
6299 it causes the local interface and port to be added to several more SMTP log
6300 lines, for example "SMTP connection from", and rejection lines.
6302 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6304 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6306 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6307 could be overwritten at the end of the current message (or the start of a
6308 new message if it was set in a HELO ACL). The value is now preserved for
6309 the duration of the SMTP connection.
6311 12. If a transport had a headers_rewrite setting, and a matching header line
6312 contained an unqualified address, that address was qualified, even if it
6313 did not match any rewriting rules. The underlying bug was that the values
6314 of the flags that permit the existence of unqualified sender and recipient
6315 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6316 non-local messages, and by -bnq for local messages) were not being
6317 preserved with the message after it was received.
6319 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6320 "next input=" as part of the text comprising the host identity instead of
6321 the correct text. The code was using the same buffer for two different
6322 strings. However, depending on which order the printing function evaluated
6323 its arguments, the bug did not always show up. Under Linux, for example, my
6324 test suite worked just fine.
6326 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6327 doesn't work with some older versions of Perl. It has been changed to "my",
6328 which in any case is probably the better facility to use.
6330 15. A really picky compiler found some instances of statements for creating
6331 error messages that either had too many or two few arguments for the format
6334 16. The size of the buffer for calls to the DNS resolver has been increased
6335 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6336 for addresses that have a lot of PTR records. This alleviates a problem; it
6337 does not fully solve it.
6339 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6340 buffer now truncates the list and logs the incident, which is the same
6341 action as happens when Exim is looking up a host name and its aliases.
6342 Previously in this situation something unpredictable would happen;
6343 sometimes it was "internal error: store_reset failed".
6345 18. If a server dropped the connection unexpectedly when an Exim client was
6346 using GnuTLS and trying to read a response, the client delivery process
6347 crashed while trying to generate an error log message.
6349 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6350 string, for example:
6352 warn message = H1: something\nH2: something
6354 the text was added as a single header line from Exim's point of view
6355 though it ended up OK in the delivered message. However, searching for the
6356 second and subsequent header lines using $h_h2: did not work. This has been
6357 fixed. Similarly, if a system filter added multiple headers in this way,
6358 the routers could not see them.
6360 20. Expanded the error message when iplsearch is called with an invalid key to
6361 suggest using net-iplsearch in a host list.
6363 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6364 ACLs are no longer actually imposed (and a message to that effect is
6367 22. If a "gecos" field in a passwd entry contained escaped characters, in
6368 particular, if it contained a \" sequence, Exim got it wrong when building
6369 a From: or a Sender: header from that name. A second bug also caused
6370 incorrect handling when an unquoted " was present following a character
6371 that needed quoting.
6373 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6374 was not being matched caselessly.
6376 24. Arranged for all hyphens in the exim.8 source to be escaped with
6379 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6380 properly. Recipient callout cache records were still being keyed to include
6381 the sender, even when use_sender was set false. This led to far more
6382 callouts that were necessary. The sender is no longer included in the key
6383 when use_sender is false.
6385 26. Added "control = submission" modifier to ACLs.
6387 27. Added the ${base62d: operator to decode base 62 numbers.
6389 28. dnsdb lookups can now access SRV records.
6391 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6392 the configuration file.
6394 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6395 (-v) mode. This makes the output for a verbose queue run more intelligible.
6397 31. Added a use_postmaster feature to recipient callouts.
6399 32. Added the $body_zerocount variable, containing the number of binary zero
6400 bytes in the message body.
6402 33. The time of last modification of the "new" subdirectory is now used as the
6403 "mailbox time last read" when there is a quota error for a maildir
6406 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6408 35. Added +ignore_unknown as a special item in host lists.
6410 36. Code for decoding IPv6 addresses in host lists is now included, even if
6411 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6412 address was recognized as an IP address, but was then not correctly decoded
6413 into binary, causing unexpected and incorrect effects when compared with
6420 1. Very minor rewording of debugging text in manualroute to say "list of
6421 hosts" instead of "hostlist".
6423 2. If verify=header_syntax was set, and a header line with an unqualified
6424 address (no domain) and a large number of spaces between the end of the
6425 name and the colon was received, the reception process suffered a buffer
6426 overflow, and (when I tested it) crashed. This was caused by some obsolete
6427 code that should have been removed. The fix is to remove it!
6429 3. When running in the test harness, delay a bit after writing a bounce
6430 message to get a bit more predictability in the log output.
6432 4. Added a call to search_tidyup() just before forking a reception process. In
6433 theory, someone could use a lookup in the expansion of smtp_accept_max_
6434 per_host which, without the tidyup, could leave open a database connection.
6436 5. Added the variables $recipient_data and $sender_data which get set from a
6437 lookup success in an ACL "recipients" or "senders" condition, or a router
6438 "senders" option, similar to $domain_data and $local_part_data.
6440 6. Moved the writing of debug_print from before to after the "senders" test
6443 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6444 problems for message scanning, either using a data ACL, or using
6445 local_scan() because the Received: header was not generated till after they
6446 were called (in order to set the time as the time of reception completion).
6447 I have revised the way this works. The header is now generated after the
6448 body is received, but before the ACL or local_scan() are called. After they
6449 are run, the timestamp in the header is updated.
6455 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6456 before starting a queue runner without re-exec. This happened only when
6457 deliver_drop_privilege was set or when the Exim user was set to root. The
6458 effect of the bug was that timeouts during subsequent deliveries caused
6459 crashes instead of being properly handled. The handler is now left at its
6460 default (and expected) setting.
6462 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6463 message, again when deliver_drop_privilege is set or Exim is run as root.
6464 The bug described in (1) was not present in this case, but the tidying up
6465 of the other signals was missing. I have made the two cases consistent.
6467 3. The ignore_target_hosts setting on a manualroute router was being ignored
6468 for hosts that were looked up using the /MX notation.
6470 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6473 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6474 operated on the sender address. After changing the $sender_address to <>
6475 for the sender address verify, Exim was re-instated it as the original
6476 (before rewriting) address, but remembering that it had rewritten it, so it
6477 wasn't rewriting it again. This bug also had the effect of breaking the
6478 sender address verification caching when the sender address was rewritten.
6480 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6481 This has been changed so that if the ip literal address matches
6482 ignore_target_hosts, the router declines.
6484 7. Added expansion conditions match_domain, match_address, and match_local_
6485 part (NOT match_host).
6487 8. The placeholder for the Received: header didn't have a length field set.
6489 9. Added code to Exim itself and to exim_lock to test for a specific race
6490 condition that could lead to file corruption when using MBX delivery. The
6491 issue is with the lockfile that is created in /tmp. If this file is removed
6492 after a process has opened it but before that process has acquired a lock,
6493 there is the potential for a second process to recreate the file and also
6494 acquire a lock. This could lead to two Exim processes writing to the file
6495 at the same time. The added code performs the same test as UW imapd; it
6496 checks after acquiring the lock that its file descriptor still refers to
6497 the same named file.
6499 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6500 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6503 11. Added the smtp_active_hostname option. If used, this will typically be made
6504 to depend on the incoming interface address. Because $interface_address is
6505 not set up until the daemon has forked a reception process, error responses
6506 that can happen earlier (such as "too many connections") no longer contain
6509 12. If an expansion in a condition on a "warn" statement fails because a lookup
6510 defers, the "warn" statement is abandoned, and the next ACL statement is
6511 processed. Previously this caused the whole ACL to be aborted.
6513 13. Added the iplsearch lookup type.
6515 14. Added ident_timeout as a log selector.
6517 15. Added tls_certificate_verified as a log selector.
6519 16. Added a global option tls_require_ciphers (compare the smtp transport
6520 option of the same name). This controls incoming TLS connections.
6522 17. I finally figured out how to make tls_require_ciphers do a similar thing
6523 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6524 before starting the TLS session.
6526 18. Tabs are now shown as \t in -bP output.
6528 19. If the log selector return_path_on_delivery was set, Exim crashed when
6529 bouncing a message because it had too many Received: header lines.
6531 20. If two routers both had headers_remove settings, and the first one included
6532 a superfluous trailing colon, the final name in the first list and the
6533 first name in the second list were incorrectly joined into one item (with a
6534 colon in the middle).
6540 1. Added -C and -D options to the exinext utility, mainly to make it easier
6541 to include in the automated testing, but these could be helpful when
6542 multiple configurations are in use.
6544 2. The exinext utility was not formatting the output nicely when there was
6545 an alternate port involved in the retry record key, nor when there was a
6546 message id as well (for retries that were specific to a specific message
6547 and a specific host). It was also confused by IPv6 addresses, because of
6548 the additional colons they contain. I have fixed the IPv4 problem, and
6549 patched it up to do a reasonable job for IPv6.
6551 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6552 delivery, the log line now contains "pipelined" if PIPELINING was used.
6554 4. An SMTP transport process used to panic and die if the bind() call to set
6555 an explicit outgoing interface failed. This has been changed; it is now
6556 treated in the same way as a connect() failure.
6558 5. A reference to $sender_host_name in the part of a conditional expansion
6559 that was being skipped was still causing a DNS lookup. This no longer
6562 6. The def: expansion condition was not recognizing references to header lines
6563 that used bh_ and bheader_.
6565 7. Added the _cache feature to named lists.
6567 8. The code for checking quota_filecount in the appendfile transport was
6568 allowing one more file than it should have been.
6570 9. For compatibility with Sendmail, the command line option
6578 and sets the incoming protocol and host name (for trusted callers). The
6579 host name and its colon can be omitted when only the protocol is to be set.
6580 Note the Exim already has two private options, -pd and -ps, that refer to
6581 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6582 "s", but I don't think that's a major issue.
6584 10. A number of refactoring changes to the code, none of which should affect
6587 (a) The number of logging options was getting close to filling up the
6588 32-bit word that was used as a bit map. I have split them into two classes:
6589 those that are passed in the argument to log_write(), and those that are
6590 only ever tested independently outside of that function. These are now in
6591 separate 32-bit words, so there is plenty of room for expansion again.
6592 There is no change in the user interface or the logging behaviour.
6594 (b) When building, for example, log lines, the code previously used a
6595 macro that called string_cat() twice, in order to add two strings. This is
6596 not really sufficiently general. Furthermore, there was one instance where
6597 it was actually wrong because one of the argument was used twice, and in
6598 one call a function was used. (As it happened, calling the function twice
6599 did not affect the overall behaviour.) The macro has been replaced by a
6600 function that can join an arbitrary number of extra strings onto a growing
6603 (c) The code for expansion conditions now uses a table and a binary chop
6604 instead of a serial search (which was left over from when there were very
6605 few conditions). Also, it now recognizes conditions like "pam" even when
6606 the relevant support is not compiled in: a suitably worded error message is
6607 given if an attempt is made to use such a condition.
6609 11. Added ${time_interval:xxxxx}.
6611 12. A bug was causing one of the ddress fields not to be passed back correctly
6612 from remote delivery subprocesses. The field in question was not being
6613 subsequently used, so this caused to problems in practice.
6615 13. Added new log selectors queue_time and deliver_time.
6617 14. Might have fixed a bug in maildirsizefile handling that threw up
6618 "unexpected character" debug warnings, and recalculated the data
6619 unnecessarily. In any case, I expanded the warning message to give more
6622 15. Added the message "Restricted characters in address" to the statements in
6623 the default ACL that block characters like @ and % in local parts.
6625 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6626 Three changes have been made:
6628 (a) There was a serious bug; a negative response to MAIL caused the whole
6629 recipient domain to be cached as invalid, thereby blocking all messages
6630 to all local parts at the same domain, from all senders. This bug has
6631 been fixed. The domain is no longer cached after a negative response to
6632 MAIL if the sender used is not empty.
6634 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6637 (c) A new callout option, "use_sender" has been added for people who want
6638 the modified behaviour.
6644 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6647 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6648 indeed breaks things for older releases.
6650 3. Added additional logging to the case where there is a problem reading data
6651 from a filter that is running in a subprocess using a pipe, in order to
6652 try to track down a specific problem.
6654 4. Testing facility fudge: when running in the test harness and attempting
6655 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6656 getting "No route to host". Convert this to a timeout.
6658 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6661 6. Some OS don't have socklen_t but use size_t instead. This affects the
6662 fifth argument of getsockopt() amongst other things. This is now
6663 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6664 can be set for individual OS. I have set it for SunOS5, OSF1, and
6665 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6666 some earlier ones do not.
6668 7. Change 4.30/15 was not doing the test caselessly.
6670 8. The standard form for an IPv6 address literal was being rejected by address
6671 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6672 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6673 this, as well as the form without the "IPv6" on the front (but only when
6674 address literals are enabled, of course).
6676 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6678 10. Exim crashed if a message with an empty sender address specified by -f
6679 encountered a router with an errors_to setting. This could be provoked only
6680 by a command such as
6684 where an empty string was supplied; "<>" did not hit this bug.
6686 11. Installed PCRE release 4.5.
6688 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6689 remained set. It is now erased.
6691 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6692 times from message ids (which are base 36 rather than the normal 62).
6694 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6695 were being counted as actual protocol errors, and logged if the log
6696 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6697 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6698 and DATA following a set of rejected RCPTs do not count as protocol errors.
6699 In other words, Exim assumes they were pipelined, though this may not
6700 actually be the case. Of course, in all cases the client gets an
6701 appropriate error code.
6703 15. If a lookup fails in an ACL condition, a message about the failure may
6704 be available; it is used if testing the ACL cannot continue, because most
6705 such messages specify what the cause of the deferral is. However, some
6706 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6707 that caused an old message to be retained and used if a later statement
6708 caused a defer, replacing the real cause of the deferral.
6710 16. If an IP address had so many PTR records that the DNS lookup buffer
6711 was not large enough to hold them, Exim could crash while trying to process
6712 the truncated data. It now detects and logs this case.
6714 17. Further to 4.21/58, another change has been made: if (and only if) the
6715 first line of a message (the first header line) ends with CRLF, a bare LF
6716 in a subsequent header line has a space inserted after it, so as not to
6717 terminate the header.
6719 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6720 unnecessarily, used atoi() instead of strtol(), and didn't check the
6721 termination when getting file sizes from file names by regex.
6723 19. Completely re-implemented the support for maildirsize files, in the light
6724 of a number of problems with the previous contributed implementation
6725 (4.30/29). In particular:
6727 . If the quota is zero, the maildirsize file is maintained, but no quota is
6730 . If the maildir directory does not exist, it is created before any attempt
6731 to write a maildirsize file.
6733 . The quota value in the file is just a cache; if the quota is changed in
6734 the transport, the new value overrides.
6736 . A regular expression is available for excluding directories from the
6739 20. The autoreply transport checks the characters in options that define the
6740 message's headers; it allows continued headers, but it was checking with
6741 isspace() after an embedded newline instead of explicitly looking for a
6744 21. If all the "regular" hosts to which an address was routed had passed their
6745 expiry times, and had not reached their retry times, the address was
6746 bounced, even if fallback hosts were defined. Now Exim should go on to try
6749 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6750 equivalent code in the SMTP transport. Some hosts send humungous responses
6751 to HELO/EHLO, more than 1024 it seems.
6753 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6754 gives compiler warnings in some environments. I've now done it "properly",
6757 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6758 (because of problems with the built-in one) was declared to return uschar *
6759 instead of char *, causing compiler failure.
6761 25. Fixed a file descriptor leak when processing alias/forward files.
6763 26. Fixed a minor format string issue in dbfn.c.
6765 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6767 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6768 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6771 29. When a sender address is verified, it is cached, to save repeating the test
6772 when there is more than one recipient in a message. However, when the
6773 verification involves a callout, it is possible for different callout
6774 options to be set for different recipients. It is too complicated to keep
6775 track of this in the cache, so now Exim always runs a verification when a
6776 callout is required, relying on the callout cache for the optimization.
6777 The overhead is duplication of the address routing, but this should not be
6780 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6781 to be verified with callout=postmaster, and the main callout worked but the
6782 postmaster check failed, the verification correctly failed. However, if a
6783 subsequent RCPT command asked for sender verification *without* the
6784 postmaster check, incorrect caching caused this verification also to fail,
6787 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6788 it was not caching the DNS options (qualify_single, search_parents) that
6789 were used when the lookup failed. A subsequent lookup with different
6790 options therefore always gave the same answer, though there were cases
6791 where it should not have. (Example: a "domains = !$mx_any" option on a
6792 dnslookup router: the "domains" option is always processed without any
6793 widening, but the router might have qualify_single set.) Now Exim uses the
6794 cached value only when the same options are set.
6796 32. Added John Jetmore's "exipick" utility to the distribution.
6798 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6799 than a timeout (e.g. a certificate is required, and is not provided), an
6800 Exim server now closes the connection immediately. Previously it waited for
6801 the client to close - but if the client is SSL, it seems that they each
6802 wait for each other, leading to a delay before one of them times out.
6804 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6805 maintained 0.8.x compatibility because I don't think many are using it, and
6806 it is clearly obsolete.
6808 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6811 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6812 client certificate was expired. A simple patch fixes this, though I don't
6813 understand the full logic of why the verify callback is called multiple
6816 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6817 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6818 which causes problems with some clients (such as the Certicom SSL Plus
6819 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6820 disables the coutermeasure allowing Eudora to connect."
6822 38. Exim was not checking that a write() to a log file succeeded. This could
6823 lead to Bad Things if a log got too big, in particular if it hit a file
6824 size limit. Exim now panics and dies if it cannot write to a log file, just
6825 as it does if it cannot open a log file.
6827 39. Modified OS/Makefile-Linux so that it now contains
6829 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6831 The two -D definitions ensure that Exim is compiled with large file
6832 support, which makes it possible to handle log files that are bigger than
6835 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6836 instance) a domain was checked against a named list that involved a lookup,
6837 causing $domain_data to be set, then another domain was checked against the
6838 same list, then the first domain was re-checked, the value of $domain_data
6839 after the final check could be wrong. In particular, if the second check
6840 failed, it could be set empty. This bug probably also applied to
6843 41. The strip_trailing_dot option was not being applied to the address given
6844 with the -f command-line option.
6846 42. The code for reading a message's header from the spool was incrementing
6847 $received_count, but never initializing it. This meant that the value was
6848 incorrect (doubled) while delivering a message in the same process in which
6849 it was received. In the most common configuration of Exim, this never
6850 happens - a fresh exec is done - but it can happen when
6851 deliver_drop_privilege is set.
6853 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6854 it now includes up to 150 characters of the unexpected data in the log
6857 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6858 and building data strings. The size of both of these buffers was 10 000
6859 bytes - far larger than anybody would *ever* want, thought I. Needless to
6860 say, somebody hit the limit. I have increased the maximum line length to
6861 20 000 and the maximum data length of concatenated lines to 100 000. I have
6862 also fixed two bugs, because there was no checking on these buffers. Tsk,
6863 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6864 buffer is too small.
6866 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6867 lsearch lookups. Now it does.
6869 46. When parsing a route_list item in a manualroute router, a fixed-length
6870 buffer was used for the list of hosts. I made this 1024 bytes long,
6871 thinking that nobody would ever have a list of hosts that long. Wrong.
6872 Somebody had a whole pile of complicated expansion conditions, and the
6873 string was silently truncated, leading to an expansion error. It turns out
6874 that it is easier to change to an unlimited length (owing to other changes
6875 that have happened since this code was originally written) than to build
6876 structure for giving a limitation error. The length of the item that
6877 expands into the list of hosts is now unlimited.
6879 47. The lsearch lookup could not handle data where the length of text line was
6880 more than 4095 characters. Such lines were truncated, leading to shortened
6881 data being returned. It should now handle lines of any length.
6883 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6884 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6887 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6888 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6889 no longer happens in comment lines. A list of replacements is now placed
6890 at the head of all of the source files, except those whose only change is
6891 to replace PERL_COMMAND in the very first #! line.
6893 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6894 messages on the queue, with a bottom-up merge sort, using code contributed
6895 by Michael Haardt. This should make operations like -bp somewhat faster on
6896 large queues. It won't affect queue runners, except when queue_run_in_order
6899 51. Installed eximstats 1.31 in the distribution.
6901 52. Added support for SRV lookups to the dnslookup router.
6903 53. If an ACL referred to $message_body or $message_body_end, the value was not
6904 reset for any messages that followed in the same SMTP session.
6906 54. The store-handling optimization for building very long strings was not
6907 differentiating between the different store pools. I don't think this
6908 actually made any difference in practice, but I've tidied it.
6910 55. While running the routers to verify a sender address, $sender_address
6911 was still set to the sender address. This is wrong, because when routing to
6912 send a bounce to the sender, it would be empty. Therefore, I have changed
6913 it so that, while verifying a sender address, $sender_address is set to <>.
6914 (There is no change to what happens when verifying a recipient address.)
6916 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6917 target A or AAAA records (if not already returned) without resetting the
6918 qualify_single or search_parents options of the DNS resolver. These are
6919 inappropriate in this case because the targets of MX and SRV records must
6920 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6921 target that, when qualified, matched something in the local domain. These
6922 two options are now turned off when doing these lookups.
6924 57. It seems that at least some releases of Reiserfs (which does not have the
6925 concept of a fixed number of inodes) returns zero and not -1 for the
6926 number of available inodes. This interacted badly with check_spool_inodes,
6927 which assumed that -1 was the "no such thing" setting. What I have done is
6928 to check that the total number of inodes is greater than zero before doing
6929 the test of how many are available.
6931 58. When a "warn" ACL statement has a log_message modifier, the message is
6932 remembered, and not repeated. This is to avoid a lot of repetition when a
6933 message has many recipients that cause the same warning to be written.
6934 However, Exim was preserving the list of already written lines for an
6935 entire SMTP session, which doesn't seem right. The memory is now reset if a
6936 new message is started.
6938 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6939 debugging output unless log_rewrite was also set.
6941 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6942 of (int)(handle) when we know that handle contains (void *)(-1).
6944 61. The Exim daemon panic-logs an error return when it closes the incoming
6945 connection. However "connection reset by peer" seems to be common, and
6946 isn't really an error worthy of noting specially, so that particular error
6949 62. When Exim is trying to find all the local interfaces, it used to panic and
6950 die if the ioctl to get the interface flags failed. However, it seems that
6951 on at least one OS (Solaris 9) it is possible to have an interface that is
6952 included in the list of interfaces, but for which you get a failure error
6953 for this call. This happens when the interface is not "plumbed" into a
6954 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6955 failure of the "get flags" call assumes that the interface is down.
6957 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6958 makes life easier for people who are doing arithmetic on fields extracted
6959 from dates, where you often get leading zeros that should not be
6960 interpreted as octal.
6962 64. Added qualify_domain to the redirect router, to override the global
6965 65. If a pathologically long header line contained very many addresses (the
6966 report of this problem mentioned 10 000) and each of them was rewritten,
6967 Exim could use up a very large amount of memory. (It kept on making new
6968 copies of the header line as it rewrote, and never released the old ones.)
6969 At the expense of a bit more processing, the header rewriting function has
6970 been changed so that it no longer eats memory in this way.
6972 66. The generation of the Received: header has been moved from the time that a
6973 message starts to be received, to the time that it finishes. The timestamp
6974 in the Received: header should now be very close to that of the <= log
6975 line. There are two side-effects of this change:
6977 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6978 logged header lines no longer include the local Received: line, because
6979 it has not yet been created. The same applies to a copy of the message
6980 that is returned to a non-SMTP sender when a message is rejected.
6982 (b) When a filter file is tested using -bf, no additional Received: header
6983 is added to the test message. After some thought, I decided that this
6986 This change does not affect the value of $received_for. It is still set
6987 after address rewriting, but before local_scan() is called.
6989 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6991 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6992 gave an unhelpful panic error message, and a defer error. I have managed to
6993 change this behaviour so that it now rejects any supplied certificate,
6994 which seems right, as the list of acceptable certificates is empty.
6996 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6997 gave an unhelpful defer error. I have not managed to make this reject any
6998 supplied certificates, but the error message it gives is "no certificate
6999 supplied", which is not helpful.
7001 70. exigrep's output now also includes lines that are not associated with any
7002 message, but which match the given pattern. Implemented by a patch from
7003 Martin Sluka, which also tidied up the Perl a bit.
7005 71. Recipient callout verification, like sender verification, was using <> in
7006 the MAIL FROM command. This isn't really the right thing, since the actual
7007 sender may affect whether the remote host accepts the recipient or not. I
7008 have changed it to use the actual sender in the callout; this means that
7009 the cache record is now keyed on a recipient/sender pair, not just the
7010 recipient address. There doesn't seem to be a real danger of callout loops,
7011 since a callout by the remote host to check the sender would use <>.
7012 [SEE ABOVE: changed after hitting problems.]
7014 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7015 temporary errors. However, in the case of such a code being given after
7016 the end of a data transmission (i.e. after ".") Exim was failing to write
7017 a retry record for the message. (Yes, there was some broken host that was
7018 actually sending 8xx at this point.)
7020 73. An unknown lookup type in a host list could cause Exim to panic-die when
7021 the list was checked. (An example that provoked this was putting <; in the
7022 middle of a list instead of at the start.) If this happened during a DATA
7023 ACL check, a -D file could be left lying around. This kind of configuration
7024 error no longer causes Exim to die; instead it causes a defer error. The
7025 incident is still logged to the main and panic logs.
7027 74. Buglet left over from Exim 3 conversion. The message "too many messages
7028 in one connection" was written to the rejectlog but not the mainlog, except
7029 when address rewriting (yes!) was being logged.
7031 75. Added write_rejectlog option.
7033 76. When a system filter was run not as root (that is, when system_filter_user
7034 was set), the values of the $n variables were not being returned to the
7035 main process; thus, they were not subsequently available in the $sn
7038 77. Added +return_path_on_delivery log selector.
7040 78. A connection timeout was being treated differently from recipients deferred
7041 when testing hosts_max_try with a message that was older than the host's
7042 retry timeout. (The host should not be counted, thus allowing all hosts to
7043 be tried at least once before bouncing.) This may have been the cause of an
7044 occasionally reported bug whereby a message would remain on the queue
7045 longer than the retry timeout, but would be bounced if a delivery was
7046 forced. I say "may" because I never totally pinned down the problem;
7047 setting up timeout/retry tests is difficult. See also the next item.
7049 79. The ultimate address timeout was not being applied to errors that involved
7050 a combination of host plus message (for example, a timeout on a MAIL
7051 command). When an address resolved to a number of possible hosts, and they
7052 were not all tried for each delivery (e.g. because of hosts_max_try), a
7053 message could remain on the queue longer than the retry timeout.
7055 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7058 81. Fixed an obscure SMTP outgoing bug which required at least the following
7059 conditions: (a) there was another message waiting for the same server;
7060 (b) the server returned 5xx to all RCPT commands in the first message so
7061 that the message was not completed; (c) the server dropped the connection
7062 or gave a negative response to the RSET that Exim sends to abort the
7063 transaction. The observed case was a dropped connection after DATA that had
7064 been sent in pipelining mode. That is, the server had advertised PIPELINING
7065 but was not implementing it correctly. The effect of the bug was incorrect
7066 behaviour, such as trying another host, and this could lead to a crash.
7072 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7073 and daemon.c were passed as pointers to ints; they should have been
7074 pointers to socklen_t variables (which are typically unsigned ints).
7076 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7079 3. Fixed a really odd bug that affected only the testing scheme; patching a
7080 certain fixed string in the binary changed the value of another string that
7081 happened to be identical to the end of the original first string.
7083 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7084 name", it returns that address as the IP address. On some operating
7085 systems (e.g. Solaris), it also passes back the IP address string as the
7086 "host name". However, on others (e.g. Linux), it passes back an empty
7087 string. Exim wasn't checking for this, and was changing the host name to an
7088 empty string, assuming it had been canonicalized.
7090 5. Although rare, it is permitted to have more than one PTR record for a given
7091 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7092 all the names associated with an address, because they do in Solaris.
7093 However, it seems that they do not in Linux for data that comes from the
7094 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7095 I found this out when I moved to a new Linux workstation and tried to run
7096 the Exim test suite.
7098 To get round this problem I have changed the code so that it now does its
7099 own call to the DNS to look up PTR records when searching for a host name.
7100 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7101 addresses that are only in /etc/hosts are still found.
7103 This behaviour is, however, controlled by an option called host_lookup_
7104 order, which defaults to "bydns:byaddr". If people want to use the other
7105 order, or indeed, just use one or the other means of lookup, they can
7106 specify it in this variable.
7108 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7109 some operating systems, this comes back from gethostbyaddr() as an empty
7110 string, and this is what Exim used to test for. However, it seems that in
7111 other systems, "." is yielded. Exim now tests for this case too.
7113 7. The values of check_spool_space and check_log_space are now held internally
7114 as a number of kilobytes instead of an absolute number of bytes. If a
7115 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7116 kilobyte. This means that much larger values can be stored.
7118 8. Exim monitor: an attempt to get the action menu when not actually pointing
7119 at a message produces an empty menu entitled "No message selected". This
7120 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7121 no entries in it ("Shell widget menu has zero width and/or height"). So I
7122 have added a single, blank menu entry in this case.
7124 9. Added ${quote_local_part.
7126 10. MIME decoding is now applied to the contents of Subject: header lines when
7129 11. Now that a reference to $sender_host_address automatically causes a reverse
7130 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7131 host lookup before query-style lookups in lists that might use this
7132 variable. This has therefore been abolished, and the "net-" prefix is no
7133 longer necessary for query-style lookups.
7135 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7136 have been a typo for LFLAGS, so it has been changed.
7138 13. The install script calls Exim with "-C /dev/null" in order to find the
7139 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7140 to be output. However, since Exim outputs its version number before the
7141 error, it didn't break the script. It just looked ugly. I fixed this by
7142 always allowing "-C /dev/null" if the caller is root.
7144 14. Ignore overlarge ACL variable number when reading spool file - insurance
7145 against a later release with more variables having written the file.
7147 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7148 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7149 this, as well as the form without the "IPv6" on the front.
7151 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7152 OS/Makefile-Darwin file.
7154 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7155 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7157 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7158 variable. However, some OS use ints for inodes. I've added cast to long int
7159 to get rid of the compiler warning.
7161 19. I had forgotten to lock out "/../" in configuration file names when
7162 ALT_CONFIG_PREFIX was set.
7164 20. Routers used for verification do not need to specify transports. However,
7165 if such a router generated a host list, and callout was configured, Exim
7166 crashed, because it could not find a port number from the (non-existent)
7167 transport. It now assumes port 25 in this circumstance.
7169 21. Added the -t option to exigrep.
7171 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7172 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7173 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7174 them is set, LOOKUP_LSEARCH is forced.
7176 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7178 24. Added sender and host information to the "rejected by local_scan()" log
7179 line; previously there was no indication of these.
7181 25. Added .include_if_exists.
7183 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7184 Linux. It turns out that not all file systems support this. Apparently some
7185 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7186 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7187 not supported on the file descriptor, is now ignored when Exim is trying to
7188 sync a directory. This applies only to Linux.
7190 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7192 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7193 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7196 29. Added support for maildirsize files from supplied patch (modified a bit).
7198 30. The use of :fail: followed by an empty string could lead Exim to respond to
7199 sender verification failures with (e.g.):
7201 550 Verification failed for <xxx>
7202 550 Sender verify failed
7204 where the first response line was missing the '-' that indicates it is not
7205 the final line of the response.
7207 31. The loop for finding the name of the user that called Exim had a hardwired
7208 limit of 10; it now uses the value of finduser_retries, which is used for
7209 all other user lookups.
7211 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7214 33. Exim was neglecting to zero errno before one call of strtol() when
7215 expanding a string and expecting an integer value. On some systems this
7216 resulted in spurious "integer overflow" errors. Also, it was casting the
7217 result into an int without checking.
7219 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7220 did not work. The code looks as if it has *never* worked, though it appears
7221 to have been documented since at least release 1.62. I have made it work.
7223 35. The "timeout_DNS" error in retry rules, also documented since at least
7224 1.62, also never worked. As it isn't clear exactly what this means, and
7225 clearly it isn't a major issue, I have abolished the feature by treating it
7226 as "timeout", and writing a warning to the main and panic logs.
7228 36. The display of retry rules for -brt wasn't always showing the error code
7231 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7232 timeout_connect_A, timeout_connect_MX.
7234 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7235 to the empty sender.
7237 39. The daemon was not analysing the content of -oX till after it had closed
7238 stderr and disconnected from the controlling terminal. This meant that any
7239 syntax errors were only noted on the panic log, and the return code from
7240 the command was 0. By re-arranging the code a little, I've made the
7241 decoding happen first, so such errors now appear on stderr, and the return
7242 code is 1. However, the actual setting up of the sockets still happens in
7243 the disconnected process, so errors there are still only recorded on the
7246 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7247 connections (as happens on some IP stacks) was logged at start up time as
7248 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7249 it from "IPv6 and IPv4", which means that two separate sockets are being
7252 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7253 says whether AF_INET or AF_INET6 was passed as an argument.
7255 42. Exiwhat output was messed up when time zones were included in log
7258 43. Exiwhat now gives more information about the daemon's listening ports,
7259 and whether -tls-on-connect was used.
7261 44. The "port" option of the smtp transport is now expanded.
7263 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7264 silently ignored. Now an error message is written to the main and panic
7267 46. There's a new ACL modifier called "logwrite" which writes to a log file
7268 as soon as it is encountered.
7270 47. Added $local_user_uid and $local_user_gid at routing time.
7272 48. Exim crashed when trying to verify a sender address that was being
7275 49. Exim was recognizing only a space character after ".include". It now also
7276 recognizes a tab character.
7278 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7279 extracting the relevant information from the specification. The man page no
7280 longer contains scrambled data for the -d option, and I've added a section
7281 at the front about calling Exim under different names.
7283 51. Added "extra_headers" argument to the "mail" command in filter files.
7285 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7288 53. Installed eximstats 1.29.
7290 54. Added transport_filter_timeout as a generic transport option.
7292 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7293 Cc: header lines. This was required by RFC 822, but it not required by RFC
7296 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7297 incoming messages that did not have them. Now it does so only if the
7298 message originates locally, that is, if there is no associated remote host
7299 address. When Resent- header lines are present, this applies to the Resent-
7300 lines rather than the non-Resent- lines.
7302 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7303 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7305 58. Messages for configuration errors now include the name of the main
7306 configuration file - useful now that there may be more than one file in a
7307 list (.included file names were always shown).
7309 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7310 for those rare installations that do not start the daemon as root or run it
7311 setuid root. I've cut out the call to initgroups() if the daemon is not
7314 60. The Exim user and group can now be bound into the binary as text strings
7315 that are looked up at the start of Exim's processing.
7317 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7319 62. Added $mailstore_basename variable.
7321 63. Installed patch to sieve.c from Michael Haardt.
7323 64. When Exim failed to open the panic log after failing to open the main log,
7324 the original message it was trying to log was written to stderr and debug
7325 output, but if they were not available (the usual case in production), it
7326 was lost. Now it is written to syslog before the two lines that record the
7327 failures to open the logs.
7329 65. Users' Exim filters run in subprocesses under the user's uid. It is
7330 possible for a "deliver" command or an alias in a "personal" command to
7331 provoke an address rewrite. If logging of address rewriting is configured,
7332 this fails because the process is not running as root or exim. There may be
7333 a better way of dealing with this, but for the moment (because 4.30 needs
7334 to be released), I have disabled address rewrite logging when running a
7335 filter in a non-root, non-exim process.
7341 1. The buildconfig auxiliary program wasn't quoting the value set for
7342 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7343 not defined. This bug was masked in 4.22 by the effect that was fixed in
7346 2. Some messages that were rejected after a message id was allocated were
7347 shown as "incomplete" by exigrep. It no longer does this for messages that
7348 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7350 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7351 have allow_domain_literals set, the ID did not get logged in the <= line.
7352 Domain literals are now always recognized in Message-ID: header lines.
7354 4. The first argument for a ${extract expansion item is the key name or field
7355 number. Leading and trailing spaces in this item were not being ignored,
7356 causing some misleading effects.
7358 5. When deliver_drop_privilege was set, single queue runner processes started
7359 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7360 same command in the process it spins off) were not dropping privilege.
7362 6. When the daemon running as "exim" started a queue runner, it always
7363 re-executed Exim in the spun-off process. This is a waste of effort when
7364 deliver_drop_privilege is set. The new process now just calls the
7365 queue-runner function directly.
7371 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7374 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7375 ignored. Though the use of -oP was forcing the writing of a pid file, it
7376 was always written to the default place.
7378 3. If the message "no IP address found for host xxxx" is generated during
7379 incoming verification, it is now followed by identification of the incoming
7380 connection (so you can more easily find what provoked it).
7382 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7384 5. Added some features to "harden" Exim a bit more against certain attacks:
7386 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7387 be put in Local/Makefile. This is like the never_users runtime option,
7388 but it cannot be overridden. The default setting is "root".
7390 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7391 prefix string with which any file named in a -C command line option
7394 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7395 is retained for -C and -D only if the caller of Exim is root. Without
7396 it, the exim user may also use -C and -D and retain privilege.
7398 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7399 command line option is disabled.
7401 6. Macro names set by the -D option must start with an upper case letter, just
7402 like macro names defined in the configuration file.
7404 7. Added "dereference=" facility to LDAP.
7406 8. Two instances of the typo "uknown" in the source files are fixed.
7408 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7409 the Configure-Makefile script screwed up while processing it.
7411 10. Incorporated PCRE 4.4.
7413 11. The SMTP synchronization check was not operating right at the start of an
7414 SMTP session. For example, it could not catch a HELO sent before the client
7415 waited for the greeting. There is now a check for outstanding input at the
7416 point when the greeting is written. Because of the duplex, asynchronous
7417 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7418 way, but not yet received, when the check is performed.
7420 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7421 on TCP/IP sockets, because this apparently causes some broken clients to
7424 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7425 unchanged) from the Cygwin maintainer.
7427 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7428 was supported instead of testing for mbx. Effectively a typo.
7430 15. The spa authenticator server code was not checking that the input it
7431 received was valid base64.
7433 16. The debug output line for the "set" modifier in ACLs was not showing the
7434 name of the variable that was being set.
7436 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7438 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7439 lookup on its own. Something else was needed to trigger the lookup. For
7440 example, a match in host_lookup or the need for a host name in a host list.
7441 Now, if $sender_host_name is referenced and the host name has not yet been
7442 looked up, a lookup is performed. If the lookup fails, the variable remains
7443 empty, and $host_lookup_failed is set to "1".
7445 19. Added "eqi" as a case-independent comparison operator.
7447 20. The saslauthd authentication condition could segfault if neither service
7448 nor realm was specified.
7450 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7451 error message that was logged was misleading, and incoming SMTP
7452 connections were dropped. The message is now more accurate, and temporary
7453 errors are given to SMTP connections.
7455 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7456 (except RSET) with a 421 error, until QUIT is received. However, it was
7457 failing to send a response to QUIT.
7459 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7460 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7461 ACL is now run after the helo_try_verify_hosts code.
7463 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7464 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7465 case-independent, so other case variants are also recognized). Apparently
7466 some systems use these upper case variants.
7468 25. If more than two messages were waiting for the same host, and a transport
7469 filter was specified for the transport, Exim sent two messages over the
7470 same TCP/IP connection, and then failed with "socket operation on non-
7471 socket" when it tried to send the third.
7473 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7475 27. The extern definition of crypt16() in expand.c was not being excluded when
7476 the OS had its own crypt16() function.
7478 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7479 as a preferred synonym for return_size_limit, both as an option and as an
7482 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7484 30. Changed the default configuration ACL to relax the local part checking rule
7485 for addresses that are not in any local domains. For these addresses,
7486 slashes and pipe symbols are allowed within local parts, but the sequence
7487 /../ is explicitly forbidden.
7489 31. SPA server authentication was not clearing the challenge buffer before
7492 32. log_message in a "warn" ACL statement was writing to the reject log as
7493 well as to the main log, which contradicts the documentation and doesn't
7494 seem right (because no rejection is happening). So I have stopped it.
7496 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7497 However, I am unable to do any testing of this.
7499 34. Fixed an infelicity in the appendfile transport. When checking directories
7500 for a mailbox, to see if any needed to be created, it was accidentally
7501 using path names with one or more superfluous leading slashes; tracing
7502 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7504 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7505 contents are added to the log line that is written for every discarded
7506 recipient. (Previously a log_message setting was ignored.)
7508 36. The ${quote: operator now quotes the string if it is empty.
7510 37. The install script runs exim in order to find its version number. If for
7511 some reason other than non-existence or emptiness, which it checks, it
7512 could not run './exim', it was installing it with an empty version number,
7513 i.e. as "exim-". This error state is now caught, and the installation is
7516 38. An argument was missing from the function that creates an error message
7517 when Exim fails to connect to the socket for saslauthd authentication.
7518 This could cause Exim to crash, or give a corrupted message.
7520 39. Added isip, isip4, and isip6 to ${if conditions.
7522 40. The ACL variables $acl_xx are now saved with the message, and can be
7523 accessed later in routers, transports, and filters.
7525 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7526 strings in the file are not string-expanded.
7528 42. If a MAIL command specified a SIZE value that was too large to fit into an
7529 int variable, the check against message_size_limit failed. Such values are
7530 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7531 day this will have to be increased, but I don't think I want to be around
7532 when emails are that large.
7539 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7540 iconv() is not standard in FreeBSD.
7542 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7543 IPv6 enabled. The observed symptom was a segmentation fault on return from
7544 the function os_common_find_running_interfaces() in src/os.c.
7546 3. In the check_special_case() function in daemon.c I had used "errno" as an
7547 argument name, which causes warnings on some systems. This was basically a
7548 typo, since it was named "eno" in the comments!
7550 4. The code that waits for the clock to tick (at a resolution of some fraction
7551 of a second) so as to ensure message-id uniqueness was always waiting for
7552 at least one whole tick, when it could have waited for less. [This is
7553 almost certainly not relevant at current processor speeds, where it is
7554 unlikely to ever wait at all. But we try to future-proof.]
7556 5. The function that sleeps for a time interval that includes fractions of a
7557 second contained a race. It did not block SIGALRM between setting the
7558 timer, and suspending (a couple of lines later). If the interval was short
7559 and the sigsuspend() was delayed until after it had expired, the suspension
7560 never ended. On busy systems this could lead to processes getting stuck for
7563 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7564 process, before it forks any delivery processes. The open lookup caching
7565 mechanism meant that the open file or database connection was passed into
7566 the delivery process. The problem was that delivery processes always tidy
7567 up cached lookup data. This could cause a problem for the next delivery
7568 process started by the queue runner, because the external queue runner
7569 process does not know about the closure. So the next delivery process
7570 still has data in the lookup cache. In the case of a file lookup, there was
7571 no problem because closing a file descriptor in a subprocess doesn't affect
7572 the parent. However, if the lookup was caching a connection to a database,
7573 the connection was closed, and the second delivery process was likely to
7574 see errors such as "PGSQL: query failed: server closed the connection
7575 unexpectedly". The problem has been fixed by closing all cached lookups
7576 in a queue runner before running a delivery process.
7578 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7579 seem to have the "const" qualifier which it has on other OS. I've
7582 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7583 *of the same type* (client or server) with the same public name that an
7584 error should be diagnosed.
7586 9. When Exim looked up a host name for an IP address, but failed to find the
7587 original IP address when looking up the host name (a safety check), it
7588 output the message "<ip address> does not match any IP for NULL", which was
7589 confusing, to say the least. The bug was that the host name should have
7590 appeared instead of "NULL".
7592 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7593 user that is built into the binary, and the -C or -D options is used, root
7594 privilege is dropped before the configuration file is read. In addition,
7595 logging is switched to stderr instead of the normal log files. If the
7596 configuration then re-defines the Exim user, the unprivileged environment
7597 is probably not what is expected, so Exim logs a panic warning message (but
7600 However, if deliver_drop_privilege is set, the unprivileged state may well
7601 be exactly what is intended, so the warning has been cut out in that case,
7602 and Exim is allowed to try to write to its normal log files.
7608 1. smtp_return_error_details was not giving details for temporary sender
7609 or receiver verification errors.
7611 2. Diagnose a configuration error if two authenticators have the same public
7614 3. Exim used not to create the message log file for a message until the first
7615 delivery attempt. This could be confusing when incoming messages were held
7616 for policy or load reasons. The message log file is now created at the time
7617 the message is received, and an initial "Received" line is written to it.
7619 4. The automatically generated man page for command line options had a minor
7620 bug that caused no ill effects; however, a more serious problem was that
7621 the procedure for building the man page automatically didn't always
7622 operate. Consequently, release 4.20 contains an out-of-date version. This
7623 shouldn't happen again.
7625 5. When building Exim with embedded Perl support, the script that builds the
7626 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7627 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7629 6. The freeze_tell option was not being used for messages that were frozen on
7630 arrival, either by an ACL or by local_scan().
7632 7. Added the smtp_incomplete_transaction log selector.
7634 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7635 was accepting AUTH without a new EHLO.
7637 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7638 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7641 10. Logging of TCP/IP connections (when configured) now happens in the main
7642 daemon process instead of the child process, so that the TCP/IP connection
7643 count is more accurate (but it can never be perfect).
7645 11. The use of "drop" in a nested ACL was not being handled correctly in the
7646 outer ACL. Now, if condition failure induced by the nested "drop" causes
7647 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7648 or "require"), the connection is dropped.
7650 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7651 that yield "discard" can now be used with an "accept" or a "discard" verb,
7652 but an error is generated for any others (because I can't see a useful way
7653 to define what should happen).
7655 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7656 are now processed in the same way as lines in the Exim configuration file.
7657 In particular, continuation lines are supported.
7659 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7661 15. Added -ti meaning -t -i.
7663 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7664 domains, and warn by logging if others are found.
7666 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7667 returned from GIFCONF when Exim is trying to find the list of interfaces on
7668 a host. The code in os.c has been modified to copy each ifreq to an aligned
7669 structure in all cases.
7671 Also, in some cases, the returned ifreq's were being copied to a 'struct
7672 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7673 means the last couple of bytes of an IPv6 address could be chopped if the
7674 ifreq contained only a normal sockaddr (14 bytes storage).
7676 18. Named domain lists were not supported in the hosts_treat_as_local option.
7677 An entry such as +xxxx was not recognized, and was treated as a literal
7680 19. Ensure that header lines added by a DATA ACL are included in the reject log
7681 if the ACL subsequently rejects the message.
7683 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7684 MD5 (which is deprecated).
7686 21. When testing a filter file using -bf, Exim was writing a message when it
7687 took the sender from a "From " line in the message, but it was not doing so
7688 when it took $return_path from a Return-Path: header line. It now does.
7690 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7691 with a valid header line field name (a series of printing characters
7692 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7694 23. Changed "disc" in the source to "disk" to conform to the documentation and
7695 the book and for uniformity.
7697 24. Ignore Sendmail's -Ooption=value command line item.
7699 25. When execve() failed while trying to run a command in a pipe transport,
7700 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7701 could be confused with a return value of 69 from the command itself. This
7702 has been changed to 127, the value the shell returns if it is asked to run
7703 a non-existent command. The wording for the related log line suggests a
7704 non-existent command as the problem.
7706 26. If received_header_text expands to an empty string, do not add a Received:
7707 header line to the message. (Well, it adds a token one on the spool, but
7708 marks it "old" so that it doesn't get used or transmitted.)
7710 27. Installed eximstats 1.28 (addition of -nt option).
7712 28. There was no check for failure on the call to getsockname() in the daemon
7713 code. This can fail if there is a shortage of resources on the system, with
7714 ENOMEM, for example. A temporary error is now given on failure.
7716 29. Contrary to the C standard, it seems that in some environments, the
7717 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7718 program. Exim now does this explicitly; it affects the formatting of
7719 timestamps using strftime().
7721 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7722 complaints. I've now initialized all the variables, to avoid this.
7724 32. Header lines added by a system filter were not being "seen" during
7725 transport-time rewrites.
7727 33. The info_callback() function passed to OpenSSL is set up with type void
7728 (*)(SSL *, int, int), as described somewhere. However, when calling the
7729 function (actually a macro) that sets it up, the type void(*)() is
7730 expected. I've put in a cast to prevent warnings from picky compilers.
7732 34. If a DNS black list lookup found a CNAME record, but there were no A
7733 records associated with the domain it pointed at, Exim crashed.
7735 35. If a DNS black list lookup returned more than one A record, Exim ignored
7736 all but the first. It now scans all returned addresses if a particular IP
7737 value is being sought. In this situation, the contents of the
7738 $dnslist_value variable are a list of all the addresses, separated by a
7741 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7742 to accept a host name and all its aliases if the forward lookup for any of
7743 them yielded the IP address of the incoming connection. Now it accepts only
7744 those names whose forward lookup yields the correct IP address. Any other
7745 names are discarded. This closes a loophole whereby a rogue DNS
7746 administrator could create reverse DNS records to break through a
7747 wildcarded host restriction in an ACL.
7749 37. If a user filter or a system filter that ran in a subprocess used any of
7750 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7751 the wrong values were passed to the pipe command ($thisaddress had the
7752 value of $0, $0 had the value of $1, etc). This bug was introduced by
7753 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7755 38. Improved the line breaking for long SMTP error messages from ACLs.
7756 Previously, if there was no break point between 40 and 75 characters, Exim
7757 left the rest of the message alone. Two changes have been made: (a) I've
7758 reduced the minimum length to 35 characters; (b) if it can't find a break
7759 point between 35 and 75 characters, it looks ahead and uses the first one
7760 that it finds. This may give the occasional overlong line, but at least the
7761 remaining text gets split now.
7763 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7764 file descriptors might be low, and that setting 1000 would always raise it.
7765 It turns out that in some environments, the limit is already over 1000 and
7766 that lowering it causes trouble. So now Exim takes care not to decrease it.
7768 40. When delivering a message, the value of $return_path is set to $sender_
7769 address at the start of routing (routers may change the value). By an
7770 oversight, this default was not being set up when an address was tested by
7771 -bt or -bv, which affected the outcome if any router or filter referred to
7774 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7775 log only when "message" or "log_message" are set. However, if one of the
7776 conditions was an address verification, or a call to a nested ACL, the
7777 messages generated by the underlying test were being passed through. This
7778 no longer happens. The underlying message is available in $acl_verify_
7779 message for both "message" and "log_message" expansions, so it can be
7780 passed through if needed.
7782 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7783 new expansion $bh_ to give the encoded byte string without charset
7784 translation. Translation happens only if iconv() is available; HAVE_ICONV
7785 indicates this at build time. HEADERS_CHARSET gives the charset to
7786 translate to; headers_charset can change it in the configuration, and
7787 "headers charset" can change it in an individual filter file.
7789 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7790 that creates RFC 2047 encoded "words" labels them as that charset instead
7791 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7792 expansion operator; (ii) when Exim creates a From: line for a local
7793 message; (iii) when a header line is rewritten to include a "phrase" part.
7795 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7796 buggy, causing it to skip the first lines of messages whose message ID
7797 ended in 'D'. This would not have bitten before Exim release 4.14, because
7798 message IDs were unlikely to end in 'D' before then. The effect was to have
7799 incorrect size information for certain domains.
7801 45. #include "config.h" was missing at the start of the crypt16.c module. This
7802 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7805 46. If there was a timeout during a "random" callout check, Exim treated it as
7806 a failure of the random address, and carried on sending RSET and the real
7807 address. If the delay was just some slowness somewhere, the response to the
7808 original RCPT would be taken as a response to RSET and so on, causing
7809 mayhem of various kinds.
7811 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7812 when I implemented it. It didn't allow for the fact that some option values
7813 may legitimately be negative (e.g. size_addition), and it didn't even do
7814 the right test for positive values.
7816 48. Domain names in DNS records are case-independent. Exim always looks them up
7817 in lower case. Some resolvers return domain names in exactly the case they
7818 appear in the zone file, that is, they may contain uppercase letters. Not
7819 all resolvers do this - some return always lower case. Exim was treating a
7820 change of case by a resolver as a change of domain, similar to a widening
7821 of a domain abbreviation. This triggered its re-routing code and so it was
7822 trying to route what was effectively the same domain again. This normally
7823 caused routing to fail (because the router wouldn't handle the domain
7824 twice). Now Exim checks for this case specially, and just changes the
7825 casing of the domain that it ultimately uses when it transmits the message
7828 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7831 50. If a filter generated a file delivery with a non-absolute name (possible if
7832 no home directory exists for the router), the forbid_file option was not
7835 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7836 the existing equality matching.
7838 52. Exim was using ints instead of ino_t variables in some places where it was
7839 dealing with inode numbers.
7841 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7842 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7843 called TMPDIR, and if it finds it is different, it changes its value.
7845 54. The smtp_printf() function is now made available to local_scan() so
7846 additional output lines can be written before returning. There is also an
7847 smtp_fflush() function to enable the detection of a dropped connection.
7848 The variables smtp_input and smtp_batched_input are exported to
7851 55. Changed the default runtime configuration: the message "Unknown user"
7852 has been removed from the ACL, and instead placed on the localuser router,
7853 using the cannot_route_message feature. This means that any verification
7854 failures that generate their own messages won't get overridden. Similarly,
7855 the "Unrouteable address" message that was in the ACL for unverifiable
7856 relay addresses has also been removed.
7858 56. Added hosts_avoid_esmtp to the smtp transport.
7860 57. The exicyclog script was not checking for the esoteric option
7861 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7862 will work only if exicyclog is run under the appropriate euid.
7864 58. Following a discussion on the list, the rules by which Exim recognises line
7865 endings on incoming messages have been changed. The -dropcr and drop_cr
7866 options are now no-ops, retained only for backwards compatibility. The
7867 following line terminators are recognized: LF CRLF CR. However, special
7868 processing applies to CR:
7870 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7871 nor a local message in the state where . is a terminator.
7873 (ii) If a bare CR is encountered in a header line, an extra space is added
7874 after the line terminator so as not to end the header. The reasoning
7875 behind this is that bare CRs in header lines are most likely either
7876 to be mistakes, or people trying to play silly games.
7878 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7879 was being incorrectly given as 18 bytes larger than it should have been.
7880 This is a VOB (very old bug).
7882 60. This may never have affected anything current, but just in case it has:
7883 When the local host is found other than at the start of a list of hosts,
7884 the local host, those with the same MX, and any that follow, are discarded.
7885 When the list in question was part of a longer list of hosts, the following
7886 hosts (not currently being processed) were also being discarded. This no
7887 longer happens. I'm not sure if this situation could ever has previously
7890 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7893 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7894 header. This is something that is recommended in a new Internet Draft, and
7895 is something that is documented as being done by Sendmail. There are two
7896 possible values. For messages generated by the autoreply transport, Exim
7899 Auto-Submitted: auto-replied
7901 whereas for all other generated messages (e.g. bounces) it adds
7903 Auto-Submitted: auto-generated
7905 63. The "personal" condition in filters now includes a test for the
7906 Auto-Submitted: header. If it contains the string "auto-" the message it
7907 not considered personal.
7909 64. Added rcpt_include_affixes as a generic transport option.
7911 65. Added queue_only_override (default true).
7913 66. Added the syslog_duplication option.
7915 67. If what should have been the first header line of a message consisted of
7916 a space followed by a colon, Exim was mis-interpreting it as a header line.
7917 It isn't of course - it is syntactically invalid and should therefore be
7918 treated as the start of the message body. The misbehaviour could have
7919 caused a number of strange effects, including loss of data in subsequent
7920 header lines, and spool format errors.
7922 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7923 client host had authenticated. This control can now be exercised by an ACL
7924 for more flexibility.
7926 69. By default, callouts do not happen when testing with -bh. There is now a
7927 variant, -bhc, which does actually run the callout code, including
7928 consulting and updating the callout cache.
7930 70. Added support for saslauthd authentication, courtesy of Alexander
7933 71. If statvfs() failed on the spool or log directories while checking their
7934 size for availability, Exim confusingly gave the error "space shortage".
7935 Furthermore, in debugging mode it crashed with a floating point exception.
7936 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7937 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7938 serious problem, Exim now writes to the main and panic logs when this
7939 happens, with details of the failure. It then refuses to accept the
7940 incoming message, giving the message "spool directory problem" or "log
7941 directory problem" with a 421 code for SMTP messages.
7943 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7944 0, 1, and 2 exist, because some OS complain for execs without them (see
7945 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7946 descriptors. However, the code omitted to check that the open succeeded,
7947 causing mysterious errors if for some reason the permissions on /dev/null
7948 got screwed. Now Exim writes a message to the main and panic logs, and
7949 bombs out if it can't open /dev/null.
7951 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7952 interact so that it is all more flexible. It is supposed to remain
7953 backwards compatible. Also added extra_local_interfaces.
7955 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7956 to bomb out with an assertion failure - to the client this appears as a
7957 connection drop. This problem occurs in the part of the code that was taken
7958 from the Samba project. Fortunately, the assertion is in a very simple
7959 function, so I have fixed this by reproducing the function inline in the
7960 one place where it is called, and arranging for authentication to fail
7961 instead of killing the process with assert().
7963 75. The SPA client code was not working when the server requested OEM rather
7964 than Unicode encoding.
7966 76. Added code to make require_files with a specific uid setting more usable in
7967 the case where statting the file as root fails - usually a non-root-mounted
7968 NFS file system. When this happens and the failure is EACCES, Exim now
7969 forks a subprocess and does the per-uid checking as the relevant uid.
7971 77. Added process_log_path.
7973 78. If log_file_path was not explicitly set, a setting of check_log_space or
7974 check_log_inodes was ignored.
7976 79. If a space check for the spool or log partitions fails, the incident is now
7977 logged. Of course, in the latter case the data may get lost...
7979 80. Added the %p formatting code to string_format() so that it can be used to
7980 print addresses in debug_print(). Adjusted all the address printing in the
7981 debugging in store.c to use %p rather than %d.
7983 81. There was a concern that a line of code in smtp_in.c could overflow a
7984 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7985 initially expressed, the concern was not well-founded, because trailing
7986 spaces are removed early. However, if the trailing spaces were followed by
7987 a NULL, they did not get removed, so the overflow was possible. Two fixes
7990 (a) I re-wrote the offending code in a cleaner fashion.
7991 (b) If an incoming SMTP command contains a NULL character, it is rejected
7994 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7995 runs initgroups(), so that if the Exim user is in any additional groups,
7996 they will be used during message reception.
8002 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob