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.
212 JH/01 Avoid costly startup code when not strictly needed. This reduces time
213 for some exim process initialisations. It does mean that the logging
214 of TLS configuration problems is only done for the daemon startup.
216 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
218 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
219 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
222 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
223 DKIM signing is being done. Previously a transport_filter would always
224 disable CHUNKING, falling back to traditional DATA.
226 JH/05 Regard command-line receipients as tainted.
228 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
230 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
231 PAM library frees one of the arguments given to it, despite the
232 documentation. Therefore a plain malloc must be used.
234 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
235 on-stack buffers were used, resulting in a taint trap when DSN information
236 copied from a received message was written into the buffer.
238 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
239 the ordering of its ARC headers. This caused a crash.
241 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
242 a new record was being constructed with information from the peer, a trap
245 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
246 installation would get error messages from DMARC verify, when it hit the
247 nonexistent file indicated by the default. Distros wanting DMARC enabled
248 should both provide the file and set the option.
249 Also enforce no DMARC verification for command-line sourced messages.
251 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
252 could, depending on the platform, hang at the STARTTLS response.
254 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
255 message on a connection. Previously if one message had ARC headers and
256 the following one did not, a crash could result when adding an
257 Authentication-Results: header.
259 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
260 Exim main code and Exim-related utities. The introduction of taint
261 tracking also did many adjustments to string handling. Since then, eximon
262 frequently terminated with an assert failure.
264 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
265 check for 452 responses. This slightly helps the inefficieny of doing
266 a large alias-expansion into a recipient-limited target. The max_rcpt
267 transport option still applies (and at the current default, will override
268 the new feature). The check is done for either cause of synch, and forces
269 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
270 connection. The new facility is not tunable at this time.
272 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
273 library live data was being used, so the results became garbage. Make
274 copies while it is still usable.
276 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
277 on delivery deferred (==) and failed (**) lines (if a delivery was
278 attemtped). Previously it was only on completion (=>) lines.
280 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
281 for the expansion of the server_scram_iter and server_scram_salt options.
283 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
284 are now specifically given a NO_DATA response without hitting the system
285 resolver. The library goes on to do the now-standard TXT lookup.
286 Use of dnsdb lookups is not affected.
288 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
289 only retrieve the errormessage once. Previously two calls to dlerror()
290 were used, and the second one (for mainlog/paniclog) retrieved null
293 JH/20 Taint checking: disallow use of tainted data for
294 - the appendfile transport file and directory options
295 - the pipe transport command
296 - the autoreply transport file, log and once options
297 - file names used by the redirect router (including filter files)
299 - paths used by single-key lookups
300 Previously this was permitted.
302 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
303 adjusted the size of a major service buffer; this failed because the
304 buffer was in use at the time. Change to a compile-time increase in the
305 buffer size, when this authenticator is compiled into exim.
307 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
308 previous fast-mode was untenable in the face of glibs using mmap to
309 support larger malloc requests.
311 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
312 New values supported, if defined on system where compiled:
313 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
314 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
316 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
317 running a limited number of proceses in parallel, a benefit is gained. The
318 amount varies with the platform hardware and load. The use of the option
319 queue_run_in_order means we cannot do this, as ordering becomes
322 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
323 had introduced a string-copy (for ensuring NUL-termination) which was not
324 appropriate for that case, which can include embedded NUL bytes in the
325 block of data. Investigation showed the copy to actually be needless, the
326 data being length-specified.
328 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
329 done during a receiving connection, and both used TLS, global info was
330 used rather than per-connection info for tracking the state of data
331 queued for transmission. This could result in a connection hang.
333 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
334 Previously, when delivering serveral messages down a single connection
335 only the first would provide a SIZE. This was due to the size information
336 not being properly tracked.
338 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
339 TAI (at 37 seconds currently), pretend to be in UTC for time-related
340 expansion and logging. Previously, spurious values such as a future
341 minute could be seen.
343 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
344 it could crash from a null-deref. This could also affect the
345 ${addresses: } operator and ${readsock } item.
347 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
348 message following a mime one, the variable was not reset.
350 JH/30 When an pipelined-connect fails at the first response, assume incorrect
351 cached capability (perhaps the peer reneged?) and immediately retry in
354 JH/31 Fix spurious detection of timeout while writing to transport filter.
356 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
357 an attempt to copy the string was made before checking it.
359 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
360 taint of the lookup key was maintained; we now regard the presence in the
361 filesystem as sufficient validation.
363 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
364 argument is supplied.
366 JH/35 The dsearch lookup now requires that the directory is an absolute path.
367 Previously this was not checked, and nonempty relative paths made an
368 access under Exim's current working directory.
370 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
371 Previously no event was raised.
373 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
374 parameter supplied by the sender MAIL FROM command. Previously it was
375 ignored, and only the check_spool_space option value for the required
378 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
379 the size of the signing public-key. Previously it was instead giving
380 the size of the signature hash.
382 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
383 the default. See the (new) dkim_verify_min_keysizes option.
385 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
386 and an ACL use a lookup for checking either the local_part or domain,
387 stale data could be accessed. Ensure that variable references are
388 dropped between messages.
390 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
391 by the client was not checked as pointing within response data before
392 being used. A malicious client could thus cause an out-of-bounds read and
393 possibly gain authentication. Fix by adding the check.
395 JH/42 Internationalisation: change the default for downconversion in the smtp
396 transport to be "if needed". Previously it was "as previously set" for
397 the message, which usually meant "if needed" for message-submission but
398 "no" for everything else. However, MTAs have been seen using SMTPUTF8
399 even when the envelope addresses did not need it, resulting in forwarding
400 failures to non-supporting MTAs. A downconvert in such cases will be
401 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
402 transport. The change does mean that addresses needing conversion will
403 be converted when previously a delivery failure would occur.
405 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
406 response was received it would be used unchecked in a fail-DSN, violating
407 standards on line-length limits. Truncate if needed.
409 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
410 doesn't work. (Additionally add a new main config option to configure the
417 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
418 side, to decode a TLS 1.3 packet capture.
420 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
421 Previously the default library behaviour applied, sending two, each in
424 JH/03 Debug output for ACL now gives the config file name and line number for
427 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
429 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
431 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
432 buffer overrun for (non-chunking) other transports.
434 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
435 TLS1.3, means that a server rejecting a client certificate is not visible
436 to the client until the first read of encrypted data (typically the
437 response to EHLO). Add detection for that case and treat it as a failed
438 TLS connection attempt, so that the normal retry-in-clear can work (if
439 suitably configured).
441 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
442 and/or domain. Found and fixed by Jason Betts.
444 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
445 configuration). If a CNAME target was not a wellformed name pattern, a
448 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
449 the OS reports them interleaved with other addresses.
451 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
452 used both for input and for a verify callout, both encrypted, SMTP
453 responses being sent by the server could be lost. This resulted in
454 dropped connections and sometimes bounces generated by a peer sending
457 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
458 string. Previously it was possible to cause undefined behaviour in a
459 library routine (usually a crash). Found by "zerons".
461 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
464 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
465 API was removed, so update to use the newer ones.
467 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
468 any timeout set, is taking a long time. Previously we would hang on to a
469 rotated logfile "forever" if the input was arriving with long gaps
470 (a previous attempt to fix addressed lack, for a long time, of initial
473 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
474 shared (NFS) environment. The length of the tempfile name is now
475 4 + 16 ("hdr.$message_exim_id") which might break on file
476 systems which restrict the file name length to lower values.
479 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
480 shared (NFS) environment.
482 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
483 did for all versions <4.90). Notably -M, -m, --invert, -I may be
486 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
487 on some platforms for bit 31.
489 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
490 to changes apparently associated with TLS1.3 handling some of the APIs
491 previously used were either nonfunctional or inappropriate. Strings
492 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
493 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
494 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
495 This affects log line X= elements, the $tls_{in,out}_cipher variables,
496 and the use of specific cipher names in the encrypted= ACL condition.
498 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
500 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
501 verification result was not updated unless hosts_require_ocsp applied.
503 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
504 queue_list_requires_admin set to false, non-admin users were denied the
507 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
508 directory-of-certs mode. Previously they were advertised despite the
511 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
512 A single TCP connection by a client will now hold a TLS connection open
513 for multiple message deliveries, by default. Previously the default was to
516 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
517 default. If built with the facility, DANE will be used. The facility
518 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
520 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
521 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
522 must be defined and you must still, unless you define DISABLE_TLS, manage
523 the the include-dir and library-file requirements that go with that
524 choice. Non-TLS builds are still supported.
526 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
529 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
530 default. If the platform supports and has the facility enabled, it will
531 be requested on all coneections.
533 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
534 controlled by the build-time option SUPPORT_PIPE_CONNECT.
536 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
538 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
539 success-DSN messages. Previously the From: header was always the default
540 one for these; the option was ignored.
542 JH/28 Fix the timeout on smtp response to apply to the whole response.
543 Previously it was reset for every read, so a teergrubing peer sending
544 single bytes within the time limit could extend the connection for a
545 long time. Credit to Qualsys Security Advisory Team for the discovery.
547 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
548 delivery address, which leaked information of the results of local
549 forwarding. Change to the original envelope recipient address, per
552 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
553 requested. Previously not bounce was generated and a log entry of
554 error ignored was made.
556 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
558 JH/32 Introduce a general tainting mechanism for values read from the input
559 channel, and values derived from them. Refuse to expand any tainted
560 values, to catch one form of exploit.
562 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
563 was unused and the unexpanded text used for the test. Found and
564 fixed by Ruben Jenster.
566 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
567 an attempt to use a TLS library read routine dereffed a nul pointer,
570 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
571 them in our smtp response.
573 JH/36 Have the general router option retry_use_local_part default to true when
574 any of the restrictive preconditions are set (to anything). Previously it
575 was only for check_local user. The change removes one item of manual
576 configuration which is required for proper retries when a remote router
577 handles a subset of addresses for a domain.
579 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
580 link count into consideration.
582 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
583 caused the extension of big_buffer, the following lines were ignored.
585 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
586 accordance with RFC 2308. Previously there was no expiry, so a longlived
587 receive process (eg. due to ACL delays) versus a short SOA value could
590 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
592 JH/39 Promote DMARC support to mainline.
594 JH/40 Bug 2452: Add a References: header to DSNs.
596 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
597 parameters. The relevant library call is documented as "Deprecated: This
598 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
599 3.6.0, DH parameters are negotiated following RFC7919."
601 HS/06 Change the default of dnssec_request_domains to "*"
603 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
604 carried on and emitted a BDAT command, even when PIPELINING was not
607 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
608 buffer was used for the filename, resulting in a trap when tainted
609 arguments (eg. $domain) were used.
611 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
612 recommended to avoid a possible server-load attack. The feature can be
613 re-enabled via the openssl_options main cofiguration option.
615 JH/45 local_scan API: documented the current smtp_printf() call. This changed
616 for version 4.90 - adding a "more data" boolean to the arguments.
617 Bumped the ABI version number also, this having been missed previously;
618 release versions 4.90 to 4.92.3 inclusive were effectively broken in
619 respect of usage of smtp_printf() by either local_scan code or libraries
620 accessed via the ${dlfunc } expansion item. Both will need coding
621 adjustment for any calls to smtp_printf() to match the new function
622 signature; a FALSE value for the new argument is always safe.
624 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
625 the file-offset (which the Linux syscall does, and exim expects); this
626 resulted in an indefinite loop.
628 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
629 to do ARC verification. The Authentication-Results: header line added
630 by the configuration then had no ARC item.
636 JH/01 Remove code calling the customisable local_scan function, unless a new
637 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
639 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
640 non-signal-safe functions being used.
642 JH/03 Bug 2269: When presented with a received message having a stupidly large
643 number of DKIM-Signature headers, disable DKIM verification to avoid
644 a resource-consumption attack. The limit is set at twenty.
646 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
647 report of oldest_pass in ${authres } in consequence, and separate out
648 some descriptions of reasons for verification fail.
650 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
651 files in the spool were present and unlocked. A queue-runner could spot
652 them, resulting in a duplicate delivery. Fix that by doing the unlock
653 after the unlink. Investigation by Tim Stewart. Take the opportunity to
654 add more error-checking on spoolfile handling while that code is being
657 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
658 No known attacks, no CVE, this is defensive hardening.
660 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
661 a queue-runner could start a delivery while other operations were ongoing.
662 Cutthrough delivery was a common victim, resulting in duplicate delivery.
663 Found and investigated by Tim Stewart. Fix by using the open message data
664 file handle rather than opening another, and not locally closing it (which
665 releases a lock) for that case, while creating the temporary .eml format
666 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
668 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
669 $sender_verify_failure/$recipient_verify_failure to "random".
671 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
674 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
675 Previously this would segfault.
677 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
680 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
681 like zero, since the resolver should be doing this for us, But we need one
682 as a CNAME but no MX presence gets the CNAME returned; we need to check
683 that doesn't point to an MX to declare it "no MX returned" rather than
684 "error, loop". A new main option is added so the older capability of
685 following some limited number of chain links is maintained.
687 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
689 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
690 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
691 now-deprecated earlier definitions used only specified the range up to TLS
692 1.2 (in the older-version library docs).
694 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
696 JH/15 Rework TLS client-side context management. Stop using a global, and
697 explicitly pass a context around. This enables future use of TLS for
698 connections to service-daemons (eg. malware scanning) while a client smtp
699 connection is using TLS; with cutthrough connections this is quite likely.
701 JH/16 Fix ARC verification to do AS checks in reverse order.
703 JH/17 Support a "tls" option on the ${readsocket } expansion item.
705 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
706 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
707 Previously the "utf8" would be re-prepended for every additional message.
709 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
710 Previously thery were accepted, resulting in issues when attempting to
711 forward messages to a non-supporting MTA.
713 PP/02 Let -n work with printing macros too, not just options.
715 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
716 one parent address was copied, and bogus data was used at delivery-logging
717 time. Either a crash (after delivery) or bogus log data could result.
718 Discovery and analysis by Tim Stewart.
720 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
721 Previously if the string ended mid-character, we did not insert the
722 promised '?' replacement.
724 PP/04 Documentation: current string operators work on bytes, not codepoints.
726 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
727 should pack well giving a smaller memory footprint so better caching and
728 therefore performance. Group the declarations where this can't be done so
729 that the byte-sized flag variables are not interspersed among pointer
730 variables, giving a better chance of good packing by the compiler.
732 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
733 non-null, to avoid issues with sites running BATV. Previously reports were
734 sent with an empty envelope sender so looked like bounces.
736 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
737 The ignore_error flag wasn't being returned from the filter subprocess so
738 was not set for later routers. Investigation and fix by Matthias Kurz.
740 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
741 and a msg:complete for the whole, when a message is manually removed using
742 -Mrm. Developement by Matthias Kurz, hacked on by JH.
744 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
745 a "Gnu special" function, asprintf() in the DB utility binary builds; I
746 hope that is portable enough.
748 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
749 requiring a known-CA anchor certificate; make it now rely entirely on the
750 TLSA as an anchor. Checking the name on the leaf cert against the name
751 on the A-record for the host is still done for TA (but not for EE mode).
753 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
754 included in delivery lines for non-proxied connections, when compiled with
755 SUPPORT_SOCKS and running with proxy logging enabled.
757 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
758 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
759 move the existing event to fire before the normal logging of message
760 failure so that custom logging is bracketed by normal logging.
762 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
763 msg:fail:internal event. Developement by Matthias Kurz.
765 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
766 far too small for todays use of crypto signatures stored there. Go all
767 the way to the max DNS message size of 64kB, even though this might be
768 overmuch for IOT constrained device use.
770 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
771 copy a string over itself. The library routine is documented as not
772 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
774 JH/32 For main options check_spool_space and check_inode_space, where the
775 platform supports 64b integers, support more than the previous 2^31 kB
776 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
777 the previous G, M, k.
779 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
780 $authenticated_fail_id variable on authentication failure. Previously
783 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
784 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
785 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
786 GNUTLS_SEC_PARAM_MEDIUM.
788 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
789 side. Previously we would continue as if no SNI had been received.
791 JH/36 Harden the handling of string-lists. When a list consisted of a sole
792 "<" character, which should be a list-separator specification, we walked
793 off past the nul-terimation.
795 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
796 causes) even when the retry time is not yet met. Previously they were
797 not, meaning that when (say) an account was over-quota and temp-rejecting,
798 and multiple senders' messages were queued, only one sender would get
799 notified on each configured delay_warning cycle.
801 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
803 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
804 cipher-suites, an error can be left on the stack even for a succeeding
805 accept; this results in impossible error messages when a later operation
808 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
809 return error codes indicating retry. Under TLS1.3 this becomes required.
811 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
812 it only wrote the new authenticators, resulting in a lack of tracking of
813 peer changes of ESMTP extensions until the next cache flush.
815 JH/41 Fix the loop reading a message header line to check for integer overflow,
816 and more-often against header_maxsize. Previously a crafted message could
817 induce a crash of the recive process; now the message is cleanly rejected.
819 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
820 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
826 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
827 When redis_servers is set to a list of > 1 element, and the Redis servers
828 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
829 case of MOVED into a DEFER case instead, thus moving the query onto the
830 next server in the list. For a cluster of N elements, all N servers must
831 be defined in redis_servers.
833 GF/02 Catch and remove uninitialized value warning in exiqsumm
834 Check for existence of @ARGV before looking at $ARGV[0]
836 JH/01 Replace the store_release() internal interface with store_newblock(),
837 which internalises the check required to safely use the old one, plus
838 the allocate and data copy operations duplicated in both (!) of the
839 extant use locations.
841 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
842 modifier. This matches the restriction on the commandline.
844 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
845 Previously only the last row was returned.
847 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
848 we assumed that tags in the header were well-formed, and parsed the
849 element content after inspecting only the first char of the tag.
850 Assumptions at that stage could crash the receive process on malformed
853 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
854 While running the DKIM ACL we operate on the Permanent memory pool so that
855 variables created with "set" persist to the DATA ACL. Also (at any time)
856 DNS lookups that fail create cache records using the Permanent pool. But
857 expansions release any allocations made on the current pool - so a dnsdb
858 lookup expansion done in the DKIM ACL releases the memory used for the
859 DNS negative-cache, and bad things result. Solution is to switch to the
860 Main pool for expansions.
861 While we're in that code, add checks on the DNS cache during store_reset,
862 active in the testsuite.
863 Problem spotted, and debugging aided, by Wolfgang Breyha.
865 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
866 When none of the hosts presented to a transport match an already-open
867 connection, close it and proceed with the list. Previously we would
868 queue the message. Spotted by Lena with Yahoo, probably involving
871 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
872 Previously a spurious "250 OK id=" response was appended to the proper
875 JH/08 The "support for" informational output now, which built with Content
876 Scanning support, has a line for the malware scanner interfaces compiled
877 in. Interface can be individually included or not at build time.
879 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
880 by the template makefile "src/EDITME". The "STREAM" support for an older
881 ClamAV interface method is removed.
883 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
884 rows affected is given instead).
886 JH/11 The runtime Berkeley DB library version is now additionally output by
887 "exim -d -bV". Previously only the compile-time version was shown.
889 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
890 SMTP connection. Previously, when one had more receipients than the
891 first, an abortive onward connection was made. Move to full support for
892 multiple onward connections in sequence, handling cutthrough connection
893 for all multi-message initiating connections.
895 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
896 routers. Previously, a multi-recipient message would fail to match the
897 onward-connection opened for the first recipient, and cause its closure.
899 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
900 a timeout on read on a GnuTLS initiating connection, resulting in the
901 initiating connection being dropped. This mattered most when the callout
902 was marked defer_ok. Fix to keep the two timeout-detection methods
905 JH/15 Relax results from ACL control request to enable cutthrough, in
906 unsupported situations, from error to silently (except under debug)
907 ignoring. This covers use with PRDR, frozen messages, queue-only and
910 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
912 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
913 metadata, resulting in a crash in free().
915 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
916 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
917 Broken also in d185889f4, with init system revamp.
919 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
920 Previously we abruptly closed the connection after reading a malware-
921 found indication; now we go on to read the "scan ok" response line,
924 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
925 ACL. Previously, a crash would result.
927 JH/19 Speed up macro lookups during configuration file read, by skipping non-
928 macro text after a replacement (previously it was only once per line) and
929 by skipping builtin macros when searching for an uppercase lead character.
931 JH/20 DANE support moved from Experimental to mainline. The Makefile control
932 for the build is renamed.
934 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
935 was allocated for every new TLS startup, meaning one per message. Fix
936 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
938 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
939 reported the original. Fix to report (as far as possible) the ACL
940 result replacing the original.
942 JH/23 Fix memory leak during multi-message connections using STARTTLS under
943 OpenSSL. Certificate information is loaded for every new TLS startup,
944 and the resources needed to be freed.
946 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
948 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
951 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
952 DATA response info to the (existing) per-recipient response info for
953 the "C=" log element. It can have useful tracking info from the
954 destination system. Patch from Simon Arlott.
956 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
957 length value. Previously this would segfault.
959 HS/02 Support Avast multiline protoocol, this allows passing flags to
960 newer versions of the scanner.
962 JH/28 Ensure that variables possibly set during message acceptance are marked
963 dead before release of memory in the daemon loop. This stops complaints
964 about them when the debug_store option is enabled. Discovered specifically
965 for sender_rate_period, but applies to a whole set of variables.
966 Do the same for the queue-runner and queue-list loops, for variables set
967 from spool message files. Do the same for the SMTP per-message loop, for
968 certain variables indirectly set in ACL operations.
970 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
971 as a multi-recipient message from a mailinglist manager). The coding had
972 an arbitrary cutoff number of characters while checking for more input;
973 enforced by writing a NUL into the buffer. This corrupted long / fast
974 input. The problem was exposed more widely when more pipelineing of SMTP
975 responses was introduced, and one Exim system was feeding another.
976 The symptom is log complaints of SMTP syntax error (NUL chars) on the
977 receiving system, and refused recipients seen by the sending system
978 (propating to people being dropped from mailing lists).
979 Discovered and pinpointed by David Carter.
981 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
982 replaced by the ${authresults } expansion.
984 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
986 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
987 allows proper process termination in container environments.
989 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
990 Previously the "final dot" had a newline after it; ensure it is CR,LF.
992 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
993 and "err_perm", deprecated since 4.83 when the RFC-defined words
994 "temperror" and "permerror" were introduced.
996 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
997 transport-filters or DKIM-signing. The restriction was lost in the
998 consolidation of verify-callout and delivery SMTP handling.
999 Extend the restriction to also cover ARC-signing.
1001 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1002 in defer=pass mode supply a 450 to the initiator. Previously the message
1005 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1006 tls_require_ciphers is used as before.
1008 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1009 "pass_unscanned". Only tmpfails from the scanner are written to
1010 the paniclog, as they may require admin intervention (permission
1011 denied, license issues). Other scanner errors (like decompression
1012 bombs) do not cause a paniclog entry.
1014 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1015 Previously it was possible to log spurious information in receive log
1018 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1019 triggered odd behaviour from Outlook Express clients.
1021 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1024 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1025 since the IETF WG has not yet settled on that versus the original
1026 "bare" representation.
1028 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1029 Previously the millisecond value corrupted the output.
1030 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1031 corrupted the output.
1037 JH/01 Rework error string handling in TLS interface so that the caller in
1038 more cases is responsible for logging. This permits library-sourced
1039 string to be attached to addresses during delivery, and collapses
1040 pairs of long lines into single ones.
1042 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1043 during configuration. Wildcards are allowed and expanded.
1045 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1046 This permits better logging.
1048 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1049 do not close it down and have the child transport start it up again on
1050 the passed-on TCP connection. Instead, proxy the child (and any
1051 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1052 affected: the continued delivery log lines do not have any DNSSEC, TLS
1053 Certificate or OCSP information. TLS cipher information is still logged.
1055 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1056 identical IP addresses on different listening ports. Will also affect
1059 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1060 add noisy ifdef guards to special-case this sillyness.
1061 Patch from Bernd Kuhls.
1063 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1064 than 255 are no longer allowed.
1066 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1067 Disable the session-cache too, which might reduce our load. Since we
1068 currrectly use a new context for every connection, both as server and
1069 client, there is no benefit for these.
1070 GnuTLS appears to not support tickets server-side by default (we don't
1071 call gnutls_session_ticket_enable_server()) but client side is enabled
1072 by default on recent versions (3.1.3 +) unless the PFS priority string
1075 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1076 <https://reproducible-builds.org/specs/source-date-epoch/>.
1078 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1079 the check for any unsuccessful recipients did not notice the limit, and
1080 erroneously found still-pending ones.
1082 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1083 MSG_MORE. Only in-clear (not on TLS connections).
1085 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1086 transport-filter is involved we can buffer the headers in memory for
1087 creating the signature, and read the spool data file once for the
1088 signature and again for transmission.
1090 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1091 4.77 as the kernel support then wasn't solid, having issues in 64bit
1092 mode. Now, it's been long enough. Add support for FreeBSD also.
1094 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1095 case where the routing stage had gathered several addresses to send to
1096 a host before calling the transport for the first, we previously failed
1097 to close down TLS in the old transport process before passing the TCP
1098 connection to the new process. The new one sent a STARTTLS command
1099 which naturally failed, giving a failed delivery and bloating the retry
1100 database. Investigation and fix prototype from Wolfgang Breyha.
1102 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1103 false-negatives in the check that the sender had not preempted a response
1104 or prompt from Exim (running as a server), due to that code's lack of
1105 awareness of the SMTP input buffering.
1107 PP/04 Add commandline_checks_require_admin option.
1108 Exim drops privileges sanely, various checks such as -be aren't a
1109 security problem, as long as you trust local users with access to their
1110 own account. When invoked by services which pass untrusted data to
1111 Exim, this might be an issue. Set this option in main configuration
1112 AND make fixes to the calling application, such as using `--` to stop
1115 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1116 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1117 responses to those, into a single TLS record each way (this usually means
1118 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1121 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1122 affects you only if you're dancing at the edge of the param size limits.
1123 If you are, and this message makes sense to you, then: raise the
1124 configured limit or use OpenSSL 1.1. Nothing we can do for older
1127 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1128 cmdline element to get the documented default one. Previously it was
1131 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1132 are unsuable for non-message-specific reasons - eg. connection timeout,
1133 banner-time rejection.
1135 JH/16 Fix logging of delivery remote port, when specified by router, under
1138 PP/06 Repair manualroute's ability to take options in any order, even if one
1139 is the name of a transport.
1142 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1144 JH/17 Change the list-building routines interface to use the expanding-string
1145 triplet model, for better allocation and copying behaviour.
1147 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1148 Previously it was constructed the first time a possibly-matching string
1149 was met in the configuration file input during startup; now it is done
1152 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1153 compatible one, to avoid the (poorly documented) possibility of a config
1154 file in the working directory redirecting the DB files, possibly correpting
1155 some existing file. CVE-2017-10140 assigned for BDB.
1157 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1158 cache-hot. Previously, although the result was properly cached, the
1159 initial verify call returned a defer.
1161 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1162 the main verify for receipient in uncached-mode.
1164 JH/22 Retire historical build files to an "unsupported" subdir. These are
1165 defined as "ones for which we have no current evidence of testing".
1167 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1168 if present. Previously it was ignored.
1170 JH/24 Start using specified-initialisers in C structure init coding. This is
1171 a C99 feature (it's 2017, so now considered safe).
1173 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1174 if was a fixed-sized field and bitmask ops via macros; it is now more
1177 PP/07 GitHub PR 56: Apply MariaDB build fix.
1178 Patch provided by Jaroslav Škarvada.
1180 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1181 during Coverity cleanups [4.87 JH/47]
1182 Diagnosis and fix provided by Michael Fischer v. Mollard.
1184 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1185 the right size to place the terminating semicolon on its own folded
1186 line, the header hash was calculated to an incorrect value thanks to
1187 the (relaxed) space the fold became.
1189 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1190 and confused the parent.
1192 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1193 which could crash as a result. This could lead to undeliverable messages.
1195 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1198 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1199 matters on fast-turnover and PID-randomising systems, which were getting
1200 out-of-order delivery.
1202 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1203 a possibly-overlapping copy. The symptom was that "Remote host closed
1204 connection in response to HELO" was logged instead of the actual 4xx
1207 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1208 Previously only that bufferd was discarded, resulting in SYMTP command
1211 JH/32 DKIM: when a message has multiple signatures matching an identity given
1212 in dkim_verify_signers, run the dkim acl once for each. Previously only
1213 one run was done. Bug 2189.
1215 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1216 "panic the current process" to "deliberately defer". The panic log is
1217 still written with the problem list name; the mail and reject logs now
1218 get a temp-reject line for the message that was being handled, saying
1219 something like "domains check lookup or other defer". The SMTP 451
1220 message is still "Temporary local problem".
1222 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1223 A crafted sequence of BDAT commands could result in in-use memory beeing
1224 freed. CVE-2017-16943.
1226 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1227 from SMTP input. Previously it was always done; now only done for DATA
1228 and not BDAT commands. CVE-2017-16944.
1230 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1231 to the message (such as an overlong header line). Previously this was
1232 not done and we did not exit BDAT mode. Followon from the previous item
1233 though a different problem.
1239 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1240 than -2003 did; needs libidn2 in addition to libidn.
1242 JH/02 The path option on a pipe transport is now expanded before use.
1244 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1245 Patch provided by "Björn", documentation fix added too.
1247 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1248 missing a wire-to-host endian conversion.
1250 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1251 close after a BDAT command line could be taken as a following command,
1252 giving a synch failure. Fix by only checking for synch immediately
1253 before acknowledging the chunk.
1255 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1256 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1257 macro. Patches provided by Josh Soref.
1259 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1260 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1261 by IANA as a service extension. However, John Klensin suggests that we
1264 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1265 the dkim code may be unix-mode line endings rather than smtp wire-format
1266 CRLF, so prepend a CR to any bare LF.
1268 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1269 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1271 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1272 the first were themselves being wrongly included in the feed into dkim
1273 processing; with most chunk sizes in use this resulted in an incorrect
1274 body hash calculated value.
1276 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1277 DKIM signature block, for verification. Although advised against by
1278 standards it is specifically not ruled illegal.
1280 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1282 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1283 missing a body hash (the bh= tag).
1285 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1286 It seems that HAProxy sends the Proxy Protocol information in clear and
1287 only then does a TLS startup, so do the same.
1289 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1290 TCP connections (such as for Spamd) unless the daemon successfully set
1291 Fast Open mode on its listening sockets. This fixes breakage seen on
1292 too-old kernels or those not configured for Fast Open, at the cost of
1293 requiring both directions being enabled for TFO, and TFO never being used
1294 by non-daemon-related Exim processes.
1296 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1297 endings, at least on the first header line. Try to canonify any that get
1298 past that check, despite the cost.
1300 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1301 now limited to an arbitrary five deep, while parsing addresses with the
1302 strip_excess_angle_brackets option enabled.
1304 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1305 instead leave the unprompted TLS handshake in socket buffer for the
1306 TLS library to consume.
1308 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1310 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1312 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1313 region in one operation, for speed, and this leaves assigned pointers
1314 dangling. Add checks run only under the testsuite which checks all
1315 variables at a store-reset and panics on a dangling pointer; add code
1316 explicitly nulling out all the variables discovered. Fixes one known
1317 bug: a transport crash, where a dangling pointer for $sending_ip_address
1318 originally assigned in a verify callout, is re-used.
1320 PP/06 Drop '.' from @INC in various Perl scripts.
1322 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1324 PP/08 Reduce a number of compilation warnings under clang; building with
1325 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1326 should be warning-free.
1328 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1330 HS/01 Fix portability problems introduced by PP/08 for platforms where
1331 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1333 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1334 chunk. This allows us to accept broken chunked messages. We need a more
1335 general solution here.
1337 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1338 already-broken messages in the queue.
1340 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1342 JH/19 Fix reference counting bug in routing-generated-address tracking.
1348 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1349 supports it and a size is available (ie. the sending peer gave us one).
1351 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1352 years of being deprecated). The replacements are the ACLs
1353 acl_smtp_mime and acl_not_smtp_mime.
1355 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1356 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1357 or even an in-clear connection were permitted. Now, if the host lookup
1358 was dnssec and dane was requested then the host is only used if the
1359 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1360 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1361 if one fails this test.
1362 This means that a poorly-configured remote DNS will make it incommunicado;
1363 but it protects against a DNS-interception attack on it.
1365 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1366 non-noisy when a race steals the message being considered.
1368 JH/05 If main configuration option tls_certificate is unset, generate a
1369 self-signed certificate for inbound TLS connections.
1371 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1372 in rewrites and routers.
1374 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1375 and logged a warning sing 4.83; now they are a configuration file error.
1377 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1378 (lacking @domain). Apply the same qualification processing as RCPT.
1380 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1382 JH/10 Support ${sha256:} applied to a string (as well as the previous
1385 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1386 a cutthrough deliver is pending, as we always want to make a connection.
1387 This also avoids re-routing the message when later placing the cutthrough
1388 connection after a verify cache hit.
1389 Do not update it with the verify result either.
1391 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1392 when routing results in more than one destination address.
1394 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1395 signing (which inhibits the cutthrough capability). Previously only
1396 the presence of an option was tested; now an expansion evaluating as
1397 empty is permissible (obviously it should depend only on data available
1398 when the cutthrough connection is made).
1400 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1401 the relevant preceding SMTP command did not note the pipelining mode.
1403 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1404 Previously they were not counted.
1406 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1407 as one having no matching records. Previously we deferred the message
1408 that needed the lookup.
1410 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1411 distinguished as "(=".
1413 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1414 for missing MX records. Previously it only worked for missing A records.
1416 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1418 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1419 after the data-go-ahead and data-ack. Patch from Jason Betts.
1421 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1422 even for a "none" policy. Patch from Tony Meyer.
1424 JH/22 Fix continued use of a connection for further deliveries. If a port was
1425 specified by a router, it must also match for the delivery to be
1428 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1429 When one of the recipients of a message was unsuitable for the connection
1430 (has no matching addresses), we lost track of needing to mark it
1431 deferred. As a result mail would be lost.
1433 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1435 JH/25 Decoding ACL controls is now done using a binary search; the source code
1436 takes up less space and should be simpler to maintain. Merge the ACL
1437 condition decode tables also, with similar effect.
1439 JH/26 Fix problem with one_time used on a redirect router which returned the
1440 parent address unchanged. A retry would see the parent address marked as
1441 delivered, so not attempt the (identical) child. As a result mail would
1444 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1445 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1446 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1449 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1450 Main config options check_{log,spool}_{inodes,space} are now
1451 100 inodes, 10MB unless set otherwise in the configuration.
1453 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1454 Previously it only applied to the main-section connection policy
1457 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1459 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1460 by me. Added RFC7919 DH primes as an alternative.
1462 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1463 are not in the system include path.
1465 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1466 GnuTLS, when a session startup failed (eg because the client disconnected)
1467 Exim did stdio operations after fclose. This was exposed by a recent
1468 change which nulled out the file handle after the fclose.
1470 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1471 signed directly by the cert-signing cert, rather than an intermediate
1472 OCSP-signing cert. This is the model used by LetsEncrypt.
1474 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1476 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1477 an incoming connection.
1479 HS/02 Bug 1802: Do not half-close the connection after sending a request
1482 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1483 fallback to "prime256v1".
1485 JH/34 SECURITY: Use proper copy of DATA command in error message.
1486 Could leak key material. Remotely exploitable. CVE-2016-9963.
1492 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1493 and 3.4.4 - once the server is enabled to respond to an OCSP request
1494 it does even when not requested, resulting in a stapling non-aware
1495 client dropping the TLS connection.
1497 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1498 support variable-length bit vectors. No functional change.
1500 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1501 The I= interface field on outgoing lines is now after the H= remote
1502 host field, same as incoming lines. There is a separate
1503 outgoing_interface log selector which allows you to disable the
1506 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1507 If not running log_selector +smtp_connection the mainlog would be held
1508 open indefinitely after a "too many connections" event, including to a
1509 deleted file after a log rotate. Leave the per net connection logging
1510 leaving it open for efficiency as that will be quickly detected by the
1511 check on the next write.
1513 HS/01 Bug 1671: Fix post transport crash.
1514 Processing the wait-<transport> messages could crash the delivery
1515 process if the message IDs didn't exist for some reason. When
1516 using 'split_spool_directory=yes' the construction of the spool
1517 file name failed already, exposing the same netto behaviour.
1519 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1520 mime_regex ACL conditions.
1522 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1523 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1524 to HELO, local diagnostic string.
1526 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1527 log line to debug. Even when configured with a tls authenticator many
1528 client connections are expected to not authenticate in this way, so
1529 an authenticate fail is not an error.
1531 HS/02 Add the Exim version string to the process info. This way exiwhat
1532 gives some more detail about the running daemon.
1534 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1535 matter for fast-change records such as DNSBLs.
1537 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1538 retry record, even if constant. There may be multiple transports with
1539 different interface settings and the retry behaviour needs to be kept
1542 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1544 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1546 JH/10 Bug 840: fix log_defer_output option of pipe transport
1548 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1549 hosts_randomize. This matters a lot when combined with mua_wrapper.
1551 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1552 ${quote_pgsql:<string>} operator.
1554 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1556 JH/14 Change header file name for freeradius-client. Relevant if compiling
1557 with Radius support; from the Gentoo tree and checked under Fedora.
1559 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1561 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1562 option transport_filter, meaning no filtering.
1564 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1566 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1567 defaults to "*" (all hosts). The variable is now available when not built
1568 with TLS, default unset, mainly to enable keeping the testsuite sane.
1569 If a server certificate is not supplied (via tls_certificate) an error is
1570 logged, and clients will find TLS connections fail on startup. Presumably
1571 they will retry in-clear.
1572 Packagers of Exim are strongly encouraged to create a server certificate
1573 at installation time.
1575 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1576 with the $config_file variable.
1578 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1579 in transport context, after the attempt, and per-recipient. The latter type
1580 is per host attempted. The event data is the error message, and the errno
1581 information encodes the lookup type (A vs. MX) used for the (first) host,
1582 and the trailing two digits of the smtp 4xx response.
1584 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1585 to write to mainlog (or rejectlog, paniclog) in the window between file
1586 creation and permissions/ownership being changed. Particularly affects
1587 installations where exicyclog is run as root, rather than exim user;
1588 result is that the running daemon panics and dies.
1590 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1592 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1593 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1594 "pri" and "weight". Note that the previous implicit priority given by the
1595 list order is no longer honoured.
1597 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1598 for DKIM processing.
1600 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1601 by defining SUPPORT_SOCKS.
1603 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1604 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1605 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1606 variables are renamed to proxy_{local,external}_{address,port}.
1608 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1609 for a build by defining SUPPORT_I18N
1611 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1612 of the query string, and make ${quote_redis:} do that quoting.
1614 JH/29 Move Events support from Experimental to mainline, enabled by default
1615 and removable for a build by defining DISABLE_EVENT.
1617 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1619 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1620 cached by the daemon.
1622 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1623 by defining LOOKUP_REDIS. The libhiredis library is required.
1625 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1626 keys are given for lookup.
1628 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1629 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1630 only supported when built with TLS support. The PolarSSL SHA routines
1631 are still used when the TLS library is too old for convenient support.
1633 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1634 openssl_options), for security. OpenSSL forces this from version 1.1.0
1635 server-side so match that on older versions.
1637 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1638 allocation for $value could be released as the expansion processing
1639 concluded, but leaving the global pointer active for it.
1641 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1642 and to use the domains and local_parts ACL conditions.
1644 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1645 incorrectly not doubled on cutthrough transmission, hence seen as a
1646 body-termination at the receiving system - resulting in truncated mails.
1647 Commonly the sender saw a TCP-level error, and retransmitted the message
1648 via the normal store-and-forward channel. This could result in duplicates
1649 received - but deduplicating mailstores were liable to retain only the
1650 initial truncated version.
1652 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1654 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1656 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1657 we're in there, support oversigning also; bug 1309.
1659 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1661 HS/04 Add support for keep_environment and add_environment options.
1663 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1664 either intentional arithmetic overflow during PRNG, or testing config-
1667 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1668 delivery resulted in actual delivery. Cancel cutthrough before DATA
1671 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1672 rejects a recipient: pass the reject to the originator.
1674 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1675 Many were false-positives and ignorable, but it's worth fixing the
1678 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1679 for the new environment-manipulation done at startup. Move the routines
1680 from being local to tls.c to being global via the os.c file.
1682 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1683 an extract embedded as result-arg for a map, the first arg for extract
1684 is unavailable so we cannot tell if this is a numbered or keyed
1685 extraction. Accept either.
1691 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1694 JH/02 The smtp transport option "multi_domain" is now expanded.
1696 JH/03 The smtp transport now requests PRDR by default, if the server offers
1699 JH/04 Certificate name checking on server certificates, when exim is a client,
1700 is now done by default. The transport option tls_verify_cert_hostnames
1701 can be used to disable this per-host. The build option
1702 EXPERIMENTAL_CERTNAMES is withdrawn.
1704 JH/05 The value of the tls_verify_certificates smtp transport and main options
1705 default to the word "system" to access the system default CA bundle.
1706 For GnuTLS, only version 3.0.20 or later.
1708 JH/06 Verification of the server certificate for a TLS connection is now tried
1709 (but not required) by default. The verification status is now logged by
1710 default, for both outbound TLS and client-certificate supplying inbound
1713 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1716 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1717 Status Notification (bounce) messages are now MIME format per RFC 3464.
1718 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1719 under the control of the dsn_advertise_hosts option, and routers may
1720 have a dsn_lasthop option.
1722 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1723 default, modifiable by a malware= option. The list separator for
1724 the options can now be changed in the usual way. Bug 68.
1726 JH/10 The smtp_receive_timeout main option is now expanded before use.
1728 JH/11 The incoming_interface log option now also enables logging of the
1729 local interface on delivery outgoing connections.
1731 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1732 if the interface and destination host and port all match.
1734 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1737 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1738 Patch from Andrew Lewis.
1740 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1741 now supports optional time-restrictions, weighting, and priority
1742 modifiers per server. Patch originally by <rommer@active.by>.
1744 JH/16 The spamd_address main option now supports a mixed list of local
1745 and remote servers. Remote servers can be IPv6 addresses, and
1746 specify a port-range.
1748 JH/17 Bug 68: The spamd_address main option now supports an optional
1749 timeout value per server.
1751 JH/18 Bug 1581: Router and transport options headers_add/remove can
1752 now have the list separator specified.
1754 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1757 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1760 JH/21 Support for the A6 type of dns record is withdrawn.
1762 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1763 rather than the verbs used.
1765 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1766 from 255 to 1024 chars.
1768 JH/24 Verification callouts now attempt to use TLS by default.
1770 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1771 are generic router options now. The defaults didn't change.
1773 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1774 Original patch from Alexander Shikoff, worked over by JH.
1776 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1777 files if ZCAT_COMMAND is not executable.
1779 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1781 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1783 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1784 Normally benign, it bites when the pair was led to by a CNAME;
1785 modern usage is to not canonicalize the domain to a CNAME target
1786 (and we were inconsistent anyway for A-only vs AAAA+A).
1788 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1790 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1791 when evaluating $sender_host_dnssec.
1793 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1794 $sender_helo_dnssec variable.
1796 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1798 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1800 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1802 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1803 documented as working, but never had. Support all but $spam_report.
1805 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1806 added for tls authenticator.
1808 HS/03 Add perl_taintmode main config option
1814 TL/01 When running the test suite, the README says that variables such as
1815 no_msglog_check are global and can be placed anywhere in a specific
1816 test's script, however it was observed that placement needed to be near
1817 the beginning for it to behave that way. Changed the runtest perl
1818 script to read through the entire script once to detect and set these
1819 variables, reset to the beginning of the script, and then run through
1820 the script parsing/test process like normal.
1822 TL/02 The BSD's have an arc4random API. One of the functions to induce
1823 adding randomness was arc4random_stir(), but it has been removed in
1824 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1825 function when detected.
1827 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1828 cause callback expansion.
1830 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1831 syntax errors in an expansion can be treated as a string instead of
1832 logging or causing an error, due to the internal use of bool_lax
1833 instead of bool when processing it.
1835 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1836 server certificates when making smtp deliveries.
1838 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1840 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1842 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1844 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1845 Merged patch from Sebastian Wiedenroth.
1847 JH/05 Fix results-pipe from transport process. Several recipients, combined
1848 with certificate use, exposed issues where response data items split
1849 over buffer boundaries were not parsed properly. This eventually
1850 resulted in duplicates being sent. This issue only became common enough
1851 to notice due to the introduction of connection certificate information,
1852 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1854 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1855 size buffer was used, resulting in syntax errors when an expansion
1858 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1859 version 3.3.6 or later.
1861 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1862 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1863 both become "event_action", the variables become $event_name, $event_data
1864 and $event_defer_errno. There is a new variable $verify_mode, usable in
1865 routers, transports and related events. The tls:cert event is now also
1866 raised for inbound connections, if the main configuration event_action
1869 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1870 early OCSP support, but no stapling (appears to be less than 1.0.0).
1872 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1873 server certificate names available under the smtp transport option
1874 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1877 JH/10 Time-related extraction expansions from certificates now use the main
1878 option "timezone" setting for output formatting, and are consistent
1879 between OpenSSL and GnuTLS compilations. Bug 1541.
1881 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1882 encoded parameter in the incoming message. Bug 1558.
1884 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1885 include certificate info, eximon was claiming there were spoolfile
1888 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1890 JH/14 Log delivery-related information more consistently, using the sequence
1891 "H=<name> [<ip>]" wherever possible.
1893 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1894 are problematic for Debian distribution, omit them from the release
1897 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1899 JH/16 Fix string representation of time values on 64bit time_t architectures.
1902 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1903 output list separator was used.
1908 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1909 checkers that were complaining about end of non-void function with no
1912 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1913 This was a regression introduced in 4.83 by another bugfix.
1915 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1917 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1918 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1924 TF/01 Correctly close the server side of TLS when forking for delivery.
1926 When a message was received over SMTP with TLS, Exim failed to clear up
1927 the incoming connection properly after forking off the child process to
1928 deliver the message. In some situations the subsequent outgoing
1929 delivery connection happened to have the same fd number as the incoming
1930 connection previously had. Exim would try to use TLS and fail, logging
1931 a "Bad file descriptor" error.
1933 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1934 utilities have not been installed.
1936 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1937 temporary space as the ACL may create new global variables.
1939 TL/01 LDAP support uses per connection or global context settings, depending
1940 upon the detected version of the libraries at build time.
1942 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1943 to extract and use the src ip:port in logging and expansions as if it
1944 were a direct connection from the outside internet. PPv2 support was
1945 updated based on HAProxy spec change in May 2014.
1947 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1949 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1950 Properly escape header and check for NULL return.
1952 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1955 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1957 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1958 characters in header names, implemented as a verify condition.
1959 Contributed by Michael Fischer v. Mollard.
1961 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1962 results permerror and temperror. Previous values are deprecated but
1963 still accepted. In a future release, err_perm and err_temp will be
1964 completely removed, which will be a backward incompatibility if the
1965 ACL tests for either of these two old results. Patch contributed by
1966 user bes-internal on the mailing list.
1968 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1970 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1971 selectors, in both main and reject logs.
1973 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1976 JH/07 Add malware type "sock" for talking to simple daemon.
1978 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1980 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1981 routers/transports under cutthrough routing.
1983 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1984 numbers. Touch up "bool" conditional to keep the same definition.
1986 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1988 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1990 JH/12 Expand items in router/transport headers_add or headers_remove lists
1991 individually rather than the list as a whole. Bug 1452.
1993 Required for reasonable handling of multiple headers_ options when
1994 they may be empty; requires that headers_remove items with embedded
1995 colons must have them doubled (or the list-separator changed).
1997 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1998 view the policy declared in the DMARC record. Currently, $dmarc_status
1999 is a combined value of both the record presence and the result of the
2002 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2004 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2005 dnslookup router and the smtp transport (applying to the forward
2008 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2009 of ldap servers used for a specific lookup. Patch provided by Heiko
2012 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2013 New variable $lookup_dnssec_authenticated for observability.
2015 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2016 Patch submitted by Lars Timman.
2018 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2020 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2021 Requires trusted mode and valid format message id, aborts otherwise.
2022 Patch contributed by Heiko Schlichting.
2024 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2025 certextract with support for various fields. Bug 1358.
2027 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2028 is requested by default, modifiable by smtp transport option
2031 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2032 operate on certificate variables to give certificate fingerprints
2033 Also new ${sha256:cert_variable}.
2035 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2037 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2038 Christian Aistleitner.
2040 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2042 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2043 file. Patch from Wolfgang Breyha.
2045 JH/25 Expand the coverage of the delivery $host and $host_address to
2046 client authenticators run in verify callout. Bug 1476.
2048 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2049 align with daemon_smtp_ports. Bug 72.
2051 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2052 support and error reporting did not work properly.
2054 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2055 and is readable. Patch from Andrew Colin Kissa.
2057 TL/14 Enhance documentation of ${run expansion and how it parses the
2058 commandline after expansion, particularly in the case when an
2059 unquoted variable expansion results in an empty value.
2061 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2063 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2064 by renaming away from it. Addresses GH issue 15, reported by
2067 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2069 TL/15 SECURITY: prevent double expansion in math comparison functions
2070 (can expand unsanitized data). Not remotely exploitable.
2077 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2079 PP/02 Make -n do something, by making it not do something.
2080 When combined with -bP, the name of an option is not output.
2082 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2085 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2086 $sender_host_name and config options to manage this, and basic check
2089 PP/05 DSCP support for outbound connections and control modifier for inbound.
2091 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2092 (Only plugin which currently uses this is kerberos4, which nobody should
2093 be using, but we should make it available and other future plugins might
2094 conceivably use it, even though it would break NAT; stuff *should* be
2095 using channel bindings instead).
2097 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2098 name; added for Sendmail compatibility; requires admin caller.
2099 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2100 just ignore it); requires trusted caller.
2101 Also parse but ignore: -Ac -Am -X<logfile>
2104 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2106 TL/02 Add +smtp_confirmation as a default logging option.
2108 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2109 Patch by Magnus Holmgren from 2007-02-20.
2111 TL/04 Bugzilla 1281 - Spec typo.
2112 Bugzilla 1283 - Spec typo.
2113 Bugzilla 1290 - Spec grammar fixes.
2115 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2117 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2119 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2120 mailing list by by Dmitry Isaikin.
2122 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2124 JH/02 Support "G" suffix to numbers in ${if comparisons.
2126 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2128 NM/01 Bugzilla 1197 - Spec typo
2129 Bugzilla 1196 - Spec examples corrections
2131 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2133 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2134 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2137 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2138 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2140 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2141 "acl {{name}{arg}...}", and optional args on acl condition
2144 JH/05 Permit multiple router/transport headers_add/remove lines.
2146 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2148 JH/07 Avoid using a waiting database for a single-message-only transport.
2149 Performance patch from Paul Fisher. Bugzilla 1262.
2151 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2154 JH/09 Add $headers_added variable, with content from use of ACL modifier
2155 add_header (but not yet added to the message). Bugzilla 199.
2157 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2158 Pulled from Bugzilla 817 by Wolfgang Breyha.
2160 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2162 (nb: this is the same fix as in Exim 4.80.1)
2164 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2167 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2170 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2172 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2173 advertise SMTP AUTH mechanism to us, instead of a generic
2174 protocol violation error. Also, make Exim more robust to bad
2175 data from the Dovecot auth socket.
2177 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2179 When a queue runner is handling a message, Exim first routes the
2180 recipient addresses, during which it prunes them based on the retry
2181 hints database. After that it attempts to deliver the message to
2182 any remaining recipients. It then updates the hints database using
2185 So if a recipient address works intermittently, it can get repeatedly
2186 deferred at routing time. The retry hints record remains fresh so the
2187 address never reaches the final cutoff time.
2189 This is a fairly common occurrence when a user is bumping up against
2190 their storage quota. Exim had some logic in its local delivery code
2191 to deal with this. However it did not apply to per-recipient defers
2192 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2194 This change adds a proper retry rule check during routing so that the
2195 final cutoff time is checked against the message's age. We only do
2196 this check if there is an address retry record and there is not a
2197 domain retry record; this implies that previous attempts to handle
2198 the address had the retry_use_local_parts option turned on. We use
2199 this as an approximation for the destination being like a local
2200 delivery, as in LMTP.
2202 I suspect this new check makes the old local delivery cutoff check
2203 redundant, but I have not verified this so I left the code in place.
2205 TF/02 Correct gecos expansion when From: is a prefix of the username.
2207 Test 0254 submits a message to Exim with the header
2211 When I ran the test suite under the user fanf2, Exim expanded
2212 the header to contain my full name, whereas it should have added
2213 a Resent-Sender: header. It erroneously treats any prefix of the
2214 username as equal to the username.
2216 This change corrects that bug.
2218 GF/01 DCC debug and logging tidyup
2219 Error conditions log to paniclog rather than rejectlog.
2220 Debug lines prefixed by "DCC: " to remove any ambiguity.
2222 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2224 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2225 Bug spotted by Jeremy Harris; was flawed since initial commit.
2226 Would have resulted in OCSP responses post-SNI triggering an Exim
2227 NULL dereference and crash.
2229 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2231 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2232 Bug detection, analysis and fix by Samuel Thibault.
2233 Bugzilla 1331, Debian bug #698092.
2235 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2237 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2238 Server implementation by Todd Lyons, client by JH.
2239 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2240 config variable "prdr_enable" controls whether the server
2241 advertises the facility. If the client requests PRDR a new
2242 acl_data_smtp_prdr ACL is called once for each recipient, after
2243 the body content is received and before the acl_smtp_data ACL.
2244 The client is controlled by both of: a hosts_try_prdr option
2245 on the smtp transport, and the server advertisement.
2246 Default client logging of deliveries and rejections involving
2247 PRDR are flagged with the string "PRDR".
2249 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2250 fclose(). Diagnosis by Todd Lyons.
2252 PP/17 Update configure.default to handle IPv6 localhost better.
2253 Patch by Alain Williams (plus minor tweaks).
2256 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2257 This is now consistent with GnuTLS, and is now documented: the
2258 previous undocumented portable approach to treating the option as
2259 unset was to force an expansion failure. That still works, and
2260 an empty string is now equivalent.
2262 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2263 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2264 not performing validation itself.
2266 PP/20 Added force_command boolean option to pipe transport.
2267 Patch from Nick Koston, of cPanel Inc.
2269 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2272 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2274 PP/21 Fix eximon continuous updating with timestamped log-files.
2275 Broken in a format-string cleanup in 4.80, missed when I repaired the
2276 other false fix of the same issue.
2277 Report and fix from Heiko Schlichting.
2280 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2281 Report from Prashanth Katuri.
2283 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2284 It's SecureTransport, so affects any MacOS clients which use the
2285 system-integrated TLS libraries, including email clients.
2287 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2288 using a MIME ACL for non-SMTP local injection.
2289 Report and assistance in diagnosis by Warren Baker.
2291 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2293 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2295 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2296 last id that failed so it may be referenced in subsequent ACL's.
2298 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2301 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2302 ldap library initialization, allowing self-signed CA's to be
2303 used. Also properly sets require_cert option later in code by
2304 using NULL (global ldap config) instead of ldap handle (per
2305 session). Bug diagnosis and testing by alxgomz.
2307 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2308 the src/util/ subdirectory.
2310 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2311 renamed to Transport Post Delivery Action by Jeremy Harris, as
2314 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2315 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2316 redis_servers = needs to be configured which will be used by the redis
2317 lookup. Patch from Warren Baker, of The Packet Hub.
2319 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2321 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2322 hostname or reverse DNS when processing a host list. Used suggestions
2323 from multiple comments on this bug.
2325 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2327 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2328 Missed a few lines, added it to make the runtest require no keyboard
2331 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2332 contains upper case chars. Make router use caseful_local_part.
2334 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2335 support when GnuTLS has been built with p11-kit.
2341 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2343 This, or similar/improved, will also be change PP/11 of 4.82.
2349 PP/01 Handle short writes when writing local log-files.
2350 In practice, only affects FreeBSD (8 onwards).
2351 Bugzilla 1053, with thanks to Dmitry Isaikin.
2353 NM/01 Bugzilla 949 - Documentation tweak
2355 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2358 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2360 PP/02 Implemented gsasl authenticator.
2362 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2364 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2365 `pkg-config foo` for cflags/libs.
2367 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2368 with rest of GSASL and with heimdal_gssapi.
2370 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2371 `pkg-config foo` for cflags/libs for the TLS implementation.
2373 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2374 properties get this fed in as external SSF. A number of robustness
2375 and debugging improvements to the cyrus_sasl authenticator.
2377 PP/08 cyrus_sasl server now expands the server_realm option.
2379 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2380 Patch by Jeremy Harris.
2382 PP/10 Added dbmjz lookup type.
2384 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2386 PP/12 MAIL args handles TAB as well as SP, for better interop with
2387 non-compliant senders.
2388 Analysis and variant patch by Todd Lyons.
2390 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2391 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2392 Patch from Dirk Mueller <dmueller@suse.com>
2394 PP/13 tls_peerdn now print-escaped for spool files.
2395 Observed some $tls_peerdn in wild which contained \n, which resulted
2396 in spool file corruption.
2398 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2399 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2400 or write after TLS renegotiation, which otherwise led to messages
2403 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2404 as a tracking header (ie: a signed header comes before the signature).
2405 Patch from Wolfgang Breyha.
2407 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2408 comma-sep list; embedded commas doubled.
2410 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2412 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2414 Report and patch from Dmitry Banschikov.
2416 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2417 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2418 failures. We appear to now support TLS1.1+ with Exim.
2420 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2421 lets Exim select keys and certificates based upon TLS SNI from client.
2422 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2423 before an outbound SMTP session. New log_selector, +tls_sni.
2425 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2426 NULL dereference. Report and patch from Alun Jones.
2428 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2429 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2430 Not seeing resolver debug output on NetBSD, but suspect this is a
2431 resolver implementation change.
2433 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2434 Left warnings. Added "eximon gdb" invocation mode.
2436 PP/21 Defaulting "accept_8bitmime" to true, not false.
2438 PP/22 Added -bw for inetd wait mode support.
2440 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2441 locate the relevant includes and libraries. Made this the default.
2443 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2444 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2446 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2447 This may cause build issues on older platforms.
2449 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2450 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2451 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2452 Added SNI support via GnuTLS too.
2453 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2455 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2457 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2458 Applied second patch from Janne, implementing suggestion to default
2459 multiple-strings-in-record handling to match SPF spec.
2461 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2463 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2464 read-only, out of scope).
2465 Patch from Wolfgang Breyha, report from Stuart Northfield.
2467 PP/29 Fix three issues highlighted by clang analyser static analysis.
2468 Only crash-plausible issue would require the Cambridge-specific
2469 iplookup router and a misconfiguration.
2470 Report from Marcin Mirosław.
2472 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2474 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2475 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2476 As part of this, removing so much warning spew let me fix some minor
2477 real issues in debug logging.
2479 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2480 assignment on my part. Fixed.
2482 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2483 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2484 Janne Snabb (who went above and beyond: thank you).
2486 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2487 string otherwise requires a connection and a bunch more work and it's
2488 relatively easy to get wrong. Should also expose TLS library linkage
2491 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2492 64-bit ${eval} (JH/03).
2494 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2495 GNU libc to support some of the 64-bit stuff, should not lead to
2496 conflicts. Defined before os.h is pulled in, so if a given platform
2497 needs to override this, it can.
2499 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2500 protection layer was required, which is not implemented.
2501 Bugzilla 1254, patch from Wolfgang Breyha.
2503 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2504 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2505 tls_dhparam take prime identifiers. Also unbreak combination of
2506 OpenSSL+DH_params+TLSSNI.
2508 PP/39 Disable SSLv2 by default in OpenSSL support.
2514 PP/01 Solaris build fix for Oracle's LDAP libraries.
2515 Bugzilla 1109, patch from Stephen Usher.
2517 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2519 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2522 TF/02 Fix a couple more cases where we did not log the error message
2523 when unlink() failed. See also change 4.74-TF/03.
2525 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2526 lock up or crash if it happened to be inside a call to libc when it
2527 got a SIGUSR1 from exiwhat.
2529 The SIGUSR1 handler appends the current process status to the process
2530 log which is later printed by exiwhat. It used to use the general
2531 purpose logging code to do this, but several functions it calls are
2532 not safe for signals.
2534 The new output code in the SIGUSR1 handler is specific to the process
2535 log, and simple enough that it's easy to inspect for signal safety.
2536 Removing some special cases also simplifies the general logging code.
2537 Removing the spurious timestamps from the process log simplifies
2540 TF/04 Improved ratelimit ACL condition.
2542 The /noupdate option has been deprecated in favour of /readonly which
2543 has clearer semantics. The /leaky, /strict, and /readonly update modes
2544 are mutually exclusive. The update mode is no longer included in the
2545 database key; it just determines when the database is updated. (This
2546 means that when you upgrade Exim will forget old rate measurements.)
2548 Exim now checks that the per_* options are used with an update mode that
2549 makes sense for the current ACL. For example, when Exim is processing a
2550 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2551 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2552 must specify per_mail/readonly. If you omit the update mode it defaults to
2553 /leaky where that makes sense (as before) or /readonly where required.
2555 The /noupdate option is now undocumented but still supported for
2556 backwards compatibility. It is equivalent to /readonly except that in
2557 ACLs where /readonly is required you may specify /leaky/noupdate or
2558 /strict/noupdate which are treated the same as /readonly.
2560 A useful new feature is the /count= option. This is a generalization
2561 of the per_byte option, so that you can measure the throughput of other
2562 aggregate values. For example, the per_byte option is now equivalent
2563 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2565 The per_rcpt option has been generalized using the /count= mechanism
2566 (though it's more complicated than the per_byte equivalence). When it is
2567 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2568 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2569 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2570 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2571 Note that using per_rcpt with a non-readonly update mode in more than
2572 one ACL will cause the recipients to be double-counted. (The per_mail
2573 and per_byte options don't have this problem.)
2575 The handling of very low rates has changed slightly. If the computed rate
2576 is less than the event's count (usually one) then this event is the first
2577 after a long gap. In this case the rate is set to the same as this event's
2578 count, so that the first message of a spam run is counted properly.
2580 The major new feature is a mechanism for counting the rate of unique
2581 events. The new per_addr option counts the number of different
2582 recipients that someone has sent messages to in the last time period. It
2583 behaves like per_rcpt if all the recipient addresses are different, but
2584 duplicate recipient addresses do not increase the measured rate. Like
2585 the /count= option this is a general mechanism, so the per_addr option
2586 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2587 example, measure the rate that a client uses different sender addresses
2588 with the options per_mail/unique=$sender_address. There are further
2589 details in the main documentation.
2591 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2593 TF/06 Removed a few PCRE remnants.
2595 TF/07 Automatically extract Exim's version number from tags in the git
2596 repository when doing development or release builds.
2598 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2599 Bugzilla 879. Patch from Paul Fisher.
2601 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2602 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2605 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2607 PP/05 Variable $av_failed, true if the AV scanner deferred.
2608 Bugzilla 1078. Patch from John Horne.
2610 PP/06 Stop make process more reliably on build failure.
2611 Bugzilla 1087. Patch from Heiko Schlittermann.
2613 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2614 Bugzilla 1089. Patch from Heiko Schlittermann.
2616 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2617 Bugzilla 1131. Patch from Holger Weiß.
2619 PP/09 Handle IPv6 addresses with SPF.
2620 Bugzilla 860. Patch from Wolfgang Breyha.
2622 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2624 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2627 PP/11 match_* no longer expand right-hand-side by default.
2628 New compile-time build option, EXPAND_LISTMATCH_RHS.
2629 New expansion conditions, "inlist", "inlisti".
2631 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2633 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2635 PP/14 fix log_write() format string regression from TF/03.
2636 Bugzilla 1152. Patch from Dmitry Isaikin.
2642 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2644 PP/02 Harmonised TLS library version reporting; only show if debugging.
2645 Layout now matches that introduced for other libraries in 4.74 PP/03.
2647 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2649 PP/04 New "dns_use_edns0" global option.
2651 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2654 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2655 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2657 TK/01 Updated PolarSSL code to 0.14.2.
2658 Bugzilla 1097. Patch from Andreas Metzler.
2660 PP/07 Catch divide-by-zero in ${eval:...}.
2661 Fixes bugzilla 1102.
2663 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2666 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2667 format-string attack -- SECURITY: remote arbitrary code execution.
2669 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2670 time unintentionally subject to list matching rules, letting the header
2671 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2672 arbitrary string expansion). This allowed for information disclosure.
2674 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2675 INT_MIN/-1 -- value coerced to INT_MAX.
2681 NM/01 Workaround for PCRE version dependency in version reporting
2684 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2685 This fixes portability to compilers other than gcc, notably
2686 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2688 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2689 makefiles for portability to HP-UX and POSIX correctness.
2691 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2692 Also via indented variable definition in the Makefile.
2693 (Debugging by Oliver Heesakkers).
2695 PP/02 Restore caching of spamd results with expanded spamd_address.
2696 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2698 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2699 Improves build reliability. Fix from: Frank Elsner
2701 NM/02 Fix wide character breakage in the rfc2047 coding
2702 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2704 NM/03 Allow underscore in dnslist lookups
2705 Fixes bug 1026. Patch from Graeme Fowler
2707 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2708 Code patches from Adam Ciarcinski of NetBSD.
2710 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2713 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2714 is logged, to avoid truncation. Patch from John Horne.
2716 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2717 Patch from Jakob Hirsch.
2719 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2720 SQL string expansion failure details.
2721 Patch from Andrey Oktyabrski.
2723 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2724 Patch from Simon Arlott.
2726 PP/09 New lookups functionality failed to compile on old gcc which rejects
2727 extern declarations in function scope.
2728 Patch from Oliver Fleischmann
2730 PP/10 Use sig_atomic_t for flags set from signal handlers.
2731 Check getgroups() return and improve debugging.
2732 Fixed developed for diagnosis in bug 927 (which turned out to be
2735 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2736 Patch from Mark Zealey.
2738 PP/12 Bugzilla 1056: Improved spamd server selection.
2739 Patch from Mark Zealey.
2741 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2742 Based on patch from Heiko Schlittermann.
2744 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2745 Patch from Uwe Doering, sign-off by Michael Haardt.
2747 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2748 variable declaration deep within a block. Bug and patch from
2751 PP/15 lookups-Makefile IRIX compatibility coercion.
2753 PP/16 Make DISABLE_DKIM build knob functional.
2755 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2756 Patch by Simon Arlott
2758 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2759 variable argument macros. Our copy now differs from upstream.
2765 TF/01 Failure to get a lock on a hints database can have serious
2766 consequences so log it to the panic log.
2768 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2769 controlled using the smtp_confirmation log selector.
2771 TF/03 Include the error message when we fail to unlink a spool file.
2773 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2774 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2775 for maintaining out-of-tree patches for some time.
2777 PP/01 Bugzilla 139: Documentation and portability issues.
2778 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2779 Handle per-OS dynamic-module compilation flags.
2781 PP/02 Let /dev/null have normal permissions.
2782 The 4.73 fixes were a little too stringent and complained about the
2783 permissions on /dev/null. Exempt it from some checks.
2784 Reported by Andreas M. Kirchwitz.
2786 PP/03 Report version information for many libraries, including
2787 Exim version information for dynamically loaded libraries. Created
2788 version.h, now support a version extension string for distributors
2789 who patch heavily. Dynamic module ABI change.
2791 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2792 privilege escalation vulnerability whereby the Exim run-time user
2793 can cause root to append content of the attacker's choosing to
2796 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2799 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2800 If dropping privileges for untrusted macros, we disabled normal logging
2801 on the basis that it would fail; for the Exim run-time user, this is not
2802 the case, and it resulted in successful deliveries going unlogged.
2803 Fixed. Reported by Andreas Metzler.
2809 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2810 only prepend for the Resent-* case. Fixes regression introduced in
2811 Exim 4.70 by NM/22 for Bugzilla 607.
2813 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2814 increasing numbers of administrators be bitten by this.
2816 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2818 PP/03 Bugzilla 994: added openssl_options main configuration option.
2820 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2822 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2824 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2826 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2827 without a peer certificate, leading to a segfault because of an
2828 assumption that peers always have certificates. Be a little more
2829 paranoid. Problem reported by Martin Tscholak.
2831 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2832 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2833 NB: ClamAV planning to remove STREAM in "middle of 2010".
2834 CL also introduces -bmalware, various -d+acl logging additions and
2835 more caution in buffer sizes.
2837 PP/09 Implemented reverse_ip expansion operator.
2839 PP/10 Bugzilla 937: provide a "debug" ACL control.
2841 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2843 PP/12 Bugzilla 973: Implement --version.
2845 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2847 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2849 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2851 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2852 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2853 ignore trailing whitespace.
2855 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2857 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2860 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2861 Notification from Dr Andrew Aitchison.
2863 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2864 ExtendedDetectionInfo response format.
2865 Notification from John Horne.
2867 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2870 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2871 XSL and documented dependency on system catalogs, with examples of how
2874 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2877 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2878 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2879 configuration file which is writeable by the Exim user or group.
2881 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2882 of configuration files to cover files specified with the -C option if
2883 they are going to be used with root privileges, not just the default
2886 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2887 option (effectively making it always true).
2889 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2890 files to be used while preserving root privileges.
2892 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2893 that rogue child processes cannot use them.
2895 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2896 run-time user, instead of root.
2898 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2899 Exim run-time user without dropping privileges.
2901 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2902 result string, instead of calling string_vformat() twice with the same
2905 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2906 for other users. Others should always drop root privileges if they use
2907 -C on the command line, even for a whitelisted configure file.
2909 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2911 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2917 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2918 $data_path, and $header_path variables; fixed documentation bugs and
2921 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2922 exipick to access non-standard spools, including the "frozen" queue
2925 NM/01 Bugzilla 965: Support mysql stored procedures.
2926 Patch from Alain Williams
2928 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2930 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2931 Patch from Andreas Metzler
2933 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2934 Patch from Kirill Miazine
2936 NM/05 Bugzilla 671: Added umask to procmail example.
2938 JJ/03 installed exipick 20100323.0, fixing doc bug
2940 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2941 directory. Notification and patch from Dan Rosenberg.
2943 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2945 TK/02 Improve log output when DKIM signing operation fails.
2947 MH/01 Treat the transport option dkim_domain as a colon separated
2948 list, not as a single string, and sign the message with each element,
2949 omitting multiple occurences of the same signer.
2951 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2952 Bugzilla 985, 986. Patch by Simon Arlott
2954 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2955 Patch by Simon Arlott
2957 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2958 MBX locking. Notification from Dan Rosenberg.
2964 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2966 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2968 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2970 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2972 NM/04 Bugzilla 915: Flag broken perl installation during build.
2978 TK/01 Added patch by Johannes Berg that expands the main option
2979 "spamd_address" if it starts with a dollar sign.
2981 TK/02 Write list of recipients to X-Envelope-Sender header when building
2982 the mbox-format spool file for content scanning (suggested by Jakob
2985 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2986 (http://www.dcc-servers.net/) support via dccifd. Activated by
2987 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2989 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2990 by Mark Daniel Reidel <mr@df.eu>.
2992 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2993 When building exim an external PCRE library is now needed -
2994 PCRE is a system library on the majority of modern systems.
2995 See entry on PCRE_LIBS in EDITME file.
2997 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2998 conversation. Added nologin parameter to request.
2999 Patch contributed by Kirill Miazine.
3001 TF/01 Do not log submission mode rewrites if they do not change the address.
3003 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3005 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3006 log files in place. Contributed by Roberto Lima.
3008 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3010 TF/03 Bugzilla 615: When checking the local_parts router precondition
3011 after a local_part_suffix or local_part_prefix option, Exim now
3012 does not use the address's named list lookup cache, since this
3013 contains cached lookups for the whole local part.
3015 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3016 Robert Millan. Documentation is in experimental-spec.txt.
3018 TF/04 Bugzilla 668: Fix parallel build (make -j).
3020 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3022 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3023 Patch provided by Jan Srzednicki.
3025 TF/05 Leading white space used to be stripped from $spam_report which
3026 wrecked the formatting. Now it is preserved.
3028 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3029 that they are available at delivery time.
3031 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3033 TF/08 TLS error reporting now respects the incoming_interface and
3034 incoming_port log selectors.
3036 TF/09 Produce a more useful error message if an SMTP transport's hosts
3037 setting expands to an empty string.
3039 NM/06 Bugzilla 744: EXPN did not work under TLS.
3040 Patch provided by Phil Pennock.
3042 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3043 Patch provided by Richard Godbee.
3045 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3046 acl_smtp_notquit, added index entry.
3048 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3049 Patch provided by Eugene Bujak.
3051 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3052 accept(). Patch provided by Maxim Dounin.
3054 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3055 Patch provided by Phil Pennock.
3057 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3059 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3060 Patch provided by Brad "anomie" Jorsch.
3062 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3063 Patch provided by Dean Brooks.
3065 TK/05 Add native DKIM support (does not depend on external libraries).
3067 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3068 Patch provided by Graeme Fowler.
3070 NM/16 Bugzilla 851: Documentation example syntax fix.
3072 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3074 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3077 NM/19 Bugzilla 745: TLS version reporting.
3078 Patch provided by Phil Pennock.
3080 NM/20 Bugzilla 167: bool: condition support.
3081 Patch provided by Phil Pennock.
3083 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3084 clients. Patch provided by Phil Pennock.
3086 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3087 Patch provided by Brad "anomie" Jorsch.
3089 NM/23 Bugzilla 687: Fix misparses in eximstats.
3090 Patch provided by Heiko Schlittermann.
3092 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3093 Patch provided by Heiko Schlittermann.
3095 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3096 plus update to original patch.
3098 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3100 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3101 Patch provided by David Brownlee.
3103 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3105 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3107 NM/30 Bugzilla 888: TLS documentation bugfixes.
3109 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3111 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3112 Unlike the original bugzilla I have changed all shell scripts in src tree.
3114 NM/33 Bugzilla 898: Transport filter timeout fix.
3115 Patch by Todd Rinaldo.
3117 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3118 Patch by Serge Demonchaux.
3120 NM/35 Bugzilla 39: Base64 decode bug fixes.
3121 Patch by Jakob Hirsch.
3123 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3125 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3127 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3129 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3135 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3136 ALT-N's libdkim that I have put here:
3137 http://duncanthrax.net/exim-experimental/
3139 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3140 were called 'true' and it seems that C99 defines that as a reserved
3141 keyword to be used with 'bool' variable types. That means you could
3142 not include C99-style headers which use bools without triggering
3143 build errors in sieve.c.
3145 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3146 as mailq or other aliases. Changed the --help handling significantly
3147 to do whats expected. exim_usage() emits usage/help information.
3149 SC/01 Added the -bylocaldomain option to eximstats.
3151 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3153 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3155 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3161 PH/01 Another patch from the Sieve maintainer.
3163 PH/02 When an IPv6 address is converted to a string for single-key lookup
3164 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3165 dots are used instead of colons so that keys in lsearch files need not
3166 contain colons. This was done some time before quoting was made available
3167 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3168 (notated using the quote facility) so as to distinguish them from IPv4
3169 keys. This meant that lookups for IP addresses in host lists did not work
3170 for iplsearch lookups.
3172 This has been fixed by arranging for IPv6 addresses to be expressed with
3173 colons if the lookup type is iplsearch. This is not incompatible, because
3174 previously such lookups could never work.
3176 The situation is now rather anomalous, since one *can* have colons in
3177 ordinary lsearch keys. However, making the change in all cases is
3178 incompatible and would probably break a number of configurations.
3180 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3183 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3184 conversion specification without a maximum field width, thereby enabling
3185 a rogue spamd server to cause a buffer overflow. While nobody in their
3186 right mind would setup Exim to query an untrusted spamd server, an
3187 attacker that gains access to a server running spamd could potentially
3188 exploit this vulnerability to run arbitrary code as the Exim user.
3190 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3191 $primary_hostname instead of what libspf2 thinks the hosts name is.
3193 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3194 a directory entry by the name of the lookup key. Previously, if a
3195 symlink pointed to a non-existing file or a file in a directory that
3196 Exim lacked permissions to read, a lookup for a key matching that
3197 symlink would fail. Now it is enough that a matching directory entry
3198 exists, symlink or not. (Bugzilla 503.)
3200 PH/03 The body_linecount and body_zerocount variables are now exported in the
3203 PH/04 Added the $dnslist_matched variable.
3205 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3206 This means they are set thereafter only if the connection becomes
3209 PH/06 Added the client_condition to authenticators so that some can be skipped
3210 by clients under certain conditions.
3212 PH/07 The error message for a badly-placed control=no_multiline_responses left
3213 "_responses" off the end of the name.
3215 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3217 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3218 (without spaces) instead of just copying the configuration text.
3220 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3222 PH/11 Added $max_received_linelength.
3224 PH/12 Added +ignore_defer and +include_defer to host lists.
3226 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3227 way in which PCRE > 7.0 is built.
3229 PH/14 Implemented queue_only_load_latch.
3231 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3232 MAIL command. The effect was to mangle the value on 64-bit systems.
3234 PH/16 Another patch from the Sieve maintainer.
3236 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3238 PH/18 If a system quota error occurred while trying to create the file for
3239 a maildir delivery, the message "Mailbox is full" was not appended to the
3240 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3241 only to a quota excession during the actual writing of the file.
3243 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3244 characters?) which causes problems in log lines. The DN values are now
3245 passed through string_printing() before being added to log lines.
3247 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3248 and InterBase are left for another time.)
3250 PH/21 Added message_body_newlines option.
3252 PH/22 Guard against possible overflow in moan_check_errorcopy().
3254 PH/23 POSIX allows open() to be a macro; guard against that.
3256 PH/24 If the recipient of an error message contained an @ in the local part
3257 (suitably quoted, of course), incorrect values were put in $domain and
3258 $local_part during the evaluation of errors_copy.
3264 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3265 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3268 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3269 issue a MAIL command.
3271 PH/02 In an ACL statement such as
3273 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3275 if a client was not listed at all, or was listed with a value other than
3276 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3277 the condition was not true (as it should be), so access was not denied.
3278 The bug was that the ! inversion was incorrectly passed on to the second
3279 item. This has been fixed.
3281 PH/03 Added additional dnslists conditions == and =& which are different from
3282 = and & when the dns lookup returns more than one IP address.
3284 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3285 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3287 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3288 FSYNC, which compiles an option called disable_fsync that allows for
3289 bypassing fsync(). The documentation is heavily laced with warnings.
3291 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3293 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3294 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3295 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3296 including adding "make clean"; (3) Added -fPIC when compiling the test
3297 dynamically loaded module, to get rid of a warning.
3299 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3300 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3301 = 0s. The bug is otherwise harmless.
3303 PH/07 There was a bug in the dovecot authenticator such that the value of
3304 $auth1 could be overwritten, and so not correctly preserved, after a
3305 successful authentication. This usually meant that the value preserved by
3306 the server_setid option was incorrect.
3308 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3310 PH/09 Installed PCRE release 7.0.
3312 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3313 run for batched SMTP input. It is now run at the start of every message
3314 in the batch. While fixing this I discovered that the process information
3315 (output by running exiwhat) was not always getting set for -bs and -bS
3316 input. This is fixed, and it now also says "batched" for BSMTP.
3318 PH/11 Added control=no_pipelining.
3320 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3321 patch, slightly modified), and move the expansion of helo_data till after
3322 the connection is made in the smtp transport (so it can use these
3325 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3327 PH/14 Added log_selector = +pid.
3329 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3331 PH/16 Add ${if forany and ${if forall.
3333 PH/17 Added dsn_from option to vary the From: line in DSNs.
3335 PH/18 Flush SMTP output before performing a callout, unless control =
3336 no_callout_flush is set.
3338 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3339 was true (the default) a successful delivery failed to delete the retry
3340 item, thus causing premature timeout of the address. The bug is now
3343 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3345 PH/21 Long custom messages for fakedefer and fakereject are now split up
3346 into multiline responses in the same way that messages for "deny" and
3347 other ACL rejections are.
3349 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3350 with slight modification.
3352 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3353 draft, changing the syntax and factoring some duplicate code".
3355 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3356 for deliveries of the second and subsequent messages over the same SMTP
3359 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3360 ${reduce, with only minor "tidies".
3362 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3364 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3365 expansion side effects.
3367 PH/27 When a message times out after an over-quota error from an Exim-imposed
3368 quota, the bounce message says "mailbox is full". This message was not
3369 being given when it was a system quota that was exceeded. It now should
3372 MH/03 Made $recipients available in local_scan(). local_scan() already has
3373 better access to the recipient list through recipients_list[], but
3374 $recipients can be useful in postmaster-provided expansion strings.
3376 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3377 in the case of a MAIL command with additional options following the
3378 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3379 were accidentally chopped off.
3381 PH/29 SMTP synchronization checks are implemented when a command is read -
3382 there is a check that no more input is waiting when there shouldn't be
3383 any. However, for some commands, a delay in an ACL can mean that it is
3384 some time before the response is written. In this time, more input might
3385 arrive, invalidly. So now there are extra checks after an ACL has run for
3386 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3387 pipelining has not been advertised.
3389 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3391 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3392 returned in the SMTP response when smtp_return_error_details was false.
3393 This has been fixed.
3395 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3396 instead of the C I/O that was originally supplied, because problems were
3397 reported on Solaris.
3399 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3400 Exim which did not show up earlier: it was assuming that a call to
3401 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3402 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3403 was a macro that became an assignment, so it seemed to work. This has
3404 changed to a proper function call with a void return, hence the compile
3405 error. Exim's code has been fixed.
3407 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3410 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3412 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3414 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3415 because it turns out to be too big or there is a timeout) there is no
3416 "Completed" line in the log. When some messages of this type were
3417 selected by exigrep, they were listed as "not completed". Others were
3418 picked up by some special patterns. I have improved the selection
3419 criteria to be more general.
3421 PH/38 The host_find_failed option in the manualroute router can now be set
3422 to "ignore", to completely ignore a host whose IP address cannot be
3423 found. If all hosts are ignored, the behaviour is controlled by the new
3424 host_all_ignored option.
3426 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3427 homing or because of multiple MX records with /mx) generated more than
3428 one IP address, and the following item turned out to be the local host,
3429 all the secondary addresses of the first item were incorrectly removed
3430 from the list, along with the local host and any following hosts (which
3431 is what is supposed to happen).
3433 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3434 whoever called Exim into the -H file. In the case of the daemon it was
3435 behaving confusingly. When first started, it used values for whoever
3436 started the daemon, but after a SIGHUP it used the Exim user (because it
3437 calls itself on a restart). I have changed the code so that it now always
3440 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3441 message are rejected with the same error (e.g. no authentication or bad
3442 sender address), and a DATA command is nevertheless sent (as can happen
3443 with PIPELINING or a stupid MUA), the error message that was given to the
3444 RCPT commands is included in the rejection of the DATA command. This is
3445 intended to be helpful for MUAs that show only the final error to their
3448 PH/42 Another patch from the Sieve maintainer.
3450 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3451 Eximstats - Fixed some broken HTML links and added missing column headers
3453 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3454 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3456 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3458 PH/43 Yet another patch from the Sieve maintainer.
3460 PH/44 I found a way to check for a TCP/IP connection going away before sending
3461 the response to the final '.' that terminates a message, but only in the
3462 case where the client has not sent further data following the '.'
3463 (unfortunately, this is allowed). However, in many cases there won't be
3464 any further data because there won't be any more messages to send. A call
3465 to select() can be used: if it shows that the input is "ready", there is
3466 either input waiting, or the socket has been closed. An attempt to read
3467 the next input character can distinguish the two cases. Previously, Exim
3468 would have sent an OK response which the client would never have see.
3469 This could lead to message repetition. This fix should cure that, at
3470 least in a lot of common cases.
3472 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3473 advertised in response to EHLO.
3479 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3480 fixed by 4.65/MH/01 (is this a record?) are fixed:
3482 (i) An empty string was always treated as zero by the numeric comparison
3483 operators. This behaviour has been restored.
3485 (ii) It is documented that the numeric comparison operators always treat
3486 their arguments as decimal numbers. This was broken in that numbers
3487 starting with 0 were being interpreted as octal.
3489 While fixing these problems I realized that there was another issue that
3490 hadn't been noticed. Values of message_size_limit (both the global option
3491 and the transport option) were treated as octal if they started with 0.
3492 The documentation was vague. These values are now always treated as
3493 decimal, and I will make that clear in the documentation.
3499 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3500 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3503 MH/01 Don't check that the operands of numeric comparison operators are
3504 integers when their expansion is in "skipping" mode (fixes bug
3505 introduced by 4.64-PH/07).
3507 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3508 child addresses, Exim now panics and dies. Previously, because the count
3509 is held in a short int, deliveries were likely to be lost. As such a
3510 large number of recipients for a single message is ridiculous
3511 (performance will be very, very poor), I have chosen to impose a limit
3512 rather than extend the field.
3518 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3519 leftover -K file (the existence of which was triggered by #402).
3520 While we were at it, introduced process PID as part of the -K
3521 filename. This should rule out race conditions when creating
3524 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3525 processing considerably. Previous code took too long for large mails,
3526 triggering a timeout which in turn triggers #401.
3528 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3529 in the DK code in transports.c. sendfile() is not really portable,
3530 hence the _LINUX specificness.
3532 TF/01 In the add_headers option to the mail command in an Exim filter,
3533 there was a bug that Exim would claim a syntax error in any
3534 header after the first one which had an odd number of characters
3537 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3538 callout verification, Exim cached a "reject" for the entire domain. This
3539 is correct for most verifications, but it is not correct for a recipient
3540 verification with use_sender or use_postmaster set, because in that case
3541 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3542 case of MAIL FROM:<> rejection from other early rejections (e.g.
3543 rejection of HELO). When verifying a recipient using a non-null MAIL
3544 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3545 Whatever the result of the callout, the value of the domain cache is
3546 left unchanged (for any other kind of callout, getting as far as trying
3547 RCPT means that the domain itself is ok).
3549 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3552 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3553 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3554 session, the daemon ignores ECONNECT errors and logs others; it now
3555 ignores EPIPE as well.
3557 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3558 (quoted-printable decoding).
3560 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3561 later the small subsequent patch to fix an introduced bug.
3563 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3565 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3567 PH/08 An error is now given if message_size_limit is specified negative.
3569 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3570 to be given (somewhat) arbitrary names.
3572 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3575 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3576 miscellaneous code fixes
3578 PH/10 Added the log_reject_target ACL modifier to specify where to log
3581 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3582 hostname. This is wrong, because it relates to the incoming message (and
3583 probably the interface on which it is arriving) and not to the outgoing
3584 callout (which could be using a different interface). This has been
3585 changed to use the value of the helo_data option from the smtp transport
3586 instead - this is what is used when a message is actually being sent. If
3587 there is no remote transport (possible with a router that sets up host
3588 addresses), $smtp_active_hostname is used.
3590 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3591 tweaks were necessary in order to get it to work (see also 21 below):
3592 (a) The code assumed that strncpy() returns a negative number on buffer
3593 overflow, which isn't the case. Replaced with Exim's string_format()
3595 (b) There were several signed/unsigned issues. I just did the minimum
3596 hacking in of casts. There is scope for a larger refactoring.
3597 (c) The code used strcasecmp() which is not a standard C function.
3598 Replaced with Exim's strcmpic() function.
3599 (d) The code set only $1; it now sets $auth1 as well.
3600 (e) A simple test gave the error "authentication client didn't specify
3601 service in request". It would seem that Dovecot has changed its
3602 interface. Fortunately there's a specification; I followed it and
3603 changed what the client sends and it appears to be working now.
3605 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3608 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3609 address A is aliased to B and C, where B exists and C does not. Without
3610 -v the output is "A verified" because verification stops after a
3611 successful redirection if more than one address is generated. However,
3612 with -v the child addresses are also verified. Exim was outputting "A
3613 failed to verify" and then showing the successful verification for C,
3614 with its parentage. It now outputs "B failed to verify", showing B's
3615 parentage before showing the successful verification of C.
3617 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3618 look up a TXT record in a specific list after matching in a combined
3621 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3622 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3623 they consult the DNS. I had assumed they would set it the way they
3624 wanted; and indeed my experiments on Linux seem to show that in some
3625 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3626 To be on the safe side, however, I have now made the interface to
3627 host_find_byname() similar to host_find_bydns(), with an argument
3628 containing the DNS resolver options. The host_find_byname() function now
3629 sets these options at its start, just as host_find_bydns() does. The smtp
3630 transport options dns_qualify_single and dns_search_parents are passed to
3631 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3632 of host_find_byname() use the default settings of RES_DEFNAMES
3633 (qualify_single) but not RES_DNSRCH (search_parents).
3635 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3636 spool_read_header() do less string testing, by means of a preliminary
3637 switch on the second character of optional "-foo" lines. (This is
3638 overdue, caused by the large number of possibilities that now exist.
3639 Originally there were few.) While I was there, I also converted the
3640 str(n)cmp tests so they don't re-test the leading "-" and the first
3641 character, in the hope this might squeeze out yet more improvement.
3643 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3644 flag allowing group syntax was set by the header_syntax check but not
3645 turned off, possible causing trouble later; (2) The flag was not being
3646 set at all for the header_verify test, causing "group"-style headers to
3647 be rejected. I have now set it in this case, and also caused header_
3648 verify to ignore an empty address taken from a group. While doing this, I
3649 came across some other cases where the code for allowing group syntax
3650 while scanning a header line wasn't quite right (mostly, not resetting
3651 the flag correctly in the right place). These bugs could have caused
3652 trouble for malformed header lines. I hope it is now all correct.
3654 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3655 with the "reply" argument non-NULL. The code, however (which originally
3656 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3657 but it didn't always do it. This confused somebody who was copying the
3658 code for some other use. I have removed all the tests.
3660 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3661 feature that was used to support insecure browsers during the U.S. crypto
3662 embargo. It requires special client support, and Exim is probably the
3663 only MTA that supported it -- and would never use it because real RSA is
3664 always available. This code has been removed, because it had the bad
3665 effect of slowing Exim down by computing (never used) parameters for the
3666 RSA_EXPORT functionality.
3668 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3669 authenticator to fail if there's a tab character in the incoming data
3670 (there should never be unless someone is messing about, as it's supposed
3671 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3672 if the connection is using TLS or if the remote IP is the same as the
3673 local IP, and the "valid-client-cert option" if a client certificate has
3676 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3677 authenticators. This can be used for authorization after authentication
3678 succeeds. (In the case of plaintext, it servers for both authentication
3681 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3682 if any retry times were supplied.
3684 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3685 connection, where there is no client IP address to check. In this
3686 situation, the verify now always succeeds.
3688 PH/25 Applied John Jetmore's -Mset patch.
3690 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3692 PH/27 In a string expansion for a processed (not raw) header when multiple
3693 headers of the same name were present, leading whitespace was being
3694 removed from all of them, but trailing whitespace was being removed only
3695 from the last one. Now trailing whitespace is removed from each header
3696 before concatenation. Completely empty headers in a concatenation (as
3697 before) are ignored.
3699 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3700 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3702 PH/29 [Removed. This was a change that I later backed out, and forgot to
3703 correct the ChangeLog entry (that I had efficiently created) before
3704 committing the later change.]
3706 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3707 address errors (4xx response to RCPT) over the same connection as other
3708 messages routed to the same hosts. Such deliveries are always "forced",
3709 so retry times are not inspected. This resulted in far too many retries
3710 for the affected addresses. The effect occurred only when there were more
3711 hosts than the hosts_max_try setting in the smtp transport when it had
3712 the 4xx errors. Those hosts that it had tried were not added to the list
3713 of hosts for which the message was waiting, so if all were tried, there
3714 was no problem. Two fixes have been applied:
3716 (i) If there are any address or message errors in an SMTP delivery, none
3717 of the hosts (tried or untried) are now added to the list of hosts
3718 for which the message is waiting, so the message should not be a
3719 candidate for sending over the same connection that was used for a
3720 successful delivery of some other message. This seems entirely
3721 reasonable: after all the message is NOT "waiting for some host".
3722 This is so "obvious" that I'm not sure why it wasn't done
3723 previously. Hope I haven't missed anything, but it can't do any
3724 harm, as the worst effect is to miss an optimization.
3726 (ii) If, despite (i), such a delivery is accidentally attempted, the
3727 routing retry time is respected, so at least it doesn't keep
3728 hammering the server.
3730 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3731 in ${readsocket because some servers need this prod.
3733 PH/32 Added some extra debug output when updating a wait-xxx database.
3735 PH/33 The hint "could be header name not terminated by colon", which has been
3736 given for certain expansion errors for a long time, was not being given
3737 for the ${if def:h_colon_omitted{... case.
3739 PH/34 The spec says: "With one important exception, whenever a domain list is
3740 being scanned, $domain contains the subject domain." There was at least
3741 one case where this was not true.
3743 PH/35 The error "getsockname() failed: connection reset by peer" was being
3744 written to the panic log as well as the main log, but it isn't really
3745 panic-worthy as it just means the connection died rather early on. I have
3746 removed the panic log writing for the ECONNRESET error when getsockname()
3749 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3750 runs only) independently of the message's sender address. This meant
3751 that, if the 4xx error was in fact related to the sender, a different
3752 message to the same recipient with a different sender could confuse
3753 things. In particular, this can happen when sending to a greylisting
3754 server, but other circumstances could also provoke similar problems.
3755 I have changed the default so that the retry time for these errors is now
3756 based a combination of the sender and recipient addresses. This change
3757 can be overridden by setting address_retry_include_sender=false in the
3760 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3761 remote server are returned as part of bounce messages. This was not
3762 happening for LMTP over a pipe (the lmtp transport), but now it is the
3763 same for both kinds of LMTP.
3765 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3766 in header lines that were in fact CNAMEs. This is no longer the case.
3768 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3769 and queue runs started by the daemon processed all messages. This has
3770 been fixed so that -R and -S can now usefully be given with -q<time>.
3772 PH/40 Import PCRE release 6.7 (fixes some bugs).
3774 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3776 PH/42 Give an error if -q is specified more than once.
3778 PH/43 Renamed the variables $interface_address and $interface_port as
3779 $received_ip_address and $received_port, to make it clear that these
3780 values apply to message reception, and not to the outgoing interface when
3781 a message is delivered. (The old names remain recognized, of course.)
3783 PH/44 There was no timeout on the connect() call when using a Unix domain
3784 socket in the ${readsocket expansion. There now is.
3786 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3787 be meaningful with "accept".
3789 SC/01 Eximstats V1.43
3790 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3792 SC/02 Eximstats V1.44
3793 Use a glob alias rather than an array ref in the generated
3794 parser. This improves both readability and performance.
3796 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3797 Collect SpamAssassin and rejection statistics.
3798 Don't display local sender or destination tables unless
3799 there is data to show.
3800 Added average volumes into the top table text output.
3802 SC/04 Eximstats V1.46
3803 Collect data on the number of addresses (recipients)
3804 as well as the number of messages.
3806 SC/05 Eximstats V1.47
3807 Added 'Message too big' to the list of mail rejection
3808 reasons (thanks to Marco Gaiarin).
3810 SC/06 Eximstats V1.48
3811 Mainlog lines which have GMT offsets and are too short to
3812 have a flag are now skipped.
3814 SC/07 Eximstats V1.49 (Alain Williams)
3815 Added the -emptyok flag.
3817 SC/08 Eximstats V1.50
3818 Fixes for obtaining the IP address from reject messages.
3820 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3821 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3822 whitespace changes from 4.64-PH/27
3824 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3827 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3828 are found, allow negative numbers in numeric criteria)
3830 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3832 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3833 to match changes made in 4.64-PH/43
3835 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3837 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3838 "domain" is actually an IP address, and force a failure. This locks out
3839 those revolvers/nameservers that support "A-for-A" lookups, in
3840 contravention of the specifications.
3842 PH/48 When a host name was looked up from an IP address, and the subsequent
3843 forward lookup of the name timed out, the host name was left in
3844 $sender_host_name, contrary to the specification.
3846 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3847 restricted to single-key lookups, Exim was not diagnosing an error if
3848 * or *@ was used with a query-style lookup.
3850 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3852 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3853 long ago, but noone interested enough thought of it. Let's just say that
3854 the "1.1" means that there are some new functions that weren't there at
3855 some point in the past.
3857 PH/51 Error processing for expansion failure of helo_data from an smtp
3858 transport during callout processing was broken.
3860 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3861 tested/used via the -bh/-bhc/-bs options.
3863 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3864 bug, fixed in subsequent PCRE releases).
3866 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3867 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3869 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3875 SC/01 Use a glob alias rather than an array ref in eximstats generated
3876 parser. This improves both readability and performance.
3878 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3879 Don't display local sender or destination tables in eximstats unless
3880 there is data to show.
3881 Added average volumes into the eximstats top table text output.
3883 SC/03 Collect data on the number of addresses (recipients) as well
3884 as the number of messages in eximstats.
3886 TF/01 Correct an error in the documentation for the redirect router. Exim
3887 does (usually) call initgroups() when daemonizing.
3889 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3890 with consistent privilege compared to when running as a daemon.
3892 TF/03 Note in the spec that $authenticated_id is not set for local
3893 submissions from trusted users.
3895 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3896 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3898 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3899 by adding some example configuration directives to the default
3900 configuration file. A little bit of work is required to uncomment the
3901 directives and define how usernames and passwords are checked, but
3902 there is now a framework to start from.
3904 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3905 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3906 without this. I don't know how relevant this is to other LDAP libraries.
3908 PH/02 Add the verb name to the "unknown ACL verb" error.
3910 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3912 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3914 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3915 directory not expanded when it should be if an expanded home directory
3916 was set for the address (which is overridden by the transport).
3918 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3921 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3922 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3923 because it is too late at that time, and has no effect.
3925 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3926 security issue with \' (bugzilla #107). I could not use the
3927 PQescapeStringConn() function, because it needs a PGconn value as one of
3930 PH/08 When testing addresses using -bt, indicate those final addresses that
3931 are duplicates that would not cause an additional delivery. At least one
3932 person was confused, thinking that -bt output corresponded to deliveries.
3933 (Suppressing duplicates isn't a good idea as you lose the information
3934 about possibly different redirections that led to the duplicates.)
3936 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3937 systems where poll() doesn't work, in particular OS X.
3939 PH/10 Added more information to debugging output for retry time not reached.
3941 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3942 operations in malware.c.
3944 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3947 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3948 syslog_duplication set false, log lines that would normally be written
3949 both the the main log and to the reject log were not written to syslog at
3952 PH/14 In the default configuration, change the use of "message" in ACL warn
3953 statements to "add_header".
3955 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3956 not followed by a command (e.g. "seen endif").
3958 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3959 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3962 PH/17 Added extra conditions to the default value of delay_warning_condition
3966 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3967 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3968 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3971 The Auto-Submitted: and various List- headers are standardised, whereas I
3972 don't think Precedence: ever was.
3974 PH/18 Refactored debugging code in route_finduser() to show more information,
3975 in particular, the error code if getpwnam() issues one.
3977 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3978 This is apparently needed in addition to the PH/07 change above to avoid
3979 any possible encoding problems.
3981 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3982 but not after initializing Perl.
3984 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3985 output them only if debugging. By default they are written stderr,
3986 apparently, which is not desirable.
3988 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3991 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3994 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3996 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3997 authenticated or an ident call has been made. Suppress the default
3998 values for $authenticated_id and $authenticated_sender (but permit -oMai
3999 and -oMas) when testing with -bh.
4001 PH/24 Re-jigged the order of the tests in the default configuration so that the
4002 tests for valid domains and recipients precede the DNS black list and CSA
4003 tests, on the grounds that those ones are more expensive.
4005 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4006 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4007 This bug exists in every version of Exim that I still have, right back to
4010 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4011 However, an attempt to turn on case-sensitivity in a regex key by
4012 including (?-i) didn't work because the subject string was already
4013 lowercased, and the effects were non-intuitive. It turns out that a
4014 one-line patch can be used to allow (?-i) to work as expected.
4020 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4021 other effects) broke the use of negated acl sub-conditions.
4023 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4026 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4027 "Deny" causes Exim to reject the incoming connection with a 554 error.
4028 Unfortunately, if there is a major crisis, such as a disk failure,
4029 tcp-wrappers gives "deny", whereas what one would like would be some
4030 kind of temporary error. A kludge has been added to help with this.
4031 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4032 554 error is used if errno is still zero or contains ENOENT (which occurs
4033 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4036 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4038 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4039 errors. Otherwise a message that provokes a temporary error (when other
4040 messages do not) can cause a whole host to time out.
4042 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4043 addresses were routed directly to files or pipes from a redirect router.
4044 File deliveries just didn't batch; pipe deliveries might have suffered
4047 PH/06 A failure to get a lock for a hints database would erroneously always say
4048 "Failed to get write lock", even when it was really a read lock.
4050 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4051 of 0600. This has been changed to use the value of the lockfile_mode
4052 option (which defaults to 0600).
4054 PH/08 Applied small patch from the Sieve maintainer.
4056 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4057 folder from quota calculations, a direct delivery into this folder messed
4058 up the contents of the maildirsize file. This was because the regex was
4059 used only to exclude .Trash (or whatever) when the size of the mailbox
4060 was calculated. There was no check that a delivery was happening into an
4061 excluded directory. This bug has been fixed by ignoring all quota
4062 processing for deliveries into excluded directories.
4064 PH/10 Added the maildirfolder_create_regex option to appendfile.
4070 PH/01 The code for finding all the local interface addresses on a FreeBSD
4071 system running IPv6 was broken. This may well have applied to all BSD
4072 systems, as well as to others that have similar system calls. The broken
4073 code found IPv4 interfaces correctly, but gave incorrect values for the
4074 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4075 that it would not match correctly against @[] and not recognize the IPv6
4078 PH/02 The ipliteral router was not recognizing addresses of the form user@
4079 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4081 PH/03 Added disable_ipv6.
4083 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4084 decoded form, because it is most often used to construct To: headers
4085 lines in autoreplies, and the decoded form may well be syntactically
4086 invalid. However, $reply_address has leading white space removed, and all
4087 newlines turned into spaces so that the autoreply transport does not
4090 PH/05 If group was specified without a user on a router, and no group or user
4091 was specified on a transport, the group from the router was ignored.
4093 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4094 for visible compile-time settings that can be used to change these
4095 numbers, for those that want even more. Backwards compatibility with old
4096 spool files has been maintained. However, going back to a previous Exim
4097 release will lost any variables that are in spool files.
4099 PH/07 Two small changes when running in the test harness: increase delay when
4100 passing a TCP/IP connection to a new process, in case the original
4101 process has to generate a bounce, and remove special handling of
4102 127.0.0.2 (sic), which is no longer necessary.
4104 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4105 be the same on different OS.
4107 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4110 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4111 whether --show-vars was specified or not
4113 JJ/02 exipick: Added support for new ACL variable spool format introduced
4116 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4117 syntactically invalid From: or Reply-to: line, and a filter used this to
4118 generate an autoreply, and therefore failed to obtain an address for the
4119 autoreply, Exim could try to deliver to a non-existent relative file
4120 name, causing unrelated and misleading errors. What now happens is that
4121 it logs this as a hard delivery error, but does not attempt to create a
4124 PH/11 The exinext utility has a -C option for testing purposes, but although
4125 the given file was scanned by exinext itself; it wasn't being passed on
4126 when Exim was called.
4128 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4129 an end-of-file indication when reading a command response.
4131 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4132 compiled. In many other places in Exim, IPv6 addresses are always
4133 recognized, so I have changed this. It also means that IPv4 domain
4134 literals of the form [IPV4:n.n.n.n] are now always recognized.
4136 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4137 used if the router is not running as root, for example, when verifying at
4138 ACL time, or when using -bh. The debugging output from this situation was
4139 non-existent - all you got was a failure to exec. I have made two
4142 (a) Failures to set uid/gid, the current directory, or a process leader
4143 in a subprocess such as that created by queryprogram now generate
4144 suitable debugging output when -d is set.
4146 (b) The queryprogram router detects when it is not running as root,
4147 outputs suitable debugging information if -d is set, and then runs
4148 the subprocess without attempting to change uid/gid.
4150 PH/15 Minor change to Makefile for building test_host (undocumented testing
4153 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4154 additional section of a DNS packet that returns MX or SRV records.
4155 Instead, it always explicitly searches for A/AAAA records. This avoids
4156 major problems that occur when a DNS server includes only records of one
4157 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4158 fixed another bug: if SRV records were looked up and the corresponding
4159 address records were *not* found in the additional section, the port
4160 values from the SRV records were lost.
4162 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4163 using the correct key (the original address) when searching the retry
4164 rules in order to find which one to use for generating the retry hint.
4166 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4167 adding the default one. Similarly, if it contains a Reply-To: header, the
4168 errors_reply_to option, if set, is not used.
4170 PH/19 When calculating a retry time, Exim used to measure the "time since
4171 failure" by looking at the "first failed" field in the retry record. Now
4172 it does not use this if it is later than than the arrival time of the
4173 message. Instead it uses the arrival time. This makes for better
4174 behaviour in cases where some deliveries succeed, thus re-setting the
4175 "first failed" field. An example is a quota failure for a huge message
4176 when small messages continue to be delivered. Without this change, the
4177 "time since failure" will always be short, possible causing more frequent
4178 delivery attempts for the huge message than are intended.
4179 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4181 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4182 $1, $2, $3) because the numerical variables can be reset during some
4183 expansion items (e.g. "match"), thereby losing the authentication data.
4185 PH/21 Make -bV show the size of off_t variables so that the test suite can
4186 decide whether to run tests for quotas > 2G.
4188 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4189 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4190 filecount value is greater than 2G or if a quota value is greater than 2G
4191 on a system where the size of off_t is not greater than 4, a panic error
4194 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4195 never match. The debug and -bh output now contains an explicit error
4196 message indicating a malformed IPv4 address or mask.
4198 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4199 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4200 PH/23 above applies.
4202 PH/25 Do not write to syslog when running in the test harness. The only
4203 occasion when this arises is a failure to open the main or panic logs
4204 (for which there is an explicit test).
4206 PH/26 Added the /no_tell option to "control=freeze".
4208 PH/27 If a host name lookup failed very early in a connection, for example, if
4209 the IP address matched host_lookup and the reverse lookup yielded a name
4210 that did not have a forward lookup, an error message of the form "no IP
4211 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4212 could be logged. Now it outputs the IP address instead of "NULL".
4214 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4215 allows the sender and the authenticated sender to be set when
4216 submitting a message from within Exim. Since child_open_exim() is
4217 documented for local_scan(), the new function should be too.
4219 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4220 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4221 results in an empty string is now treated as unset.
4223 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4225 PH/31 Added sender_verify_fail logging option.
4227 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4228 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4229 tidied the source and removed it altogether.
4231 PH/33 When a queue run was abandoned because the load average was too high, a
4232 log line was always written; now it is written only if the queue_run log
4233 selector is set. In addition, the log line for abandonment now contains
4234 information about the queue run such as the pid. This is always present
4235 in "start" and "stop" lines but was omitted from the "abandon" line.
4237 PH/34 Omit spaces between a header name and the colon in the error message that
4238 is given when verify = headers_syntax fails (if there are lots of them,
4239 the message gets confusing).
4241 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4242 names, as there are now some PTR records that contain slashes. This check
4243 is only to protect against broken name servers that fall over on strange
4244 characters, so the fact that it applies to all lookups doesn't matter.
4246 PH/36 Now that the new test suite is complete, we can remove some of the
4247 special code in Exim that was needed for the old test suite. For example,
4248 sorting DNS records because real resolvers return them in an arbitrary
4249 order. The new test suite's fake resolver always returns records in the
4252 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4253 bounces) except when queue_only is set, to avoid logging races between
4254 the different processes.
4256 PH/38 Panic-die if .include specifies a non-absolute path.
4258 PH/39 A tweak to the "H" retry rule from its user.
4260 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4261 a label. They prevented compilation on older perls.
4263 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4264 a warning to be raised on newish perls.
4266 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4267 on queue. Changes to match documented behaviour of showing count of
4268 messages matching specified criteria.
4270 PH/40 Changed the default ident timeout from 30s to 5s.
4272 PH/41 Added support for the use of login_cap features, on those BSD systems
4273 that have them, for controlling the resources used by pipe deliveries.
4275 PH/42 The content-scanning code uses fopen() to create files in which to put
4276 message data. Previously it was not paying any attention to the mode of
4277 the files. Exim runs with umask(0) because the rest of the code creates
4278 files with open(), and sets the required mode explicitly. Thus, these
4279 files were ending up world-writeable. This was not a big issue, because,
4280 being within the spool directory, they were not world-accessible. I have
4281 created a function called modefopen, which takes an additional mode
4282 argument. It sets umask(777), creates the file, chmods it to the required
4283 mode, then resets the umask. All the relevant calls to fopen() in the
4284 content scanning code have been changed to use this function.
4286 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4287 to 24 hours. This avoids potential overflow problems when processing G
4288 and H retry rules. I suspect nobody ever tinkers with this value.
4290 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4292 PH/45 When the plaintext authenticator is running as a client, the server's
4293 challenges are checked to ensure they are valid base64 strings. By
4294 default, the authentication attempt is cancelled if an invalid string is
4295 received. Setting client_ignore_invalid_base64 true ignores these errors.
4296 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4297 they are received. Thus, the responses can be made to depend on the
4298 challenges. If an invalid string is ignored, an empty string is placed in
4301 PH/46 Messages that are created by the autoreply transport now contains a
4302 References: header, in accordance with RFCs 2822 and 3834.
4304 PH/47 Added authenticated_sender_force to the smtp transport.
4306 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4308 PH/49 Installed latest patch from the Sieve maintainer.
4310 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4311 was also set, the appendfile transport was unnecessarily scanning a
4312 directory of message files (e.g. for maildir delivery) to find the count
4313 of files (along with the size), even though it did not need this
4314 information. It now does the scan only if it needs to find either the
4315 size of the count of files.
4317 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4319 PH/52 Two bugs concerned with error handling when the smtp transport is
4322 (i) Exim was not creating retry information for temporary errors given
4323 for individual recipients after the DATA command when the smtp transport
4324 was used in LMTP mode. This meant that they could be retried too
4325 frequently, and not timed out correctly.
4327 (ii) Exim was setting the flag that allows error details to be returned
4328 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4329 recipients that were returned after the DATA command.
4331 PH/53 This is related to PH/52, but is more general: for any failing address,
4332 when detailed error information was permitted to be returned to the
4333 sender, but the error was temporary, then after the final timeout, only
4334 "retry timeout exceeded" was returned. Now it returns the full error as
4335 well as "retry timeout exceeded".
4337 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4338 do this, and (what is worse) MTAs that accept it.
4340 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4341 will now be deprecated.
4343 PH/56 New os.c-cygwin from the Cygwin maintainer.
4345 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4346 formats (previously only available in exim formats via -bpr, -bpru,
4347 and -bpra. Now also available in native and exiqgrep formats)
4349 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4350 with very large, slow to parse queues
4352 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4354 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4356 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4357 responses to authentication challenges, though it was showing the
4358 challenges; (ii) I've removed the CR characters from the debug output for
4361 PH/58 Allow for the insertion of a newline as well as a space when a string
4362 is turned into more than one encoded-word during RFC 2047 encoding. The
4363 Sieve code now uses this.
4365 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4366 data_4xx, lost_connection, tls_required.
4368 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4369 message was being sent as an SMTP response.
4371 PH/61 Add -l and -k options to exicyclog.
4373 PH/62 When verifying, if an address was redirected to one new address, so that
4374 verification continued, and the new address failed or deferred after
4375 having set something in $address_data, the value of $address_data was not
4376 passed back to the ACL. This was different to the case when no
4377 redirection occurred. The value is now passed back in both cases.
4379 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4380 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4381 use login_cap.h, so on its own it isn't the distinguishing feature. The
4382 new name refers directly to the setclassresources() function.
4384 PH/65 Added configuration files for NetBSD3.
4386 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4388 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4389 is preferred over IPv4.
4391 PH/68 The bounce_return_message and bounce_return_body options were not being
4392 honoured for bounces generated during the reception of non-SMTP messages.
4393 In particular, this applied to messages rejected by the ACL. This bug has
4394 been fixed. However, if bounce_return_message is true and bounce_return_
4395 body is false, the headers that are returned for a non-SMTP message
4396 include only those that have been read before the error was detected.
4397 (In the case of an ACL rejection, they have all been read.)
4399 PH/69 The HTML version of the specification is now built in a directory called
4400 spec_html instead of spec.html, because the latter looks like a path with
4401 a MIME-type, and this confuses some software.
4403 PH/70 Catch two compiler warnings in sieve.c.
4405 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4406 function verify_get_ident() calls ip_connect() to connect a socket, but
4407 if the "connect()" function timed out, ip_connect() used to close the
4408 socket. However, verify_get_ident() also closes the socket later, and in
4409 between Exim writes to the log, which may get opened at this point. When
4410 the socket was closed in ip_connect(), the log could get the same file
4411 descriptor number as the socket. This naturally causes chaos. The fix is
4412 not to close the socket in ip_connect(); the socket should be closed by
4413 the function that creates it. There was only one place in the code where
4414 this was missing, in the iplookup router, which I don't think anybody now
4415 uses, but I've fixed it anyway.
4417 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4418 well as to direct DNS lookups. Otherwise the handling of names in host
4419 lists is inconsistent and therefore confusing.
4425 PH/01 Two changes to the default runtime configuration:
4427 (1) Move the checks for relay_from_hosts and authenticated clients from
4428 after to before the (commented out) DNS black list checks.
4430 (2) Add control=submission to the relay_from_hosts and authenticated
4431 clients checks, on the grounds that messages accepted by these
4432 statements are most likely to be submissions.
4434 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4436 (1) Generate an error if the third argument for the ${prvs expansion is
4439 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4442 (3) Reset the variables that are obtained from the first argument of
4443 ${prvscheck and used in the second argument before leaving the code,
4444 because their memory is reclaimed, so using them afterwards may do
4447 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4448 one (it's much easier than Tom thought :-).
4450 (5) Because of (4), we can now allow for the use of $prvscheck_result
4451 inside the third argument.
4453 PH/03 For some reason, the default setting of PATH when running a command from
4454 a pipe transport was just "/usr/bin". I have changed it to
4457 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4458 anything to be listed in the output from -bV.
4460 PH/05 When a filter generated an autoreply, the entire To: header line was
4461 quoted in the delivery log line, like this:
4463 => >A.N.Other <ano@some.domain> <original@ddress> ...
4465 This has been changed so that it extracts the operative address. There
4466 may be more than one such address. If so, they are comma-separated, like
4469 => >ano@some.domain,ona@other.domain <original@ddress> ...
4471 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4472 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4473 not being looked up in the rDNS to get a host name, Exim was showing the
4474 IP address twice in Received: lines, even though the IP addresses were
4475 identical. For example:
4477 Received: from [1.2.3.4] (helo=[1.2.3.4])
4479 However, if the real host name was known, it was omitting the HELO data
4480 if it matched the actual IP address. This has been tidied up so that it
4481 doesn't show the same IP address twice.
4483 PH/07 When both +timestamp and +memory debugging was on, the value given by
4484 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4485 called by the debug printing, thereby overwriting the timestamp buffer.
4486 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4488 PH/08 When the original message was included in an autoreply transport, it
4489 always said "this is a copy of the message, including all the headers",
4490 even if body_only or headers_only was set. It now gives an appropriate
4493 PH/09 Applied a patch from the Sieve maintainer which:
4495 o fixes some comments
4496 o adds the (disabled) notify extension core
4497 o adds some debug output for the result of if/elsif tests
4498 o points to the current vacation draft in the documentation
4499 and documents the missing references header update
4503 o fixes a bug in processing the envelope test (when testing
4504 multiple envelope elements, the last element determined the
4507 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4508 Electronic Mail") by including:
4510 Auto-submitted: auto-generated
4512 in the messages that it generates (bounce messages and others, such as
4513 warnings). In the case of bounce messages for non-SMTP messages, there was
4514 also a typo: it was using "Auto_submitted" (underscore instead of
4515 hyphen). Since every message generated by Exim is necessarily in response
4516 to another message, thes have all been changed to:
4518 Auto-Submitted: auto-replied
4520 in accordance with these statements in the RFC:
4522 The auto-replied keyword:
4524 - SHOULD be used on messages sent in direct response to another
4525 message by an automatic process,
4527 - MUST NOT be used on manually-generated messages,
4529 - MAY be used on Delivery Status Notifications (DSNs) and Message
4530 Disposition Notifications (MDNs),
4532 - MUST NOT be used on messages generated by automatic or periodic
4533 processes, except for messages which are automatic responses to
4536 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4537 to the default Received: header definition.
4539 PH/12 Added log selector acl_warn_skipped (default on).
4541 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4542 variables because (a) they are in the wrong storage pool and (b) even if
4543 they were copied, it wouldn't work properly because of the caching.
4545 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4546 checking when decoding. Apparently there are clients that generate
4547 overlong encoded strings. Why am I not surprised?
4549 PH/15 If the first argument of "${if match_address" was not empty, but did not
4550 contain an "@" character, Exim crashed. Now it writes a panic log message
4551 and treats the condition as false.
4553 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4555 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4556 extension "envelope-auth". The code is finished and in agreement with
4557 other implementations, but there is no documentation so far and in fact,
4558 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4559 not changing the active code.
4561 Print executed "if" and "elsif" statements when debugging is used. This
4562 helps a great deal to understand what a filter does.
4564 Document more things not specified clearly in RFC3028. I had all this
4565 sorted out, when out of a sudden new issues came to my mind. Oops."
4567 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4570 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4571 canonical form (as documented). However, after a host name lookup from
4572 the IP address, check_host() was doing a simple string comparison with
4573 addresses acquired from the DNS when checking that the found name did
4574 have the original IP as one of its addresses. Since any found IPv6
4575 addresses are likely to be in abbreviated form, the comparison could
4576 fail. Luckily, there already exists a function for doing the comparison
4577 by converting both addresses to binary, so now that is used instead of
4578 the text comparison.
4580 PH/20 There was another similar case to PH/19, when a complete host name was
4581 given in a host list; looking up its IP address could give an abbreviated
4582 form, whereas the current host's name might or might not be abbreviated.
4583 The same fix has been applied.
4589 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4590 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4593 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4594 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4596 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4598 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4599 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4600 Also, for "Auto-Submitted", treat anything other than "no" as
4601 non-personal, in accordance with RFC 3834. (Previously it treated
4602 anything starting "auto-" as non-personal.)
4604 TF/01 The control=submission/name=... option had a problem with syntax
4605 errors if the name included a slash character. The /name= option
4606 now slurps the rest of the string, so it can include any characters
4607 but it must come last in the list of options (after /sender_retain
4610 PH/05 Some modifications to the interface to the fake nameserver for the new
4618 TK/01 Added the "success_on_redirect" address verification option. See
4619 NewStuff for rationale and an example.
4621 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4623 PH/02 Patch to exigrep to allow it to work on syslog lines.
4625 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4626 fread() to skip over the body file's header line, because in Cygwin the
4627 header line is locked and is inaccessible.
4629 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4630 co-exist for some time) to make it clear that it is the Exim ID that is
4631 referenced, not the Message-ID: header line.
4633 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4634 string_format() function, because snprintf() does not exist on all
4637 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4638 ${stat: expansion item.
4640 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4641 protocol synchronization error", to keep the pedants happy.
4643 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4644 well as for IRIX systems, when gcc is being used. See the host.c source
4647 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4649 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4652 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4653 transport and to the smtp transport in LMTP mode.
4655 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4657 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4658 run a filter in a subprocess. This could lead to confusion in subsequent
4659 lookups in the parent process. There should also be a search_tidyup() at
4660 the end of the subprocess.
4662 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4663 only if the host matched helo_try_verify_hosts, which caused the
4664 verification to occur when the EHLO/HELO command was issued. The ACL just
4665 tested the remembered result. Now, if a previous verification attempt has
4666 not happened, "verify = helo" does it there and then.
4668 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4670 TK/03 Fix log output including CR from clamd.
4672 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4673 exist provoked a memory error which could cause a segfault.
4675 PH/15 Installed PCRE 6.2
4677 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4679 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4680 of the problem. Specifically, suggested +O2 rather than +O1 for the
4683 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4685 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4686 record, the port was not correctly logged when the outgoing_port log
4687 selector was set (it logged the transort's default port).
4689 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4690 fallback_hosts, and "hosts" in the smtp transport.
4692 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4693 host errors such as "Connection refused".
4695 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4696 authentication with radiusclient 0.4.9:
4698 - Error returned from rc_read_config was caught wrongly
4699 - Username/password not passed on to radius server due to wrong length.
4701 The presumption is that some radiusclient API changes for 4.51/PH/17
4702 were not taken care of correctly. The code is still untested by me (my
4703 Linux distribution still has 0.3.2 of radiusclient), but it was
4704 contributed by a Radius user.
4706 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4707 expanding the "port" option of the smtp transport.
4709 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4710 while reading a MIME header. Thanks to Tom Hughes for a patch.
4712 PH/24 Include config.h inside local_scan.h so that configuration settings are
4715 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4716 that in an ACL for RCPT (for example), you can examine exactly what was
4719 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4720 commands, but it was not correctly comparing the address with the actual
4721 client host address. Thus, it would show the EHLO address in Received:
4722 header lines when this was not necessary.
4724 PH/27 Added the % operator to ${eval:}.
4726 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4727 it should be ignoring failures (because with -C, for example, it has lost
4728 privilege). It wasn't ignoring creation failures other than "already
4731 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4734 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4735 because an input line was too long, either on its own, or by virtue of
4736 too many continuations, the temporary file was not being removed, and the
4737 return code was incorrect.
4739 PH/31 Missing "BOOL" in function definition in filtertest.c.
4741 PH/32 Applied Sieve patches from the maintainer.
4743 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4745 PH/33 Added "verify = not_blind".
4747 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4748 Local/Makefile (with some defaults set). These are used in built scripts
4749 such as exicyclog, but they have never been used in the exim_install
4750 script (though there are many overriding facilities there). I have
4751 arranged that the exim_install script now takes note of these two
4754 PH/35 Installed configuration files for Dragonfly.
4756 PH/36 When a locally submitted message by a trusted user did not contain a
4757 From: header, and the sender address was obtained from -f or from an SMTP
4758 MAIL command, and the trusted user did not use -F to supply a sender
4759 name, $originator_name was incorrectly used when constructing a From:
4760 header. Furthermore, $originator_name was used for submission mode
4761 messages from external hosts without From: headers in a similar way,
4762 which is clearly wrong.
4764 PH/37 Added control=suppress_local_fixups.
4766 PH/38 When log_selector = +received_sender was set, and the addition of the
4767 sender made the log line's construction buffer exactly full, or one byte
4768 less than full, an overflow happened when the terminating "\n" was
4771 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4772 when the result of a list match is failure because a DNS lookup failed.
4774 PH/40 RM_COMMAND is now used in the building process.
4776 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4777 the "build-* directories that it finds.
4779 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4780 address in a domain literal was a prefix of an interface address.
4782 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4783 when verifying a sender address, unless rewrite_headers is false.
4785 PH/44 Wrote a long comment about why errors_to addresses are verified as
4786 recipients, not senders.
4788 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4789 the ratelimit ACL was added.
4791 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4793 PH/46 Added extra information about PostgreSQL errors to the error string.
4795 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4796 avoiding the need to install special zones in a real server. This is
4797 backwards compatible; if it can't find the fake resolver, it drops back.
4798 Thus, both old and new test suites can be run.
4800 TF/02 Added util/ratelimit.pl
4802 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4803 clock is set back in time.
4805 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4806 Candler <B.Candler@pobox.com>.
4808 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4809 OK for addresses that are the result of redirections.
4811 PH/48 A number of further additions for the benefit of the new test suite,
4812 including a fake gethostbyname() that interfaces to the fake DNS resolver
4815 TF/06 The fix for widen_domains has also been applied to qualify_single and
4816 search_parents which are the other dnslookup options that can cause
4819 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4822 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4824 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4825 service" if it encountered an MX record with an empty target hostname.
4826 The message is now "an MX or SRV record indicated no SMTP service".
4828 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4829 if the DNS of the sending site is misconfigured. This is quite a
4830 common situation. This change restores the behaviour of treating a
4831 helo verification defer as a failure.
4833 PH/51 If self=fail was set on a router, the bounce message did not include the
4834 actual error message.
4840 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4842 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4843 command itself ended in error, the underlying message about the transport
4844 filter timeout was being overwritten with the pipe command error. Now the
4845 underlying error message should be appended to the second error message.
4847 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4849 PH/02 Reduce the amount of output that "make" produces by default. Full output
4850 can still be requested.
4852 PH/03 The warning log line about a condition test deferring for a "warn" verb
4853 was being output only once per connection, rather than after each
4854 occurrence (because it was using the same function as for successful
4855 "warn" verbs). This seems wrong, so I have changed it.
4857 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4858 it should not have, which might have caused a crash in the right
4859 circumstances, but probably never did.
4861 PH/04 Installed a modified version of Tony Finch's patch to make submission
4862 mode fix the return path as well as the Sender: header line, and to
4863 add a /name= option so that you can make the user's friendly name appear
4866 TF/03 Added the control = fakedefer ACL modifier.
4868 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4869 Mark Lowes for thorough testing.
4871 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4873 TK/03 Merged latest SRS patch from Miles Wilton.
4875 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4876 with the definition in sysexits.h (which is #included earlier).
4877 Fortunately, Exim does not actually use EX_OK. The code used to try to
4878 preserve the sysexits.h value, by assuming that macro definitions were
4879 scanned for macro replacements. I have been disabused of this notion,
4880 so now the code just undefines EX_OK before #including unistd.h.
4882 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4883 in the smtp transport. When a block could not be written in a single
4884 write() function, the timeout was being re-applied to each part-write.
4885 This seems wrong - if the receiver was accepting one byte at a time it
4886 would take for ever. The timeout is now adjusted when this happens. It
4887 doesn't have to be particularly precise.
4889 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4890 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4892 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4893 without a domain if the check to <postmaster@domain> fails.
4895 SC/01 Eximstats: added -xls and the ability to specify output files
4896 (patch written by Frank Heydlauf).
4898 SC/02 Eximstats: use FileHandles for outputting results.
4900 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4902 SC/04 Eximstats: fixed display of large numbers with -nvr option
4904 SC/05 Eximstats: fixed merging of reports with empty tables.
4906 SC/06 Eximstats: added the -include_original_destination flag
4908 SC/07 Eximstats: removed tabs and trailing whitespace.
4910 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4912 TK/06 MBOX spool code: Add real "From " MBOX separator line
4913 so the .eml file is really in mbox format (even though
4914 most programs do not really care). Patch from Alex Miller.
4916 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4917 The latter is generated from $received_to and is only set if the
4918 message has one envelope recipient. SA can use these headers,
4919 obviously out-of-the-box. Patch from Alex Miller.
4921 PH/08 The ${def test on a variable was returning false if the variable's
4922 value was "0", contrary to what the specification has always said!
4923 The result should be true unless the variable is empty.
4925 PH/09 The syntax error of a character other than { following "${if
4926 def:variable_name" (after optional whitespace) was not being diagnosed.
4927 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4928 accidental colon was present, for example, could give incorrect results.
4930 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4931 result is used (not including appendfile, where other changes are about
4934 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4935 This involved changing a lot of size variables from int to off_t. It
4936 should work with maildirs and everything.
4938 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4939 spamd dying while we are connected to it.
4941 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4944 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4945 function for BDB 4.3.
4947 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4949 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4950 the sources for PCRE have been re-arranged and more files are now
4953 PH/15 The code I had for printing potentially long long variables in PH/11
4954 above was not the best (it lost precision). The length of off_t variables
4955 is now inspected at build time, and an appropriate printing format (%ld
4956 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4957 to be "long long int" or "long int". This is needed for the internal
4958 formatting function string_vformat().
4960 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4961 the configuration file to be ":syslog", then the script "guesses" where
4962 the logs files are, rather than using the compiled in default. In our
4963 case the guess is not the same as the compiled default, so the script
4964 suddenly stopped working when I started to use syslog. The patch checks
4965 to see if log_file_path is "". If so, it attempts to read it from exim
4966 with no configuration file to get the compiled in version, before it
4967 falls back to the previous guessing code."
4969 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4970 implementing BATV in an Exim configuration. See NewStuff for the gory
4973 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4974 Makefile that are specific to HP-UX.
4976 PH/18 If the "use_postmaster" option was set for a recipient callout together
4977 with the "random" option, the postmaster address was used as the MAIL
4978 FROM address for the random test, but not for the subsequent recipient
4979 test. It is now used for both.
4981 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4982 patch removes a few documentation additions to RFC 3028, because the
4983 latest draft now contains them. It adds the new en;ascii-case comparator
4984 and a new error check for 8bit text in MIME parts. Comparator and
4985 require names are now matched exactly. I enabled the subaddress
4986 extension, but it is not well tested yet (read: it works for me)."
4988 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4989 rework some of the code of TK/09 above to avoid the hardwired use of
4990 "%lld" and "long long". Replaced the call to snprintf() with a call to
4993 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4994 records point to non-existent hosts", "retry timeout exceeded", and
4995 "retry time not reached for any host after a long failure period".
4997 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4998 experimental DomainKeys support:
5000 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5001 (2) On an error such as an illegally used "control", the wrong name for
5002 the control was given.
5004 These problems did NOT occur unless DomainKeys support was compiled.
5006 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5008 PH/24 Added ${if match_ip condition.
5010 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5011 databases so that it will be absolutely obvious if a crash occurs in the
5012 DB library. This is a regular occurrence (often caused by mis-matched
5015 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5016 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5017 picked up on a user's system that detects such things. There doesn't seem
5018 to be a gcc warning option for this - only an attribute that has to be
5019 put on the function's prototype. It seems that in Fedora Core 4 they have
5020 set this on a number of new functions. No doubt there will be more in due
5023 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5024 specify a transport. However, if an address that was verified by such a
5025 router was the subject of a callout, Exim crashed because it tried to
5026 read the rcpt_include_affixes from the non-existent transport. Now it
5027 just assumes that the setting of that option is false. This bug was
5028 introduced by 4.51/PH/31.
5030 PH/28 Changed -d+all to exclude +memory, because that information is very
5031 rarely of interest, but it makes the output a lot bigger. People tend to
5032 do -d+all out of habit.
5034 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5035 code in os-type was giving problems when libc.so lives in lib64, like on
5038 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5039 aren't the modern standard, and it seems that some systems' include files
5040 don't always have them. Exim was already checking for some of the newer
5041 ones like T_AAAA, and defining it itself. I've added checks for all the
5042 record types that Exim uses.
5044 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5045 not automatically generating a new one, as it is supposed to. This
5046 prevented TLS from working. If the file did exist, but contained invalid
5047 data, a new version was generated, as expected. It was only the case of a
5048 non-existent file that was broken.
5050 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5051 with a change in libdomainkeys > 0.64.
5053 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5054 from DNS. If the selector record carries the flag, it now has
5055 precedence over the domain-wide flag.
5057 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5059 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5060 the use of a port name that isn't defined in /etc/services) occurred, the
5061 message was deferred as in a normal delivery, and thus remained on the
5062 spool, instead of being failed because of the mua_wrapper setting. This
5063 is now fixed, and I tidied up some of the mua_wrapper messages at the
5066 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5067 the messages in a hash of arrays rather than using individual hashes.
5068 This is a bit cleaner and results in dramatic memory savings, albeit
5069 at a slight CPU cost.
5071 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5072 as requested by Marc Sherman.
5074 SC/10 Eximstats: added histograms for user specified patterns as requested
5077 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5079 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5080 fopen() in the content-scanning modules that did not already have it.
5086 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5087 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5089 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5091 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5093 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5094 to test Sieve filters that use "vacation".
5096 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5097 that changes the way the GnuTLS parameters are stored in the cache file.
5098 The new format can be generated externally. For backward compatibility,
5099 if the data in the cache doesn't make sense, Exim assumes it has read an
5100 old-format file, and it generates new data and writes a new file. This
5101 means that you can't go back to an older release without removing the
5104 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5105 work if there are any delivery delays because "one_time" forces the
5106 parent to be marked "delivered", so its unseen clone is never tried
5107 again. For this reason, Exim now forbids the simultaneous setting of
5110 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5111 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5112 entry if you want to know the details. The fix, however, neglected to
5113 consider the case when local delivery batching is involved. The test for
5114 "previously delivered" was not happening when checking to see if an
5115 address could be batched with a previous (undelivered) one; under
5116 certain circumstances this could lead to multiple deliveries to the same
5119 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5120 in its include files, and this causes problems building Exim.
5122 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5123 header_syntax) but Exim was just ignoring anything given after a slash.
5124 In particular, this caused confusion with an attempt to use "verify =
5125 reverse_host_lookup/defer_ok". An error is now given when options are
5126 supplied for verify items that do not have them. (Maybe reverse_host_
5127 lookup should have a defer_ok option, but that's a different point.)
5129 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5130 defined by RFC 821) to 2048, because there were problems with some AUTH
5131 commands, and RFC 1869 says the size should be increased for extended
5132 SMTP commands that take arguments.
5134 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5137 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5138 "unknown" error; now it says that the functionality isn't in the binary.
5140 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5141 an address' error message when a string expansion fails (syntax or
5142 whatever). Otherwise the password may appear in the log. Following change
5143 PH/42 below, there is no longer a chance of it appearing in a bounce
5146 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5148 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5149 of its addresses was ever tried. (Bugzilla bug #2.)
5151 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5152 the result incorrectly in the debug output. (It correctly added a newline
5153 to what was transported.)
5155 TF/01 Added $received_time.
5157 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5158 commented out examples of how to interface to a virus scanner and to
5159 SpamAssassin. Also added commented examples of av_scanner and
5160 spamd_address settings.
5162 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5163 and controls are allowed in which ACLs. There were a couple of minor
5164 errors. Some of the entries in the conditions table (which is a table of
5165 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5166 negation of where the condition IS allowed.
5168 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5170 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5171 header file does not have a version number, so I've had to invent a new
5172 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5173 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5174 radiusclient), but it was contributed by a Radius user.
5176 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5177 files or directories, for OpenSSL.
5179 PH/19 When an Exim process that is running as root has to create an Exim log
5180 file, it does so in a subprocess that runs as exim:exim so as to get the
5181 ownership right at creation (otherwise, other Exim processes might see
5182 the file with the wrong ownership). There was no test for failure of this
5183 fork() call, which would lead to the process getting stuck as it waited
5184 for a non-existent subprocess. Forks do occasionally fail when resources
5185 run out. I reviewed all the other calls to fork(); they all seem to check
5188 PH/20 When checking for unexpected SMTP input at connect time (before writing
5189 the banner), Exim was not dealing correctly with a non-positive return
5190 from the read() function. If the client had disconnected by this time,
5191 the result was a log entry for a synchronization error with an empty
5192 string after "input=" when read() returned zero. If read() returned -1
5193 (an event I could not check), uninitialized data bytes were printed.
5194 There were reports of junk text (parts of files, etc) appearing after
5197 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5199 PH/22 Added support for macro redefinition, and (re)definition in between
5200 driver and ACL definitions.
5202 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5203 forgetting to use the resulting value; it was using the unexpanded value.
5205 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5206 hadn't been configured. The fix is from Juergen Kreileder, who
5207 understands it better than I do:
5209 "Here's what I see happening with three configured cyrus_sasl
5210 authenticators configured (plain, login, cram-md5):
5212 On startup auth_cyrus_sasl_init() gets called for each of these.
5213 This means three calls to sasl_listmech() without a specified mech_list.
5214 => SASL tests which mechs of all available mechs actually work
5215 => three warnings about OTP not working
5216 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5218 With the patch, sasl_listmech() also gets called three times. But now
5219 SASL's mech_list option is set to the server_mech specified in the the
5220 authenticator. Or in other words, the answer from sasl_listmech()
5221 gets limited to just the mech you're testing for (which is different
5223 => the return list contains just 'plain' or 'login', 'cram-md5' or
5224 nothing depending on the value of ob->server_mech.
5226 I've just tested the patch: Authentication still works fine,
5227 unavailable mechs specified in the exim configuration are still
5228 caught, and the auth.log warnings about OTP are gone."
5230 PH/25 When debugging is enabled, the contents of the command line are added
5231 to the debugging output, even when log_selector=+arguments is not
5234 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5235 answer is "GNU", and only if the return is "GNU/something" is the answer
5238 PH/27 $acl_verify_message is now set immediately after the failure of a
5239 verification in an ACL, and so is available in subsequent modifiers. In
5240 particular, the message can be preserved by coding like this:
5242 warn !verify = sender
5243 set acl_m0 = $acl_verify_message
5245 Previously, $acl_verify_message was set only while expanding "message"
5246 and "log_message" when a very denied access.
5248 PH/28 Modified OS/os.c-Linux with
5250 -#ifndef OS_LOAD_AVERAGE
5251 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5253 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5254 nomenclature these days.)
5256 PH/29 Installed patch from the Sieve maintainer that adds the options
5257 sieve_useraddress and sieve_subaddress to the redirect router.
5259 PH/30 In these circumstances:
5260 . Two addresses routed to the same list of hosts;
5261 . First host does not offer TLS;
5262 . First host accepts first address;
5263 . First host gives temporary error to second address;
5264 . Second host offers TLS and a TLS session is established;
5265 . Second host accepts second address.
5266 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5267 and peerdn, if requested) that were in fact used only for the second
5270 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5271 attention to any local part prefix or suffix that was matched by the
5272 router that accepted the address. It now behaves in the same way as it
5273 does for delivery: the affixes are removed from the local part unless
5274 rcpt_include_affixes is set on the transport.
5276 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5277 timeout during the DATA phase of an incoming message.
5279 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5280 applied a patch sanctioned by the Sieve maintainer.
5282 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5283 the uid or gid is negative. A case of a negative gid caused this to be
5284 noticed. The fix allows for either to be negative.
5286 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5287 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5290 PH/36 The change PH/12 above was broken. Fixed it.
5292 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5293 the grounds that routing the same address twice would always produce the
5294 same answer. This might have been true once, but it is certainly no
5295 longer true now. Routing a child address may depend on the previous
5296 routing that produced that child. Some complicated redirection strategies
5297 went wrong when messages had multiple recipients, and made Exim's
5298 behaviour dependent on the order in which the addresses were given.
5300 I have moved the duplicate checking until after the routing is complete.
5301 Exim scans the addresses that are assigned to local and remote
5302 transports, and removes any duplicates. This means that more work will be
5303 done, as duplicates will always all be routed, but duplicates are
5304 presumably rare, so I don't expect this is of any significance.
5306 For deliveries to pipes, files, and autoreplies, the duplicate checking
5307 still happens during the routing process, since they are not going to be
5310 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5311 It corrects a timeout issue with spamd. This is Ian's comment: "The
5312 background is that sometimes spamd either never reads data from a
5313 connection it has accepted, or it never writes response data. The exiscan
5314 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5315 blindly assumes that writes won't block so it may never time out."
5317 PH/39 Allow G after quota size as well as K and M.
5319 PH/40 The value set for $authenticated_id in an authenticator may not contain
5320 binary zeroes or newlines because the value is written to log lines and
5321 to spool files. There was no check on this. Now the value is run through
5322 the string_printing() function so that such characters are converted to
5323 printable escape sequences.
5325 PH/41 $message_linecount is a new variable that contains the total number of
5326 lines in the message. Compare $body_linecount, which is the count for the
5329 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5330 bounce and delay warning messages, except in certain special cases, which
5333 (a) An SMTP error message from a remote host;
5334 (b) A message specified in a :fail: redirection;
5335 (c) A message specified in a "fail" command in a system filter;
5336 (d) A message specified in a FAIL return from the queryprogram router;
5337 (e) A message specified by the cannot_route_message router option.
5339 In these cases only, Exim does include the error details in bounce and
5340 warning messages. There are also a few cases where bland messages such
5341 as "unrouteable address" or "local delivery error" are given.
5343 PH/43 $value is now also set for the "else" part of a ${run expansion.
5345 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5346 being worked on, but at least Exim now implements the latest version to
5349 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5350 process to complete was treated as a delivery failure, a timeout while
5351 writing the message to the pipe was logged, but erroneously treated as a
5352 successful delivery. Such timeouts include transport filter timeouts. For
5353 consistency with the overall process timeout, these timeouts are now
5354 treated as errors, giving rise to delivery failures by default. However,
5355 there is now a new Boolean option for the pipe transport called
5356 timeout_defer, which, if set TRUE, converts the failures into defers for
5357 both kinds of timeout. A transport filter timeout is now identified in
5360 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5361 systems where "make" and "gmake" are different, calling "gmake" at top
5362 level broke things. I've arranged for the value of $(MAKE) to be passed
5363 from the Makefile to this script so that it can call the same version of
5367 A note about Exim versions 4.44 and 4.50
5368 ----------------------------------------
5370 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5371 changes of various kinds. As a consequence, a big documentation update was
5372 needed. This delayed the release for rather longer than seemed good, especially
5373 in the light of a couple of (minor) security issues. Therefore, the changes
5374 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5375 release. So 4.44 and 4.50 are in effect two different branches that both start
5378 I have left the 4.50 change log unchanged; it contains all the changes since
5379 4.43. The change log for 4.44 is below; many of its items are identical to
5380 those for 4.50. This seems to be the most sensible way to preserve the
5381 historical information.
5387 1. Minor wording change to the doc/README.SIEVE file.
5389 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5390 computation of the current number of files was incorrect.
5392 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5393 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5396 4. Give more explanation in the error message when the command for a transport
5397 filter fails to execute.
5399 5. There are several places where Exim runs a non-Exim command in a
5400 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5401 was being done only for the command run by the queryprogram router. It is
5402 now done for all such subprocesses. The other cases are: ${run, transport
5403 filters, and the commands run by the lmtp and pipe transports.
5405 6. Added CONFIGURE_GROUP build-time option.
5407 7. Some older OS have a limit of 256 on the maximum number of file
5408 descriptors. Exim was using setrlimit() to set 1000 as a large value
5409 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5410 systems. I've change it so that if it can't get 1000, it tries for 256.
5412 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5413 was an oversight, and furthermore, ever since the addition of extra
5414 controls (e.g. 4.43/32), the checks on when to allow different forms of
5415 "control" were broken. There should now be diagnostics for all cases when a
5416 control that does not make sense is encountered.
5418 9. Added the /retain_sender option to "control=submission".
5420 10. $recipients is now available in the predata ACL (oversight).
5422 11. Tidy the search cache before the fork to do a delivery from a message
5423 received from the command line. Otherwise the child will trigger a lookup
5424 failure and thereby defer the delivery if it tries to use (for example) a
5425 cached ldap connection that the parent has called unbind on.
5427 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5428 of $address_data from the recipient verification was clobbered by the
5429 sender verification.
5431 13. The value of address_data from a sender verification is now available in
5432 $sender_address_data in subsequent conditions in the ACL statement.
5434 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5436 15. Added a new option "connect=<time>" to callout options, to set a different
5439 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5440 was its contents. (It was OK if the option was not defined at all.)
5442 17. A "Completed" log line is now written for messages that are removed from
5443 the spool by the -Mrm option.
5445 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5446 information about exactly what failed.
5448 19. Added -dd to debug only the daemon process.
5450 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5451 handles timeouts, both on the server side and network timeouts. Renamed the
5452 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5454 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5455 It is now set to "smtps".
5457 22. $host_address is now set to the target address during the checking of
5458 ignore_target_hosts.
5460 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5461 being passed; this would have caused $sender_host_name to have been used if
5462 matching the list had actually called for a host name (not very likely,
5463 since this list is usually IP addresses). A host name is now passed as
5466 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5467 code that specifies a non-restarting handler (typically sigaction() in
5468 modern systems) in an attempt to fix a rare and obscure crash bug.
5470 25. Narrowed the window for a race in the daemon that could cause it to ignore
5471 SIGCHLD signals. This is not a major problem, because they are used only to
5472 wake it up if nothing else does.
5474 26. A malformed maildirsize file could cause Exim to calculate negative values
5475 for the mailbox size or file count. Odd effects could occur as a result.
5476 The maildirsize information is now recalculated if the size or filecount
5479 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5480 support for a long time. Removed HAVE_SYS_VFS_H.
5482 28. Installed the latest version of exipick from John Jetmore.
5484 29. In an address list, if the pattern was not a regular expression, an empty
5485 subject address (from a bounce message) matched only if the pattern was an
5486 empty string. Non-empty patterns were not even tested. This was the wrong
5487 because it is perfectly reasonable to use an empty address as part of a
5488 database query. An empty address is now tested by patterns that are
5489 lookups. However, all the other forms of pattern expect the subject to
5490 contain a local part and a domain, and therefore, for them, an empty
5491 address still always fails if the pattern is not itself empty.
5493 30. Exim went into a mad DNS loop when attempting to do a callout where the
5494 host was specified on an smtp transport, and looking it up yielded more
5495 than one IP address.
5497 31. Re-factored the code for checking spool and log partition space into a
5498 function that finds that data and another that does the check. The former
5499 is then used to implement four new variables: $spool_space, $log_space,
5500 $spool_inodes, and $log_inodes.
5502 32. The RFC2047 encoding function was originally intended for short strings
5503 such as real names; it was not keeping to the 75-character limit for
5504 encoded words that the RFC imposes. It now respects the limit, and
5505 generates multiple encoded words if necessary. To be on the safe side, I
5506 have increased the buffer size for the ${rfc2047: expansion operator from
5509 33. It is now permitted to omit both strings after an "if" condition; if the
5510 condition is true, the result is "true". As before, when the second string
5511 is omitted, a false condition yields an empty string. This makes it less
5512 cumbersome to write custom ACL and router conditions.
5514 34. Failure to deliver a bounce message always caused it to be frozen, even if
5515 there was an errors_to setting on the router. The errors_to setting is now
5518 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5519 canonical form (fully expanded) before being placed in
5520 $sender_host_address.
5522 36. The table in the code that translates DNS record types into text (T_A to
5523 "A" for instance) was missing entries for NS and CNAME. It is just possible
5524 that this could have caused confusion if both these types were looked up
5525 for the same domain, because the text type is used as part of Exim's
5526 per-process caching. But the chance of anyone hitting this buglet seems
5529 37. The dnsdb lookup has been extended in a number of ways.
5531 (1) There is a new type, "zns", which walks up the domain tree until it
5532 finds some nameserver records. It should be used with care.
5534 (2) There is a new type, "mxh", which is like "mx" except that it returns
5535 just the host names, not the priorities.
5537 (3) It is now possible to give a list of domains (or IP addresses) to be
5538 looked up. The behaviour when one of the lookups defers can be
5539 controlled by a keyword.
5541 (4) It is now possible to specify the separator character for use when
5542 multiple records are returned.
5544 38. The dnslists ACL condition has been extended: it is now possible to supply
5545 a list of IP addresses and/or domains to be looked up in a particular DNS
5548 39. Added log_selector=+queue_time_overall.
5550 40. When running the queue in the test harness, wait just a tad after forking a
5551 delivery process, to get repeatability of debugging output.
5553 41. Include certificate and key file names in error message when GnuTLS fails
5554 to set them up, because the GnuTLS error message doesn't include the name
5555 of the failing file when there is a problem reading it.
5557 42. Allow both -bf and -bF in the same test run.
5559 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5561 44. The "Exiscan patch" is now merged into the mainline Exim source.
5563 45. Sometimes the final signoff response after QUIT could fail to get
5564 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5565 < 0 before doing a fflush(). This bug looks as though it goes back to the
5566 introduction of TLS in release 3.20, but "sometimes" must have been rare
5567 because the tests only now provoked it.
5569 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5570 (this can affect the format of dates).
5572 47. exim_tidydb, when checking for the continued existence of a message for
5573 which it has found a message-specific retry record, was not finding
5574 messages that were in split spool directories. Consequently, it was
5575 deleting retry records that should have stayed in existence.
5577 48. Steve fixed some bugs in eximstats.
5579 49. The SPA authentication driver was not abandoning authentication and moving
5580 on to the next authenticator when an expansion was forced to fail,
5581 contradicting the general specification for all authenticators. Instead it
5582 was generating a temporary error. It now behaves as specified.
5584 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5585 (the order specifies the preference for clients). The order is now AES256,
5586 AES128, 3DES, ARCFOUR128.
5588 51. Small patch to Sieve code - explicitly set From: when generating an
5591 52. Exim crashed if a remote delivery caused a very long error message to be
5592 recorded - for instance if somebody sent an entire SpamAssassin report back
5593 as a large number of 550 error lines. This bug was coincidentally fixed by
5594 increasing the size of one of Exim's internal buffers (big_buffer) that
5595 happened as part of the Exiscan merge. However, to be on the safe side, I
5596 have made the code more robust (and fixed the comments that describe what
5599 53. Now that there can be additional text after "Completed" in log lines (if
5600 the queue_time_overall log selector is set), a one-byte patch to exigrep
5601 was needed to allow it to recognize "Completed" as not the last thing in
5604 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5605 patch that reportedly fixes this has been added. I am not expert enough to
5606 create a test for it. This is what the patch creator wrote:
5608 "I found a little strange behaviour of ldap code when working with
5609 Windows 2003 AD Domain, where users was placed in more than one
5610 Organization Units. When I tried to give exim partial DN, the exit code
5611 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5612 But simultaneously result of request was absolutely normal ldap result,
5613 so I produce this patch..."
5615 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5616 REFERENCE, so I have modified the code to exclude the patch when that macro
5619 55. Some experimental protocols are using DNS PTR records for new purposes. The
5620 keys for these records are domain names, not reversed IP addresses. The
5621 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5622 leaves it alone. Component reversal etc. now happens only for IP addresses.
5625 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5627 57. Double the size of the debug message buffer (to 2048) so that more of very
5628 long debug lines gets shown.
5630 58. The exicyclog utility now does better if the number of log files to keep
5631 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5633 59. Two changes related to the smtp_active_hostname option:
5635 (1) $smtp_active_hostname is now available as a variable.
5636 (2) The default for smtp_banner uses $smtp_active_hostname instead
5637 of $primary_hostname.
5639 60. The host_aton() function is supposed to be passed a string that is known
5640 to be a valid IP address. However, in the case of IPv6 addresses, it was
5641 not checking this. This is a hostage to fortune. Exim now panics and dies
5642 if the condition is not met. A case was found where this could be provoked
5643 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5644 components; fortuitously, this particular loophole had already been fixed
5645 by change 4.50/55 above.
5647 If there are any other similar loopholes, the new check in host_aton()
5648 itself should stop them being exploited. The report I received stated that
5649 data on the command line could provoke the exploit when Exim was running as
5650 exim, but did not say which command line option was involved. All I could
5651 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5652 running as the user.
5655 61. There was a buffer overflow vulnerability in the SPA authentication code
5656 (which came originally from the Samba project). I have added a test to the
5657 spa_base64_to_bits() function which I hope fixes it.
5660 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5661 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5663 63. The daemon start-up calls getloadavg() while still root for those OS that
5664 need the first call to be done as root, but it missed one case: when
5665 deliver_queue_load_max is set with deliver_drop_privilege. This is
5666 necessary for the benefit of the queue runner, because there is no re-exec
5667 when deliver_drop_privilege is set.
5669 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5670 This has been fixed.
5672 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5673 was in use, was not putting the data itself into the right store pool;
5674 consequently, it could be overwritten for a subsequent message in the same
5675 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5678 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5680 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5681 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5682 the function were treating the return as a boolean value, which happened to
5683 work because 0=false and not-0=true, but is not correct code.
5685 68. The host_aton() function was not handling scoped IPv6 addresses (those
5686 with, for example, "%eth0" on the end) correctly.
5688 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5689 negated items (that is, ~something) in unsigned ints. Some compilers
5690 apparently mutter when there is no cast.
5692 70. If an address verification called from an ACL failed, and did not produce a
5693 user-specific message (i.e. there was only a "system" message), nothing was
5694 put in $acl_verify_message. In this situation, it now puts the system
5697 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5698 session; change 4.31/43 added the unwanted input to the log line - except
5699 that it did not do this in the start of session case. It now does.
5701 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5702 This is wrong and can cause the other end to generate a synchronization
5703 error if it is another Exim or anything else that does the synchronization
5704 check. A QUIT command is no longer sent after a timeout.
5706 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5707 during host lookups.
5709 74. The defer_ok option of callout verification was not working if it was used
5710 when verifying addresses in header lines, that is, for this case:
5712 verify = header_sender/callout=defer_ok
5714 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5715 those file descriptors could be used for SMTP connections. If anything
5716 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5717 could be sent to the SMTP client, causing chaos. The daemon now opens
5718 stdin, stdout, and stderr to /dev/null when it puts itself into the
5721 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5722 log by default. The user can override this with suitable Perl magic.
5724 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5725 the log message when discard triggers, was not working for the DATA ACL or
5726 for the non-SMTP ACL.
5728 78. Error message wording change in sieve.c.
5730 79. If smtp_accept_max_per_host was set, the number of connections could be
5731 restricted to fewer than expected, because the daemon was trying to set up
5732 a new connection before checking whether the processes handling previous
5733 connections had finished. The check for completed processes is now done
5734 earlier. On busy systems, this bug wouldn't be noticed because something
5735 else would have woken the daemon, and it would have reaped the completed
5738 80. If a message was submitted locally by a user whose login name contained one
5739 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5740 It caused a spool format error. I have fixed the spool reading code. A
5741 related problem was that the "from" clause in the Received: line became
5742 illegal because of the space(s). It is now covered by ${quote_local_part.
5744 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5747 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5749 83. Patch from Sieve maintainer to fix unterminated string problem in
5750 "vacation" handling.
5752 84. Some minor changes to the Linux configuration files to help with other
5753 OS variants using glibc.
5755 85. One more patch for Sieve to update vacation handling to latest spec.
5758 ----------------------------------------------------
5759 See the note above about the 4.44 and 4.50 releases.
5760 ----------------------------------------------------
5766 1. Change 4.43/35 introduced a bug that caused file counts to be
5767 incorrectly computed when quota_filecount was set in an appendfile
5770 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5771 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5774 3. Give more explanation in the error message when the command for a transport
5775 filter fails to execute.
5777 4. There are several places where Exim runs a non-Exim command in a
5778 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5779 was being done only for the command run by the queryprogram router. It is
5780 now done for all such subprocesses. The other cases are: ${run, transport
5781 filters, and the commands run by the lmtp and pipe transports.
5783 5. Some older OS have a limit of 256 on the maximum number of file
5784 descriptors. Exim was using setrlimit() to set 1000 as a large value
5785 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5786 systems. I've change it so that if it can't get 1000, it tries for 256.
5788 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5789 was an oversight, and furthermore, ever since the addition of extra
5790 controls (e.g. 4.43/32), the checks on when to allow different forms of
5791 "control" were broken. There should now be diagnostics for all cases when a
5792 control that does not make sense is encountered.
5794 7. $recipients is now available in the predata ACL (oversight).
5796 8. Tidy the search cache before the fork to do a delivery from a message
5797 received from the command line. Otherwise the child will trigger a lookup
5798 failure and thereby defer the delivery if it tries to use (for example) a
5799 cached ldap connection that the parent has called unbind on.
5801 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5802 of $address_data from the recipient verification was clobbered by the
5803 sender verification.
5805 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5806 was its contents. (It was OK if the option was not defined at all.)
5808 11. A "Completed" log line is now written for messages that are removed from
5809 the spool by the -Mrm option.
5811 12. $host_address is now set to the target address during the checking of
5812 ignore_target_hosts.
5814 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5815 being passed; this would have caused $sender_host_name to have been used if
5816 matching the list had actually called for a host name (not very likely,
5817 since this list is usually IP addresses). A host name is now passed as
5820 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5821 code that specifies a non-restarting handler (typically sigaction() in
5822 modern systems) in an attempt to fix a rare and obscure crash bug.
5824 15. Narrowed the window for a race in the daemon that could cause it to ignore
5825 SIGCHLD signals. This is not a major problem, because they are used only to
5826 wake it up if nothing else does.
5828 16. A malformed maildirsize file could cause Exim to calculate negative values
5829 for the mailbox size or file count. Odd effects could occur as a result.
5830 The maildirsize information is now recalculated if the size or filecount
5833 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5834 support for a long time. Removed HAVE_SYS_VFS_H.
5836 18. Updated exipick to current release from John Jetmore.
5838 19. Allow an empty sender to be matched against a lookup in an address list.
5839 Previously the only cases considered were a regular expression, or an
5842 20. Exim went into a mad DNS lookup loop when doing a callout where the
5843 host was specified on the transport, if the DNS lookup yielded more than
5846 21. The RFC2047 encoding function was originally intended for short strings
5847 such as real names; it was not keeping to the 75-character limit for
5848 encoded words that the RFC imposes. It now respects the limit, and
5849 generates multiple encoded words if necessary. To be on the safe side, I
5850 have increased the buffer size for the ${rfc2047: expansion operator from
5853 22. Failure to deliver a bounce message always caused it to be frozen, even if
5854 there was an errors_to setting on the router. The errors_to setting is now
5857 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5858 canonical form (fully expanded) before being placed in
5859 $sender_host_address.
5861 24. Updated eximstats to version 1.33
5863 25. Include certificate and key file names in error message when GnuTLS fails
5864 to set them up, because the GnuTLS error message doesn't include the name
5865 of the failing file when there is a problem reading it.
5867 26. Expand error message when OpenSSL has problems setting up cert/key files.
5870 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5871 (this can affect the format of dates).
5873 28. exim_tidydb, when checking for the continued existence of a message for
5874 which it has found a message-specific retry record, was not finding
5875 messages that were in split spool directories. Consequently, it was
5876 deleting retry records that should have stayed in existence.
5878 29. eximstats updated to version 1.35
5879 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5880 1.35 - bugfix such that pie charts by volume are generated correctly
5882 30. The SPA authentication driver was not abandoning authentication and moving
5883 on to the next authenticator when an expansion was forced to fail,
5884 contradicting the general specification for all authenticators. Instead it
5885 was generating a temporary error. It now behaves as specified.
5887 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5888 (the order specifies the preference for clients). The order is now AES256,
5889 AES128, 3DES, ARCFOUR128.
5891 31. Small patch to Sieve code - explicitly set From: when generating an
5894 32. Exim crashed if a remote delivery caused a very long error message to be
5895 recorded - for instance if somebody sent an entire SpamAssassin report back
5896 as a large number of 550 error lines. This bug was coincidentally fixed by
5897 increasing the size of one of Exim's internal buffers (big_buffer) that
5898 happened as part of the Exiscan merge. However, to be on the safe side, I
5899 have made the code more robust (and fixed the comments that describe what
5902 33. Some experimental protocols are using DNS PTR records for new purposes. The
5903 keys for these records are domain names, not reversed IP addresses. The
5904 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5905 leaves it alone. Component reversal etc. now happens only for IP addresses.
5908 34. The host_aton() function is supposed to be passed a string that is known
5909 to be a valid IP address. However, in the case of IPv6 addresses, it was
5910 not checking this. This is a hostage to fortune. Exim now panics and dies
5911 if the condition is not met. A case was found where this could be provoked
5912 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5913 components; fortuitously, this particular loophole had already been fixed
5914 by change 4.50/55 or 4.44/33 above.
5916 If there are any other similar loopholes, the new check in host_aton()
5917 itself should stop them being exploited. The report I received stated that
5918 data on the command line could provoke the exploit when Exim was running as
5919 exim, but did not say which command line option was involved. All I could
5920 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5921 running as the user.
5924 35. There was a buffer overflow vulnerability in the SPA authentication code
5925 (which came originally from the Samba project). I have added a test to the
5926 spa_base64_to_bits() function which I hope fixes it.
5929 36. The daemon start-up calls getloadavg() while still root for those OS that
5930 need the first call to be done as root, but it missed one case: when
5931 deliver_queue_load_max is set with deliver_drop_privilege. This is
5932 necessary for the benefit of the queue runner, because there is no re-exec
5933 when deliver_drop_privilege is set.
5935 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5936 was in use, was not putting the data itself into the right store pool;
5937 consequently, it could be overwritten for a subsequent message in the same
5938 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5941 38. Sometimes the final signoff response after QUIT could fail to get
5942 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5943 < 0 before doing a fflush(). This bug looks as though it goes back to the
5944 introduction of TLS in release 3.20, but "sometimes" must have been rare
5945 because the tests only now provoked it.
5951 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5952 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5953 Now it returns an integer. A number of calls were still expecting a T/F
5954 return. Fortuitously, in all cases, the tests worked in OK situations,
5955 which is the norm. However, things would have gone wrong on any write
5956 failures on the smtp file descriptor. This function is used when sending
5957 messages over SMTP and also when doing verify callouts.
5959 2. When Exim is called to do synchronous delivery of a locally submitted
5960 message (the -odf or -odi options), it no longer closes stderr before doing
5963 3. Implemented the mua_wrapper option.
5965 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5967 5. Implemented the functions header_remove(), header_testname(),
5968 header_add_at_position(), and receive_remove_recipient(), and exported them
5971 6. If an ACL "warn" statement specified the addition of headers, Exim already
5972 inserted X-ACL-Warn: at the start if there was no header name. However, it
5973 was not making this test for the second and subsequent header lines if
5974 there were newlines in the string. This meant that an invalid header could
5975 be inserted if Exim was badly configured.
5977 7. Allow an ACL "warn" statement to add header lines at the start or after all
5978 the Received: headers, as well as at the end.
5980 8. Added the rcpt_4xx retry error code.
5982 9. Added postmaster_mailfrom=xxx to callout verification option.
5984 10. Added mailfrom=xxxx to the callout verification option, for verify=
5987 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5988 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5990 12. Inserted some casts to stop certain compilers warning when using pointer
5991 differences as field lengths or precisions in printf-type calls (mostly
5992 affecting debugging statements).
5994 13. Added optional readline() support for -be (dynamically loaded).
5996 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5997 same clock tick as a message's arrival, so that its received time was the
5998 same as the "first fail" time on the retry record, and that message
5999 remained on the queue past the ultimate address timeout, every queue runner
6000 would try a delivery (because it was past the ultimate address timeout) but
6001 after another failure, the ultimate address timeout, which should have then
6002 bounced the address, did not kick in. This was a "< instead of <=" error;
6003 in most cases the first failure would have been in the next clock tick
6004 after the received time, and all would be well.
6006 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6007 being recognized when the domain list was tested by the match_domain
6008 condition in an expansion string.
6010 16. Added the ${str2b64: operator.
6012 17. Exim was always calling setrlimit() to set a large limit for the number of
6013 processes, without checking whether the existing limit was already
6014 adequate. (It did check for the limit on file descriptors.) Furthermore,
6015 errors from getrlimit() and setrlimit() were being ignored. Now they are
6016 logged to the main and panic logs, but Exim does carry on, to try to do its
6017 job under whatever limits there are.
6019 18. Imported PCRE 5.0.
6021 19. Trivial typo in log message " temporarily refused connection" (the leading
6024 20. If the log selector return_path_on_delivery was set and an address was
6025 redirected to /dev/null, the delivery process crashed because it assumed
6026 that a return path would always be set for a "successful" delivery. In this
6027 case, the whole delivery is bypassed as an optimization, and therefore no
6030 21. Internal re-arrangement: the function for sending a challenge and reading
6031 a response while authentication was assuming a zero-terminated challenge
6032 string. It's now changed to take a pointer and a length, to allow for
6033 binary data in such strings.
6035 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6037 23. Exim was not respecting finduser_retries when seeking the login of the
6038 uid under which it was called; it was always trying 10 times. (The default
6039 setting of finduser_retries is zero.) Also, it was sleeping after the final
6040 failure, which is pointless.
6042 24. Implemented tls_on_connect_ports.
6044 25. Implemented acl_smtp_predata.
6046 26. If the domain in control=submission is set empty, Exim assumes that the
6047 authenticated id is a complete email address when it generates From: or
6048 Sender: header lines.
6050 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6051 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6052 chown and chgrp in /bin and hostname in /usr/bin.
6054 28. Exim was keeping the "process log" file open after each use, just as it
6055 does for the main log. This opens the possibility of it remaining open for
6056 long periods when the USR1 signal hits a daemon. Occasional processlog
6057 errors were reported, that could have been caused by this. Anyway, it seems
6058 much more sensible not to leave this file open at all, so that is what now
6061 29. The long-running daemon process does not normally write to the log once it
6062 has entered its main loop, and it closes the log before doing so. This is
6063 so that log files can straightforwardly be renamed and moved. However,
6064 there are a couple of unusual error situations where the daemon does write
6065 log entries, and I had neglected to close the log afterwards.
6067 30. The text of an SMTP error response that was received during a remote
6068 delivery was being truncated at 512 bytes. This is too short for some of
6069 the long messages that one sometimes sees. I've increased the limit to
6072 31. It is now possible to make retry rules that apply only when a message has a
6073 specific sender, in particular, an empty sender.
6075 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6076 it possible to be selective about when SMTP synchronization is enforced.
6078 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6080 32. Implemented hosts_connection_nolog.
6082 33. Added an ACL for QUIT.
6084 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6087 35. Added mailbox_size and mailbox_filecount to appendfile.
6089 36. Added control = no_multiline_responses to ACLs.
6091 37. There was a bug in the logic of the code that waits for the clock to tick
6092 in the case where the clock went backwards by a substantial amount such
6093 that the microsecond fraction of "now" was more than the microsecond
6094 fraction of "then" (but the whole seconds number was less).
6096 38. Added support for the libradius Radius client library this is found on
6097 FreeBSD (previously only the radiusclient library was supported).
6103 1. When certain lookups returned multiple values in the form name=value, the
6104 quoting of the values was not always being done properly. Specifically:
6105 (a) If the value started with a double quote, but contained no whitespace,
6107 (b) If the value contained whitespace other than a space character (i.e.
6108 tabs or newlines or carriage returns) it was not quoted.
6109 This fix has been applied to the mysql and pgsql lookups by writing a
6110 separate quoting function and calling it from the lookup code. The fix
6111 should probably also be applied to nisplus, ibase and oracle lookups, but
6112 since I cannot test any of those, I have not disturbed their existing code.
6114 2. A hit in the callout cache for a specific address caused a log line with no
6115 reason for rejecting RCPT. Now it says "Previous (cached) callout
6116 verification failure".
6118 3. There was an off-by-one bug in the queryprogram router. An over-long
6119 return line was truncated at 256 instead of 255 characters, thereby
6120 overflowing its buffer with the terminating zero. As well as fixing this, I
6121 have increased the buffer size to 1024 (and made a note to document this).
6123 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6124 when Exim is waiting for an SMTP response from a remote server, Exim
6125 restarts its select() call on the socket, thereby resetting its timeout.
6126 This is not a problem when such interrupts are rare. Somebody set up a cron
6127 job to run exiwhat every 2 minutes, which is less than the normal select()
6128 timeout (5 or 10 minutes). This meant that the select() timeout never
6129 kicked in because it was always reset. I have fixed this by comparing the
6130 time when an interrupt arrives with the time at the start of the first call
6131 to select(). If more time than the timeout has elapsed, the interrupt is
6132 treated as a timeout.
6134 5. Some internal re-factoring in preparation for the addition of Sieve
6135 extensions (by MH). In particular, the "personal" test is moved to a
6136 separate function, and given an option for scanning Cc: and Bcc: (which is
6137 not set for Exim filters).
6139 6. When Exim created an email address using the login of the caller as the
6140 local part (e.g. when creating a From: or Sender: header line), it was not
6141 quoting the local part when it contained special characters such as @.
6143 7. Installed new OpenBSD configuration files.
6145 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6146 try to make them clearer.
6148 9. Callout options, other than the timeout value, were being ignored when
6149 verifying sender addresses in header lines. For example, when using
6151 verify = header_sender/callout=no_cache
6153 the cache was (incorrectly) being used.
6155 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6156 only the Cygwin environment.
6158 11. When return_path_on_delivery was set as a log selector, if different remote
6159 addresses in the same message used different return paths and parallel
6160 remote delivery occurred, the wrong values would sometimes be logged.
6161 (Whenever a remote delivery process finished, the return path value from
6162 the most recently started remote delivery process was logged.)
6164 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6165 lines when AUTH and/or TLS are in use. This is the "received protocol"
6166 field. Exim used to use "asmtp" for authenticated SMTP, without any
6167 indication (in the protocol name) for TLS use. Now it follows the RFC and
6168 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6169 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6170 names appear in log lines as well as in Received: header lines.
6172 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6173 extensions, and comparison tests, and to fix some bugs.
6175 14. Changes to the "personal" filter test:
6177 (1) The test was buggy in that it was just doing the equivalent of
6178 "contains" tests on header lines. For example, if a user's address was
6179 anne@some.where, the "personal" test would incorrectly be true for
6181 To: susanne@some.where
6183 This test is now done by extracting each address from the header in turn,
6184 and checking the entire address. Other tests that are part of "personal"
6185 are now done using regular expressions (for example, to check local parts
6186 of addresses in From: header lines).
6188 (2) The list of non-personal local parts in From: addresses has been
6189 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6190 taken from the Sieve specification recommendations.
6192 (3) If the message contains any header line starting with "List-" it is
6193 treated as non-personal.
6195 (4) The test for "circular" in the Subject: header line has been removed
6196 because it now seems ill-conceived.
6198 15. Minor typos in src/EDITME comments corrected.
6200 16. Installed latest exipick from John Jetmore.
6202 17. If headers_add on a router specified a text string that was too long for
6203 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6204 of string_sprintf() is now avoided.
6206 18. $message_body_size was not set (it was always zero) when running the DATA
6207 ACL and the local_scan() function.
6209 19. For the "mail" command in an Exim filter, no default was being set for
6210 the once_repeat time, causing a random time value to be used if "once" was
6211 specified. (If the value happened to be <= 0, no repeat happened.) The
6212 default is now 0s, meaning "never repeat". The "vacation" command was OK
6213 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6214 (I found it when inspecting the code).
6216 20. There is now an overall timeout for performing a callout verification. It
6217 defaults to 4 times the callout timeout, which applies to individual SMTP
6218 commands during the callout. The overall timeout applies when there is more
6219 than one host that can be tried. The timeout is checked before trying the
6220 next host. This prevents very long delays if there are a large number of
6221 hosts and all are timing out (e.g. when the network connections are timing
6222 out). The value of the overall timeout can be changed by specifying an
6223 additional sub-option for "callout", called "maxwait". For example:
6225 verify = sender/callout=5s,maxwait=20s
6227 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6228 to the end before writing, but this should make it even safer).
6230 22. Exim was forgetting that it had advertised PIPELINING for the second and
6231 subsequent messages on an SMTP connection. It was also not resetting its
6232 memory on STARTTLS and an internal HELO.
6234 23. When Exim logs an SMTP synchronization error within a session, it now
6235 records whether PIPELINING has been advertised or not.
6237 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6238 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6239 rather than long int.
6241 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6243 26. Added the never_mail option to autoreply.
6249 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6250 crash if the getsockname() call failed; this can happen if a connection is
6251 closed very soon after it is established. The problem was simply in the
6252 order in which certain operations were done, causing Exim to try to write
6253 to the SMTP stream before it had set up the file descriptor. The bug has
6254 been fixed by making things happen in the correct order.
6260 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6261 before the connection was closed, thus losing the rejection response.
6263 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6264 some early Solaris releases, but causes trouble in current releases where
6265 socklen_t is defined.
6267 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6270 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6273 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6274 that releasing the top of it at the end releases what was used for sub-
6275 expansions (unless the block got too big). However, discard this block if
6276 the first thing is a variable or header, so that we can use its block when
6277 it is dynamic (useful for very large $message_headers, for example).
6279 6. Lookups now cache *every* query, not just the most recent. A new, separate
6280 store pool is used for this. It can be recovered when all lookup caches are
6281 flushed. Lookups now release memory at the end of their result strings.
6282 This has involved some general refactoring of the lookup sources.
6284 7. Some code has been added to the store_xxx() functions to reduce the amount
6285 of flapping under certain conditions.
6287 8. log_incoming_interface used to affect only the <= reception log lines. Now
6288 it causes the local interface and port to be added to several more SMTP log
6289 lines, for example "SMTP connection from", and rejection lines.
6291 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6293 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6295 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6296 could be overwritten at the end of the current message (or the start of a
6297 new message if it was set in a HELO ACL). The value is now preserved for
6298 the duration of the SMTP connection.
6300 12. If a transport had a headers_rewrite setting, and a matching header line
6301 contained an unqualified address, that address was qualified, even if it
6302 did not match any rewriting rules. The underlying bug was that the values
6303 of the flags that permit the existence of unqualified sender and recipient
6304 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6305 non-local messages, and by -bnq for local messages) were not being
6306 preserved with the message after it was received.
6308 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6309 "next input=" as part of the text comprising the host identity instead of
6310 the correct text. The code was using the same buffer for two different
6311 strings. However, depending on which order the printing function evaluated
6312 its arguments, the bug did not always show up. Under Linux, for example, my
6313 test suite worked just fine.
6315 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6316 doesn't work with some older versions of Perl. It has been changed to "my",
6317 which in any case is probably the better facility to use.
6319 15. A really picky compiler found some instances of statements for creating
6320 error messages that either had too many or two few arguments for the format
6323 16. The size of the buffer for calls to the DNS resolver has been increased
6324 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6325 for addresses that have a lot of PTR records. This alleviates a problem; it
6326 does not fully solve it.
6328 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6329 buffer now truncates the list and logs the incident, which is the same
6330 action as happens when Exim is looking up a host name and its aliases.
6331 Previously in this situation something unpredictable would happen;
6332 sometimes it was "internal error: store_reset failed".
6334 18. If a server dropped the connection unexpectedly when an Exim client was
6335 using GnuTLS and trying to read a response, the client delivery process
6336 crashed while trying to generate an error log message.
6338 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6339 string, for example:
6341 warn message = H1: something\nH2: something
6343 the text was added as a single header line from Exim's point of view
6344 though it ended up OK in the delivered message. However, searching for the
6345 second and subsequent header lines using $h_h2: did not work. This has been
6346 fixed. Similarly, if a system filter added multiple headers in this way,
6347 the routers could not see them.
6349 20. Expanded the error message when iplsearch is called with an invalid key to
6350 suggest using net-iplsearch in a host list.
6352 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6353 ACLs are no longer actually imposed (and a message to that effect is
6356 22. If a "gecos" field in a passwd entry contained escaped characters, in
6357 particular, if it contained a \" sequence, Exim got it wrong when building
6358 a From: or a Sender: header from that name. A second bug also caused
6359 incorrect handling when an unquoted " was present following a character
6360 that needed quoting.
6362 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6363 was not being matched caselessly.
6365 24. Arranged for all hyphens in the exim.8 source to be escaped with
6368 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6369 properly. Recipient callout cache records were still being keyed to include
6370 the sender, even when use_sender was set false. This led to far more
6371 callouts that were necessary. The sender is no longer included in the key
6372 when use_sender is false.
6374 26. Added "control = submission" modifier to ACLs.
6376 27. Added the ${base62d: operator to decode base 62 numbers.
6378 28. dnsdb lookups can now access SRV records.
6380 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6381 the configuration file.
6383 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6384 (-v) mode. This makes the output for a verbose queue run more intelligible.
6386 31. Added a use_postmaster feature to recipient callouts.
6388 32. Added the $body_zerocount variable, containing the number of binary zero
6389 bytes in the message body.
6391 33. The time of last modification of the "new" subdirectory is now used as the
6392 "mailbox time last read" when there is a quota error for a maildir
6395 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6397 35. Added +ignore_unknown as a special item in host lists.
6399 36. Code for decoding IPv6 addresses in host lists is now included, even if
6400 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6401 address was recognized as an IP address, but was then not correctly decoded
6402 into binary, causing unexpected and incorrect effects when compared with
6409 1. Very minor rewording of debugging text in manualroute to say "list of
6410 hosts" instead of "hostlist".
6412 2. If verify=header_syntax was set, and a header line with an unqualified
6413 address (no domain) and a large number of spaces between the end of the
6414 name and the colon was received, the reception process suffered a buffer
6415 overflow, and (when I tested it) crashed. This was caused by some obsolete
6416 code that should have been removed. The fix is to remove it!
6418 3. When running in the test harness, delay a bit after writing a bounce
6419 message to get a bit more predictability in the log output.
6421 4. Added a call to search_tidyup() just before forking a reception process. In
6422 theory, someone could use a lookup in the expansion of smtp_accept_max_
6423 per_host which, without the tidyup, could leave open a database connection.
6425 5. Added the variables $recipient_data and $sender_data which get set from a
6426 lookup success in an ACL "recipients" or "senders" condition, or a router
6427 "senders" option, similar to $domain_data and $local_part_data.
6429 6. Moved the writing of debug_print from before to after the "senders" test
6432 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6433 problems for message scanning, either using a data ACL, or using
6434 local_scan() because the Received: header was not generated till after they
6435 were called (in order to set the time as the time of reception completion).
6436 I have revised the way this works. The header is now generated after the
6437 body is received, but before the ACL or local_scan() are called. After they
6438 are run, the timestamp in the header is updated.
6444 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6445 before starting a queue runner without re-exec. This happened only when
6446 deliver_drop_privilege was set or when the Exim user was set to root. The
6447 effect of the bug was that timeouts during subsequent deliveries caused
6448 crashes instead of being properly handled. The handler is now left at its
6449 default (and expected) setting.
6451 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6452 message, again when deliver_drop_privilege is set or Exim is run as root.
6453 The bug described in (1) was not present in this case, but the tidying up
6454 of the other signals was missing. I have made the two cases consistent.
6456 3. The ignore_target_hosts setting on a manualroute router was being ignored
6457 for hosts that were looked up using the /MX notation.
6459 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6462 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6463 operated on the sender address. After changing the $sender_address to <>
6464 for the sender address verify, Exim was re-instated it as the original
6465 (before rewriting) address, but remembering that it had rewritten it, so it
6466 wasn't rewriting it again. This bug also had the effect of breaking the
6467 sender address verification caching when the sender address was rewritten.
6469 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6470 This has been changed so that if the ip literal address matches
6471 ignore_target_hosts, the router declines.
6473 7. Added expansion conditions match_domain, match_address, and match_local_
6474 part (NOT match_host).
6476 8. The placeholder for the Received: header didn't have a length field set.
6478 9. Added code to Exim itself and to exim_lock to test for a specific race
6479 condition that could lead to file corruption when using MBX delivery. The
6480 issue is with the lockfile that is created in /tmp. If this file is removed
6481 after a process has opened it but before that process has acquired a lock,
6482 there is the potential for a second process to recreate the file and also
6483 acquire a lock. This could lead to two Exim processes writing to the file
6484 at the same time. The added code performs the same test as UW imapd; it
6485 checks after acquiring the lock that its file descriptor still refers to
6486 the same named file.
6488 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6489 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6492 11. Added the smtp_active_hostname option. If used, this will typically be made
6493 to depend on the incoming interface address. Because $interface_address is
6494 not set up until the daemon has forked a reception process, error responses
6495 that can happen earlier (such as "too many connections") no longer contain
6498 12. If an expansion in a condition on a "warn" statement fails because a lookup
6499 defers, the "warn" statement is abandoned, and the next ACL statement is
6500 processed. Previously this caused the whole ACL to be aborted.
6502 13. Added the iplsearch lookup type.
6504 14. Added ident_timeout as a log selector.
6506 15. Added tls_certificate_verified as a log selector.
6508 16. Added a global option tls_require_ciphers (compare the smtp transport
6509 option of the same name). This controls incoming TLS connections.
6511 17. I finally figured out how to make tls_require_ciphers do a similar thing
6512 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6513 before starting the TLS session.
6515 18. Tabs are now shown as \t in -bP output.
6517 19. If the log selector return_path_on_delivery was set, Exim crashed when
6518 bouncing a message because it had too many Received: header lines.
6520 20. If two routers both had headers_remove settings, and the first one included
6521 a superfluous trailing colon, the final name in the first list and the
6522 first name in the second list were incorrectly joined into one item (with a
6523 colon in the middle).
6529 1. Added -C and -D options to the exinext utility, mainly to make it easier
6530 to include in the automated testing, but these could be helpful when
6531 multiple configurations are in use.
6533 2. The exinext utility was not formatting the output nicely when there was
6534 an alternate port involved in the retry record key, nor when there was a
6535 message id as well (for retries that were specific to a specific message
6536 and a specific host). It was also confused by IPv6 addresses, because of
6537 the additional colons they contain. I have fixed the IPv4 problem, and
6538 patched it up to do a reasonable job for IPv6.
6540 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6541 delivery, the log line now contains "pipelined" if PIPELINING was used.
6543 4. An SMTP transport process used to panic and die if the bind() call to set
6544 an explicit outgoing interface failed. This has been changed; it is now
6545 treated in the same way as a connect() failure.
6547 5. A reference to $sender_host_name in the part of a conditional expansion
6548 that was being skipped was still causing a DNS lookup. This no longer
6551 6. The def: expansion condition was not recognizing references to header lines
6552 that used bh_ and bheader_.
6554 7. Added the _cache feature to named lists.
6556 8. The code for checking quota_filecount in the appendfile transport was
6557 allowing one more file than it should have been.
6559 9. For compatibility with Sendmail, the command line option
6567 and sets the incoming protocol and host name (for trusted callers). The
6568 host name and its colon can be omitted when only the protocol is to be set.
6569 Note the Exim already has two private options, -pd and -ps, that refer to
6570 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6571 "s", but I don't think that's a major issue.
6573 10. A number of refactoring changes to the code, none of which should affect
6576 (a) The number of logging options was getting close to filling up the
6577 32-bit word that was used as a bit map. I have split them into two classes:
6578 those that are passed in the argument to log_write(), and those that are
6579 only ever tested independently outside of that function. These are now in
6580 separate 32-bit words, so there is plenty of room for expansion again.
6581 There is no change in the user interface or the logging behaviour.
6583 (b) When building, for example, log lines, the code previously used a
6584 macro that called string_cat() twice, in order to add two strings. This is
6585 not really sufficiently general. Furthermore, there was one instance where
6586 it was actually wrong because one of the argument was used twice, and in
6587 one call a function was used. (As it happened, calling the function twice
6588 did not affect the overall behaviour.) The macro has been replaced by a
6589 function that can join an arbitrary number of extra strings onto a growing
6592 (c) The code for expansion conditions now uses a table and a binary chop
6593 instead of a serial search (which was left over from when there were very
6594 few conditions). Also, it now recognizes conditions like "pam" even when
6595 the relevant support is not compiled in: a suitably worded error message is
6596 given if an attempt is made to use such a condition.
6598 11. Added ${time_interval:xxxxx}.
6600 12. A bug was causing one of the ddress fields not to be passed back correctly
6601 from remote delivery subprocesses. The field in question was not being
6602 subsequently used, so this caused to problems in practice.
6604 13. Added new log selectors queue_time and deliver_time.
6606 14. Might have fixed a bug in maildirsizefile handling that threw up
6607 "unexpected character" debug warnings, and recalculated the data
6608 unnecessarily. In any case, I expanded the warning message to give more
6611 15. Added the message "Restricted characters in address" to the statements in
6612 the default ACL that block characters like @ and % in local parts.
6614 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6615 Three changes have been made:
6617 (a) There was a serious bug; a negative response to MAIL caused the whole
6618 recipient domain to be cached as invalid, thereby blocking all messages
6619 to all local parts at the same domain, from all senders. This bug has
6620 been fixed. The domain is no longer cached after a negative response to
6621 MAIL if the sender used is not empty.
6623 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6626 (c) A new callout option, "use_sender" has been added for people who want
6627 the modified behaviour.
6633 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6636 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6637 indeed breaks things for older releases.
6639 3. Added additional logging to the case where there is a problem reading data
6640 from a filter that is running in a subprocess using a pipe, in order to
6641 try to track down a specific problem.
6643 4. Testing facility fudge: when running in the test harness and attempting
6644 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6645 getting "No route to host". Convert this to a timeout.
6647 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6650 6. Some OS don't have socklen_t but use size_t instead. This affects the
6651 fifth argument of getsockopt() amongst other things. This is now
6652 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6653 can be set for individual OS. I have set it for SunOS5, OSF1, and
6654 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6655 some earlier ones do not.
6657 7. Change 4.30/15 was not doing the test caselessly.
6659 8. The standard form for an IPv6 address literal was being rejected by address
6660 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6661 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6662 this, as well as the form without the "IPv6" on the front (but only when
6663 address literals are enabled, of course).
6665 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6667 10. Exim crashed if a message with an empty sender address specified by -f
6668 encountered a router with an errors_to setting. This could be provoked only
6669 by a command such as
6673 where an empty string was supplied; "<>" did not hit this bug.
6675 11. Installed PCRE release 4.5.
6677 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6678 remained set. It is now erased.
6680 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6681 times from message ids (which are base 36 rather than the normal 62).
6683 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6684 were being counted as actual protocol errors, and logged if the log
6685 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6686 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6687 and DATA following a set of rejected RCPTs do not count as protocol errors.
6688 In other words, Exim assumes they were pipelined, though this may not
6689 actually be the case. Of course, in all cases the client gets an
6690 appropriate error code.
6692 15. If a lookup fails in an ACL condition, a message about the failure may
6693 be available; it is used if testing the ACL cannot continue, because most
6694 such messages specify what the cause of the deferral is. However, some
6695 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6696 that caused an old message to be retained and used if a later statement
6697 caused a defer, replacing the real cause of the deferral.
6699 16. If an IP address had so many PTR records that the DNS lookup buffer
6700 was not large enough to hold them, Exim could crash while trying to process
6701 the truncated data. It now detects and logs this case.
6703 17. Further to 4.21/58, another change has been made: if (and only if) the
6704 first line of a message (the first header line) ends with CRLF, a bare LF
6705 in a subsequent header line has a space inserted after it, so as not to
6706 terminate the header.
6708 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6709 unnecessarily, used atoi() instead of strtol(), and didn't check the
6710 termination when getting file sizes from file names by regex.
6712 19. Completely re-implemented the support for maildirsize files, in the light
6713 of a number of problems with the previous contributed implementation
6714 (4.30/29). In particular:
6716 . If the quota is zero, the maildirsize file is maintained, but no quota is
6719 . If the maildir directory does not exist, it is created before any attempt
6720 to write a maildirsize file.
6722 . The quota value in the file is just a cache; if the quota is changed in
6723 the transport, the new value overrides.
6725 . A regular expression is available for excluding directories from the
6728 20. The autoreply transport checks the characters in options that define the
6729 message's headers; it allows continued headers, but it was checking with
6730 isspace() after an embedded newline instead of explicitly looking for a
6733 21. If all the "regular" hosts to which an address was routed had passed their
6734 expiry times, and had not reached their retry times, the address was
6735 bounced, even if fallback hosts were defined. Now Exim should go on to try
6738 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6739 equivalent code in the SMTP transport. Some hosts send humungous responses
6740 to HELO/EHLO, more than 1024 it seems.
6742 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6743 gives compiler warnings in some environments. I've now done it "properly",
6746 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6747 (because of problems with the built-in one) was declared to return uschar *
6748 instead of char *, causing compiler failure.
6750 25. Fixed a file descriptor leak when processing alias/forward files.
6752 26. Fixed a minor format string issue in dbfn.c.
6754 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6756 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6757 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6760 29. When a sender address is verified, it is cached, to save repeating the test
6761 when there is more than one recipient in a message. However, when the
6762 verification involves a callout, it is possible for different callout
6763 options to be set for different recipients. It is too complicated to keep
6764 track of this in the cache, so now Exim always runs a verification when a
6765 callout is required, relying on the callout cache for the optimization.
6766 The overhead is duplication of the address routing, but this should not be
6769 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6770 to be verified with callout=postmaster, and the main callout worked but the
6771 postmaster check failed, the verification correctly failed. However, if a
6772 subsequent RCPT command asked for sender verification *without* the
6773 postmaster check, incorrect caching caused this verification also to fail,
6776 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6777 it was not caching the DNS options (qualify_single, search_parents) that
6778 were used when the lookup failed. A subsequent lookup with different
6779 options therefore always gave the same answer, though there were cases
6780 where it should not have. (Example: a "domains = !$mx_any" option on a
6781 dnslookup router: the "domains" option is always processed without any
6782 widening, but the router might have qualify_single set.) Now Exim uses the
6783 cached value only when the same options are set.
6785 32. Added John Jetmore's "exipick" utility to the distribution.
6787 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6788 than a timeout (e.g. a certificate is required, and is not provided), an
6789 Exim server now closes the connection immediately. Previously it waited for
6790 the client to close - but if the client is SSL, it seems that they each
6791 wait for each other, leading to a delay before one of them times out.
6793 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6794 maintained 0.8.x compatibility because I don't think many are using it, and
6795 it is clearly obsolete.
6797 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6800 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6801 client certificate was expired. A simple patch fixes this, though I don't
6802 understand the full logic of why the verify callback is called multiple
6805 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6806 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6807 which causes problems with some clients (such as the Certicom SSL Plus
6808 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6809 disables the coutermeasure allowing Eudora to connect."
6811 38. Exim was not checking that a write() to a log file succeeded. This could
6812 lead to Bad Things if a log got too big, in particular if it hit a file
6813 size limit. Exim now panics and dies if it cannot write to a log file, just
6814 as it does if it cannot open a log file.
6816 39. Modified OS/Makefile-Linux so that it now contains
6818 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6820 The two -D definitions ensure that Exim is compiled with large file
6821 support, which makes it possible to handle log files that are bigger than
6824 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6825 instance) a domain was checked against a named list that involved a lookup,
6826 causing $domain_data to be set, then another domain was checked against the
6827 same list, then the first domain was re-checked, the value of $domain_data
6828 after the final check could be wrong. In particular, if the second check
6829 failed, it could be set empty. This bug probably also applied to
6832 41. The strip_trailing_dot option was not being applied to the address given
6833 with the -f command-line option.
6835 42. The code for reading a message's header from the spool was incrementing
6836 $received_count, but never initializing it. This meant that the value was
6837 incorrect (doubled) while delivering a message in the same process in which
6838 it was received. In the most common configuration of Exim, this never
6839 happens - a fresh exec is done - but it can happen when
6840 deliver_drop_privilege is set.
6842 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6843 it now includes up to 150 characters of the unexpected data in the log
6846 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6847 and building data strings. The size of both of these buffers was 10 000
6848 bytes - far larger than anybody would *ever* want, thought I. Needless to
6849 say, somebody hit the limit. I have increased the maximum line length to
6850 20 000 and the maximum data length of concatenated lines to 100 000. I have
6851 also fixed two bugs, because there was no checking on these buffers. Tsk,
6852 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6853 buffer is too small.
6855 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6856 lsearch lookups. Now it does.
6858 46. When parsing a route_list item in a manualroute router, a fixed-length
6859 buffer was used for the list of hosts. I made this 1024 bytes long,
6860 thinking that nobody would ever have a list of hosts that long. Wrong.
6861 Somebody had a whole pile of complicated expansion conditions, and the
6862 string was silently truncated, leading to an expansion error. It turns out
6863 that it is easier to change to an unlimited length (owing to other changes
6864 that have happened since this code was originally written) than to build
6865 structure for giving a limitation error. The length of the item that
6866 expands into the list of hosts is now unlimited.
6868 47. The lsearch lookup could not handle data where the length of text line was
6869 more than 4095 characters. Such lines were truncated, leading to shortened
6870 data being returned. It should now handle lines of any length.
6872 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6873 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6876 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6877 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6878 no longer happens in comment lines. A list of replacements is now placed
6879 at the head of all of the source files, except those whose only change is
6880 to replace PERL_COMMAND in the very first #! line.
6882 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6883 messages on the queue, with a bottom-up merge sort, using code contributed
6884 by Michael Haardt. This should make operations like -bp somewhat faster on
6885 large queues. It won't affect queue runners, except when queue_run_in_order
6888 51. Installed eximstats 1.31 in the distribution.
6890 52. Added support for SRV lookups to the dnslookup router.
6892 53. If an ACL referred to $message_body or $message_body_end, the value was not
6893 reset for any messages that followed in the same SMTP session.
6895 54. The store-handling optimization for building very long strings was not
6896 differentiating between the different store pools. I don't think this
6897 actually made any difference in practice, but I've tidied it.
6899 55. While running the routers to verify a sender address, $sender_address
6900 was still set to the sender address. This is wrong, because when routing to
6901 send a bounce to the sender, it would be empty. Therefore, I have changed
6902 it so that, while verifying a sender address, $sender_address is set to <>.
6903 (There is no change to what happens when verifying a recipient address.)
6905 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6906 target A or AAAA records (if not already returned) without resetting the
6907 qualify_single or search_parents options of the DNS resolver. These are
6908 inappropriate in this case because the targets of MX and SRV records must
6909 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6910 target that, when qualified, matched something in the local domain. These
6911 two options are now turned off when doing these lookups.
6913 57. It seems that at least some releases of Reiserfs (which does not have the
6914 concept of a fixed number of inodes) returns zero and not -1 for the
6915 number of available inodes. This interacted badly with check_spool_inodes,
6916 which assumed that -1 was the "no such thing" setting. What I have done is
6917 to check that the total number of inodes is greater than zero before doing
6918 the test of how many are available.
6920 58. When a "warn" ACL statement has a log_message modifier, the message is
6921 remembered, and not repeated. This is to avoid a lot of repetition when a
6922 message has many recipients that cause the same warning to be written.
6923 However, Exim was preserving the list of already written lines for an
6924 entire SMTP session, which doesn't seem right. The memory is now reset if a
6925 new message is started.
6927 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6928 debugging output unless log_rewrite was also set.
6930 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6931 of (int)(handle) when we know that handle contains (void *)(-1).
6933 61. The Exim daemon panic-logs an error return when it closes the incoming
6934 connection. However "connection reset by peer" seems to be common, and
6935 isn't really an error worthy of noting specially, so that particular error
6938 62. When Exim is trying to find all the local interfaces, it used to panic and
6939 die if the ioctl to get the interface flags failed. However, it seems that
6940 on at least one OS (Solaris 9) it is possible to have an interface that is
6941 included in the list of interfaces, but for which you get a failure error
6942 for this call. This happens when the interface is not "plumbed" into a
6943 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6944 failure of the "get flags" call assumes that the interface is down.
6946 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6947 makes life easier for people who are doing arithmetic on fields extracted
6948 from dates, where you often get leading zeros that should not be
6949 interpreted as octal.
6951 64. Added qualify_domain to the redirect router, to override the global
6954 65. If a pathologically long header line contained very many addresses (the
6955 report of this problem mentioned 10 000) and each of them was rewritten,
6956 Exim could use up a very large amount of memory. (It kept on making new
6957 copies of the header line as it rewrote, and never released the old ones.)
6958 At the expense of a bit more processing, the header rewriting function has
6959 been changed so that it no longer eats memory in this way.
6961 66. The generation of the Received: header has been moved from the time that a
6962 message starts to be received, to the time that it finishes. The timestamp
6963 in the Received: header should now be very close to that of the <= log
6964 line. There are two side-effects of this change:
6966 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6967 logged header lines no longer include the local Received: line, because
6968 it has not yet been created. The same applies to a copy of the message
6969 that is returned to a non-SMTP sender when a message is rejected.
6971 (b) When a filter file is tested using -bf, no additional Received: header
6972 is added to the test message. After some thought, I decided that this
6975 This change does not affect the value of $received_for. It is still set
6976 after address rewriting, but before local_scan() is called.
6978 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6980 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6981 gave an unhelpful panic error message, and a defer error. I have managed to
6982 change this behaviour so that it now rejects any supplied certificate,
6983 which seems right, as the list of acceptable certificates is empty.
6985 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6986 gave an unhelpful defer error. I have not managed to make this reject any
6987 supplied certificates, but the error message it gives is "no certificate
6988 supplied", which is not helpful.
6990 70. exigrep's output now also includes lines that are not associated with any
6991 message, but which match the given pattern. Implemented by a patch from
6992 Martin Sluka, which also tidied up the Perl a bit.
6994 71. Recipient callout verification, like sender verification, was using <> in
6995 the MAIL FROM command. This isn't really the right thing, since the actual
6996 sender may affect whether the remote host accepts the recipient or not. I
6997 have changed it to use the actual sender in the callout; this means that
6998 the cache record is now keyed on a recipient/sender pair, not just the
6999 recipient address. There doesn't seem to be a real danger of callout loops,
7000 since a callout by the remote host to check the sender would use <>.
7001 [SEE ABOVE: changed after hitting problems.]
7003 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7004 temporary errors. However, in the case of such a code being given after
7005 the end of a data transmission (i.e. after ".") Exim was failing to write
7006 a retry record for the message. (Yes, there was some broken host that was
7007 actually sending 8xx at this point.)
7009 73. An unknown lookup type in a host list could cause Exim to panic-die when
7010 the list was checked. (An example that provoked this was putting <; in the
7011 middle of a list instead of at the start.) If this happened during a DATA
7012 ACL check, a -D file could be left lying around. This kind of configuration
7013 error no longer causes Exim to die; instead it causes a defer error. The
7014 incident is still logged to the main and panic logs.
7016 74. Buglet left over from Exim 3 conversion. The message "too many messages
7017 in one connection" was written to the rejectlog but not the mainlog, except
7018 when address rewriting (yes!) was being logged.
7020 75. Added write_rejectlog option.
7022 76. When a system filter was run not as root (that is, when system_filter_user
7023 was set), the values of the $n variables were not being returned to the
7024 main process; thus, they were not subsequently available in the $sn
7027 77. Added +return_path_on_delivery log selector.
7029 78. A connection timeout was being treated differently from recipients deferred
7030 when testing hosts_max_try with a message that was older than the host's
7031 retry timeout. (The host should not be counted, thus allowing all hosts to
7032 be tried at least once before bouncing.) This may have been the cause of an
7033 occasionally reported bug whereby a message would remain on the queue
7034 longer than the retry timeout, but would be bounced if a delivery was
7035 forced. I say "may" because I never totally pinned down the problem;
7036 setting up timeout/retry tests is difficult. See also the next item.
7038 79. The ultimate address timeout was not being applied to errors that involved
7039 a combination of host plus message (for example, a timeout on a MAIL
7040 command). When an address resolved to a number of possible hosts, and they
7041 were not all tried for each delivery (e.g. because of hosts_max_try), a
7042 message could remain on the queue longer than the retry timeout.
7044 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7047 81. Fixed an obscure SMTP outgoing bug which required at least the following
7048 conditions: (a) there was another message waiting for the same server;
7049 (b) the server returned 5xx to all RCPT commands in the first message so
7050 that the message was not completed; (c) the server dropped the connection
7051 or gave a negative response to the RSET that Exim sends to abort the
7052 transaction. The observed case was a dropped connection after DATA that had
7053 been sent in pipelining mode. That is, the server had advertised PIPELINING
7054 but was not implementing it correctly. The effect of the bug was incorrect
7055 behaviour, such as trying another host, and this could lead to a crash.
7061 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7062 and daemon.c were passed as pointers to ints; they should have been
7063 pointers to socklen_t variables (which are typically unsigned ints).
7065 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7068 3. Fixed a really odd bug that affected only the testing scheme; patching a
7069 certain fixed string in the binary changed the value of another string that
7070 happened to be identical to the end of the original first string.
7072 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7073 name", it returns that address as the IP address. On some operating
7074 systems (e.g. Solaris), it also passes back the IP address string as the
7075 "host name". However, on others (e.g. Linux), it passes back an empty
7076 string. Exim wasn't checking for this, and was changing the host name to an
7077 empty string, assuming it had been canonicalized.
7079 5. Although rare, it is permitted to have more than one PTR record for a given
7080 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7081 all the names associated with an address, because they do in Solaris.
7082 However, it seems that they do not in Linux for data that comes from the
7083 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7084 I found this out when I moved to a new Linux workstation and tried to run
7085 the Exim test suite.
7087 To get round this problem I have changed the code so that it now does its
7088 own call to the DNS to look up PTR records when searching for a host name.
7089 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7090 addresses that are only in /etc/hosts are still found.
7092 This behaviour is, however, controlled by an option called host_lookup_
7093 order, which defaults to "bydns:byaddr". If people want to use the other
7094 order, or indeed, just use one or the other means of lookup, they can
7095 specify it in this variable.
7097 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7098 some operating systems, this comes back from gethostbyaddr() as an empty
7099 string, and this is what Exim used to test for. However, it seems that in
7100 other systems, "." is yielded. Exim now tests for this case too.
7102 7. The values of check_spool_space and check_log_space are now held internally
7103 as a number of kilobytes instead of an absolute number of bytes. If a
7104 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7105 kilobyte. This means that much larger values can be stored.
7107 8. Exim monitor: an attempt to get the action menu when not actually pointing
7108 at a message produces an empty menu entitled "No message selected". This
7109 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7110 no entries in it ("Shell widget menu has zero width and/or height"). So I
7111 have added a single, blank menu entry in this case.
7113 9. Added ${quote_local_part.
7115 10. MIME decoding is now applied to the contents of Subject: header lines when
7118 11. Now that a reference to $sender_host_address automatically causes a reverse
7119 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7120 host lookup before query-style lookups in lists that might use this
7121 variable. This has therefore been abolished, and the "net-" prefix is no
7122 longer necessary for query-style lookups.
7124 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7125 have been a typo for LFLAGS, so it has been changed.
7127 13. The install script calls Exim with "-C /dev/null" in order to find the
7128 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7129 to be output. However, since Exim outputs its version number before the
7130 error, it didn't break the script. It just looked ugly. I fixed this by
7131 always allowing "-C /dev/null" if the caller is root.
7133 14. Ignore overlarge ACL variable number when reading spool file - insurance
7134 against a later release with more variables having written the file.
7136 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7137 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7138 this, as well as the form without the "IPv6" on the front.
7140 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7141 OS/Makefile-Darwin file.
7143 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7144 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7146 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7147 variable. However, some OS use ints for inodes. I've added cast to long int
7148 to get rid of the compiler warning.
7150 19. I had forgotten to lock out "/../" in configuration file names when
7151 ALT_CONFIG_PREFIX was set.
7153 20. Routers used for verification do not need to specify transports. However,
7154 if such a router generated a host list, and callout was configured, Exim
7155 crashed, because it could not find a port number from the (non-existent)
7156 transport. It now assumes port 25 in this circumstance.
7158 21. Added the -t option to exigrep.
7160 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7161 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7162 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7163 them is set, LOOKUP_LSEARCH is forced.
7165 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7167 24. Added sender and host information to the "rejected by local_scan()" log
7168 line; previously there was no indication of these.
7170 25. Added .include_if_exists.
7172 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7173 Linux. It turns out that not all file systems support this. Apparently some
7174 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7175 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7176 not supported on the file descriptor, is now ignored when Exim is trying to
7177 sync a directory. This applies only to Linux.
7179 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7181 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7182 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7185 29. Added support for maildirsize files from supplied patch (modified a bit).
7187 30. The use of :fail: followed by an empty string could lead Exim to respond to
7188 sender verification failures with (e.g.):
7190 550 Verification failed for <xxx>
7191 550 Sender verify failed
7193 where the first response line was missing the '-' that indicates it is not
7194 the final line of the response.
7196 31. The loop for finding the name of the user that called Exim had a hardwired
7197 limit of 10; it now uses the value of finduser_retries, which is used for
7198 all other user lookups.
7200 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7203 33. Exim was neglecting to zero errno before one call of strtol() when
7204 expanding a string and expecting an integer value. On some systems this
7205 resulted in spurious "integer overflow" errors. Also, it was casting the
7206 result into an int without checking.
7208 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7209 did not work. The code looks as if it has *never* worked, though it appears
7210 to have been documented since at least release 1.62. I have made it work.
7212 35. The "timeout_DNS" error in retry rules, also documented since at least
7213 1.62, also never worked. As it isn't clear exactly what this means, and
7214 clearly it isn't a major issue, I have abolished the feature by treating it
7215 as "timeout", and writing a warning to the main and panic logs.
7217 36. The display of retry rules for -brt wasn't always showing the error code
7220 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7221 timeout_connect_A, timeout_connect_MX.
7223 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7224 to the empty sender.
7226 39. The daemon was not analysing the content of -oX till after it had closed
7227 stderr and disconnected from the controlling terminal. This meant that any
7228 syntax errors were only noted on the panic log, and the return code from
7229 the command was 0. By re-arranging the code a little, I've made the
7230 decoding happen first, so such errors now appear on stderr, and the return
7231 code is 1. However, the actual setting up of the sockets still happens in
7232 the disconnected process, so errors there are still only recorded on the
7235 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7236 connections (as happens on some IP stacks) was logged at start up time as
7237 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7238 it from "IPv6 and IPv4", which means that two separate sockets are being
7241 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7242 says whether AF_INET or AF_INET6 was passed as an argument.
7244 42. Exiwhat output was messed up when time zones were included in log
7247 43. Exiwhat now gives more information about the daemon's listening ports,
7248 and whether -tls-on-connect was used.
7250 44. The "port" option of the smtp transport is now expanded.
7252 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7253 silently ignored. Now an error message is written to the main and panic
7256 46. There's a new ACL modifier called "logwrite" which writes to a log file
7257 as soon as it is encountered.
7259 47. Added $local_user_uid and $local_user_gid at routing time.
7261 48. Exim crashed when trying to verify a sender address that was being
7264 49. Exim was recognizing only a space character after ".include". It now also
7265 recognizes a tab character.
7267 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7268 extracting the relevant information from the specification. The man page no
7269 longer contains scrambled data for the -d option, and I've added a section
7270 at the front about calling Exim under different names.
7272 51. Added "extra_headers" argument to the "mail" command in filter files.
7274 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7277 53. Installed eximstats 1.29.
7279 54. Added transport_filter_timeout as a generic transport option.
7281 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7282 Cc: header lines. This was required by RFC 822, but it not required by RFC
7285 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7286 incoming messages that did not have them. Now it does so only if the
7287 message originates locally, that is, if there is no associated remote host
7288 address. When Resent- header lines are present, this applies to the Resent-
7289 lines rather than the non-Resent- lines.
7291 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7292 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7294 58. Messages for configuration errors now include the name of the main
7295 configuration file - useful now that there may be more than one file in a
7296 list (.included file names were always shown).
7298 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7299 for those rare installations that do not start the daemon as root or run it
7300 setuid root. I've cut out the call to initgroups() if the daemon is not
7303 60. The Exim user and group can now be bound into the binary as text strings
7304 that are looked up at the start of Exim's processing.
7306 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7308 62. Added $mailstore_basename variable.
7310 63. Installed patch to sieve.c from Michael Haardt.
7312 64. When Exim failed to open the panic log after failing to open the main log,
7313 the original message it was trying to log was written to stderr and debug
7314 output, but if they were not available (the usual case in production), it
7315 was lost. Now it is written to syslog before the two lines that record the
7316 failures to open the logs.
7318 65. Users' Exim filters run in subprocesses under the user's uid. It is
7319 possible for a "deliver" command or an alias in a "personal" command to
7320 provoke an address rewrite. If logging of address rewriting is configured,
7321 this fails because the process is not running as root or exim. There may be
7322 a better way of dealing with this, but for the moment (because 4.30 needs
7323 to be released), I have disabled address rewrite logging when running a
7324 filter in a non-root, non-exim process.
7330 1. The buildconfig auxiliary program wasn't quoting the value set for
7331 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7332 not defined. This bug was masked in 4.22 by the effect that was fixed in
7335 2. Some messages that were rejected after a message id was allocated were
7336 shown as "incomplete" by exigrep. It no longer does this for messages that
7337 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7339 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7340 have allow_domain_literals set, the ID did not get logged in the <= line.
7341 Domain literals are now always recognized in Message-ID: header lines.
7343 4. The first argument for a ${extract expansion item is the key name or field
7344 number. Leading and trailing spaces in this item were not being ignored,
7345 causing some misleading effects.
7347 5. When deliver_drop_privilege was set, single queue runner processes started
7348 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7349 same command in the process it spins off) were not dropping privilege.
7351 6. When the daemon running as "exim" started a queue runner, it always
7352 re-executed Exim in the spun-off process. This is a waste of effort when
7353 deliver_drop_privilege is set. The new process now just calls the
7354 queue-runner function directly.
7360 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7363 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7364 ignored. Though the use of -oP was forcing the writing of a pid file, it
7365 was always written to the default place.
7367 3. If the message "no IP address found for host xxxx" is generated during
7368 incoming verification, it is now followed by identification of the incoming
7369 connection (so you can more easily find what provoked it).
7371 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7373 5. Added some features to "harden" Exim a bit more against certain attacks:
7375 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7376 be put in Local/Makefile. This is like the never_users runtime option,
7377 but it cannot be overridden. The default setting is "root".
7379 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7380 prefix string with which any file named in a -C command line option
7383 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7384 is retained for -C and -D only if the caller of Exim is root. Without
7385 it, the exim user may also use -C and -D and retain privilege.
7387 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7388 command line option is disabled.
7390 6. Macro names set by the -D option must start with an upper case letter, just
7391 like macro names defined in the configuration file.
7393 7. Added "dereference=" facility to LDAP.
7395 8. Two instances of the typo "uknown" in the source files are fixed.
7397 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7398 the Configure-Makefile script screwed up while processing it.
7400 10. Incorporated PCRE 4.4.
7402 11. The SMTP synchronization check was not operating right at the start of an
7403 SMTP session. For example, it could not catch a HELO sent before the client
7404 waited for the greeting. There is now a check for outstanding input at the
7405 point when the greeting is written. Because of the duplex, asynchronous
7406 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7407 way, but not yet received, when the check is performed.
7409 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7410 on TCP/IP sockets, because this apparently causes some broken clients to
7413 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7414 unchanged) from the Cygwin maintainer.
7416 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7417 was supported instead of testing for mbx. Effectively a typo.
7419 15. The spa authenticator server code was not checking that the input it
7420 received was valid base64.
7422 16. The debug output line for the "set" modifier in ACLs was not showing the
7423 name of the variable that was being set.
7425 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7427 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7428 lookup on its own. Something else was needed to trigger the lookup. For
7429 example, a match in host_lookup or the need for a host name in a host list.
7430 Now, if $sender_host_name is referenced and the host name has not yet been
7431 looked up, a lookup is performed. If the lookup fails, the variable remains
7432 empty, and $host_lookup_failed is set to "1".
7434 19. Added "eqi" as a case-independent comparison operator.
7436 20. The saslauthd authentication condition could segfault if neither service
7437 nor realm was specified.
7439 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7440 error message that was logged was misleading, and incoming SMTP
7441 connections were dropped. The message is now more accurate, and temporary
7442 errors are given to SMTP connections.
7444 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7445 (except RSET) with a 421 error, until QUIT is received. However, it was
7446 failing to send a response to QUIT.
7448 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7449 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7450 ACL is now run after the helo_try_verify_hosts code.
7452 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7453 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7454 case-independent, so other case variants are also recognized). Apparently
7455 some systems use these upper case variants.
7457 25. If more than two messages were waiting for the same host, and a transport
7458 filter was specified for the transport, Exim sent two messages over the
7459 same TCP/IP connection, and then failed with "socket operation on non-
7460 socket" when it tried to send the third.
7462 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7464 27. The extern definition of crypt16() in expand.c was not being excluded when
7465 the OS had its own crypt16() function.
7467 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7468 as a preferred synonym for return_size_limit, both as an option and as an
7471 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7473 30. Changed the default configuration ACL to relax the local part checking rule
7474 for addresses that are not in any local domains. For these addresses,
7475 slashes and pipe symbols are allowed within local parts, but the sequence
7476 /../ is explicitly forbidden.
7478 31. SPA server authentication was not clearing the challenge buffer before
7481 32. log_message in a "warn" ACL statement was writing to the reject log as
7482 well as to the main log, which contradicts the documentation and doesn't
7483 seem right (because no rejection is happening). So I have stopped it.
7485 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7486 However, I am unable to do any testing of this.
7488 34. Fixed an infelicity in the appendfile transport. When checking directories
7489 for a mailbox, to see if any needed to be created, it was accidentally
7490 using path names with one or more superfluous leading slashes; tracing
7491 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7493 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7494 contents are added to the log line that is written for every discarded
7495 recipient. (Previously a log_message setting was ignored.)
7497 36. The ${quote: operator now quotes the string if it is empty.
7499 37. The install script runs exim in order to find its version number. If for
7500 some reason other than non-existence or emptiness, which it checks, it
7501 could not run './exim', it was installing it with an empty version number,
7502 i.e. as "exim-". This error state is now caught, and the installation is
7505 38. An argument was missing from the function that creates an error message
7506 when Exim fails to connect to the socket for saslauthd authentication.
7507 This could cause Exim to crash, or give a corrupted message.
7509 39. Added isip, isip4, and isip6 to ${if conditions.
7511 40. The ACL variables $acl_xx are now saved with the message, and can be
7512 accessed later in routers, transports, and filters.
7514 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7515 strings in the file are not string-expanded.
7517 42. If a MAIL command specified a SIZE value that was too large to fit into an
7518 int variable, the check against message_size_limit failed. Such values are
7519 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7520 day this will have to be increased, but I don't think I want to be around
7521 when emails are that large.
7528 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7529 iconv() is not standard in FreeBSD.
7531 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7532 IPv6 enabled. The observed symptom was a segmentation fault on return from
7533 the function os_common_find_running_interfaces() in src/os.c.
7535 3. In the check_special_case() function in daemon.c I had used "errno" as an
7536 argument name, which causes warnings on some systems. This was basically a
7537 typo, since it was named "eno" in the comments!
7539 4. The code that waits for the clock to tick (at a resolution of some fraction
7540 of a second) so as to ensure message-id uniqueness was always waiting for
7541 at least one whole tick, when it could have waited for less. [This is
7542 almost certainly not relevant at current processor speeds, where it is
7543 unlikely to ever wait at all. But we try to future-proof.]
7545 5. The function that sleeps for a time interval that includes fractions of a
7546 second contained a race. It did not block SIGALRM between setting the
7547 timer, and suspending (a couple of lines later). If the interval was short
7548 and the sigsuspend() was delayed until after it had expired, the suspension
7549 never ended. On busy systems this could lead to processes getting stuck for
7552 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7553 process, before it forks any delivery processes. The open lookup caching
7554 mechanism meant that the open file or database connection was passed into
7555 the delivery process. The problem was that delivery processes always tidy
7556 up cached lookup data. This could cause a problem for the next delivery
7557 process started by the queue runner, because the external queue runner
7558 process does not know about the closure. So the next delivery process
7559 still has data in the lookup cache. In the case of a file lookup, there was
7560 no problem because closing a file descriptor in a subprocess doesn't affect
7561 the parent. However, if the lookup was caching a connection to a database,
7562 the connection was closed, and the second delivery process was likely to
7563 see errors such as "PGSQL: query failed: server closed the connection
7564 unexpectedly". The problem has been fixed by closing all cached lookups
7565 in a queue runner before running a delivery process.
7567 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7568 seem to have the "const" qualifier which it has on other OS. I've
7571 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7572 *of the same type* (client or server) with the same public name that an
7573 error should be diagnosed.
7575 9. When Exim looked up a host name for an IP address, but failed to find the
7576 original IP address when looking up the host name (a safety check), it
7577 output the message "<ip address> does not match any IP for NULL", which was
7578 confusing, to say the least. The bug was that the host name should have
7579 appeared instead of "NULL".
7581 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7582 user that is built into the binary, and the -C or -D options is used, root
7583 privilege is dropped before the configuration file is read. In addition,
7584 logging is switched to stderr instead of the normal log files. If the
7585 configuration then re-defines the Exim user, the unprivileged environment
7586 is probably not what is expected, so Exim logs a panic warning message (but
7589 However, if deliver_drop_privilege is set, the unprivileged state may well
7590 be exactly what is intended, so the warning has been cut out in that case,
7591 and Exim is allowed to try to write to its normal log files.
7597 1. smtp_return_error_details was not giving details for temporary sender
7598 or receiver verification errors.
7600 2. Diagnose a configuration error if two authenticators have the same public
7603 3. Exim used not to create the message log file for a message until the first
7604 delivery attempt. This could be confusing when incoming messages were held
7605 for policy or load reasons. The message log file is now created at the time
7606 the message is received, and an initial "Received" line is written to it.
7608 4. The automatically generated man page for command line options had a minor
7609 bug that caused no ill effects; however, a more serious problem was that
7610 the procedure for building the man page automatically didn't always
7611 operate. Consequently, release 4.20 contains an out-of-date version. This
7612 shouldn't happen again.
7614 5. When building Exim with embedded Perl support, the script that builds the
7615 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7616 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7618 6. The freeze_tell option was not being used for messages that were frozen on
7619 arrival, either by an ACL or by local_scan().
7621 7. Added the smtp_incomplete_transaction log selector.
7623 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7624 was accepting AUTH without a new EHLO.
7626 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7627 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7630 10. Logging of TCP/IP connections (when configured) now happens in the main
7631 daemon process instead of the child process, so that the TCP/IP connection
7632 count is more accurate (but it can never be perfect).
7634 11. The use of "drop" in a nested ACL was not being handled correctly in the
7635 outer ACL. Now, if condition failure induced by the nested "drop" causes
7636 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7637 or "require"), the connection is dropped.
7639 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7640 that yield "discard" can now be used with an "accept" or a "discard" verb,
7641 but an error is generated for any others (because I can't see a useful way
7642 to define what should happen).
7644 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7645 are now processed in the same way as lines in the Exim configuration file.
7646 In particular, continuation lines are supported.
7648 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7650 15. Added -ti meaning -t -i.
7652 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7653 domains, and warn by logging if others are found.
7655 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7656 returned from GIFCONF when Exim is trying to find the list of interfaces on
7657 a host. The code in os.c has been modified to copy each ifreq to an aligned
7658 structure in all cases.
7660 Also, in some cases, the returned ifreq's were being copied to a 'struct
7661 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7662 means the last couple of bytes of an IPv6 address could be chopped if the
7663 ifreq contained only a normal sockaddr (14 bytes storage).
7665 18. Named domain lists were not supported in the hosts_treat_as_local option.
7666 An entry such as +xxxx was not recognized, and was treated as a literal
7669 19. Ensure that header lines added by a DATA ACL are included in the reject log
7670 if the ACL subsequently rejects the message.
7672 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7673 MD5 (which is deprecated).
7675 21. When testing a filter file using -bf, Exim was writing a message when it
7676 took the sender from a "From " line in the message, but it was not doing so
7677 when it took $return_path from a Return-Path: header line. It now does.
7679 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7680 with a valid header line field name (a series of printing characters
7681 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7683 23. Changed "disc" in the source to "disk" to conform to the documentation and
7684 the book and for uniformity.
7686 24. Ignore Sendmail's -Ooption=value command line item.
7688 25. When execve() failed while trying to run a command in a pipe transport,
7689 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7690 could be confused with a return value of 69 from the command itself. This
7691 has been changed to 127, the value the shell returns if it is asked to run
7692 a non-existent command. The wording for the related log line suggests a
7693 non-existent command as the problem.
7695 26. If received_header_text expands to an empty string, do not add a Received:
7696 header line to the message. (Well, it adds a token one on the spool, but
7697 marks it "old" so that it doesn't get used or transmitted.)
7699 27. Installed eximstats 1.28 (addition of -nt option).
7701 28. There was no check for failure on the call to getsockname() in the daemon
7702 code. This can fail if there is a shortage of resources on the system, with
7703 ENOMEM, for example. A temporary error is now given on failure.
7705 29. Contrary to the C standard, it seems that in some environments, the
7706 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7707 program. Exim now does this explicitly; it affects the formatting of
7708 timestamps using strftime().
7710 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7711 complaints. I've now initialized all the variables, to avoid this.
7713 32. Header lines added by a system filter were not being "seen" during
7714 transport-time rewrites.
7716 33. The info_callback() function passed to OpenSSL is set up with type void
7717 (*)(SSL *, int, int), as described somewhere. However, when calling the
7718 function (actually a macro) that sets it up, the type void(*)() is
7719 expected. I've put in a cast to prevent warnings from picky compilers.
7721 34. If a DNS black list lookup found a CNAME record, but there were no A
7722 records associated with the domain it pointed at, Exim crashed.
7724 35. If a DNS black list lookup returned more than one A record, Exim ignored
7725 all but the first. It now scans all returned addresses if a particular IP
7726 value is being sought. In this situation, the contents of the
7727 $dnslist_value variable are a list of all the addresses, separated by a
7730 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7731 to accept a host name and all its aliases if the forward lookup for any of
7732 them yielded the IP address of the incoming connection. Now it accepts only
7733 those names whose forward lookup yields the correct IP address. Any other
7734 names are discarded. This closes a loophole whereby a rogue DNS
7735 administrator could create reverse DNS records to break through a
7736 wildcarded host restriction in an ACL.
7738 37. If a user filter or a system filter that ran in a subprocess used any of
7739 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7740 the wrong values were passed to the pipe command ($thisaddress had the
7741 value of $0, $0 had the value of $1, etc). This bug was introduced by
7742 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7744 38. Improved the line breaking for long SMTP error messages from ACLs.
7745 Previously, if there was no break point between 40 and 75 characters, Exim
7746 left the rest of the message alone. Two changes have been made: (a) I've
7747 reduced the minimum length to 35 characters; (b) if it can't find a break
7748 point between 35 and 75 characters, it looks ahead and uses the first one
7749 that it finds. This may give the occasional overlong line, but at least the
7750 remaining text gets split now.
7752 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7753 file descriptors might be low, and that setting 1000 would always raise it.
7754 It turns out that in some environments, the limit is already over 1000 and
7755 that lowering it causes trouble. So now Exim takes care not to decrease it.
7757 40. When delivering a message, the value of $return_path is set to $sender_
7758 address at the start of routing (routers may change the value). By an
7759 oversight, this default was not being set up when an address was tested by
7760 -bt or -bv, which affected the outcome if any router or filter referred to
7763 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7764 log only when "message" or "log_message" are set. However, if one of the
7765 conditions was an address verification, or a call to a nested ACL, the
7766 messages generated by the underlying test were being passed through. This
7767 no longer happens. The underlying message is available in $acl_verify_
7768 message for both "message" and "log_message" expansions, so it can be
7769 passed through if needed.
7771 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7772 new expansion $bh_ to give the encoded byte string without charset
7773 translation. Translation happens only if iconv() is available; HAVE_ICONV
7774 indicates this at build time. HEADERS_CHARSET gives the charset to
7775 translate to; headers_charset can change it in the configuration, and
7776 "headers charset" can change it in an individual filter file.
7778 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7779 that creates RFC 2047 encoded "words" labels them as that charset instead
7780 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7781 expansion operator; (ii) when Exim creates a From: line for a local
7782 message; (iii) when a header line is rewritten to include a "phrase" part.
7784 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7785 buggy, causing it to skip the first lines of messages whose message ID
7786 ended in 'D'. This would not have bitten before Exim release 4.14, because
7787 message IDs were unlikely to end in 'D' before then. The effect was to have
7788 incorrect size information for certain domains.
7790 45. #include "config.h" was missing at the start of the crypt16.c module. This
7791 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7794 46. If there was a timeout during a "random" callout check, Exim treated it as
7795 a failure of the random address, and carried on sending RSET and the real
7796 address. If the delay was just some slowness somewhere, the response to the
7797 original RCPT would be taken as a response to RSET and so on, causing
7798 mayhem of various kinds.
7800 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7801 when I implemented it. It didn't allow for the fact that some option values
7802 may legitimately be negative (e.g. size_addition), and it didn't even do
7803 the right test for positive values.
7805 48. Domain names in DNS records are case-independent. Exim always looks them up
7806 in lower case. Some resolvers return domain names in exactly the case they
7807 appear in the zone file, that is, they may contain uppercase letters. Not
7808 all resolvers do this - some return always lower case. Exim was treating a
7809 change of case by a resolver as a change of domain, similar to a widening
7810 of a domain abbreviation. This triggered its re-routing code and so it was
7811 trying to route what was effectively the same domain again. This normally
7812 caused routing to fail (because the router wouldn't handle the domain
7813 twice). Now Exim checks for this case specially, and just changes the
7814 casing of the domain that it ultimately uses when it transmits the message
7817 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7820 50. If a filter generated a file delivery with a non-absolute name (possible if
7821 no home directory exists for the router), the forbid_file option was not
7824 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7825 the existing equality matching.
7827 52. Exim was using ints instead of ino_t variables in some places where it was
7828 dealing with inode numbers.
7830 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7831 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7832 called TMPDIR, and if it finds it is different, it changes its value.
7834 54. The smtp_printf() function is now made available to local_scan() so
7835 additional output lines can be written before returning. There is also an
7836 smtp_fflush() function to enable the detection of a dropped connection.
7837 The variables smtp_input and smtp_batched_input are exported to
7840 55. Changed the default runtime configuration: the message "Unknown user"
7841 has been removed from the ACL, and instead placed on the localuser router,
7842 using the cannot_route_message feature. This means that any verification
7843 failures that generate their own messages won't get overridden. Similarly,
7844 the "Unrouteable address" message that was in the ACL for unverifiable
7845 relay addresses has also been removed.
7847 56. Added hosts_avoid_esmtp to the smtp transport.
7849 57. The exicyclog script was not checking for the esoteric option
7850 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7851 will work only if exicyclog is run under the appropriate euid.
7853 58. Following a discussion on the list, the rules by which Exim recognises line
7854 endings on incoming messages have been changed. The -dropcr and drop_cr
7855 options are now no-ops, retained only for backwards compatibility. The
7856 following line terminators are recognized: LF CRLF CR. However, special
7857 processing applies to CR:
7859 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7860 nor a local message in the state where . is a terminator.
7862 (ii) If a bare CR is encountered in a header line, an extra space is added
7863 after the line terminator so as not to end the header. The reasoning
7864 behind this is that bare CRs in header lines are most likely either
7865 to be mistakes, or people trying to play silly games.
7867 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7868 was being incorrectly given as 18 bytes larger than it should have been.
7869 This is a VOB (very old bug).
7871 60. This may never have affected anything current, but just in case it has:
7872 When the local host is found other than at the start of a list of hosts,
7873 the local host, those with the same MX, and any that follow, are discarded.
7874 When the list in question was part of a longer list of hosts, the following
7875 hosts (not currently being processed) were also being discarded. This no
7876 longer happens. I'm not sure if this situation could ever has previously
7879 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7882 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7883 header. This is something that is recommended in a new Internet Draft, and
7884 is something that is documented as being done by Sendmail. There are two
7885 possible values. For messages generated by the autoreply transport, Exim
7888 Auto-Submitted: auto-replied
7890 whereas for all other generated messages (e.g. bounces) it adds
7892 Auto-Submitted: auto-generated
7894 63. The "personal" condition in filters now includes a test for the
7895 Auto-Submitted: header. If it contains the string "auto-" the message it
7896 not considered personal.
7898 64. Added rcpt_include_affixes as a generic transport option.
7900 65. Added queue_only_override (default true).
7902 66. Added the syslog_duplication option.
7904 67. If what should have been the first header line of a message consisted of
7905 a space followed by a colon, Exim was mis-interpreting it as a header line.
7906 It isn't of course - it is syntactically invalid and should therefore be
7907 treated as the start of the message body. The misbehaviour could have
7908 caused a number of strange effects, including loss of data in subsequent
7909 header lines, and spool format errors.
7911 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7912 client host had authenticated. This control can now be exercised by an ACL
7913 for more flexibility.
7915 69. By default, callouts do not happen when testing with -bh. There is now a
7916 variant, -bhc, which does actually run the callout code, including
7917 consulting and updating the callout cache.
7919 70. Added support for saslauthd authentication, courtesy of Alexander
7922 71. If statvfs() failed on the spool or log directories while checking their
7923 size for availability, Exim confusingly gave the error "space shortage".
7924 Furthermore, in debugging mode it crashed with a floating point exception.
7925 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7926 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7927 serious problem, Exim now writes to the main and panic logs when this
7928 happens, with details of the failure. It then refuses to accept the
7929 incoming message, giving the message "spool directory problem" or "log
7930 directory problem" with a 421 code for SMTP messages.
7932 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7933 0, 1, and 2 exist, because some OS complain for execs without them (see
7934 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7935 descriptors. However, the code omitted to check that the open succeeded,
7936 causing mysterious errors if for some reason the permissions on /dev/null
7937 got screwed. Now Exim writes a message to the main and panic logs, and
7938 bombs out if it can't open /dev/null.
7940 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7941 interact so that it is all more flexible. It is supposed to remain
7942 backwards compatible. Also added extra_local_interfaces.
7944 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7945 to bomb out with an assertion failure - to the client this appears as a
7946 connection drop. This problem occurs in the part of the code that was taken
7947 from the Samba project. Fortunately, the assertion is in a very simple
7948 function, so I have fixed this by reproducing the function inline in the
7949 one place where it is called, and arranging for authentication to fail
7950 instead of killing the process with assert().
7952 75. The SPA client code was not working when the server requested OEM rather
7953 than Unicode encoding.
7955 76. Added code to make require_files with a specific uid setting more usable in
7956 the case where statting the file as root fails - usually a non-root-mounted
7957 NFS file system. When this happens and the failure is EACCES, Exim now
7958 forks a subprocess and does the per-uid checking as the relevant uid.
7960 77. Added process_log_path.
7962 78. If log_file_path was not explicitly set, a setting of check_log_space or
7963 check_log_inodes was ignored.
7965 79. If a space check for the spool or log partitions fails, the incident is now
7966 logged. Of course, in the latter case the data may get lost...
7968 80. Added the %p formatting code to string_format() so that it can be used to
7969 print addresses in debug_print(). Adjusted all the address printing in the
7970 debugging in store.c to use %p rather than %d.
7972 81. There was a concern that a line of code in smtp_in.c could overflow a
7973 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7974 initially expressed, the concern was not well-founded, because trailing
7975 spaces are removed early. However, if the trailing spaces were followed by
7976 a NULL, they did not get removed, so the overflow was possible. Two fixes
7979 (a) I re-wrote the offending code in a cleaner fashion.
7980 (b) If an incoming SMTP command contains a NULL character, it is rejected
7983 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7984 runs initgroups(), so that if the Exim user is in any additional groups,
7985 they will be used during message reception.
7991 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob