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 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
178 better. Reported by Qualys.
180 PP/04 Impose security length checks on various command-line options.
181 Fixes CVE-2020-SPRSS reported by Qualys.
187 JH/01 Avoid costly startup code when not strictly needed. This reduces time
188 for some exim process initialisations. It does mean that the logging
189 of TLS configuration problems is only done for the daemon startup.
191 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
193 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
194 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
197 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
198 DKIM signing is being done. Previously a transport_filter would always
199 disable CHUNKING, falling back to traditional DATA.
201 JH/05 Regard command-line receipients as tainted.
203 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
205 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
206 PAM library frees one of the arguments given to it, despite the
207 documentation. Therefore a plain malloc must be used.
209 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
210 on-stack buffers were used, resulting in a taint trap when DSN information
211 copied from a received message was written into the buffer.
213 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
214 the ordering of its ARC headers. This caused a crash.
216 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
217 a new record was being constructed with information from the peer, a trap
220 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
221 installation would get error messages from DMARC verify, when it hit the
222 nonexistent file indicated by the default. Distros wanting DMARC enabled
223 should both provide the file and set the option.
224 Also enforce no DMARC verification for command-line sourced messages.
226 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
227 could, depending on the platform, hang at the STARTTLS response.
229 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
230 message on a connection. Previously if one message had ARC headers and
231 the following one did not, a crash could result when adding an
232 Authentication-Results: header.
234 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
235 Exim main code and Exim-related utities. The introduction of taint
236 tracking also did many adjustments to string handling. Since then, eximon
237 frequently terminated with an assert failure.
239 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
240 check for 452 responses. This slightly helps the inefficieny of doing
241 a large alias-expansion into a recipient-limited target. The max_rcpt
242 transport option still applies (and at the current default, will override
243 the new feature). The check is done for either cause of synch, and forces
244 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
245 connection. The new facility is not tunable at this time.
247 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
248 library live data was being used, so the results became garbage. Make
249 copies while it is still usable.
251 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
252 on delivery deferred (==) and failed (**) lines (if a delivery was
253 attemtped). Previously it was only on completion (=>) lines.
255 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
256 for the expansion of the server_scram_iter and server_scram_salt options.
258 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
259 are now specifically given a NO_DATA response without hitting the system
260 resolver. The library goes on to do the now-standard TXT lookup.
261 Use of dnsdb lookups is not affected.
263 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
264 only retrieve the errormessage once. Previously two calls to dlerror()
265 were used, and the second one (for mainlog/paniclog) retrieved null
268 JH/20 Taint checking: disallow use of tainted data for
269 - the appendfile transport file and directory options
270 - the pipe transport command
271 - the autoreply transport file, log and once options
272 - file names used by the redirect router (including filter files)
274 - paths used by single-key lookups
275 Previously this was permitted.
277 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
278 adjusted the size of a major service buffer; this failed because the
279 buffer was in use at the time. Change to a compile-time increase in the
280 buffer size, when this authenticator is compiled into exim.
282 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
283 previous fast-mode was untenable in the face of glibs using mmap to
284 support larger malloc requests.
286 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
287 New values supported, if defined on system where compiled:
288 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
289 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
291 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
292 running a limited number of proceses in parallel, a benefit is gained. The
293 amount varies with the platform hardware and load. The use of the option
294 queue_run_in_order means we cannot do this, as ordering becomes
297 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
298 had introduced a string-copy (for ensuring NUL-termination) which was not
299 appropriate for that case, which can include embedded NUL bytes in the
300 block of data. Investigation showed the copy to actually be needless, the
301 data being length-specified.
303 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
304 done during a receiving connection, and both used TLS, global info was
305 used rather than per-connection info for tracking the state of data
306 queued for transmission. This could result in a connection hang.
308 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
309 Previously, when delivering serveral messages down a single connection
310 only the first would provide a SIZE. This was due to the size information
311 not being properly tracked.
313 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
314 TAI (at 37 seconds currently), pretend to be in UTC for time-related
315 expansion and logging. Previously, spurious values such as a future
316 minute could be seen.
318 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
319 it could crash from a null-deref. This could also affect the
320 ${addresses: } operator and ${readsock } item.
322 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
323 message following a mime one, the variable was not reset.
325 JH/30 When an pipelined-connect fails at the first response, assume incorrect
326 cached capability (perhaps the peer reneged?) and immediately retry in
329 JH/31 Fix spurious detection of timeout while writing to transport filter.
331 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
332 an attempt to copy the string was made before checking it.
334 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
335 taint of the lookup key was maintained; we now regard the presence in the
336 filesystem as sufficient validation.
338 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
339 argument is supplied.
341 JH/35 The dsearch lookup now requires that the directory is an absolute path.
342 Previously this was not checked, and nonempty relative paths made an
343 access under Exim's current working directory.
345 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
346 Previously no event was raised.
348 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
349 parameter supplied by the sender MAIL FROM command. Previously it was
350 ignored, and only the check_spool_space option value for the required
353 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
354 the size of the signing public-key. Previously it was instead giving
355 the size of the signature hash.
357 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
358 the default. See the (new) dkim_verify_min_keysizes option.
360 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
361 and an ACL use a lookup for checking either the local_part or domain,
362 stale data could be accessed. Ensure that variable references are
363 dropped between messages.
365 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
366 by the client was not checked as pointing within response data before
367 being used. A malicious client could thus cause an out-of-bounds read and
368 possibly gain authentication. Fix by adding the check.
370 JH/42 Internationalisation: change the default for downconversion in the smtp
371 transport to be "if needed". Previously it was "as previously set" for
372 the message, which usually meant "if needed" for message-submission but
373 "no" for everything else. However, MTAs have been seen using SMTPUTF8
374 even when the envelope addresses did not need it, resulting in forwarding
375 failures to non-supporting MTAs. A downconvert in such cases will be
376 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
377 transport. The change does mean that addresses needing conversion will
378 be converted when previously a delivery failure would occur.
380 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
381 response was received it would be used unchecked in a fail-DSN, violating
382 standards on line-length limits. Truncate if needed.
384 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
385 doesn't work. (Additionally add a new main config option to configure the
392 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
393 side, to decode a TLS 1.3 packet capture.
395 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
396 Previously the default library behaviour applied, sending two, each in
399 JH/03 Debug output for ACL now gives the config file name and line number for
402 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
404 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
406 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
407 buffer overrun for (non-chunking) other transports.
409 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
410 TLS1.3, means that a server rejecting a client certificate is not visible
411 to the client until the first read of encrypted data (typically the
412 response to EHLO). Add detection for that case and treat it as a failed
413 TLS connection attempt, so that the normal retry-in-clear can work (if
414 suitably configured).
416 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
417 and/or domain. Found and fixed by Jason Betts.
419 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
420 configuration). If a CNAME target was not a wellformed name pattern, a
423 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
424 the OS reports them interleaved with other addresses.
426 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
427 used both for input and for a verify callout, both encrypted, SMTP
428 responses being sent by the server could be lost. This resulted in
429 dropped connections and sometimes bounces generated by a peer sending
432 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
433 string. Previously it was possible to cause undefined behaviour in a
434 library routine (usually a crash). Found by "zerons".
436 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
439 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
440 API was removed, so update to use the newer ones.
442 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
443 any timeout set, is taking a long time. Previously we would hang on to a
444 rotated logfile "forever" if the input was arriving with long gaps
445 (a previous attempt to fix addressed lack, for a long time, of initial
448 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
449 shared (NFS) environment. The length of the tempfile name is now
450 4 + 16 ("hdr.$message_exim_id") which might break on file
451 systems which restrict the file name length to lower values.
454 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
455 shared (NFS) environment.
457 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
458 did for all versions <4.90). Notably -M, -m, --invert, -I may be
461 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
462 on some platforms for bit 31.
464 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
465 to changes apparently associated with TLS1.3 handling some of the APIs
466 previously used were either nonfunctional or inappropriate. Strings
467 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
468 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
469 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
470 This affects log line X= elements, the $tls_{in,out}_cipher variables,
471 and the use of specific cipher names in the encrypted= ACL condition.
473 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
475 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
476 verification result was not updated unless hosts_require_ocsp applied.
478 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
479 queue_list_requires_admin set to false, non-admin users were denied the
482 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
483 directory-of-certs mode. Previously they were advertised despite the
486 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
487 A single TCP connection by a client will now hold a TLS connection open
488 for multiple message deliveries, by default. Previously the default was to
491 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
492 default. If built with the facility, DANE will be used. The facility
493 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
495 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
496 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
497 must be defined and you must still, unless you define DISABLE_TLS, manage
498 the the include-dir and library-file requirements that go with that
499 choice. Non-TLS builds are still supported.
501 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
504 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
505 default. If the platform supports and has the facility enabled, it will
506 be requested on all coneections.
508 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
509 controlled by the build-time option SUPPORT_PIPE_CONNECT.
511 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
513 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
514 success-DSN messages. Previously the From: header was always the default
515 one for these; the option was ignored.
517 JH/28 Fix the timeout on smtp response to apply to the whole response.
518 Previously it was reset for every read, so a teergrubing peer sending
519 single bytes within the time limit could extend the connection for a
520 long time. Credit to Qualsys Security Advisory Team for the discovery.
522 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
523 delivery address, which leaked information of the results of local
524 forwarding. Change to the original envelope recipient address, per
527 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
528 requested. Previously not bounce was generated and a log entry of
529 error ignored was made.
531 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
533 JH/32 Introduce a general tainting mechanism for values read from the input
534 channel, and values derived from them. Refuse to expand any tainted
535 values, to catch one form of exploit.
537 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
538 was unused and the unexpanded text used for the test. Found and
539 fixed by Ruben Jenster.
541 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
542 an attempt to use a TLS library read routine dereffed a nul pointer,
545 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
546 them in our smtp response.
548 JH/36 Have the general router option retry_use_local_part default to true when
549 any of the restrictive preconditions are set (to anything). Previously it
550 was only for check_local user. The change removes one item of manual
551 configuration which is required for proper retries when a remote router
552 handles a subset of addresses for a domain.
554 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
555 link count into consideration.
557 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
558 caused the extension of big_buffer, the following lines were ignored.
560 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
561 accordance with RFC 2308. Previously there was no expiry, so a longlived
562 receive process (eg. due to ACL delays) versus a short SOA value could
565 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
567 JH/39 Promote DMARC support to mainline.
569 JH/40 Bug 2452: Add a References: header to DSNs.
571 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
572 parameters. The relevant library call is documented as "Deprecated: This
573 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
574 3.6.0, DH parameters are negotiated following RFC7919."
576 HS/06 Change the default of dnssec_request_domains to "*"
578 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
579 carried on and emitted a BDAT command, even when PIPELINING was not
582 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
583 buffer was used for the filename, resulting in a trap when tainted
584 arguments (eg. $domain) were used.
586 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
587 recommended to avoid a possible server-load attack. The feature can be
588 re-enabled via the openssl_options main cofiguration option.
590 JH/45 local_scan API: documented the current smtp_printf() call. This changed
591 for version 4.90 - adding a "more data" boolean to the arguments.
592 Bumped the ABI version number also, this having been missed previously;
593 release versions 4.90 to 4.92.3 inclusive were effectively broken in
594 respect of usage of smtp_printf() by either local_scan code or libraries
595 accessed via the ${dlfunc } expansion item. Both will need coding
596 adjustment for any calls to smtp_printf() to match the new function
597 signature; a FALSE value for the new argument is always safe.
599 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
600 the file-offset (which the Linux syscall does, and exim expects); this
601 resulted in an indefinite loop.
603 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
604 to do ARC verification. The Authentication-Results: header line added
605 by the configuration then had no ARC item.
611 JH/01 Remove code calling the customisable local_scan function, unless a new
612 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
614 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
615 non-signal-safe functions being used.
617 JH/03 Bug 2269: When presented with a received message having a stupidly large
618 number of DKIM-Signature headers, disable DKIM verification to avoid
619 a resource-consumption attack. The limit is set at twenty.
621 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
622 report of oldest_pass in ${authres } in consequence, and separate out
623 some descriptions of reasons for verification fail.
625 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
626 files in the spool were present and unlocked. A queue-runner could spot
627 them, resulting in a duplicate delivery. Fix that by doing the unlock
628 after the unlink. Investigation by Tim Stewart. Take the opportunity to
629 add more error-checking on spoolfile handling while that code is being
632 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
633 No known attacks, no CVE, this is defensive hardening.
635 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
636 a queue-runner could start a delivery while other operations were ongoing.
637 Cutthrough delivery was a common victim, resulting in duplicate delivery.
638 Found and investigated by Tim Stewart. Fix by using the open message data
639 file handle rather than opening another, and not locally closing it (which
640 releases a lock) for that case, while creating the temporary .eml format
641 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
643 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
644 $sender_verify_failure/$recipient_verify_failure to "random".
646 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
649 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
650 Previously this would segfault.
652 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
655 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
656 like zero, since the resolver should be doing this for us, But we need one
657 as a CNAME but no MX presence gets the CNAME returned; we need to check
658 that doesn't point to an MX to declare it "no MX returned" rather than
659 "error, loop". A new main option is added so the older capability of
660 following some limited number of chain links is maintained.
662 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
664 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
665 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
666 now-deprecated earlier definitions used only specified the range up to TLS
667 1.2 (in the older-version library docs).
669 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
671 JH/15 Rework TLS client-side context management. Stop using a global, and
672 explicitly pass a context around. This enables future use of TLS for
673 connections to service-daemons (eg. malware scanning) while a client smtp
674 connection is using TLS; with cutthrough connections this is quite likely.
676 JH/16 Fix ARC verification to do AS checks in reverse order.
678 JH/17 Support a "tls" option on the ${readsocket } expansion item.
680 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
681 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
682 Previously the "utf8" would be re-prepended for every additional message.
684 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
685 Previously thery were accepted, resulting in issues when attempting to
686 forward messages to a non-supporting MTA.
688 PP/02 Let -n work with printing macros too, not just options.
690 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
691 one parent address was copied, and bogus data was used at delivery-logging
692 time. Either a crash (after delivery) or bogus log data could result.
693 Discovery and analysis by Tim Stewart.
695 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
696 Previously if the string ended mid-character, we did not insert the
697 promised '?' replacement.
699 PP/04 Documentation: current string operators work on bytes, not codepoints.
701 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
702 should pack well giving a smaller memory footprint so better caching and
703 therefore performance. Group the declarations where this can't be done so
704 that the byte-sized flag variables are not interspersed among pointer
705 variables, giving a better chance of good packing by the compiler.
707 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
708 non-null, to avoid issues with sites running BATV. Previously reports were
709 sent with an empty envelope sender so looked like bounces.
711 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
712 The ignore_error flag wasn't being returned from the filter subprocess so
713 was not set for later routers. Investigation and fix by Matthias Kurz.
715 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
716 and a msg:complete for the whole, when a message is manually removed using
717 -Mrm. Developement by Matthias Kurz, hacked on by JH.
719 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
720 a "Gnu special" function, asprintf() in the DB utility binary builds; I
721 hope that is portable enough.
723 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
724 requiring a known-CA anchor certificate; make it now rely entirely on the
725 TLSA as an anchor. Checking the name on the leaf cert against the name
726 on the A-record for the host is still done for TA (but not for EE mode).
728 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
729 included in delivery lines for non-proxied connections, when compiled with
730 SUPPORT_SOCKS and running with proxy logging enabled.
732 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
733 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
734 move the existing event to fire before the normal logging of message
735 failure so that custom logging is bracketed by normal logging.
737 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
738 msg:fail:internal event. Developement by Matthias Kurz.
740 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
741 far too small for todays use of crypto signatures stored there. Go all
742 the way to the max DNS message size of 64kB, even though this might be
743 overmuch for IOT constrained device use.
745 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
746 copy a string over itself. The library routine is documented as not
747 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
749 JH/32 For main options check_spool_space and check_inode_space, where the
750 platform supports 64b integers, support more than the previous 2^31 kB
751 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
752 the previous G, M, k.
754 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
755 $authenticated_fail_id variable on authentication failure. Previously
758 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
759 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
760 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
761 GNUTLS_SEC_PARAM_MEDIUM.
763 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
764 side. Previously we would continue as if no SNI had been received.
766 JH/36 Harden the handling of string-lists. When a list consisted of a sole
767 "<" character, which should be a list-separator specification, we walked
768 off past the nul-terimation.
770 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
771 causes) even when the retry time is not yet met. Previously they were
772 not, meaning that when (say) an account was over-quota and temp-rejecting,
773 and multiple senders' messages were queued, only one sender would get
774 notified on each configured delay_warning cycle.
776 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
778 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
779 cipher-suites, an error can be left on the stack even for a succeeding
780 accept; this results in impossible error messages when a later operation
783 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
784 return error codes indicating retry. Under TLS1.3 this becomes required.
786 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
787 it only wrote the new authenticators, resulting in a lack of tracking of
788 peer changes of ESMTP extensions until the next cache flush.
790 JH/41 Fix the loop reading a message header line to check for integer overflow,
791 and more-often against header_maxsize. Previously a crafted message could
792 induce a crash of the recive process; now the message is cleanly rejected.
794 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
795 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
801 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
802 When redis_servers is set to a list of > 1 element, and the Redis servers
803 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
804 case of MOVED into a DEFER case instead, thus moving the query onto the
805 next server in the list. For a cluster of N elements, all N servers must
806 be defined in redis_servers.
808 GF/02 Catch and remove uninitialized value warning in exiqsumm
809 Check for existence of @ARGV before looking at $ARGV[0]
811 JH/01 Replace the store_release() internal interface with store_newblock(),
812 which internalises the check required to safely use the old one, plus
813 the allocate and data copy operations duplicated in both (!) of the
814 extant use locations.
816 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
817 modifier. This matches the restriction on the commandline.
819 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
820 Previously only the last row was returned.
822 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
823 we assumed that tags in the header were well-formed, and parsed the
824 element content after inspecting only the first char of the tag.
825 Assumptions at that stage could crash the receive process on malformed
828 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
829 While running the DKIM ACL we operate on the Permanent memory pool so that
830 variables created with "set" persist to the DATA ACL. Also (at any time)
831 DNS lookups that fail create cache records using the Permanent pool. But
832 expansions release any allocations made on the current pool - so a dnsdb
833 lookup expansion done in the DKIM ACL releases the memory used for the
834 DNS negative-cache, and bad things result. Solution is to switch to the
835 Main pool for expansions.
836 While we're in that code, add checks on the DNS cache during store_reset,
837 active in the testsuite.
838 Problem spotted, and debugging aided, by Wolfgang Breyha.
840 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
841 When none of the hosts presented to a transport match an already-open
842 connection, close it and proceed with the list. Previously we would
843 queue the message. Spotted by Lena with Yahoo, probably involving
846 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
847 Previously a spurious "250 OK id=" response was appended to the proper
850 JH/08 The "support for" informational output now, which built with Content
851 Scanning support, has a line for the malware scanner interfaces compiled
852 in. Interface can be individually included or not at build time.
854 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
855 by the template makefile "src/EDITME". The "STREAM" support for an older
856 ClamAV interface method is removed.
858 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
859 rows affected is given instead).
861 JH/11 The runtime Berkeley DB library version is now additionally output by
862 "exim -d -bV". Previously only the compile-time version was shown.
864 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
865 SMTP connection. Previously, when one had more receipients than the
866 first, an abortive onward connection was made. Move to full support for
867 multiple onward connections in sequence, handling cutthrough connection
868 for all multi-message initiating connections.
870 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
871 routers. Previously, a multi-recipient message would fail to match the
872 onward-connection opened for the first recipient, and cause its closure.
874 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
875 a timeout on read on a GnuTLS initiating connection, resulting in the
876 initiating connection being dropped. This mattered most when the callout
877 was marked defer_ok. Fix to keep the two timeout-detection methods
880 JH/15 Relax results from ACL control request to enable cutthrough, in
881 unsupported situations, from error to silently (except under debug)
882 ignoring. This covers use with PRDR, frozen messages, queue-only and
885 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
887 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
888 metadata, resulting in a crash in free().
890 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
891 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
892 Broken also in d185889f4, with init system revamp.
894 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
895 Previously we abruptly closed the connection after reading a malware-
896 found indication; now we go on to read the "scan ok" response line,
899 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
900 ACL. Previously, a crash would result.
902 JH/19 Speed up macro lookups during configuration file read, by skipping non-
903 macro text after a replacement (previously it was only once per line) and
904 by skipping builtin macros when searching for an uppercase lead character.
906 JH/20 DANE support moved from Experimental to mainline. The Makefile control
907 for the build is renamed.
909 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
910 was allocated for every new TLS startup, meaning one per message. Fix
911 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
913 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
914 reported the original. Fix to report (as far as possible) the ACL
915 result replacing the original.
917 JH/23 Fix memory leak during multi-message connections using STARTTLS under
918 OpenSSL. Certificate information is loaded for every new TLS startup,
919 and the resources needed to be freed.
921 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
923 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
926 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
927 DATA response info to the (existing) per-recipient response info for
928 the "C=" log element. It can have useful tracking info from the
929 destination system. Patch from Simon Arlott.
931 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
932 length value. Previously this would segfault.
934 HS/02 Support Avast multiline protoocol, this allows passing flags to
935 newer versions of the scanner.
937 JH/28 Ensure that variables possibly set during message acceptance are marked
938 dead before release of memory in the daemon loop. This stops complaints
939 about them when the debug_store option is enabled. Discovered specifically
940 for sender_rate_period, but applies to a whole set of variables.
941 Do the same for the queue-runner and queue-list loops, for variables set
942 from spool message files. Do the same for the SMTP per-message loop, for
943 certain variables indirectly set in ACL operations.
945 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
946 as a multi-recipient message from a mailinglist manager). The coding had
947 an arbitrary cutoff number of characters while checking for more input;
948 enforced by writing a NUL into the buffer. This corrupted long / fast
949 input. The problem was exposed more widely when more pipelineing of SMTP
950 responses was introduced, and one Exim system was feeding another.
951 The symptom is log complaints of SMTP syntax error (NUL chars) on the
952 receiving system, and refused recipients seen by the sending system
953 (propating to people being dropped from mailing lists).
954 Discovered and pinpointed by David Carter.
956 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
957 replaced by the ${authresults } expansion.
959 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
961 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
962 allows proper process termination in container environments.
964 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
965 Previously the "final dot" had a newline after it; ensure it is CR,LF.
967 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
968 and "err_perm", deprecated since 4.83 when the RFC-defined words
969 "temperror" and "permerror" were introduced.
971 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
972 transport-filters or DKIM-signing. The restriction was lost in the
973 consolidation of verify-callout and delivery SMTP handling.
974 Extend the restriction to also cover ARC-signing.
976 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
977 in defer=pass mode supply a 450 to the initiator. Previously the message
980 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
981 tls_require_ciphers is used as before.
983 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
984 "pass_unscanned". Only tmpfails from the scanner are written to
985 the paniclog, as they may require admin intervention (permission
986 denied, license issues). Other scanner errors (like decompression
987 bombs) do not cause a paniclog entry.
989 JH/36 Fix reinitialisation of DKIM logging variable between messages.
990 Previously it was possible to log spurious information in receive log
993 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
994 triggered odd behaviour from Outlook Express clients.
996 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
999 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1000 since the IETF WG has not yet settled on that versus the original
1001 "bare" representation.
1003 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1004 Previously the millisecond value corrupted the output.
1005 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1006 corrupted the output.
1012 JH/01 Rework error string handling in TLS interface so that the caller in
1013 more cases is responsible for logging. This permits library-sourced
1014 string to be attached to addresses during delivery, and collapses
1015 pairs of long lines into single ones.
1017 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1018 during configuration. Wildcards are allowed and expanded.
1020 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1021 This permits better logging.
1023 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1024 do not close it down and have the child transport start it up again on
1025 the passed-on TCP connection. Instead, proxy the child (and any
1026 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1027 affected: the continued delivery log lines do not have any DNSSEC, TLS
1028 Certificate or OCSP information. TLS cipher information is still logged.
1030 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1031 identical IP addresses on different listening ports. Will also affect
1034 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1035 add noisy ifdef guards to special-case this sillyness.
1036 Patch from Bernd Kuhls.
1038 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1039 than 255 are no longer allowed.
1041 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1042 Disable the session-cache too, which might reduce our load. Since we
1043 currrectly use a new context for every connection, both as server and
1044 client, there is no benefit for these.
1045 GnuTLS appears to not support tickets server-side by default (we don't
1046 call gnutls_session_ticket_enable_server()) but client side is enabled
1047 by default on recent versions (3.1.3 +) unless the PFS priority string
1050 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1051 <https://reproducible-builds.org/specs/source-date-epoch/>.
1053 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1054 the check for any unsuccessful recipients did not notice the limit, and
1055 erroneously found still-pending ones.
1057 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1058 MSG_MORE. Only in-clear (not on TLS connections).
1060 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1061 transport-filter is involved we can buffer the headers in memory for
1062 creating the signature, and read the spool data file once for the
1063 signature and again for transmission.
1065 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1066 4.77 as the kernel support then wasn't solid, having issues in 64bit
1067 mode. Now, it's been long enough. Add support for FreeBSD also.
1069 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1070 case where the routing stage had gathered several addresses to send to
1071 a host before calling the transport for the first, we previously failed
1072 to close down TLS in the old transport process before passing the TCP
1073 connection to the new process. The new one sent a STARTTLS command
1074 which naturally failed, giving a failed delivery and bloating the retry
1075 database. Investigation and fix prototype from Wolfgang Breyha.
1077 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1078 false-negatives in the check that the sender had not preempted a response
1079 or prompt from Exim (running as a server), due to that code's lack of
1080 awareness of the SMTP input buffering.
1082 PP/04 Add commandline_checks_require_admin option.
1083 Exim drops privileges sanely, various checks such as -be aren't a
1084 security problem, as long as you trust local users with access to their
1085 own account. When invoked by services which pass untrusted data to
1086 Exim, this might be an issue. Set this option in main configuration
1087 AND make fixes to the calling application, such as using `--` to stop
1090 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1091 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1092 responses to those, into a single TLS record each way (this usually means
1093 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1096 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1097 affects you only if you're dancing at the edge of the param size limits.
1098 If you are, and this message makes sense to you, then: raise the
1099 configured limit or use OpenSSL 1.1. Nothing we can do for older
1102 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1103 cmdline element to get the documented default one. Previously it was
1106 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1107 are unsuable for non-message-specific reasons - eg. connection timeout,
1108 banner-time rejection.
1110 JH/16 Fix logging of delivery remote port, when specified by router, under
1113 PP/06 Repair manualroute's ability to take options in any order, even if one
1114 is the name of a transport.
1117 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1119 JH/17 Change the list-building routines interface to use the expanding-string
1120 triplet model, for better allocation and copying behaviour.
1122 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1123 Previously it was constructed the first time a possibly-matching string
1124 was met in the configuration file input during startup; now it is done
1127 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1128 compatible one, to avoid the (poorly documented) possibility of a config
1129 file in the working directory redirecting the DB files, possibly correpting
1130 some existing file. CVE-2017-10140 assigned for BDB.
1132 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1133 cache-hot. Previously, although the result was properly cached, the
1134 initial verify call returned a defer.
1136 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1137 the main verify for receipient in uncached-mode.
1139 JH/22 Retire historical build files to an "unsupported" subdir. These are
1140 defined as "ones for which we have no current evidence of testing".
1142 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1143 if present. Previously it was ignored.
1145 JH/24 Start using specified-initialisers in C structure init coding. This is
1146 a C99 feature (it's 2017, so now considered safe).
1148 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1149 if was a fixed-sized field and bitmask ops via macros; it is now more
1152 PP/07 GitHub PR 56: Apply MariaDB build fix.
1153 Patch provided by Jaroslav Škarvada.
1155 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1156 during Coverity cleanups [4.87 JH/47]
1157 Diagnosis and fix provided by Michael Fischer v. Mollard.
1159 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1160 the right size to place the terminating semicolon on its own folded
1161 line, the header hash was calculated to an incorrect value thanks to
1162 the (relaxed) space the fold became.
1164 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1165 and confused the parent.
1167 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1168 which could crash as a result. This could lead to undeliverable messages.
1170 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1173 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1174 matters on fast-turnover and PID-randomising systems, which were getting
1175 out-of-order delivery.
1177 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1178 a possibly-overlapping copy. The symptom was that "Remote host closed
1179 connection in response to HELO" was logged instead of the actual 4xx
1182 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1183 Previously only that bufferd was discarded, resulting in SYMTP command
1186 JH/32 DKIM: when a message has multiple signatures matching an identity given
1187 in dkim_verify_signers, run the dkim acl once for each. Previously only
1188 one run was done. Bug 2189.
1190 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1191 "panic the current process" to "deliberately defer". The panic log is
1192 still written with the problem list name; the mail and reject logs now
1193 get a temp-reject line for the message that was being handled, saying
1194 something like "domains check lookup or other defer". The SMTP 451
1195 message is still "Temporary local problem".
1197 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1198 A crafted sequence of BDAT commands could result in in-use memory beeing
1199 freed. CVE-2017-16943.
1201 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1202 from SMTP input. Previously it was always done; now only done for DATA
1203 and not BDAT commands. CVE-2017-16944.
1205 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1206 to the message (such as an overlong header line). Previously this was
1207 not done and we did not exit BDAT mode. Followon from the previous item
1208 though a different problem.
1214 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1215 than -2003 did; needs libidn2 in addition to libidn.
1217 JH/02 The path option on a pipe transport is now expanded before use.
1219 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1220 Patch provided by "Björn", documentation fix added too.
1222 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1223 missing a wire-to-host endian conversion.
1225 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1226 close after a BDAT command line could be taken as a following command,
1227 giving a synch failure. Fix by only checking for synch immediately
1228 before acknowledging the chunk.
1230 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1231 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1232 macro. Patches provided by Josh Soref.
1234 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1235 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1236 by IANA as a service extension. However, John Klensin suggests that we
1239 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1240 the dkim code may be unix-mode line endings rather than smtp wire-format
1241 CRLF, so prepend a CR to any bare LF.
1243 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1244 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1246 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1247 the first were themselves being wrongly included in the feed into dkim
1248 processing; with most chunk sizes in use this resulted in an incorrect
1249 body hash calculated value.
1251 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1252 DKIM signature block, for verification. Although advised against by
1253 standards it is specifically not ruled illegal.
1255 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1257 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1258 missing a body hash (the bh= tag).
1260 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1261 It seems that HAProxy sends the Proxy Protocol information in clear and
1262 only then does a TLS startup, so do the same.
1264 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1265 TCP connections (such as for Spamd) unless the daemon successfully set
1266 Fast Open mode on its listening sockets. This fixes breakage seen on
1267 too-old kernels or those not configured for Fast Open, at the cost of
1268 requiring both directions being enabled for TFO, and TFO never being used
1269 by non-daemon-related Exim processes.
1271 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1272 endings, at least on the first header line. Try to canonify any that get
1273 past that check, despite the cost.
1275 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1276 now limited to an arbitrary five deep, while parsing addresses with the
1277 strip_excess_angle_brackets option enabled.
1279 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1280 instead leave the unprompted TLS handshake in socket buffer for the
1281 TLS library to consume.
1283 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1285 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1287 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1288 region in one operation, for speed, and this leaves assigned pointers
1289 dangling. Add checks run only under the testsuite which checks all
1290 variables at a store-reset and panics on a dangling pointer; add code
1291 explicitly nulling out all the variables discovered. Fixes one known
1292 bug: a transport crash, where a dangling pointer for $sending_ip_address
1293 originally assigned in a verify callout, is re-used.
1295 PP/06 Drop '.' from @INC in various Perl scripts.
1297 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1299 PP/08 Reduce a number of compilation warnings under clang; building with
1300 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1301 should be warning-free.
1303 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1305 HS/01 Fix portability problems introduced by PP/08 for platforms where
1306 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1308 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1309 chunk. This allows us to accept broken chunked messages. We need a more
1310 general solution here.
1312 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1313 already-broken messages in the queue.
1315 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1317 JH/19 Fix reference counting bug in routing-generated-address tracking.
1323 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1324 supports it and a size is available (ie. the sending peer gave us one).
1326 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1327 years of being deprecated). The replacements are the ACLs
1328 acl_smtp_mime and acl_not_smtp_mime.
1330 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1331 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1332 or even an in-clear connection were permitted. Now, if the host lookup
1333 was dnssec and dane was requested then the host is only used if the
1334 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1335 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1336 if one fails this test.
1337 This means that a poorly-configured remote DNS will make it incommunicado;
1338 but it protects against a DNS-interception attack on it.
1340 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1341 non-noisy when a race steals the message being considered.
1343 JH/05 If main configuration option tls_certificate is unset, generate a
1344 self-signed certificate for inbound TLS connections.
1346 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1347 in rewrites and routers.
1349 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1350 and logged a warning sing 4.83; now they are a configuration file error.
1352 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1353 (lacking @domain). Apply the same qualification processing as RCPT.
1355 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1357 JH/10 Support ${sha256:} applied to a string (as well as the previous
1360 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1361 a cutthrough deliver is pending, as we always want to make a connection.
1362 This also avoids re-routing the message when later placing the cutthrough
1363 connection after a verify cache hit.
1364 Do not update it with the verify result either.
1366 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1367 when routing results in more than one destination address.
1369 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1370 signing (which inhibits the cutthrough capability). Previously only
1371 the presence of an option was tested; now an expansion evaluating as
1372 empty is permissible (obviously it should depend only on data available
1373 when the cutthrough connection is made).
1375 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1376 the relevant preceding SMTP command did not note the pipelining mode.
1378 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1379 Previously they were not counted.
1381 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1382 as one having no matching records. Previously we deferred the message
1383 that needed the lookup.
1385 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1386 distinguished as "(=".
1388 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1389 for missing MX records. Previously it only worked for missing A records.
1391 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1393 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1394 after the data-go-ahead and data-ack. Patch from Jason Betts.
1396 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1397 even for a "none" policy. Patch from Tony Meyer.
1399 JH/22 Fix continued use of a connection for further deliveries. If a port was
1400 specified by a router, it must also match for the delivery to be
1403 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1404 When one of the recipients of a message was unsuitable for the connection
1405 (has no matching addresses), we lost track of needing to mark it
1406 deferred. As a result mail would be lost.
1408 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1410 JH/25 Decoding ACL controls is now done using a binary search; the source code
1411 takes up less space and should be simpler to maintain. Merge the ACL
1412 condition decode tables also, with similar effect.
1414 JH/26 Fix problem with one_time used on a redirect router which returned the
1415 parent address unchanged. A retry would see the parent address marked as
1416 delivered, so not attempt the (identical) child. As a result mail would
1419 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1420 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1421 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1424 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1425 Main config options check_{log,spool}_{inodes,space} are now
1426 100 inodes, 10MB unless set otherwise in the configuration.
1428 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1429 Previously it only applied to the main-section connection policy
1432 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1434 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1435 by me. Added RFC7919 DH primes as an alternative.
1437 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1438 are not in the system include path.
1440 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1441 GnuTLS, when a session startup failed (eg because the client disconnected)
1442 Exim did stdio operations after fclose. This was exposed by a recent
1443 change which nulled out the file handle after the fclose.
1445 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1446 signed directly by the cert-signing cert, rather than an intermediate
1447 OCSP-signing cert. This is the model used by LetsEncrypt.
1449 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1451 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1452 an incoming connection.
1454 HS/02 Bug 1802: Do not half-close the connection after sending a request
1457 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1458 fallback to "prime256v1".
1460 JH/34 SECURITY: Use proper copy of DATA command in error message.
1461 Could leak key material. Remotely exploitable. CVE-2016-9963.
1467 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1468 and 3.4.4 - once the server is enabled to respond to an OCSP request
1469 it does even when not requested, resulting in a stapling non-aware
1470 client dropping the TLS connection.
1472 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1473 support variable-length bit vectors. No functional change.
1475 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1476 The I= interface field on outgoing lines is now after the H= remote
1477 host field, same as incoming lines. There is a separate
1478 outgoing_interface log selector which allows you to disable the
1481 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1482 If not running log_selector +smtp_connection the mainlog would be held
1483 open indefinitely after a "too many connections" event, including to a
1484 deleted file after a log rotate. Leave the per net connection logging
1485 leaving it open for efficiency as that will be quickly detected by the
1486 check on the next write.
1488 HS/01 Bug 1671: Fix post transport crash.
1489 Processing the wait-<transport> messages could crash the delivery
1490 process if the message IDs didn't exist for some reason. When
1491 using 'split_spool_directory=yes' the construction of the spool
1492 file name failed already, exposing the same netto behaviour.
1494 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1495 mime_regex ACL conditions.
1497 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1498 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1499 to HELO, local diagnostic string.
1501 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1502 log line to debug. Even when configured with a tls authenticator many
1503 client connections are expected to not authenticate in this way, so
1504 an authenticate fail is not an error.
1506 HS/02 Add the Exim version string to the process info. This way exiwhat
1507 gives some more detail about the running daemon.
1509 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1510 matter for fast-change records such as DNSBLs.
1512 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1513 retry record, even if constant. There may be multiple transports with
1514 different interface settings and the retry behaviour needs to be kept
1517 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1519 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1521 JH/10 Bug 840: fix log_defer_output option of pipe transport
1523 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1524 hosts_randomize. This matters a lot when combined with mua_wrapper.
1526 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1527 ${quote_pgsql:<string>} operator.
1529 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1531 JH/14 Change header file name for freeradius-client. Relevant if compiling
1532 with Radius support; from the Gentoo tree and checked under Fedora.
1534 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1536 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1537 option transport_filter, meaning no filtering.
1539 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1541 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1542 defaults to "*" (all hosts). The variable is now available when not built
1543 with TLS, default unset, mainly to enable keeping the testsuite sane.
1544 If a server certificate is not supplied (via tls_certificate) an error is
1545 logged, and clients will find TLS connections fail on startup. Presumably
1546 they will retry in-clear.
1547 Packagers of Exim are strongly encouraged to create a server certificate
1548 at installation time.
1550 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1551 with the $config_file variable.
1553 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1554 in transport context, after the attempt, and per-recipient. The latter type
1555 is per host attempted. The event data is the error message, and the errno
1556 information encodes the lookup type (A vs. MX) used for the (first) host,
1557 and the trailing two digits of the smtp 4xx response.
1559 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1560 to write to mainlog (or rejectlog, paniclog) in the window between file
1561 creation and permissions/ownership being changed. Particularly affects
1562 installations where exicyclog is run as root, rather than exim user;
1563 result is that the running daemon panics and dies.
1565 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1567 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1568 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1569 "pri" and "weight". Note that the previous implicit priority given by the
1570 list order is no longer honoured.
1572 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1573 for DKIM processing.
1575 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1576 by defining SUPPORT_SOCKS.
1578 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1579 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1580 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1581 variables are renamed to proxy_{local,external}_{address,port}.
1583 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1584 for a build by defining SUPPORT_I18N
1586 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1587 of the query string, and make ${quote_redis:} do that quoting.
1589 JH/29 Move Events support from Experimental to mainline, enabled by default
1590 and removable for a build by defining DISABLE_EVENT.
1592 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1594 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1595 cached by the daemon.
1597 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1598 by defining LOOKUP_REDIS. The libhiredis library is required.
1600 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1601 keys are given for lookup.
1603 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1604 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1605 only supported when built with TLS support. The PolarSSL SHA routines
1606 are still used when the TLS library is too old for convenient support.
1608 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1609 openssl_options), for security. OpenSSL forces this from version 1.1.0
1610 server-side so match that on older versions.
1612 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1613 allocation for $value could be released as the expansion processing
1614 concluded, but leaving the global pointer active for it.
1616 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1617 and to use the domains and local_parts ACL conditions.
1619 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1620 incorrectly not doubled on cutthrough transmission, hence seen as a
1621 body-termination at the receiving system - resulting in truncated mails.
1622 Commonly the sender saw a TCP-level error, and retransmitted the message
1623 via the normal store-and-forward channel. This could result in duplicates
1624 received - but deduplicating mailstores were liable to retain only the
1625 initial truncated version.
1627 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1629 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1631 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1632 we're in there, support oversigning also; bug 1309.
1634 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1636 HS/04 Add support for keep_environment and add_environment options.
1638 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1639 either intentional arithmetic overflow during PRNG, or testing config-
1642 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1643 delivery resulted in actual delivery. Cancel cutthrough before DATA
1646 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1647 rejects a recipient: pass the reject to the originator.
1649 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1650 Many were false-positives and ignorable, but it's worth fixing the
1653 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1654 for the new environment-manipulation done at startup. Move the routines
1655 from being local to tls.c to being global via the os.c file.
1657 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1658 an extract embedded as result-arg for a map, the first arg for extract
1659 is unavailable so we cannot tell if this is a numbered or keyed
1660 extraction. Accept either.
1666 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1669 JH/02 The smtp transport option "multi_domain" is now expanded.
1671 JH/03 The smtp transport now requests PRDR by default, if the server offers
1674 JH/04 Certificate name checking on server certificates, when exim is a client,
1675 is now done by default. The transport option tls_verify_cert_hostnames
1676 can be used to disable this per-host. The build option
1677 EXPERIMENTAL_CERTNAMES is withdrawn.
1679 JH/05 The value of the tls_verify_certificates smtp transport and main options
1680 default to the word "system" to access the system default CA bundle.
1681 For GnuTLS, only version 3.0.20 or later.
1683 JH/06 Verification of the server certificate for a TLS connection is now tried
1684 (but not required) by default. The verification status is now logged by
1685 default, for both outbound TLS and client-certificate supplying inbound
1688 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1691 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1692 Status Notification (bounce) messages are now MIME format per RFC 3464.
1693 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1694 under the control of the dsn_advertise_hosts option, and routers may
1695 have a dsn_lasthop option.
1697 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1698 default, modifiable by a malware= option. The list separator for
1699 the options can now be changed in the usual way. Bug 68.
1701 JH/10 The smtp_receive_timeout main option is now expanded before use.
1703 JH/11 The incoming_interface log option now also enables logging of the
1704 local interface on delivery outgoing connections.
1706 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1707 if the interface and destination host and port all match.
1709 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1712 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1713 Patch from Andrew Lewis.
1715 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1716 now supports optional time-restrictions, weighting, and priority
1717 modifiers per server. Patch originally by <rommer@active.by>.
1719 JH/16 The spamd_address main option now supports a mixed list of local
1720 and remote servers. Remote servers can be IPv6 addresses, and
1721 specify a port-range.
1723 JH/17 Bug 68: The spamd_address main option now supports an optional
1724 timeout value per server.
1726 JH/18 Bug 1581: Router and transport options headers_add/remove can
1727 now have the list separator specified.
1729 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1732 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1735 JH/21 Support for the A6 type of dns record is withdrawn.
1737 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1738 rather than the verbs used.
1740 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1741 from 255 to 1024 chars.
1743 JH/24 Verification callouts now attempt to use TLS by default.
1745 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1746 are generic router options now. The defaults didn't change.
1748 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1749 Original patch from Alexander Shikoff, worked over by JH.
1751 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1752 files if ZCAT_COMMAND is not executable.
1754 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1756 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1758 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1759 Normally benign, it bites when the pair was led to by a CNAME;
1760 modern usage is to not canonicalize the domain to a CNAME target
1761 (and we were inconsistent anyway for A-only vs AAAA+A).
1763 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1765 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1766 when evaluating $sender_host_dnssec.
1768 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1769 $sender_helo_dnssec variable.
1771 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1773 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1775 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1777 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1778 documented as working, but never had. Support all but $spam_report.
1780 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1781 added for tls authenticator.
1783 HS/03 Add perl_taintmode main config option
1789 TL/01 When running the test suite, the README says that variables such as
1790 no_msglog_check are global and can be placed anywhere in a specific
1791 test's script, however it was observed that placement needed to be near
1792 the beginning for it to behave that way. Changed the runtest perl
1793 script to read through the entire script once to detect and set these
1794 variables, reset to the beginning of the script, and then run through
1795 the script parsing/test process like normal.
1797 TL/02 The BSD's have an arc4random API. One of the functions to induce
1798 adding randomness was arc4random_stir(), but it has been removed in
1799 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1800 function when detected.
1802 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1803 cause callback expansion.
1805 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1806 syntax errors in an expansion can be treated as a string instead of
1807 logging or causing an error, due to the internal use of bool_lax
1808 instead of bool when processing it.
1810 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1811 server certificates when making smtp deliveries.
1813 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1815 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1817 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1819 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1820 Merged patch from Sebastian Wiedenroth.
1822 JH/05 Fix results-pipe from transport process. Several recipients, combined
1823 with certificate use, exposed issues where response data items split
1824 over buffer boundaries were not parsed properly. This eventually
1825 resulted in duplicates being sent. This issue only became common enough
1826 to notice due to the introduction of connection certificate information,
1827 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1829 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1830 size buffer was used, resulting in syntax errors when an expansion
1833 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1834 version 3.3.6 or later.
1836 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1837 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1838 both become "event_action", the variables become $event_name, $event_data
1839 and $event_defer_errno. There is a new variable $verify_mode, usable in
1840 routers, transports and related events. The tls:cert event is now also
1841 raised for inbound connections, if the main configuration event_action
1844 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1845 early OCSP support, but no stapling (appears to be less than 1.0.0).
1847 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1848 server certificate names available under the smtp transport option
1849 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1852 JH/10 Time-related extraction expansions from certificates now use the main
1853 option "timezone" setting for output formatting, and are consistent
1854 between OpenSSL and GnuTLS compilations. Bug 1541.
1856 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1857 encoded parameter in the incoming message. Bug 1558.
1859 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1860 include certificate info, eximon was claiming there were spoolfile
1863 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1865 JH/14 Log delivery-related information more consistently, using the sequence
1866 "H=<name> [<ip>]" wherever possible.
1868 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1869 are problematic for Debian distribution, omit them from the release
1872 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1874 JH/16 Fix string representation of time values on 64bit time_t architectures.
1877 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1878 output list separator was used.
1883 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1884 checkers that were complaining about end of non-void function with no
1887 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1888 This was a regression introduced in 4.83 by another bugfix.
1890 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1892 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1893 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1899 TF/01 Correctly close the server side of TLS when forking for delivery.
1901 When a message was received over SMTP with TLS, Exim failed to clear up
1902 the incoming connection properly after forking off the child process to
1903 deliver the message. In some situations the subsequent outgoing
1904 delivery connection happened to have the same fd number as the incoming
1905 connection previously had. Exim would try to use TLS and fail, logging
1906 a "Bad file descriptor" error.
1908 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1909 utilities have not been installed.
1911 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1912 temporary space as the ACL may create new global variables.
1914 TL/01 LDAP support uses per connection or global context settings, depending
1915 upon the detected version of the libraries at build time.
1917 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1918 to extract and use the src ip:port in logging and expansions as if it
1919 were a direct connection from the outside internet. PPv2 support was
1920 updated based on HAProxy spec change in May 2014.
1922 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1924 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1925 Properly escape header and check for NULL return.
1927 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1930 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1932 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1933 characters in header names, implemented as a verify condition.
1934 Contributed by Michael Fischer v. Mollard.
1936 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1937 results permerror and temperror. Previous values are deprecated but
1938 still accepted. In a future release, err_perm and err_temp will be
1939 completely removed, which will be a backward incompatibility if the
1940 ACL tests for either of these two old results. Patch contributed by
1941 user bes-internal on the mailing list.
1943 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1945 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1946 selectors, in both main and reject logs.
1948 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1951 JH/07 Add malware type "sock" for talking to simple daemon.
1953 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1955 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1956 routers/transports under cutthrough routing.
1958 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1959 numbers. Touch up "bool" conditional to keep the same definition.
1961 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1963 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1965 JH/12 Expand items in router/transport headers_add or headers_remove lists
1966 individually rather than the list as a whole. Bug 1452.
1968 Required for reasonable handling of multiple headers_ options when
1969 they may be empty; requires that headers_remove items with embedded
1970 colons must have them doubled (or the list-separator changed).
1972 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1973 view the policy declared in the DMARC record. Currently, $dmarc_status
1974 is a combined value of both the record presence and the result of the
1977 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1979 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1980 dnslookup router and the smtp transport (applying to the forward
1983 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1984 of ldap servers used for a specific lookup. Patch provided by Heiko
1987 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1988 New variable $lookup_dnssec_authenticated for observability.
1990 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1991 Patch submitted by Lars Timman.
1993 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1995 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1996 Requires trusted mode and valid format message id, aborts otherwise.
1997 Patch contributed by Heiko Schlichting.
1999 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2000 certextract with support for various fields. Bug 1358.
2002 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2003 is requested by default, modifiable by smtp transport option
2006 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2007 operate on certificate variables to give certificate fingerprints
2008 Also new ${sha256:cert_variable}.
2010 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2012 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2013 Christian Aistleitner.
2015 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2017 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2018 file. Patch from Wolfgang Breyha.
2020 JH/25 Expand the coverage of the delivery $host and $host_address to
2021 client authenticators run in verify callout. Bug 1476.
2023 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2024 align with daemon_smtp_ports. Bug 72.
2026 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2027 support and error reporting did not work properly.
2029 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2030 and is readable. Patch from Andrew Colin Kissa.
2032 TL/14 Enhance documentation of ${run expansion and how it parses the
2033 commandline after expansion, particularly in the case when an
2034 unquoted variable expansion results in an empty value.
2036 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2038 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2039 by renaming away from it. Addresses GH issue 15, reported by
2042 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2044 TL/15 SECURITY: prevent double expansion in math comparison functions
2045 (can expand unsanitized data). Not remotely exploitable.
2052 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2054 PP/02 Make -n do something, by making it not do something.
2055 When combined with -bP, the name of an option is not output.
2057 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2060 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2061 $sender_host_name and config options to manage this, and basic check
2064 PP/05 DSCP support for outbound connections and control modifier for inbound.
2066 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2067 (Only plugin which currently uses this is kerberos4, which nobody should
2068 be using, but we should make it available and other future plugins might
2069 conceivably use it, even though it would break NAT; stuff *should* be
2070 using channel bindings instead).
2072 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2073 name; added for Sendmail compatibility; requires admin caller.
2074 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2075 just ignore it); requires trusted caller.
2076 Also parse but ignore: -Ac -Am -X<logfile>
2079 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2081 TL/02 Add +smtp_confirmation as a default logging option.
2083 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2084 Patch by Magnus Holmgren from 2007-02-20.
2086 TL/04 Bugzilla 1281 - Spec typo.
2087 Bugzilla 1283 - Spec typo.
2088 Bugzilla 1290 - Spec grammar fixes.
2090 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2092 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2094 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2095 mailing list by by Dmitry Isaikin.
2097 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2099 JH/02 Support "G" suffix to numbers in ${if comparisons.
2101 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2103 NM/01 Bugzilla 1197 - Spec typo
2104 Bugzilla 1196 - Spec examples corrections
2106 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2108 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2109 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2112 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2113 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2115 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2116 "acl {{name}{arg}...}", and optional args on acl condition
2119 JH/05 Permit multiple router/transport headers_add/remove lines.
2121 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2123 JH/07 Avoid using a waiting database for a single-message-only transport.
2124 Performance patch from Paul Fisher. Bugzilla 1262.
2126 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2129 JH/09 Add $headers_added variable, with content from use of ACL modifier
2130 add_header (but not yet added to the message). Bugzilla 199.
2132 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2133 Pulled from Bugzilla 817 by Wolfgang Breyha.
2135 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2137 (nb: this is the same fix as in Exim 4.80.1)
2139 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2142 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2145 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2147 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2148 advertise SMTP AUTH mechanism to us, instead of a generic
2149 protocol violation error. Also, make Exim more robust to bad
2150 data from the Dovecot auth socket.
2152 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2154 When a queue runner is handling a message, Exim first routes the
2155 recipient addresses, during which it prunes them based on the retry
2156 hints database. After that it attempts to deliver the message to
2157 any remaining recipients. It then updates the hints database using
2160 So if a recipient address works intermittently, it can get repeatedly
2161 deferred at routing time. The retry hints record remains fresh so the
2162 address never reaches the final cutoff time.
2164 This is a fairly common occurrence when a user is bumping up against
2165 their storage quota. Exim had some logic in its local delivery code
2166 to deal with this. However it did not apply to per-recipient defers
2167 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2169 This change adds a proper retry rule check during routing so that the
2170 final cutoff time is checked against the message's age. We only do
2171 this check if there is an address retry record and there is not a
2172 domain retry record; this implies that previous attempts to handle
2173 the address had the retry_use_local_parts option turned on. We use
2174 this as an approximation for the destination being like a local
2175 delivery, as in LMTP.
2177 I suspect this new check makes the old local delivery cutoff check
2178 redundant, but I have not verified this so I left the code in place.
2180 TF/02 Correct gecos expansion when From: is a prefix of the username.
2182 Test 0254 submits a message to Exim with the header
2186 When I ran the test suite under the user fanf2, Exim expanded
2187 the header to contain my full name, whereas it should have added
2188 a Resent-Sender: header. It erroneously treats any prefix of the
2189 username as equal to the username.
2191 This change corrects that bug.
2193 GF/01 DCC debug and logging tidyup
2194 Error conditions log to paniclog rather than rejectlog.
2195 Debug lines prefixed by "DCC: " to remove any ambiguity.
2197 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2199 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2200 Bug spotted by Jeremy Harris; was flawed since initial commit.
2201 Would have resulted in OCSP responses post-SNI triggering an Exim
2202 NULL dereference and crash.
2204 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2206 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2207 Bug detection, analysis and fix by Samuel Thibault.
2208 Bugzilla 1331, Debian bug #698092.
2210 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2212 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2213 Server implementation by Todd Lyons, client by JH.
2214 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2215 config variable "prdr_enable" controls whether the server
2216 advertises the facility. If the client requests PRDR a new
2217 acl_data_smtp_prdr ACL is called once for each recipient, after
2218 the body content is received and before the acl_smtp_data ACL.
2219 The client is controlled by both of: a hosts_try_prdr option
2220 on the smtp transport, and the server advertisement.
2221 Default client logging of deliveries and rejections involving
2222 PRDR are flagged with the string "PRDR".
2224 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2225 fclose(). Diagnosis by Todd Lyons.
2227 PP/17 Update configure.default to handle IPv6 localhost better.
2228 Patch by Alain Williams (plus minor tweaks).
2231 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2232 This is now consistent with GnuTLS, and is now documented: the
2233 previous undocumented portable approach to treating the option as
2234 unset was to force an expansion failure. That still works, and
2235 an empty string is now equivalent.
2237 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2238 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2239 not performing validation itself.
2241 PP/20 Added force_command boolean option to pipe transport.
2242 Patch from Nick Koston, of cPanel Inc.
2244 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2247 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2249 PP/21 Fix eximon continuous updating with timestamped log-files.
2250 Broken in a format-string cleanup in 4.80, missed when I repaired the
2251 other false fix of the same issue.
2252 Report and fix from Heiko Schlichting.
2255 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2256 Report from Prashanth Katuri.
2258 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2259 It's SecureTransport, so affects any MacOS clients which use the
2260 system-integrated TLS libraries, including email clients.
2262 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2263 using a MIME ACL for non-SMTP local injection.
2264 Report and assistance in diagnosis by Warren Baker.
2266 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2268 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2270 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2271 last id that failed so it may be referenced in subsequent ACL's.
2273 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2276 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2277 ldap library initialization, allowing self-signed CA's to be
2278 used. Also properly sets require_cert option later in code by
2279 using NULL (global ldap config) instead of ldap handle (per
2280 session). Bug diagnosis and testing by alxgomz.
2282 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2283 the src/util/ subdirectory.
2285 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2286 renamed to Transport Post Delivery Action by Jeremy Harris, as
2289 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2290 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2291 redis_servers = needs to be configured which will be used by the redis
2292 lookup. Patch from Warren Baker, of The Packet Hub.
2294 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2296 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2297 hostname or reverse DNS when processing a host list. Used suggestions
2298 from multiple comments on this bug.
2300 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2302 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2303 Missed a few lines, added it to make the runtest require no keyboard
2306 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2307 contains upper case chars. Make router use caseful_local_part.
2309 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2310 support when GnuTLS has been built with p11-kit.
2316 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2318 This, or similar/improved, will also be change PP/11 of 4.82.
2324 PP/01 Handle short writes when writing local log-files.
2325 In practice, only affects FreeBSD (8 onwards).
2326 Bugzilla 1053, with thanks to Dmitry Isaikin.
2328 NM/01 Bugzilla 949 - Documentation tweak
2330 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2333 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2335 PP/02 Implemented gsasl authenticator.
2337 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2339 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2340 `pkg-config foo` for cflags/libs.
2342 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2343 with rest of GSASL and with heimdal_gssapi.
2345 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2346 `pkg-config foo` for cflags/libs for the TLS implementation.
2348 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2349 properties get this fed in as external SSF. A number of robustness
2350 and debugging improvements to the cyrus_sasl authenticator.
2352 PP/08 cyrus_sasl server now expands the server_realm option.
2354 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2355 Patch by Jeremy Harris.
2357 PP/10 Added dbmjz lookup type.
2359 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2361 PP/12 MAIL args handles TAB as well as SP, for better interop with
2362 non-compliant senders.
2363 Analysis and variant patch by Todd Lyons.
2365 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2366 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2367 Patch from Dirk Mueller <dmueller@suse.com>
2369 PP/13 tls_peerdn now print-escaped for spool files.
2370 Observed some $tls_peerdn in wild which contained \n, which resulted
2371 in spool file corruption.
2373 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2374 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2375 or write after TLS renegotiation, which otherwise led to messages
2378 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2379 as a tracking header (ie: a signed header comes before the signature).
2380 Patch from Wolfgang Breyha.
2382 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2383 comma-sep list; embedded commas doubled.
2385 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2387 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2389 Report and patch from Dmitry Banschikov.
2391 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2392 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2393 failures. We appear to now support TLS1.1+ with Exim.
2395 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2396 lets Exim select keys and certificates based upon TLS SNI from client.
2397 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2398 before an outbound SMTP session. New log_selector, +tls_sni.
2400 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2401 NULL dereference. Report and patch from Alun Jones.
2403 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2404 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2405 Not seeing resolver debug output on NetBSD, but suspect this is a
2406 resolver implementation change.
2408 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2409 Left warnings. Added "eximon gdb" invocation mode.
2411 PP/21 Defaulting "accept_8bitmime" to true, not false.
2413 PP/22 Added -bw for inetd wait mode support.
2415 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2416 locate the relevant includes and libraries. Made this the default.
2418 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2419 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2421 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2422 This may cause build issues on older platforms.
2424 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2425 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2426 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2427 Added SNI support via GnuTLS too.
2428 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2430 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2432 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2433 Applied second patch from Janne, implementing suggestion to default
2434 multiple-strings-in-record handling to match SPF spec.
2436 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2438 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2439 read-only, out of scope).
2440 Patch from Wolfgang Breyha, report from Stuart Northfield.
2442 PP/29 Fix three issues highlighted by clang analyser static analysis.
2443 Only crash-plausible issue would require the Cambridge-specific
2444 iplookup router and a misconfiguration.
2445 Report from Marcin Mirosław.
2447 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2449 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2450 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2451 As part of this, removing so much warning spew let me fix some minor
2452 real issues in debug logging.
2454 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2455 assignment on my part. Fixed.
2457 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2458 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2459 Janne Snabb (who went above and beyond: thank you).
2461 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2462 string otherwise requires a connection and a bunch more work and it's
2463 relatively easy to get wrong. Should also expose TLS library linkage
2466 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2467 64-bit ${eval} (JH/03).
2469 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2470 GNU libc to support some of the 64-bit stuff, should not lead to
2471 conflicts. Defined before os.h is pulled in, so if a given platform
2472 needs to override this, it can.
2474 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2475 protection layer was required, which is not implemented.
2476 Bugzilla 1254, patch from Wolfgang Breyha.
2478 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2479 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2480 tls_dhparam take prime identifiers. Also unbreak combination of
2481 OpenSSL+DH_params+TLSSNI.
2483 PP/39 Disable SSLv2 by default in OpenSSL support.
2489 PP/01 Solaris build fix for Oracle's LDAP libraries.
2490 Bugzilla 1109, patch from Stephen Usher.
2492 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2494 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2497 TF/02 Fix a couple more cases where we did not log the error message
2498 when unlink() failed. See also change 4.74-TF/03.
2500 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2501 lock up or crash if it happened to be inside a call to libc when it
2502 got a SIGUSR1 from exiwhat.
2504 The SIGUSR1 handler appends the current process status to the process
2505 log which is later printed by exiwhat. It used to use the general
2506 purpose logging code to do this, but several functions it calls are
2507 not safe for signals.
2509 The new output code in the SIGUSR1 handler is specific to the process
2510 log, and simple enough that it's easy to inspect for signal safety.
2511 Removing some special cases also simplifies the general logging code.
2512 Removing the spurious timestamps from the process log simplifies
2515 TF/04 Improved ratelimit ACL condition.
2517 The /noupdate option has been deprecated in favour of /readonly which
2518 has clearer semantics. The /leaky, /strict, and /readonly update modes
2519 are mutually exclusive. The update mode is no longer included in the
2520 database key; it just determines when the database is updated. (This
2521 means that when you upgrade Exim will forget old rate measurements.)
2523 Exim now checks that the per_* options are used with an update mode that
2524 makes sense for the current ACL. For example, when Exim is processing a
2525 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2526 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2527 must specify per_mail/readonly. If you omit the update mode it defaults to
2528 /leaky where that makes sense (as before) or /readonly where required.
2530 The /noupdate option is now undocumented but still supported for
2531 backwards compatibility. It is equivalent to /readonly except that in
2532 ACLs where /readonly is required you may specify /leaky/noupdate or
2533 /strict/noupdate which are treated the same as /readonly.
2535 A useful new feature is the /count= option. This is a generalization
2536 of the per_byte option, so that you can measure the throughput of other
2537 aggregate values. For example, the per_byte option is now equivalent
2538 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2540 The per_rcpt option has been generalized using the /count= mechanism
2541 (though it's more complicated than the per_byte equivalence). When it is
2542 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2543 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2544 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2545 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2546 Note that using per_rcpt with a non-readonly update mode in more than
2547 one ACL will cause the recipients to be double-counted. (The per_mail
2548 and per_byte options don't have this problem.)
2550 The handling of very low rates has changed slightly. If the computed rate
2551 is less than the event's count (usually one) then this event is the first
2552 after a long gap. In this case the rate is set to the same as this event's
2553 count, so that the first message of a spam run is counted properly.
2555 The major new feature is a mechanism for counting the rate of unique
2556 events. The new per_addr option counts the number of different
2557 recipients that someone has sent messages to in the last time period. It
2558 behaves like per_rcpt if all the recipient addresses are different, but
2559 duplicate recipient addresses do not increase the measured rate. Like
2560 the /count= option this is a general mechanism, so the per_addr option
2561 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2562 example, measure the rate that a client uses different sender addresses
2563 with the options per_mail/unique=$sender_address. There are further
2564 details in the main documentation.
2566 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2568 TF/06 Removed a few PCRE remnants.
2570 TF/07 Automatically extract Exim's version number from tags in the git
2571 repository when doing development or release builds.
2573 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2574 Bugzilla 879. Patch from Paul Fisher.
2576 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2577 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2580 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2582 PP/05 Variable $av_failed, true if the AV scanner deferred.
2583 Bugzilla 1078. Patch from John Horne.
2585 PP/06 Stop make process more reliably on build failure.
2586 Bugzilla 1087. Patch from Heiko Schlittermann.
2588 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2589 Bugzilla 1089. Patch from Heiko Schlittermann.
2591 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2592 Bugzilla 1131. Patch from Holger Weiß.
2594 PP/09 Handle IPv6 addresses with SPF.
2595 Bugzilla 860. Patch from Wolfgang Breyha.
2597 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2599 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2602 PP/11 match_* no longer expand right-hand-side by default.
2603 New compile-time build option, EXPAND_LISTMATCH_RHS.
2604 New expansion conditions, "inlist", "inlisti".
2606 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2608 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2610 PP/14 fix log_write() format string regression from TF/03.
2611 Bugzilla 1152. Patch from Dmitry Isaikin.
2617 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2619 PP/02 Harmonised TLS library version reporting; only show if debugging.
2620 Layout now matches that introduced for other libraries in 4.74 PP/03.
2622 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2624 PP/04 New "dns_use_edns0" global option.
2626 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2629 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2630 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2632 TK/01 Updated PolarSSL code to 0.14.2.
2633 Bugzilla 1097. Patch from Andreas Metzler.
2635 PP/07 Catch divide-by-zero in ${eval:...}.
2636 Fixes bugzilla 1102.
2638 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2641 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2642 format-string attack -- SECURITY: remote arbitrary code execution.
2644 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2645 time unintentionally subject to list matching rules, letting the header
2646 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2647 arbitrary string expansion). This allowed for information disclosure.
2649 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2650 INT_MIN/-1 -- value coerced to INT_MAX.
2656 NM/01 Workaround for PCRE version dependency in version reporting
2659 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2660 This fixes portability to compilers other than gcc, notably
2661 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2663 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2664 makefiles for portability to HP-UX and POSIX correctness.
2666 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2667 Also via indented variable definition in the Makefile.
2668 (Debugging by Oliver Heesakkers).
2670 PP/02 Restore caching of spamd results with expanded spamd_address.
2671 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2673 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2674 Improves build reliability. Fix from: Frank Elsner
2676 NM/02 Fix wide character breakage in the rfc2047 coding
2677 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2679 NM/03 Allow underscore in dnslist lookups
2680 Fixes bug 1026. Patch from Graeme Fowler
2682 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2683 Code patches from Adam Ciarcinski of NetBSD.
2685 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2688 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2689 is logged, to avoid truncation. Patch from John Horne.
2691 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2692 Patch from Jakob Hirsch.
2694 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2695 SQL string expansion failure details.
2696 Patch from Andrey Oktyabrski.
2698 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2699 Patch from Simon Arlott.
2701 PP/09 New lookups functionality failed to compile on old gcc which rejects
2702 extern declarations in function scope.
2703 Patch from Oliver Fleischmann
2705 PP/10 Use sig_atomic_t for flags set from signal handlers.
2706 Check getgroups() return and improve debugging.
2707 Fixed developed for diagnosis in bug 927 (which turned out to be
2710 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2711 Patch from Mark Zealey.
2713 PP/12 Bugzilla 1056: Improved spamd server selection.
2714 Patch from Mark Zealey.
2716 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2717 Based on patch from Heiko Schlittermann.
2719 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2720 Patch from Uwe Doering, sign-off by Michael Haardt.
2722 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2723 variable declaration deep within a block. Bug and patch from
2726 PP/15 lookups-Makefile IRIX compatibility coercion.
2728 PP/16 Make DISABLE_DKIM build knob functional.
2730 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2731 Patch by Simon Arlott
2733 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2734 variable argument macros. Our copy now differs from upstream.
2740 TF/01 Failure to get a lock on a hints database can have serious
2741 consequences so log it to the panic log.
2743 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2744 controlled using the smtp_confirmation log selector.
2746 TF/03 Include the error message when we fail to unlink a spool file.
2748 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2749 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2750 for maintaining out-of-tree patches for some time.
2752 PP/01 Bugzilla 139: Documentation and portability issues.
2753 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2754 Handle per-OS dynamic-module compilation flags.
2756 PP/02 Let /dev/null have normal permissions.
2757 The 4.73 fixes were a little too stringent and complained about the
2758 permissions on /dev/null. Exempt it from some checks.
2759 Reported by Andreas M. Kirchwitz.
2761 PP/03 Report version information for many libraries, including
2762 Exim version information for dynamically loaded libraries. Created
2763 version.h, now support a version extension string for distributors
2764 who patch heavily. Dynamic module ABI change.
2766 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2767 privilege escalation vulnerability whereby the Exim run-time user
2768 can cause root to append content of the attacker's choosing to
2771 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2774 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2775 If dropping privileges for untrusted macros, we disabled normal logging
2776 on the basis that it would fail; for the Exim run-time user, this is not
2777 the case, and it resulted in successful deliveries going unlogged.
2778 Fixed. Reported by Andreas Metzler.
2784 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2785 only prepend for the Resent-* case. Fixes regression introduced in
2786 Exim 4.70 by NM/22 for Bugzilla 607.
2788 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2789 increasing numbers of administrators be bitten by this.
2791 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2793 PP/03 Bugzilla 994: added openssl_options main configuration option.
2795 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2797 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2799 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2801 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2802 without a peer certificate, leading to a segfault because of an
2803 assumption that peers always have certificates. Be a little more
2804 paranoid. Problem reported by Martin Tscholak.
2806 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2807 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2808 NB: ClamAV planning to remove STREAM in "middle of 2010".
2809 CL also introduces -bmalware, various -d+acl logging additions and
2810 more caution in buffer sizes.
2812 PP/09 Implemented reverse_ip expansion operator.
2814 PP/10 Bugzilla 937: provide a "debug" ACL control.
2816 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2818 PP/12 Bugzilla 973: Implement --version.
2820 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2822 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2824 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2826 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2827 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2828 ignore trailing whitespace.
2830 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2832 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2835 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2836 Notification from Dr Andrew Aitchison.
2838 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2839 ExtendedDetectionInfo response format.
2840 Notification from John Horne.
2842 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2845 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2846 XSL and documented dependency on system catalogs, with examples of how
2849 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2852 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2853 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2854 configuration file which is writeable by the Exim user or group.
2856 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2857 of configuration files to cover files specified with the -C option if
2858 they are going to be used with root privileges, not just the default
2861 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2862 option (effectively making it always true).
2864 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2865 files to be used while preserving root privileges.
2867 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2868 that rogue child processes cannot use them.
2870 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2871 run-time user, instead of root.
2873 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2874 Exim run-time user without dropping privileges.
2876 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2877 result string, instead of calling string_vformat() twice with the same
2880 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2881 for other users. Others should always drop root privileges if they use
2882 -C on the command line, even for a whitelisted configure file.
2884 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2886 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2892 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2893 $data_path, and $header_path variables; fixed documentation bugs and
2896 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2897 exipick to access non-standard spools, including the "frozen" queue
2900 NM/01 Bugzilla 965: Support mysql stored procedures.
2901 Patch from Alain Williams
2903 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2905 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2906 Patch from Andreas Metzler
2908 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2909 Patch from Kirill Miazine
2911 NM/05 Bugzilla 671: Added umask to procmail example.
2913 JJ/03 installed exipick 20100323.0, fixing doc bug
2915 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2916 directory. Notification and patch from Dan Rosenberg.
2918 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2920 TK/02 Improve log output when DKIM signing operation fails.
2922 MH/01 Treat the transport option dkim_domain as a colon separated
2923 list, not as a single string, and sign the message with each element,
2924 omitting multiple occurences of the same signer.
2926 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2927 Bugzilla 985, 986. Patch by Simon Arlott
2929 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2930 Patch by Simon Arlott
2932 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2933 MBX locking. Notification from Dan Rosenberg.
2939 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2941 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2943 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2945 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2947 NM/04 Bugzilla 915: Flag broken perl installation during build.
2953 TK/01 Added patch by Johannes Berg that expands the main option
2954 "spamd_address" if it starts with a dollar sign.
2956 TK/02 Write list of recipients to X-Envelope-Sender header when building
2957 the mbox-format spool file for content scanning (suggested by Jakob
2960 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2961 (http://www.dcc-servers.net/) support via dccifd. Activated by
2962 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2964 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2965 by Mark Daniel Reidel <mr@df.eu>.
2967 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2968 When building exim an external PCRE library is now needed -
2969 PCRE is a system library on the majority of modern systems.
2970 See entry on PCRE_LIBS in EDITME file.
2972 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2973 conversation. Added nologin parameter to request.
2974 Patch contributed by Kirill Miazine.
2976 TF/01 Do not log submission mode rewrites if they do not change the address.
2978 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2980 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2981 log files in place. Contributed by Roberto Lima.
2983 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2985 TF/03 Bugzilla 615: When checking the local_parts router precondition
2986 after a local_part_suffix or local_part_prefix option, Exim now
2987 does not use the address's named list lookup cache, since this
2988 contains cached lookups for the whole local part.
2990 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2991 Robert Millan. Documentation is in experimental-spec.txt.
2993 TF/04 Bugzilla 668: Fix parallel build (make -j).
2995 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2997 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2998 Patch provided by Jan Srzednicki.
3000 TF/05 Leading white space used to be stripped from $spam_report which
3001 wrecked the formatting. Now it is preserved.
3003 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3004 that they are available at delivery time.
3006 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3008 TF/08 TLS error reporting now respects the incoming_interface and
3009 incoming_port log selectors.
3011 TF/09 Produce a more useful error message if an SMTP transport's hosts
3012 setting expands to an empty string.
3014 NM/06 Bugzilla 744: EXPN did not work under TLS.
3015 Patch provided by Phil Pennock.
3017 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3018 Patch provided by Richard Godbee.
3020 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3021 acl_smtp_notquit, added index entry.
3023 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3024 Patch provided by Eugene Bujak.
3026 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3027 accept(). Patch provided by Maxim Dounin.
3029 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3030 Patch provided by Phil Pennock.
3032 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3034 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3035 Patch provided by Brad "anomie" Jorsch.
3037 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3038 Patch provided by Dean Brooks.
3040 TK/05 Add native DKIM support (does not depend on external libraries).
3042 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3043 Patch provided by Graeme Fowler.
3045 NM/16 Bugzilla 851: Documentation example syntax fix.
3047 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3049 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3052 NM/19 Bugzilla 745: TLS version reporting.
3053 Patch provided by Phil Pennock.
3055 NM/20 Bugzilla 167: bool: condition support.
3056 Patch provided by Phil Pennock.
3058 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3059 clients. Patch provided by Phil Pennock.
3061 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3062 Patch provided by Brad "anomie" Jorsch.
3064 NM/23 Bugzilla 687: Fix misparses in eximstats.
3065 Patch provided by Heiko Schlittermann.
3067 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3068 Patch provided by Heiko Schlittermann.
3070 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3071 plus update to original patch.
3073 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3075 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3076 Patch provided by David Brownlee.
3078 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3080 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3082 NM/30 Bugzilla 888: TLS documentation bugfixes.
3084 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3086 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3087 Unlike the original bugzilla I have changed all shell scripts in src tree.
3089 NM/33 Bugzilla 898: Transport filter timeout fix.
3090 Patch by Todd Rinaldo.
3092 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3093 Patch by Serge Demonchaux.
3095 NM/35 Bugzilla 39: Base64 decode bug fixes.
3096 Patch by Jakob Hirsch.
3098 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3100 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3102 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3104 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3110 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3111 ALT-N's libdkim that I have put here:
3112 http://duncanthrax.net/exim-experimental/
3114 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3115 were called 'true' and it seems that C99 defines that as a reserved
3116 keyword to be used with 'bool' variable types. That means you could
3117 not include C99-style headers which use bools without triggering
3118 build errors in sieve.c.
3120 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3121 as mailq or other aliases. Changed the --help handling significantly
3122 to do whats expected. exim_usage() emits usage/help information.
3124 SC/01 Added the -bylocaldomain option to eximstats.
3126 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3128 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3130 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3136 PH/01 Another patch from the Sieve maintainer.
3138 PH/02 When an IPv6 address is converted to a string for single-key lookup
3139 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3140 dots are used instead of colons so that keys in lsearch files need not
3141 contain colons. This was done some time before quoting was made available
3142 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3143 (notated using the quote facility) so as to distinguish them from IPv4
3144 keys. This meant that lookups for IP addresses in host lists did not work
3145 for iplsearch lookups.
3147 This has been fixed by arranging for IPv6 addresses to be expressed with
3148 colons if the lookup type is iplsearch. This is not incompatible, because
3149 previously such lookups could never work.
3151 The situation is now rather anomalous, since one *can* have colons in
3152 ordinary lsearch keys. However, making the change in all cases is
3153 incompatible and would probably break a number of configurations.
3155 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3158 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3159 conversion specification without a maximum field width, thereby enabling
3160 a rogue spamd server to cause a buffer overflow. While nobody in their
3161 right mind would setup Exim to query an untrusted spamd server, an
3162 attacker that gains access to a server running spamd could potentially
3163 exploit this vulnerability to run arbitrary code as the Exim user.
3165 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3166 $primary_hostname instead of what libspf2 thinks the hosts name is.
3168 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3169 a directory entry by the name of the lookup key. Previously, if a
3170 symlink pointed to a non-existing file or a file in a directory that
3171 Exim lacked permissions to read, a lookup for a key matching that
3172 symlink would fail. Now it is enough that a matching directory entry
3173 exists, symlink or not. (Bugzilla 503.)
3175 PH/03 The body_linecount and body_zerocount variables are now exported in the
3178 PH/04 Added the $dnslist_matched variable.
3180 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3181 This means they are set thereafter only if the connection becomes
3184 PH/06 Added the client_condition to authenticators so that some can be skipped
3185 by clients under certain conditions.
3187 PH/07 The error message for a badly-placed control=no_multiline_responses left
3188 "_responses" off the end of the name.
3190 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3192 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3193 (without spaces) instead of just copying the configuration text.
3195 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3197 PH/11 Added $max_received_linelength.
3199 PH/12 Added +ignore_defer and +include_defer to host lists.
3201 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3202 way in which PCRE > 7.0 is built.
3204 PH/14 Implemented queue_only_load_latch.
3206 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3207 MAIL command. The effect was to mangle the value on 64-bit systems.
3209 PH/16 Another patch from the Sieve maintainer.
3211 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3213 PH/18 If a system quota error occurred while trying to create the file for
3214 a maildir delivery, the message "Mailbox is full" was not appended to the
3215 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3216 only to a quota excession during the actual writing of the file.
3218 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3219 characters?) which causes problems in log lines. The DN values are now
3220 passed through string_printing() before being added to log lines.
3222 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3223 and InterBase are left for another time.)
3225 PH/21 Added message_body_newlines option.
3227 PH/22 Guard against possible overflow in moan_check_errorcopy().
3229 PH/23 POSIX allows open() to be a macro; guard against that.
3231 PH/24 If the recipient of an error message contained an @ in the local part
3232 (suitably quoted, of course), incorrect values were put in $domain and
3233 $local_part during the evaluation of errors_copy.
3239 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3240 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3243 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3244 issue a MAIL command.
3246 PH/02 In an ACL statement such as
3248 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3250 if a client was not listed at all, or was listed with a value other than
3251 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3252 the condition was not true (as it should be), so access was not denied.
3253 The bug was that the ! inversion was incorrectly passed on to the second
3254 item. This has been fixed.
3256 PH/03 Added additional dnslists conditions == and =& which are different from
3257 = and & when the dns lookup returns more than one IP address.
3259 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3260 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3262 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3263 FSYNC, which compiles an option called disable_fsync that allows for
3264 bypassing fsync(). The documentation is heavily laced with warnings.
3266 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3268 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3269 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3270 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3271 including adding "make clean"; (3) Added -fPIC when compiling the test
3272 dynamically loaded module, to get rid of a warning.
3274 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3275 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3276 = 0s. The bug is otherwise harmless.
3278 PH/07 There was a bug in the dovecot authenticator such that the value of
3279 $auth1 could be overwritten, and so not correctly preserved, after a
3280 successful authentication. This usually meant that the value preserved by
3281 the server_setid option was incorrect.
3283 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3285 PH/09 Installed PCRE release 7.0.
3287 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3288 run for batched SMTP input. It is now run at the start of every message
3289 in the batch. While fixing this I discovered that the process information
3290 (output by running exiwhat) was not always getting set for -bs and -bS
3291 input. This is fixed, and it now also says "batched" for BSMTP.
3293 PH/11 Added control=no_pipelining.
3295 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3296 patch, slightly modified), and move the expansion of helo_data till after
3297 the connection is made in the smtp transport (so it can use these
3300 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3302 PH/14 Added log_selector = +pid.
3304 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3306 PH/16 Add ${if forany and ${if forall.
3308 PH/17 Added dsn_from option to vary the From: line in DSNs.
3310 PH/18 Flush SMTP output before performing a callout, unless control =
3311 no_callout_flush is set.
3313 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3314 was true (the default) a successful delivery failed to delete the retry
3315 item, thus causing premature timeout of the address. The bug is now
3318 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3320 PH/21 Long custom messages for fakedefer and fakereject are now split up
3321 into multiline responses in the same way that messages for "deny" and
3322 other ACL rejections are.
3324 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3325 with slight modification.
3327 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3328 draft, changing the syntax and factoring some duplicate code".
3330 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3331 for deliveries of the second and subsequent messages over the same SMTP
3334 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3335 ${reduce, with only minor "tidies".
3337 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3339 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3340 expansion side effects.
3342 PH/27 When a message times out after an over-quota error from an Exim-imposed
3343 quota, the bounce message says "mailbox is full". This message was not
3344 being given when it was a system quota that was exceeded. It now should
3347 MH/03 Made $recipients available in local_scan(). local_scan() already has
3348 better access to the recipient list through recipients_list[], but
3349 $recipients can be useful in postmaster-provided expansion strings.
3351 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3352 in the case of a MAIL command with additional options following the
3353 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3354 were accidentally chopped off.
3356 PH/29 SMTP synchronization checks are implemented when a command is read -
3357 there is a check that no more input is waiting when there shouldn't be
3358 any. However, for some commands, a delay in an ACL can mean that it is
3359 some time before the response is written. In this time, more input might
3360 arrive, invalidly. So now there are extra checks after an ACL has run for
3361 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3362 pipelining has not been advertised.
3364 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3366 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3367 returned in the SMTP response when smtp_return_error_details was false.
3368 This has been fixed.
3370 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3371 instead of the C I/O that was originally supplied, because problems were
3372 reported on Solaris.
3374 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3375 Exim which did not show up earlier: it was assuming that a call to
3376 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3377 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3378 was a macro that became an assignment, so it seemed to work. This has
3379 changed to a proper function call with a void return, hence the compile
3380 error. Exim's code has been fixed.
3382 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3385 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3387 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3389 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3390 because it turns out to be too big or there is a timeout) there is no
3391 "Completed" line in the log. When some messages of this type were
3392 selected by exigrep, they were listed as "not completed". Others were
3393 picked up by some special patterns. I have improved the selection
3394 criteria to be more general.
3396 PH/38 The host_find_failed option in the manualroute router can now be set
3397 to "ignore", to completely ignore a host whose IP address cannot be
3398 found. If all hosts are ignored, the behaviour is controlled by the new
3399 host_all_ignored option.
3401 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3402 homing or because of multiple MX records with /mx) generated more than
3403 one IP address, and the following item turned out to be the local host,
3404 all the secondary addresses of the first item were incorrectly removed
3405 from the list, along with the local host and any following hosts (which
3406 is what is supposed to happen).
3408 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3409 whoever called Exim into the -H file. In the case of the daemon it was
3410 behaving confusingly. When first started, it used values for whoever
3411 started the daemon, but after a SIGHUP it used the Exim user (because it
3412 calls itself on a restart). I have changed the code so that it now always
3415 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3416 message are rejected with the same error (e.g. no authentication or bad
3417 sender address), and a DATA command is nevertheless sent (as can happen
3418 with PIPELINING or a stupid MUA), the error message that was given to the
3419 RCPT commands is included in the rejection of the DATA command. This is
3420 intended to be helpful for MUAs that show only the final error to their
3423 PH/42 Another patch from the Sieve maintainer.
3425 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3426 Eximstats - Fixed some broken HTML links and added missing column headers
3428 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3429 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3431 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3433 PH/43 Yet another patch from the Sieve maintainer.
3435 PH/44 I found a way to check for a TCP/IP connection going away before sending
3436 the response to the final '.' that terminates a message, but only in the
3437 case where the client has not sent further data following the '.'
3438 (unfortunately, this is allowed). However, in many cases there won't be
3439 any further data because there won't be any more messages to send. A call
3440 to select() can be used: if it shows that the input is "ready", there is
3441 either input waiting, or the socket has been closed. An attempt to read
3442 the next input character can distinguish the two cases. Previously, Exim
3443 would have sent an OK response which the client would never have see.
3444 This could lead to message repetition. This fix should cure that, at
3445 least in a lot of common cases.
3447 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3448 advertised in response to EHLO.
3454 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3455 fixed by 4.65/MH/01 (is this a record?) are fixed:
3457 (i) An empty string was always treated as zero by the numeric comparison
3458 operators. This behaviour has been restored.
3460 (ii) It is documented that the numeric comparison operators always treat
3461 their arguments as decimal numbers. This was broken in that numbers
3462 starting with 0 were being interpreted as octal.
3464 While fixing these problems I realized that there was another issue that
3465 hadn't been noticed. Values of message_size_limit (both the global option
3466 and the transport option) were treated as octal if they started with 0.
3467 The documentation was vague. These values are now always treated as
3468 decimal, and I will make that clear in the documentation.
3474 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3475 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3478 MH/01 Don't check that the operands of numeric comparison operators are
3479 integers when their expansion is in "skipping" mode (fixes bug
3480 introduced by 4.64-PH/07).
3482 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3483 child addresses, Exim now panics and dies. Previously, because the count
3484 is held in a short int, deliveries were likely to be lost. As such a
3485 large number of recipients for a single message is ridiculous
3486 (performance will be very, very poor), I have chosen to impose a limit
3487 rather than extend the field.
3493 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3494 leftover -K file (the existence of which was triggered by #402).
3495 While we were at it, introduced process PID as part of the -K
3496 filename. This should rule out race conditions when creating
3499 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3500 processing considerably. Previous code took too long for large mails,
3501 triggering a timeout which in turn triggers #401.
3503 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3504 in the DK code in transports.c. sendfile() is not really portable,
3505 hence the _LINUX specificness.
3507 TF/01 In the add_headers option to the mail command in an Exim filter,
3508 there was a bug that Exim would claim a syntax error in any
3509 header after the first one which had an odd number of characters
3512 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3513 callout verification, Exim cached a "reject" for the entire domain. This
3514 is correct for most verifications, but it is not correct for a recipient
3515 verification with use_sender or use_postmaster set, because in that case
3516 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3517 case of MAIL FROM:<> rejection from other early rejections (e.g.
3518 rejection of HELO). When verifying a recipient using a non-null MAIL
3519 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3520 Whatever the result of the callout, the value of the domain cache is
3521 left unchanged (for any other kind of callout, getting as far as trying
3522 RCPT means that the domain itself is ok).
3524 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3527 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3528 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3529 session, the daemon ignores ECONNECT errors and logs others; it now
3530 ignores EPIPE as well.
3532 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3533 (quoted-printable decoding).
3535 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3536 later the small subsequent patch to fix an introduced bug.
3538 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3540 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3542 PH/08 An error is now given if message_size_limit is specified negative.
3544 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3545 to be given (somewhat) arbitrary names.
3547 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3550 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3551 miscellaneous code fixes
3553 PH/10 Added the log_reject_target ACL modifier to specify where to log
3556 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3557 hostname. This is wrong, because it relates to the incoming message (and
3558 probably the interface on which it is arriving) and not to the outgoing
3559 callout (which could be using a different interface). This has been
3560 changed to use the value of the helo_data option from the smtp transport
3561 instead - this is what is used when a message is actually being sent. If
3562 there is no remote transport (possible with a router that sets up host
3563 addresses), $smtp_active_hostname is used.
3565 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3566 tweaks were necessary in order to get it to work (see also 21 below):
3567 (a) The code assumed that strncpy() returns a negative number on buffer
3568 overflow, which isn't the case. Replaced with Exim's string_format()
3570 (b) There were several signed/unsigned issues. I just did the minimum
3571 hacking in of casts. There is scope for a larger refactoring.
3572 (c) The code used strcasecmp() which is not a standard C function.
3573 Replaced with Exim's strcmpic() function.
3574 (d) The code set only $1; it now sets $auth1 as well.
3575 (e) A simple test gave the error "authentication client didn't specify
3576 service in request". It would seem that Dovecot has changed its
3577 interface. Fortunately there's a specification; I followed it and
3578 changed what the client sends and it appears to be working now.
3580 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3583 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3584 address A is aliased to B and C, where B exists and C does not. Without
3585 -v the output is "A verified" because verification stops after a
3586 successful redirection if more than one address is generated. However,
3587 with -v the child addresses are also verified. Exim was outputting "A
3588 failed to verify" and then showing the successful verification for C,
3589 with its parentage. It now outputs "B failed to verify", showing B's
3590 parentage before showing the successful verification of C.
3592 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3593 look up a TXT record in a specific list after matching in a combined
3596 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3597 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3598 they consult the DNS. I had assumed they would set it the way they
3599 wanted; and indeed my experiments on Linux seem to show that in some
3600 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3601 To be on the safe side, however, I have now made the interface to
3602 host_find_byname() similar to host_find_bydns(), with an argument
3603 containing the DNS resolver options. The host_find_byname() function now
3604 sets these options at its start, just as host_find_bydns() does. The smtp
3605 transport options dns_qualify_single and dns_search_parents are passed to
3606 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3607 of host_find_byname() use the default settings of RES_DEFNAMES
3608 (qualify_single) but not RES_DNSRCH (search_parents).
3610 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3611 spool_read_header() do less string testing, by means of a preliminary
3612 switch on the second character of optional "-foo" lines. (This is
3613 overdue, caused by the large number of possibilities that now exist.
3614 Originally there were few.) While I was there, I also converted the
3615 str(n)cmp tests so they don't re-test the leading "-" and the first
3616 character, in the hope this might squeeze out yet more improvement.
3618 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3619 flag allowing group syntax was set by the header_syntax check but not
3620 turned off, possible causing trouble later; (2) The flag was not being
3621 set at all for the header_verify test, causing "group"-style headers to
3622 be rejected. I have now set it in this case, and also caused header_
3623 verify to ignore an empty address taken from a group. While doing this, I
3624 came across some other cases where the code for allowing group syntax
3625 while scanning a header line wasn't quite right (mostly, not resetting
3626 the flag correctly in the right place). These bugs could have caused
3627 trouble for malformed header lines. I hope it is now all correct.
3629 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3630 with the "reply" argument non-NULL. The code, however (which originally
3631 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3632 but it didn't always do it. This confused somebody who was copying the
3633 code for some other use. I have removed all the tests.
3635 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3636 feature that was used to support insecure browsers during the U.S. crypto
3637 embargo. It requires special client support, and Exim is probably the
3638 only MTA that supported it -- and would never use it because real RSA is
3639 always available. This code has been removed, because it had the bad
3640 effect of slowing Exim down by computing (never used) parameters for the
3641 RSA_EXPORT functionality.
3643 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3644 authenticator to fail if there's a tab character in the incoming data
3645 (there should never be unless someone is messing about, as it's supposed
3646 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3647 if the connection is using TLS or if the remote IP is the same as the
3648 local IP, and the "valid-client-cert option" if a client certificate has
3651 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3652 authenticators. This can be used for authorization after authentication
3653 succeeds. (In the case of plaintext, it servers for both authentication
3656 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3657 if any retry times were supplied.
3659 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3660 connection, where there is no client IP address to check. In this
3661 situation, the verify now always succeeds.
3663 PH/25 Applied John Jetmore's -Mset patch.
3665 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3667 PH/27 In a string expansion for a processed (not raw) header when multiple
3668 headers of the same name were present, leading whitespace was being
3669 removed from all of them, but trailing whitespace was being removed only
3670 from the last one. Now trailing whitespace is removed from each header
3671 before concatenation. Completely empty headers in a concatenation (as
3672 before) are ignored.
3674 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3675 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3677 PH/29 [Removed. This was a change that I later backed out, and forgot to
3678 correct the ChangeLog entry (that I had efficiently created) before
3679 committing the later change.]
3681 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3682 address errors (4xx response to RCPT) over the same connection as other
3683 messages routed to the same hosts. Such deliveries are always "forced",
3684 so retry times are not inspected. This resulted in far too many retries
3685 for the affected addresses. The effect occurred only when there were more
3686 hosts than the hosts_max_try setting in the smtp transport when it had
3687 the 4xx errors. Those hosts that it had tried were not added to the list
3688 of hosts for which the message was waiting, so if all were tried, there
3689 was no problem. Two fixes have been applied:
3691 (i) If there are any address or message errors in an SMTP delivery, none
3692 of the hosts (tried or untried) are now added to the list of hosts
3693 for which the message is waiting, so the message should not be a
3694 candidate for sending over the same connection that was used for a
3695 successful delivery of some other message. This seems entirely
3696 reasonable: after all the message is NOT "waiting for some host".
3697 This is so "obvious" that I'm not sure why it wasn't done
3698 previously. Hope I haven't missed anything, but it can't do any
3699 harm, as the worst effect is to miss an optimization.
3701 (ii) If, despite (i), such a delivery is accidentally attempted, the
3702 routing retry time is respected, so at least it doesn't keep
3703 hammering the server.
3705 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3706 in ${readsocket because some servers need this prod.
3708 PH/32 Added some extra debug output when updating a wait-xxx database.
3710 PH/33 The hint "could be header name not terminated by colon", which has been
3711 given for certain expansion errors for a long time, was not being given
3712 for the ${if def:h_colon_omitted{... case.
3714 PH/34 The spec says: "With one important exception, whenever a domain list is
3715 being scanned, $domain contains the subject domain." There was at least
3716 one case where this was not true.
3718 PH/35 The error "getsockname() failed: connection reset by peer" was being
3719 written to the panic log as well as the main log, but it isn't really
3720 panic-worthy as it just means the connection died rather early on. I have
3721 removed the panic log writing for the ECONNRESET error when getsockname()
3724 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3725 runs only) independently of the message's sender address. This meant
3726 that, if the 4xx error was in fact related to the sender, a different
3727 message to the same recipient with a different sender could confuse
3728 things. In particular, this can happen when sending to a greylisting
3729 server, but other circumstances could also provoke similar problems.
3730 I have changed the default so that the retry time for these errors is now
3731 based a combination of the sender and recipient addresses. This change
3732 can be overridden by setting address_retry_include_sender=false in the
3735 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3736 remote server are returned as part of bounce messages. This was not
3737 happening for LMTP over a pipe (the lmtp transport), but now it is the
3738 same for both kinds of LMTP.
3740 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3741 in header lines that were in fact CNAMEs. This is no longer the case.
3743 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3744 and queue runs started by the daemon processed all messages. This has
3745 been fixed so that -R and -S can now usefully be given with -q<time>.
3747 PH/40 Import PCRE release 6.7 (fixes some bugs).
3749 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3751 PH/42 Give an error if -q is specified more than once.
3753 PH/43 Renamed the variables $interface_address and $interface_port as
3754 $received_ip_address and $received_port, to make it clear that these
3755 values apply to message reception, and not to the outgoing interface when
3756 a message is delivered. (The old names remain recognized, of course.)
3758 PH/44 There was no timeout on the connect() call when using a Unix domain
3759 socket in the ${readsocket expansion. There now is.
3761 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3762 be meaningful with "accept".
3764 SC/01 Eximstats V1.43
3765 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3767 SC/02 Eximstats V1.44
3768 Use a glob alias rather than an array ref in the generated
3769 parser. This improves both readability and performance.
3771 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3772 Collect SpamAssassin and rejection statistics.
3773 Don't display local sender or destination tables unless
3774 there is data to show.
3775 Added average volumes into the top table text output.
3777 SC/04 Eximstats V1.46
3778 Collect data on the number of addresses (recipients)
3779 as well as the number of messages.
3781 SC/05 Eximstats V1.47
3782 Added 'Message too big' to the list of mail rejection
3783 reasons (thanks to Marco Gaiarin).
3785 SC/06 Eximstats V1.48
3786 Mainlog lines which have GMT offsets and are too short to
3787 have a flag are now skipped.
3789 SC/07 Eximstats V1.49 (Alain Williams)
3790 Added the -emptyok flag.
3792 SC/08 Eximstats V1.50
3793 Fixes for obtaining the IP address from reject messages.
3795 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3796 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3797 whitespace changes from 4.64-PH/27
3799 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3802 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3803 are found, allow negative numbers in numeric criteria)
3805 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3807 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3808 to match changes made in 4.64-PH/43
3810 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3812 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3813 "domain" is actually an IP address, and force a failure. This locks out
3814 those revolvers/nameservers that support "A-for-A" lookups, in
3815 contravention of the specifications.
3817 PH/48 When a host name was looked up from an IP address, and the subsequent
3818 forward lookup of the name timed out, the host name was left in
3819 $sender_host_name, contrary to the specification.
3821 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3822 restricted to single-key lookups, Exim was not diagnosing an error if
3823 * or *@ was used with a query-style lookup.
3825 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3827 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3828 long ago, but noone interested enough thought of it. Let's just say that
3829 the "1.1" means that there are some new functions that weren't there at
3830 some point in the past.
3832 PH/51 Error processing for expansion failure of helo_data from an smtp
3833 transport during callout processing was broken.
3835 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3836 tested/used via the -bh/-bhc/-bs options.
3838 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3839 bug, fixed in subsequent PCRE releases).
3841 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3842 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3844 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3850 SC/01 Use a glob alias rather than an array ref in eximstats generated
3851 parser. This improves both readability and performance.
3853 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3854 Don't display local sender or destination tables in eximstats unless
3855 there is data to show.
3856 Added average volumes into the eximstats top table text output.
3858 SC/03 Collect data on the number of addresses (recipients) as well
3859 as the number of messages in eximstats.
3861 TF/01 Correct an error in the documentation for the redirect router. Exim
3862 does (usually) call initgroups() when daemonizing.
3864 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3865 with consistent privilege compared to when running as a daemon.
3867 TF/03 Note in the spec that $authenticated_id is not set for local
3868 submissions from trusted users.
3870 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3871 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3873 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3874 by adding some example configuration directives to the default
3875 configuration file. A little bit of work is required to uncomment the
3876 directives and define how usernames and passwords are checked, but
3877 there is now a framework to start from.
3879 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3880 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3881 without this. I don't know how relevant this is to other LDAP libraries.
3883 PH/02 Add the verb name to the "unknown ACL verb" error.
3885 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3887 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3889 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3890 directory not expanded when it should be if an expanded home directory
3891 was set for the address (which is overridden by the transport).
3893 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3896 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3897 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3898 because it is too late at that time, and has no effect.
3900 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3901 security issue with \' (bugzilla #107). I could not use the
3902 PQescapeStringConn() function, because it needs a PGconn value as one of
3905 PH/08 When testing addresses using -bt, indicate those final addresses that
3906 are duplicates that would not cause an additional delivery. At least one
3907 person was confused, thinking that -bt output corresponded to deliveries.
3908 (Suppressing duplicates isn't a good idea as you lose the information
3909 about possibly different redirections that led to the duplicates.)
3911 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3912 systems where poll() doesn't work, in particular OS X.
3914 PH/10 Added more information to debugging output for retry time not reached.
3916 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3917 operations in malware.c.
3919 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3922 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3923 syslog_duplication set false, log lines that would normally be written
3924 both the the main log and to the reject log were not written to syslog at
3927 PH/14 In the default configuration, change the use of "message" in ACL warn
3928 statements to "add_header".
3930 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3931 not followed by a command (e.g. "seen endif").
3933 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3934 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3937 PH/17 Added extra conditions to the default value of delay_warning_condition
3941 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3942 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3943 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3946 The Auto-Submitted: and various List- headers are standardised, whereas I
3947 don't think Precedence: ever was.
3949 PH/18 Refactored debugging code in route_finduser() to show more information,
3950 in particular, the error code if getpwnam() issues one.
3952 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3953 This is apparently needed in addition to the PH/07 change above to avoid
3954 any possible encoding problems.
3956 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3957 but not after initializing Perl.
3959 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3960 output them only if debugging. By default they are written stderr,
3961 apparently, which is not desirable.
3963 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3966 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3969 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3971 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3972 authenticated or an ident call has been made. Suppress the default
3973 values for $authenticated_id and $authenticated_sender (but permit -oMai
3974 and -oMas) when testing with -bh.
3976 PH/24 Re-jigged the order of the tests in the default configuration so that the
3977 tests for valid domains and recipients precede the DNS black list and CSA
3978 tests, on the grounds that those ones are more expensive.
3980 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3981 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3982 This bug exists in every version of Exim that I still have, right back to
3985 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3986 However, an attempt to turn on case-sensitivity in a regex key by
3987 including (?-i) didn't work because the subject string was already
3988 lowercased, and the effects were non-intuitive. It turns out that a
3989 one-line patch can be used to allow (?-i) to work as expected.
3995 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3996 other effects) broke the use of negated acl sub-conditions.
3998 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4001 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4002 "Deny" causes Exim to reject the incoming connection with a 554 error.
4003 Unfortunately, if there is a major crisis, such as a disk failure,
4004 tcp-wrappers gives "deny", whereas what one would like would be some
4005 kind of temporary error. A kludge has been added to help with this.
4006 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4007 554 error is used if errno is still zero or contains ENOENT (which occurs
4008 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4011 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4013 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4014 errors. Otherwise a message that provokes a temporary error (when other
4015 messages do not) can cause a whole host to time out.
4017 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4018 addresses were routed directly to files or pipes from a redirect router.
4019 File deliveries just didn't batch; pipe deliveries might have suffered
4022 PH/06 A failure to get a lock for a hints database would erroneously always say
4023 "Failed to get write lock", even when it was really a read lock.
4025 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4026 of 0600. This has been changed to use the value of the lockfile_mode
4027 option (which defaults to 0600).
4029 PH/08 Applied small patch from the Sieve maintainer.
4031 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4032 folder from quota calculations, a direct delivery into this folder messed
4033 up the contents of the maildirsize file. This was because the regex was
4034 used only to exclude .Trash (or whatever) when the size of the mailbox
4035 was calculated. There was no check that a delivery was happening into an
4036 excluded directory. This bug has been fixed by ignoring all quota
4037 processing for deliveries into excluded directories.
4039 PH/10 Added the maildirfolder_create_regex option to appendfile.
4045 PH/01 The code for finding all the local interface addresses on a FreeBSD
4046 system running IPv6 was broken. This may well have applied to all BSD
4047 systems, as well as to others that have similar system calls. The broken
4048 code found IPv4 interfaces correctly, but gave incorrect values for the
4049 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4050 that it would not match correctly against @[] and not recognize the IPv6
4053 PH/02 The ipliteral router was not recognizing addresses of the form user@
4054 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4056 PH/03 Added disable_ipv6.
4058 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4059 decoded form, because it is most often used to construct To: headers
4060 lines in autoreplies, and the decoded form may well be syntactically
4061 invalid. However, $reply_address has leading white space removed, and all
4062 newlines turned into spaces so that the autoreply transport does not
4065 PH/05 If group was specified without a user on a router, and no group or user
4066 was specified on a transport, the group from the router was ignored.
4068 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4069 for visible compile-time settings that can be used to change these
4070 numbers, for those that want even more. Backwards compatibility with old
4071 spool files has been maintained. However, going back to a previous Exim
4072 release will lost any variables that are in spool files.
4074 PH/07 Two small changes when running in the test harness: increase delay when
4075 passing a TCP/IP connection to a new process, in case the original
4076 process has to generate a bounce, and remove special handling of
4077 127.0.0.2 (sic), which is no longer necessary.
4079 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4080 be the same on different OS.
4082 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4085 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4086 whether --show-vars was specified or not
4088 JJ/02 exipick: Added support for new ACL variable spool format introduced
4091 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4092 syntactically invalid From: or Reply-to: line, and a filter used this to
4093 generate an autoreply, and therefore failed to obtain an address for the
4094 autoreply, Exim could try to deliver to a non-existent relative file
4095 name, causing unrelated and misleading errors. What now happens is that
4096 it logs this as a hard delivery error, but does not attempt to create a
4099 PH/11 The exinext utility has a -C option for testing purposes, but although
4100 the given file was scanned by exinext itself; it wasn't being passed on
4101 when Exim was called.
4103 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4104 an end-of-file indication when reading a command response.
4106 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4107 compiled. In many other places in Exim, IPv6 addresses are always
4108 recognized, so I have changed this. It also means that IPv4 domain
4109 literals of the form [IPV4:n.n.n.n] are now always recognized.
4111 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4112 used if the router is not running as root, for example, when verifying at
4113 ACL time, or when using -bh. The debugging output from this situation was
4114 non-existent - all you got was a failure to exec. I have made two
4117 (a) Failures to set uid/gid, the current directory, or a process leader
4118 in a subprocess such as that created by queryprogram now generate
4119 suitable debugging output when -d is set.
4121 (b) The queryprogram router detects when it is not running as root,
4122 outputs suitable debugging information if -d is set, and then runs
4123 the subprocess without attempting to change uid/gid.
4125 PH/15 Minor change to Makefile for building test_host (undocumented testing
4128 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4129 additional section of a DNS packet that returns MX or SRV records.
4130 Instead, it always explicitly searches for A/AAAA records. This avoids
4131 major problems that occur when a DNS server includes only records of one
4132 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4133 fixed another bug: if SRV records were looked up and the corresponding
4134 address records were *not* found in the additional section, the port
4135 values from the SRV records were lost.
4137 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4138 using the correct key (the original address) when searching the retry
4139 rules in order to find which one to use for generating the retry hint.
4141 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4142 adding the default one. Similarly, if it contains a Reply-To: header, the
4143 errors_reply_to option, if set, is not used.
4145 PH/19 When calculating a retry time, Exim used to measure the "time since
4146 failure" by looking at the "first failed" field in the retry record. Now
4147 it does not use this if it is later than than the arrival time of the
4148 message. Instead it uses the arrival time. This makes for better
4149 behaviour in cases where some deliveries succeed, thus re-setting the
4150 "first failed" field. An example is a quota failure for a huge message
4151 when small messages continue to be delivered. Without this change, the
4152 "time since failure" will always be short, possible causing more frequent
4153 delivery attempts for the huge message than are intended.
4154 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4156 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4157 $1, $2, $3) because the numerical variables can be reset during some
4158 expansion items (e.g. "match"), thereby losing the authentication data.
4160 PH/21 Make -bV show the size of off_t variables so that the test suite can
4161 decide whether to run tests for quotas > 2G.
4163 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4164 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4165 filecount value is greater than 2G or if a quota value is greater than 2G
4166 on a system where the size of off_t is not greater than 4, a panic error
4169 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4170 never match. The debug and -bh output now contains an explicit error
4171 message indicating a malformed IPv4 address or mask.
4173 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4174 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4175 PH/23 above applies.
4177 PH/25 Do not write to syslog when running in the test harness. The only
4178 occasion when this arises is a failure to open the main or panic logs
4179 (for which there is an explicit test).
4181 PH/26 Added the /no_tell option to "control=freeze".
4183 PH/27 If a host name lookup failed very early in a connection, for example, if
4184 the IP address matched host_lookup and the reverse lookup yielded a name
4185 that did not have a forward lookup, an error message of the form "no IP
4186 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4187 could be logged. Now it outputs the IP address instead of "NULL".
4189 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4190 allows the sender and the authenticated sender to be set when
4191 submitting a message from within Exim. Since child_open_exim() is
4192 documented for local_scan(), the new function should be too.
4194 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4195 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4196 results in an empty string is now treated as unset.
4198 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4200 PH/31 Added sender_verify_fail logging option.
4202 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4203 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4204 tidied the source and removed it altogether.
4206 PH/33 When a queue run was abandoned because the load average was too high, a
4207 log line was always written; now it is written only if the queue_run log
4208 selector is set. In addition, the log line for abandonment now contains
4209 information about the queue run such as the pid. This is always present
4210 in "start" and "stop" lines but was omitted from the "abandon" line.
4212 PH/34 Omit spaces between a header name and the colon in the error message that
4213 is given when verify = headers_syntax fails (if there are lots of them,
4214 the message gets confusing).
4216 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4217 names, as there are now some PTR records that contain slashes. This check
4218 is only to protect against broken name servers that fall over on strange
4219 characters, so the fact that it applies to all lookups doesn't matter.
4221 PH/36 Now that the new test suite is complete, we can remove some of the
4222 special code in Exim that was needed for the old test suite. For example,
4223 sorting DNS records because real resolvers return them in an arbitrary
4224 order. The new test suite's fake resolver always returns records in the
4227 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4228 bounces) except when queue_only is set, to avoid logging races between
4229 the different processes.
4231 PH/38 Panic-die if .include specifies a non-absolute path.
4233 PH/39 A tweak to the "H" retry rule from its user.
4235 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4236 a label. They prevented compilation on older perls.
4238 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4239 a warning to be raised on newish perls.
4241 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4242 on queue. Changes to match documented behaviour of showing count of
4243 messages matching specified criteria.
4245 PH/40 Changed the default ident timeout from 30s to 5s.
4247 PH/41 Added support for the use of login_cap features, on those BSD systems
4248 that have them, for controlling the resources used by pipe deliveries.
4250 PH/42 The content-scanning code uses fopen() to create files in which to put
4251 message data. Previously it was not paying any attention to the mode of
4252 the files. Exim runs with umask(0) because the rest of the code creates
4253 files with open(), and sets the required mode explicitly. Thus, these
4254 files were ending up world-writeable. This was not a big issue, because,
4255 being within the spool directory, they were not world-accessible. I have
4256 created a function called modefopen, which takes an additional mode
4257 argument. It sets umask(777), creates the file, chmods it to the required
4258 mode, then resets the umask. All the relevant calls to fopen() in the
4259 content scanning code have been changed to use this function.
4261 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4262 to 24 hours. This avoids potential overflow problems when processing G
4263 and H retry rules. I suspect nobody ever tinkers with this value.
4265 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4267 PH/45 When the plaintext authenticator is running as a client, the server's
4268 challenges are checked to ensure they are valid base64 strings. By
4269 default, the authentication attempt is cancelled if an invalid string is
4270 received. Setting client_ignore_invalid_base64 true ignores these errors.
4271 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4272 they are received. Thus, the responses can be made to depend on the
4273 challenges. If an invalid string is ignored, an empty string is placed in
4276 PH/46 Messages that are created by the autoreply transport now contains a
4277 References: header, in accordance with RFCs 2822 and 3834.
4279 PH/47 Added authenticated_sender_force to the smtp transport.
4281 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4283 PH/49 Installed latest patch from the Sieve maintainer.
4285 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4286 was also set, the appendfile transport was unnecessarily scanning a
4287 directory of message files (e.g. for maildir delivery) to find the count
4288 of files (along with the size), even though it did not need this
4289 information. It now does the scan only if it needs to find either the
4290 size of the count of files.
4292 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4294 PH/52 Two bugs concerned with error handling when the smtp transport is
4297 (i) Exim was not creating retry information for temporary errors given
4298 for individual recipients after the DATA command when the smtp transport
4299 was used in LMTP mode. This meant that they could be retried too
4300 frequently, and not timed out correctly.
4302 (ii) Exim was setting the flag that allows error details to be returned
4303 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4304 recipients that were returned after the DATA command.
4306 PH/53 This is related to PH/52, but is more general: for any failing address,
4307 when detailed error information was permitted to be returned to the
4308 sender, but the error was temporary, then after the final timeout, only
4309 "retry timeout exceeded" was returned. Now it returns the full error as
4310 well as "retry timeout exceeded".
4312 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4313 do this, and (what is worse) MTAs that accept it.
4315 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4316 will now be deprecated.
4318 PH/56 New os.c-cygwin from the Cygwin maintainer.
4320 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4321 formats (previously only available in exim formats via -bpr, -bpru,
4322 and -bpra. Now also available in native and exiqgrep formats)
4324 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4325 with very large, slow to parse queues
4327 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4329 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4331 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4332 responses to authentication challenges, though it was showing the
4333 challenges; (ii) I've removed the CR characters from the debug output for
4336 PH/58 Allow for the insertion of a newline as well as a space when a string
4337 is turned into more than one encoded-word during RFC 2047 encoding. The
4338 Sieve code now uses this.
4340 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4341 data_4xx, lost_connection, tls_required.
4343 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4344 message was being sent as an SMTP response.
4346 PH/61 Add -l and -k options to exicyclog.
4348 PH/62 When verifying, if an address was redirected to one new address, so that
4349 verification continued, and the new address failed or deferred after
4350 having set something in $address_data, the value of $address_data was not
4351 passed back to the ACL. This was different to the case when no
4352 redirection occurred. The value is now passed back in both cases.
4354 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4355 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4356 use login_cap.h, so on its own it isn't the distinguishing feature. The
4357 new name refers directly to the setclassresources() function.
4359 PH/65 Added configuration files for NetBSD3.
4361 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4363 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4364 is preferred over IPv4.
4366 PH/68 The bounce_return_message and bounce_return_body options were not being
4367 honoured for bounces generated during the reception of non-SMTP messages.
4368 In particular, this applied to messages rejected by the ACL. This bug has
4369 been fixed. However, if bounce_return_message is true and bounce_return_
4370 body is false, the headers that are returned for a non-SMTP message
4371 include only those that have been read before the error was detected.
4372 (In the case of an ACL rejection, they have all been read.)
4374 PH/69 The HTML version of the specification is now built in a directory called
4375 spec_html instead of spec.html, because the latter looks like a path with
4376 a MIME-type, and this confuses some software.
4378 PH/70 Catch two compiler warnings in sieve.c.
4380 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4381 function verify_get_ident() calls ip_connect() to connect a socket, but
4382 if the "connect()" function timed out, ip_connect() used to close the
4383 socket. However, verify_get_ident() also closes the socket later, and in
4384 between Exim writes to the log, which may get opened at this point. When
4385 the socket was closed in ip_connect(), the log could get the same file
4386 descriptor number as the socket. This naturally causes chaos. The fix is
4387 not to close the socket in ip_connect(); the socket should be closed by
4388 the function that creates it. There was only one place in the code where
4389 this was missing, in the iplookup router, which I don't think anybody now
4390 uses, but I've fixed it anyway.
4392 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4393 well as to direct DNS lookups. Otherwise the handling of names in host
4394 lists is inconsistent and therefore confusing.
4400 PH/01 Two changes to the default runtime configuration:
4402 (1) Move the checks for relay_from_hosts and authenticated clients from
4403 after to before the (commented out) DNS black list checks.
4405 (2) Add control=submission to the relay_from_hosts and authenticated
4406 clients checks, on the grounds that messages accepted by these
4407 statements are most likely to be submissions.
4409 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4411 (1) Generate an error if the third argument for the ${prvs expansion is
4414 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4417 (3) Reset the variables that are obtained from the first argument of
4418 ${prvscheck and used in the second argument before leaving the code,
4419 because their memory is reclaimed, so using them afterwards may do
4422 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4423 one (it's much easier than Tom thought :-).
4425 (5) Because of (4), we can now allow for the use of $prvscheck_result
4426 inside the third argument.
4428 PH/03 For some reason, the default setting of PATH when running a command from
4429 a pipe transport was just "/usr/bin". I have changed it to
4432 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4433 anything to be listed in the output from -bV.
4435 PH/05 When a filter generated an autoreply, the entire To: header line was
4436 quoted in the delivery log line, like this:
4438 => >A.N.Other <ano@some.domain> <original@ddress> ...
4440 This has been changed so that it extracts the operative address. There
4441 may be more than one such address. If so, they are comma-separated, like
4444 => >ano@some.domain,ona@other.domain <original@ddress> ...
4446 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4447 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4448 not being looked up in the rDNS to get a host name, Exim was showing the
4449 IP address twice in Received: lines, even though the IP addresses were
4450 identical. For example:
4452 Received: from [1.2.3.4] (helo=[1.2.3.4])
4454 However, if the real host name was known, it was omitting the HELO data
4455 if it matched the actual IP address. This has been tidied up so that it
4456 doesn't show the same IP address twice.
4458 PH/07 When both +timestamp and +memory debugging was on, the value given by
4459 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4460 called by the debug printing, thereby overwriting the timestamp buffer.
4461 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4463 PH/08 When the original message was included in an autoreply transport, it
4464 always said "this is a copy of the message, including all the headers",
4465 even if body_only or headers_only was set. It now gives an appropriate
4468 PH/09 Applied a patch from the Sieve maintainer which:
4470 o fixes some comments
4471 o adds the (disabled) notify extension core
4472 o adds some debug output for the result of if/elsif tests
4473 o points to the current vacation draft in the documentation
4474 and documents the missing references header update
4478 o fixes a bug in processing the envelope test (when testing
4479 multiple envelope elements, the last element determined the
4482 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4483 Electronic Mail") by including:
4485 Auto-submitted: auto-generated
4487 in the messages that it generates (bounce messages and others, such as
4488 warnings). In the case of bounce messages for non-SMTP messages, there was
4489 also a typo: it was using "Auto_submitted" (underscore instead of
4490 hyphen). Since every message generated by Exim is necessarily in response
4491 to another message, thes have all been changed to:
4493 Auto-Submitted: auto-replied
4495 in accordance with these statements in the RFC:
4497 The auto-replied keyword:
4499 - SHOULD be used on messages sent in direct response to another
4500 message by an automatic process,
4502 - MUST NOT be used on manually-generated messages,
4504 - MAY be used on Delivery Status Notifications (DSNs) and Message
4505 Disposition Notifications (MDNs),
4507 - MUST NOT be used on messages generated by automatic or periodic
4508 processes, except for messages which are automatic responses to
4511 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4512 to the default Received: header definition.
4514 PH/12 Added log selector acl_warn_skipped (default on).
4516 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4517 variables because (a) they are in the wrong storage pool and (b) even if
4518 they were copied, it wouldn't work properly because of the caching.
4520 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4521 checking when decoding. Apparently there are clients that generate
4522 overlong encoded strings. Why am I not surprised?
4524 PH/15 If the first argument of "${if match_address" was not empty, but did not
4525 contain an "@" character, Exim crashed. Now it writes a panic log message
4526 and treats the condition as false.
4528 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4530 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4531 extension "envelope-auth". The code is finished and in agreement with
4532 other implementations, but there is no documentation so far and in fact,
4533 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4534 not changing the active code.
4536 Print executed "if" and "elsif" statements when debugging is used. This
4537 helps a great deal to understand what a filter does.
4539 Document more things not specified clearly in RFC3028. I had all this
4540 sorted out, when out of a sudden new issues came to my mind. Oops."
4542 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4545 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4546 canonical form (as documented). However, after a host name lookup from
4547 the IP address, check_host() was doing a simple string comparison with
4548 addresses acquired from the DNS when checking that the found name did
4549 have the original IP as one of its addresses. Since any found IPv6
4550 addresses are likely to be in abbreviated form, the comparison could
4551 fail. Luckily, there already exists a function for doing the comparison
4552 by converting both addresses to binary, so now that is used instead of
4553 the text comparison.
4555 PH/20 There was another similar case to PH/19, when a complete host name was
4556 given in a host list; looking up its IP address could give an abbreviated
4557 form, whereas the current host's name might or might not be abbreviated.
4558 The same fix has been applied.
4564 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4565 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4568 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4569 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4571 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4573 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4574 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4575 Also, for "Auto-Submitted", treat anything other than "no" as
4576 non-personal, in accordance with RFC 3834. (Previously it treated
4577 anything starting "auto-" as non-personal.)
4579 TF/01 The control=submission/name=... option had a problem with syntax
4580 errors if the name included a slash character. The /name= option
4581 now slurps the rest of the string, so it can include any characters
4582 but it must come last in the list of options (after /sender_retain
4585 PH/05 Some modifications to the interface to the fake nameserver for the new
4593 TK/01 Added the "success_on_redirect" address verification option. See
4594 NewStuff for rationale and an example.
4596 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4598 PH/02 Patch to exigrep to allow it to work on syslog lines.
4600 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4601 fread() to skip over the body file's header line, because in Cygwin the
4602 header line is locked and is inaccessible.
4604 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4605 co-exist for some time) to make it clear that it is the Exim ID that is
4606 referenced, not the Message-ID: header line.
4608 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4609 string_format() function, because snprintf() does not exist on all
4612 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4613 ${stat: expansion item.
4615 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4616 protocol synchronization error", to keep the pedants happy.
4618 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4619 well as for IRIX systems, when gcc is being used. See the host.c source
4622 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4624 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4627 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4628 transport and to the smtp transport in LMTP mode.
4630 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4632 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4633 run a filter in a subprocess. This could lead to confusion in subsequent
4634 lookups in the parent process. There should also be a search_tidyup() at
4635 the end of the subprocess.
4637 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4638 only if the host matched helo_try_verify_hosts, which caused the
4639 verification to occur when the EHLO/HELO command was issued. The ACL just
4640 tested the remembered result. Now, if a previous verification attempt has
4641 not happened, "verify = helo" does it there and then.
4643 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4645 TK/03 Fix log output including CR from clamd.
4647 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4648 exist provoked a memory error which could cause a segfault.
4650 PH/15 Installed PCRE 6.2
4652 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4654 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4655 of the problem. Specifically, suggested +O2 rather than +O1 for the
4658 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4660 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4661 record, the port was not correctly logged when the outgoing_port log
4662 selector was set (it logged the transort's default port).
4664 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4665 fallback_hosts, and "hosts" in the smtp transport.
4667 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4668 host errors such as "Connection refused".
4670 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4671 authentication with radiusclient 0.4.9:
4673 - Error returned from rc_read_config was caught wrongly
4674 - Username/password not passed on to radius server due to wrong length.
4676 The presumption is that some radiusclient API changes for 4.51/PH/17
4677 were not taken care of correctly. The code is still untested by me (my
4678 Linux distribution still has 0.3.2 of radiusclient), but it was
4679 contributed by a Radius user.
4681 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4682 expanding the "port" option of the smtp transport.
4684 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4685 while reading a MIME header. Thanks to Tom Hughes for a patch.
4687 PH/24 Include config.h inside local_scan.h so that configuration settings are
4690 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4691 that in an ACL for RCPT (for example), you can examine exactly what was
4694 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4695 commands, but it was not correctly comparing the address with the actual
4696 client host address. Thus, it would show the EHLO address in Received:
4697 header lines when this was not necessary.
4699 PH/27 Added the % operator to ${eval:}.
4701 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4702 it should be ignoring failures (because with -C, for example, it has lost
4703 privilege). It wasn't ignoring creation failures other than "already
4706 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4709 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4710 because an input line was too long, either on its own, or by virtue of
4711 too many continuations, the temporary file was not being removed, and the
4712 return code was incorrect.
4714 PH/31 Missing "BOOL" in function definition in filtertest.c.
4716 PH/32 Applied Sieve patches from the maintainer.
4718 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4720 PH/33 Added "verify = not_blind".
4722 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4723 Local/Makefile (with some defaults set). These are used in built scripts
4724 such as exicyclog, but they have never been used in the exim_install
4725 script (though there are many overriding facilities there). I have
4726 arranged that the exim_install script now takes note of these two
4729 PH/35 Installed configuration files for Dragonfly.
4731 PH/36 When a locally submitted message by a trusted user did not contain a
4732 From: header, and the sender address was obtained from -f or from an SMTP
4733 MAIL command, and the trusted user did not use -F to supply a sender
4734 name, $originator_name was incorrectly used when constructing a From:
4735 header. Furthermore, $originator_name was used for submission mode
4736 messages from external hosts without From: headers in a similar way,
4737 which is clearly wrong.
4739 PH/37 Added control=suppress_local_fixups.
4741 PH/38 When log_selector = +received_sender was set, and the addition of the
4742 sender made the log line's construction buffer exactly full, or one byte
4743 less than full, an overflow happened when the terminating "\n" was
4746 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4747 when the result of a list match is failure because a DNS lookup failed.
4749 PH/40 RM_COMMAND is now used in the building process.
4751 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4752 the "build-* directories that it finds.
4754 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4755 address in a domain literal was a prefix of an interface address.
4757 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4758 when verifying a sender address, unless rewrite_headers is false.
4760 PH/44 Wrote a long comment about why errors_to addresses are verified as
4761 recipients, not senders.
4763 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4764 the ratelimit ACL was added.
4766 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4768 PH/46 Added extra information about PostgreSQL errors to the error string.
4770 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4771 avoiding the need to install special zones in a real server. This is
4772 backwards compatible; if it can't find the fake resolver, it drops back.
4773 Thus, both old and new test suites can be run.
4775 TF/02 Added util/ratelimit.pl
4777 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4778 clock is set back in time.
4780 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4781 Candler <B.Candler@pobox.com>.
4783 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4784 OK for addresses that are the result of redirections.
4786 PH/48 A number of further additions for the benefit of the new test suite,
4787 including a fake gethostbyname() that interfaces to the fake DNS resolver
4790 TF/06 The fix for widen_domains has also been applied to qualify_single and
4791 search_parents which are the other dnslookup options that can cause
4794 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4797 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4799 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4800 service" if it encountered an MX record with an empty target hostname.
4801 The message is now "an MX or SRV record indicated no SMTP service".
4803 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4804 if the DNS of the sending site is misconfigured. This is quite a
4805 common situation. This change restores the behaviour of treating a
4806 helo verification defer as a failure.
4808 PH/51 If self=fail was set on a router, the bounce message did not include the
4809 actual error message.
4815 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4817 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4818 command itself ended in error, the underlying message about the transport
4819 filter timeout was being overwritten with the pipe command error. Now the
4820 underlying error message should be appended to the second error message.
4822 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4824 PH/02 Reduce the amount of output that "make" produces by default. Full output
4825 can still be requested.
4827 PH/03 The warning log line about a condition test deferring for a "warn" verb
4828 was being output only once per connection, rather than after each
4829 occurrence (because it was using the same function as for successful
4830 "warn" verbs). This seems wrong, so I have changed it.
4832 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4833 it should not have, which might have caused a crash in the right
4834 circumstances, but probably never did.
4836 PH/04 Installed a modified version of Tony Finch's patch to make submission
4837 mode fix the return path as well as the Sender: header line, and to
4838 add a /name= option so that you can make the user's friendly name appear
4841 TF/03 Added the control = fakedefer ACL modifier.
4843 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4844 Mark Lowes for thorough testing.
4846 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4848 TK/03 Merged latest SRS patch from Miles Wilton.
4850 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4851 with the definition in sysexits.h (which is #included earlier).
4852 Fortunately, Exim does not actually use EX_OK. The code used to try to
4853 preserve the sysexits.h value, by assuming that macro definitions were
4854 scanned for macro replacements. I have been disabused of this notion,
4855 so now the code just undefines EX_OK before #including unistd.h.
4857 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4858 in the smtp transport. When a block could not be written in a single
4859 write() function, the timeout was being re-applied to each part-write.
4860 This seems wrong - if the receiver was accepting one byte at a time it
4861 would take for ever. The timeout is now adjusted when this happens. It
4862 doesn't have to be particularly precise.
4864 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4865 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4867 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4868 without a domain if the check to <postmaster@domain> fails.
4870 SC/01 Eximstats: added -xls and the ability to specify output files
4871 (patch written by Frank Heydlauf).
4873 SC/02 Eximstats: use FileHandles for outputting results.
4875 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4877 SC/04 Eximstats: fixed display of large numbers with -nvr option
4879 SC/05 Eximstats: fixed merging of reports with empty tables.
4881 SC/06 Eximstats: added the -include_original_destination flag
4883 SC/07 Eximstats: removed tabs and trailing whitespace.
4885 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4887 TK/06 MBOX spool code: Add real "From " MBOX separator line
4888 so the .eml file is really in mbox format (even though
4889 most programs do not really care). Patch from Alex Miller.
4891 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4892 The latter is generated from $received_to and is only set if the
4893 message has one envelope recipient. SA can use these headers,
4894 obviously out-of-the-box. Patch from Alex Miller.
4896 PH/08 The ${def test on a variable was returning false if the variable's
4897 value was "0", contrary to what the specification has always said!
4898 The result should be true unless the variable is empty.
4900 PH/09 The syntax error of a character other than { following "${if
4901 def:variable_name" (after optional whitespace) was not being diagnosed.
4902 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4903 accidental colon was present, for example, could give incorrect results.
4905 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4906 result is used (not including appendfile, where other changes are about
4909 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4910 This involved changing a lot of size variables from int to off_t. It
4911 should work with maildirs and everything.
4913 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4914 spamd dying while we are connected to it.
4916 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4919 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4920 function for BDB 4.3.
4922 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4924 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4925 the sources for PCRE have been re-arranged and more files are now
4928 PH/15 The code I had for printing potentially long long variables in PH/11
4929 above was not the best (it lost precision). The length of off_t variables
4930 is now inspected at build time, and an appropriate printing format (%ld
4931 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4932 to be "long long int" or "long int". This is needed for the internal
4933 formatting function string_vformat().
4935 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4936 the configuration file to be ":syslog", then the script "guesses" where
4937 the logs files are, rather than using the compiled in default. In our
4938 case the guess is not the same as the compiled default, so the script
4939 suddenly stopped working when I started to use syslog. The patch checks
4940 to see if log_file_path is "". If so, it attempts to read it from exim
4941 with no configuration file to get the compiled in version, before it
4942 falls back to the previous guessing code."
4944 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4945 implementing BATV in an Exim configuration. See NewStuff for the gory
4948 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4949 Makefile that are specific to HP-UX.
4951 PH/18 If the "use_postmaster" option was set for a recipient callout together
4952 with the "random" option, the postmaster address was used as the MAIL
4953 FROM address for the random test, but not for the subsequent recipient
4954 test. It is now used for both.
4956 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4957 patch removes a few documentation additions to RFC 3028, because the
4958 latest draft now contains them. It adds the new en;ascii-case comparator
4959 and a new error check for 8bit text in MIME parts. Comparator and
4960 require names are now matched exactly. I enabled the subaddress
4961 extension, but it is not well tested yet (read: it works for me)."
4963 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4964 rework some of the code of TK/09 above to avoid the hardwired use of
4965 "%lld" and "long long". Replaced the call to snprintf() with a call to
4968 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4969 records point to non-existent hosts", "retry timeout exceeded", and
4970 "retry time not reached for any host after a long failure period".
4972 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4973 experimental DomainKeys support:
4975 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4976 (2) On an error such as an illegally used "control", the wrong name for
4977 the control was given.
4979 These problems did NOT occur unless DomainKeys support was compiled.
4981 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4983 PH/24 Added ${if match_ip condition.
4985 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4986 databases so that it will be absolutely obvious if a crash occurs in the
4987 DB library. This is a regular occurrence (often caused by mis-matched
4990 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4991 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4992 picked up on a user's system that detects such things. There doesn't seem
4993 to be a gcc warning option for this - only an attribute that has to be
4994 put on the function's prototype. It seems that in Fedora Core 4 they have
4995 set this on a number of new functions. No doubt there will be more in due
4998 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4999 specify a transport. However, if an address that was verified by such a
5000 router was the subject of a callout, Exim crashed because it tried to
5001 read the rcpt_include_affixes from the non-existent transport. Now it
5002 just assumes that the setting of that option is false. This bug was
5003 introduced by 4.51/PH/31.
5005 PH/28 Changed -d+all to exclude +memory, because that information is very
5006 rarely of interest, but it makes the output a lot bigger. People tend to
5007 do -d+all out of habit.
5009 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5010 code in os-type was giving problems when libc.so lives in lib64, like on
5013 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5014 aren't the modern standard, and it seems that some systems' include files
5015 don't always have them. Exim was already checking for some of the newer
5016 ones like T_AAAA, and defining it itself. I've added checks for all the
5017 record types that Exim uses.
5019 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5020 not automatically generating a new one, as it is supposed to. This
5021 prevented TLS from working. If the file did exist, but contained invalid
5022 data, a new version was generated, as expected. It was only the case of a
5023 non-existent file that was broken.
5025 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5026 with a change in libdomainkeys > 0.64.
5028 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5029 from DNS. If the selector record carries the flag, it now has
5030 precedence over the domain-wide flag.
5032 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5034 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5035 the use of a port name that isn't defined in /etc/services) occurred, the
5036 message was deferred as in a normal delivery, and thus remained on the
5037 spool, instead of being failed because of the mua_wrapper setting. This
5038 is now fixed, and I tidied up some of the mua_wrapper messages at the
5041 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5042 the messages in a hash of arrays rather than using individual hashes.
5043 This is a bit cleaner and results in dramatic memory savings, albeit
5044 at a slight CPU cost.
5046 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5047 as requested by Marc Sherman.
5049 SC/10 Eximstats: added histograms for user specified patterns as requested
5052 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5054 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5055 fopen() in the content-scanning modules that did not already have it.
5061 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5062 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5064 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5066 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5068 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5069 to test Sieve filters that use "vacation".
5071 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5072 that changes the way the GnuTLS parameters are stored in the cache file.
5073 The new format can be generated externally. For backward compatibility,
5074 if the data in the cache doesn't make sense, Exim assumes it has read an
5075 old-format file, and it generates new data and writes a new file. This
5076 means that you can't go back to an older release without removing the
5079 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5080 work if there are any delivery delays because "one_time" forces the
5081 parent to be marked "delivered", so its unseen clone is never tried
5082 again. For this reason, Exim now forbids the simultaneous setting of
5085 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5086 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5087 entry if you want to know the details. The fix, however, neglected to
5088 consider the case when local delivery batching is involved. The test for
5089 "previously delivered" was not happening when checking to see if an
5090 address could be batched with a previous (undelivered) one; under
5091 certain circumstances this could lead to multiple deliveries to the same
5094 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5095 in its include files, and this causes problems building Exim.
5097 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5098 header_syntax) but Exim was just ignoring anything given after a slash.
5099 In particular, this caused confusion with an attempt to use "verify =
5100 reverse_host_lookup/defer_ok". An error is now given when options are
5101 supplied for verify items that do not have them. (Maybe reverse_host_
5102 lookup should have a defer_ok option, but that's a different point.)
5104 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5105 defined by RFC 821) to 2048, because there were problems with some AUTH
5106 commands, and RFC 1869 says the size should be increased for extended
5107 SMTP commands that take arguments.
5109 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5112 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5113 "unknown" error; now it says that the functionality isn't in the binary.
5115 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5116 an address' error message when a string expansion fails (syntax or
5117 whatever). Otherwise the password may appear in the log. Following change
5118 PH/42 below, there is no longer a chance of it appearing in a bounce
5121 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5123 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5124 of its addresses was ever tried. (Bugzilla bug #2.)
5126 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5127 the result incorrectly in the debug output. (It correctly added a newline
5128 to what was transported.)
5130 TF/01 Added $received_time.
5132 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5133 commented out examples of how to interface to a virus scanner and to
5134 SpamAssassin. Also added commented examples of av_scanner and
5135 spamd_address settings.
5137 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5138 and controls are allowed in which ACLs. There were a couple of minor
5139 errors. Some of the entries in the conditions table (which is a table of
5140 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5141 negation of where the condition IS allowed.
5143 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5145 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5146 header file does not have a version number, so I've had to invent a new
5147 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5148 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5149 radiusclient), but it was contributed by a Radius user.
5151 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5152 files or directories, for OpenSSL.
5154 PH/19 When an Exim process that is running as root has to create an Exim log
5155 file, it does so in a subprocess that runs as exim:exim so as to get the
5156 ownership right at creation (otherwise, other Exim processes might see
5157 the file with the wrong ownership). There was no test for failure of this
5158 fork() call, which would lead to the process getting stuck as it waited
5159 for a non-existent subprocess. Forks do occasionally fail when resources
5160 run out. I reviewed all the other calls to fork(); they all seem to check
5163 PH/20 When checking for unexpected SMTP input at connect time (before writing
5164 the banner), Exim was not dealing correctly with a non-positive return
5165 from the read() function. If the client had disconnected by this time,
5166 the result was a log entry for a synchronization error with an empty
5167 string after "input=" when read() returned zero. If read() returned -1
5168 (an event I could not check), uninitialized data bytes were printed.
5169 There were reports of junk text (parts of files, etc) appearing after
5172 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5174 PH/22 Added support for macro redefinition, and (re)definition in between
5175 driver and ACL definitions.
5177 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5178 forgetting to use the resulting value; it was using the unexpanded value.
5180 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5181 hadn't been configured. The fix is from Juergen Kreileder, who
5182 understands it better than I do:
5184 "Here's what I see happening with three configured cyrus_sasl
5185 authenticators configured (plain, login, cram-md5):
5187 On startup auth_cyrus_sasl_init() gets called for each of these.
5188 This means three calls to sasl_listmech() without a specified mech_list.
5189 => SASL tests which mechs of all available mechs actually work
5190 => three warnings about OTP not working
5191 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5193 With the patch, sasl_listmech() also gets called three times. But now
5194 SASL's mech_list option is set to the server_mech specified in the the
5195 authenticator. Or in other words, the answer from sasl_listmech()
5196 gets limited to just the mech you're testing for (which is different
5198 => the return list contains just 'plain' or 'login', 'cram-md5' or
5199 nothing depending on the value of ob->server_mech.
5201 I've just tested the patch: Authentication still works fine,
5202 unavailable mechs specified in the exim configuration are still
5203 caught, and the auth.log warnings about OTP are gone."
5205 PH/25 When debugging is enabled, the contents of the command line are added
5206 to the debugging output, even when log_selector=+arguments is not
5209 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5210 answer is "GNU", and only if the return is "GNU/something" is the answer
5213 PH/27 $acl_verify_message is now set immediately after the failure of a
5214 verification in an ACL, and so is available in subsequent modifiers. In
5215 particular, the message can be preserved by coding like this:
5217 warn !verify = sender
5218 set acl_m0 = $acl_verify_message
5220 Previously, $acl_verify_message was set only while expanding "message"
5221 and "log_message" when a very denied access.
5223 PH/28 Modified OS/os.c-Linux with
5225 -#ifndef OS_LOAD_AVERAGE
5226 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5228 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5229 nomenclature these days.)
5231 PH/29 Installed patch from the Sieve maintainer that adds the options
5232 sieve_useraddress and sieve_subaddress to the redirect router.
5234 PH/30 In these circumstances:
5235 . Two addresses routed to the same list of hosts;
5236 . First host does not offer TLS;
5237 . First host accepts first address;
5238 . First host gives temporary error to second address;
5239 . Second host offers TLS and a TLS session is established;
5240 . Second host accepts second address.
5241 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5242 and peerdn, if requested) that were in fact used only for the second
5245 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5246 attention to any local part prefix or suffix that was matched by the
5247 router that accepted the address. It now behaves in the same way as it
5248 does for delivery: the affixes are removed from the local part unless
5249 rcpt_include_affixes is set on the transport.
5251 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5252 timeout during the DATA phase of an incoming message.
5254 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5255 applied a patch sanctioned by the Sieve maintainer.
5257 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5258 the uid or gid is negative. A case of a negative gid caused this to be
5259 noticed. The fix allows for either to be negative.
5261 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5262 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5265 PH/36 The change PH/12 above was broken. Fixed it.
5267 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5268 the grounds that routing the same address twice would always produce the
5269 same answer. This might have been true once, but it is certainly no
5270 longer true now. Routing a child address may depend on the previous
5271 routing that produced that child. Some complicated redirection strategies
5272 went wrong when messages had multiple recipients, and made Exim's
5273 behaviour dependent on the order in which the addresses were given.
5275 I have moved the duplicate checking until after the routing is complete.
5276 Exim scans the addresses that are assigned to local and remote
5277 transports, and removes any duplicates. This means that more work will be
5278 done, as duplicates will always all be routed, but duplicates are
5279 presumably rare, so I don't expect this is of any significance.
5281 For deliveries to pipes, files, and autoreplies, the duplicate checking
5282 still happens during the routing process, since they are not going to be
5285 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5286 It corrects a timeout issue with spamd. This is Ian's comment: "The
5287 background is that sometimes spamd either never reads data from a
5288 connection it has accepted, or it never writes response data. The exiscan
5289 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5290 blindly assumes that writes won't block so it may never time out."
5292 PH/39 Allow G after quota size as well as K and M.
5294 PH/40 The value set for $authenticated_id in an authenticator may not contain
5295 binary zeroes or newlines because the value is written to log lines and
5296 to spool files. There was no check on this. Now the value is run through
5297 the string_printing() function so that such characters are converted to
5298 printable escape sequences.
5300 PH/41 $message_linecount is a new variable that contains the total number of
5301 lines in the message. Compare $body_linecount, which is the count for the
5304 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5305 bounce and delay warning messages, except in certain special cases, which
5308 (a) An SMTP error message from a remote host;
5309 (b) A message specified in a :fail: redirection;
5310 (c) A message specified in a "fail" command in a system filter;
5311 (d) A message specified in a FAIL return from the queryprogram router;
5312 (e) A message specified by the cannot_route_message router option.
5314 In these cases only, Exim does include the error details in bounce and
5315 warning messages. There are also a few cases where bland messages such
5316 as "unrouteable address" or "local delivery error" are given.
5318 PH/43 $value is now also set for the "else" part of a ${run expansion.
5320 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5321 being worked on, but at least Exim now implements the latest version to
5324 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5325 process to complete was treated as a delivery failure, a timeout while
5326 writing the message to the pipe was logged, but erroneously treated as a
5327 successful delivery. Such timeouts include transport filter timeouts. For
5328 consistency with the overall process timeout, these timeouts are now
5329 treated as errors, giving rise to delivery failures by default. However,
5330 there is now a new Boolean option for the pipe transport called
5331 timeout_defer, which, if set TRUE, converts the failures into defers for
5332 both kinds of timeout. A transport filter timeout is now identified in
5335 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5336 systems where "make" and "gmake" are different, calling "gmake" at top
5337 level broke things. I've arranged for the value of $(MAKE) to be passed
5338 from the Makefile to this script so that it can call the same version of
5342 A note about Exim versions 4.44 and 4.50
5343 ----------------------------------------
5345 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5346 changes of various kinds. As a consequence, a big documentation update was
5347 needed. This delayed the release for rather longer than seemed good, especially
5348 in the light of a couple of (minor) security issues. Therefore, the changes
5349 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5350 release. So 4.44 and 4.50 are in effect two different branches that both start
5353 I have left the 4.50 change log unchanged; it contains all the changes since
5354 4.43. The change log for 4.44 is below; many of its items are identical to
5355 those for 4.50. This seems to be the most sensible way to preserve the
5356 historical information.
5362 1. Minor wording change to the doc/README.SIEVE file.
5364 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5365 computation of the current number of files was incorrect.
5367 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5368 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5371 4. Give more explanation in the error message when the command for a transport
5372 filter fails to execute.
5374 5. There are several places where Exim runs a non-Exim command in a
5375 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5376 was being done only for the command run by the queryprogram router. It is
5377 now done for all such subprocesses. The other cases are: ${run, transport
5378 filters, and the commands run by the lmtp and pipe transports.
5380 6. Added CONFIGURE_GROUP build-time option.
5382 7. Some older OS have a limit of 256 on the maximum number of file
5383 descriptors. Exim was using setrlimit() to set 1000 as a large value
5384 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5385 systems. I've change it so that if it can't get 1000, it tries for 256.
5387 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5388 was an oversight, and furthermore, ever since the addition of extra
5389 controls (e.g. 4.43/32), the checks on when to allow different forms of
5390 "control" were broken. There should now be diagnostics for all cases when a
5391 control that does not make sense is encountered.
5393 9. Added the /retain_sender option to "control=submission".
5395 10. $recipients is now available in the predata ACL (oversight).
5397 11. Tidy the search cache before the fork to do a delivery from a message
5398 received from the command line. Otherwise the child will trigger a lookup
5399 failure and thereby defer the delivery if it tries to use (for example) a
5400 cached ldap connection that the parent has called unbind on.
5402 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5403 of $address_data from the recipient verification was clobbered by the
5404 sender verification.
5406 13. The value of address_data from a sender verification is now available in
5407 $sender_address_data in subsequent conditions in the ACL statement.
5409 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5411 15. Added a new option "connect=<time>" to callout options, to set a different
5414 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5415 was its contents. (It was OK if the option was not defined at all.)
5417 17. A "Completed" log line is now written for messages that are removed from
5418 the spool by the -Mrm option.
5420 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5421 information about exactly what failed.
5423 19. Added -dd to debug only the daemon process.
5425 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5426 handles timeouts, both on the server side and network timeouts. Renamed the
5427 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5429 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5430 It is now set to "smtps".
5432 22. $host_address is now set to the target address during the checking of
5433 ignore_target_hosts.
5435 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5436 being passed; this would have caused $sender_host_name to have been used if
5437 matching the list had actually called for a host name (not very likely,
5438 since this list is usually IP addresses). A host name is now passed as
5441 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5442 code that specifies a non-restarting handler (typically sigaction() in
5443 modern systems) in an attempt to fix a rare and obscure crash bug.
5445 25. Narrowed the window for a race in the daemon that could cause it to ignore
5446 SIGCHLD signals. This is not a major problem, because they are used only to
5447 wake it up if nothing else does.
5449 26. A malformed maildirsize file could cause Exim to calculate negative values
5450 for the mailbox size or file count. Odd effects could occur as a result.
5451 The maildirsize information is now recalculated if the size or filecount
5454 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5455 support for a long time. Removed HAVE_SYS_VFS_H.
5457 28. Installed the latest version of exipick from John Jetmore.
5459 29. In an address list, if the pattern was not a regular expression, an empty
5460 subject address (from a bounce message) matched only if the pattern was an
5461 empty string. Non-empty patterns were not even tested. This was the wrong
5462 because it is perfectly reasonable to use an empty address as part of a
5463 database query. An empty address is now tested by patterns that are
5464 lookups. However, all the other forms of pattern expect the subject to
5465 contain a local part and a domain, and therefore, for them, an empty
5466 address still always fails if the pattern is not itself empty.
5468 30. Exim went into a mad DNS loop when attempting to do a callout where the
5469 host was specified on an smtp transport, and looking it up yielded more
5470 than one IP address.
5472 31. Re-factored the code for checking spool and log partition space into a
5473 function that finds that data and another that does the check. The former
5474 is then used to implement four new variables: $spool_space, $log_space,
5475 $spool_inodes, and $log_inodes.
5477 32. The RFC2047 encoding function was originally intended for short strings
5478 such as real names; it was not keeping to the 75-character limit for
5479 encoded words that the RFC imposes. It now respects the limit, and
5480 generates multiple encoded words if necessary. To be on the safe side, I
5481 have increased the buffer size for the ${rfc2047: expansion operator from
5484 33. It is now permitted to omit both strings after an "if" condition; if the
5485 condition is true, the result is "true". As before, when the second string
5486 is omitted, a false condition yields an empty string. This makes it less
5487 cumbersome to write custom ACL and router conditions.
5489 34. Failure to deliver a bounce message always caused it to be frozen, even if
5490 there was an errors_to setting on the router. The errors_to setting is now
5493 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5494 canonical form (fully expanded) before being placed in
5495 $sender_host_address.
5497 36. The table in the code that translates DNS record types into text (T_A to
5498 "A" for instance) was missing entries for NS and CNAME. It is just possible
5499 that this could have caused confusion if both these types were looked up
5500 for the same domain, because the text type is used as part of Exim's
5501 per-process caching. But the chance of anyone hitting this buglet seems
5504 37. The dnsdb lookup has been extended in a number of ways.
5506 (1) There is a new type, "zns", which walks up the domain tree until it
5507 finds some nameserver records. It should be used with care.
5509 (2) There is a new type, "mxh", which is like "mx" except that it returns
5510 just the host names, not the priorities.
5512 (3) It is now possible to give a list of domains (or IP addresses) to be
5513 looked up. The behaviour when one of the lookups defers can be
5514 controlled by a keyword.
5516 (4) It is now possible to specify the separator character for use when
5517 multiple records are returned.
5519 38. The dnslists ACL condition has been extended: it is now possible to supply
5520 a list of IP addresses and/or domains to be looked up in a particular DNS
5523 39. Added log_selector=+queue_time_overall.
5525 40. When running the queue in the test harness, wait just a tad after forking a
5526 delivery process, to get repeatability of debugging output.
5528 41. Include certificate and key file names in error message when GnuTLS fails
5529 to set them up, because the GnuTLS error message doesn't include the name
5530 of the failing file when there is a problem reading it.
5532 42. Allow both -bf and -bF in the same test run.
5534 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5536 44. The "Exiscan patch" is now merged into the mainline Exim source.
5538 45. Sometimes the final signoff response after QUIT could fail to get
5539 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5540 < 0 before doing a fflush(). This bug looks as though it goes back to the
5541 introduction of TLS in release 3.20, but "sometimes" must have been rare
5542 because the tests only now provoked it.
5544 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5545 (this can affect the format of dates).
5547 47. exim_tidydb, when checking for the continued existence of a message for
5548 which it has found a message-specific retry record, was not finding
5549 messages that were in split spool directories. Consequently, it was
5550 deleting retry records that should have stayed in existence.
5552 48. Steve fixed some bugs in eximstats.
5554 49. The SPA authentication driver was not abandoning authentication and moving
5555 on to the next authenticator when an expansion was forced to fail,
5556 contradicting the general specification for all authenticators. Instead it
5557 was generating a temporary error. It now behaves as specified.
5559 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5560 (the order specifies the preference for clients). The order is now AES256,
5561 AES128, 3DES, ARCFOUR128.
5563 51. Small patch to Sieve code - explicitly set From: when generating an
5566 52. Exim crashed if a remote delivery caused a very long error message to be
5567 recorded - for instance if somebody sent an entire SpamAssassin report back
5568 as a large number of 550 error lines. This bug was coincidentally fixed by
5569 increasing the size of one of Exim's internal buffers (big_buffer) that
5570 happened as part of the Exiscan merge. However, to be on the safe side, I
5571 have made the code more robust (and fixed the comments that describe what
5574 53. Now that there can be additional text after "Completed" in log lines (if
5575 the queue_time_overall log selector is set), a one-byte patch to exigrep
5576 was needed to allow it to recognize "Completed" as not the last thing in
5579 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5580 patch that reportedly fixes this has been added. I am not expert enough to
5581 create a test for it. This is what the patch creator wrote:
5583 "I found a little strange behaviour of ldap code when working with
5584 Windows 2003 AD Domain, where users was placed in more than one
5585 Organization Units. When I tried to give exim partial DN, the exit code
5586 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5587 But simultaneously result of request was absolutely normal ldap result,
5588 so I produce this patch..."
5590 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5591 REFERENCE, so I have modified the code to exclude the patch when that macro
5594 55. Some experimental protocols are using DNS PTR records for new purposes. The
5595 keys for these records are domain names, not reversed IP addresses. The
5596 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5597 leaves it alone. Component reversal etc. now happens only for IP addresses.
5600 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5602 57. Double the size of the debug message buffer (to 2048) so that more of very
5603 long debug lines gets shown.
5605 58. The exicyclog utility now does better if the number of log files to keep
5606 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5608 59. Two changes related to the smtp_active_hostname option:
5610 (1) $smtp_active_hostname is now available as a variable.
5611 (2) The default for smtp_banner uses $smtp_active_hostname instead
5612 of $primary_hostname.
5614 60. The host_aton() function is supposed to be passed a string that is known
5615 to be a valid IP address. However, in the case of IPv6 addresses, it was
5616 not checking this. This is a hostage to fortune. Exim now panics and dies
5617 if the condition is not met. A case was found where this could be provoked
5618 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5619 components; fortuitously, this particular loophole had already been fixed
5620 by change 4.50/55 above.
5622 If there are any other similar loopholes, the new check in host_aton()
5623 itself should stop them being exploited. The report I received stated that
5624 data on the command line could provoke the exploit when Exim was running as
5625 exim, but did not say which command line option was involved. All I could
5626 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5627 running as the user.
5630 61. There was a buffer overflow vulnerability in the SPA authentication code
5631 (which came originally from the Samba project). I have added a test to the
5632 spa_base64_to_bits() function which I hope fixes it.
5635 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5636 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5638 63. The daemon start-up calls getloadavg() while still root for those OS that
5639 need the first call to be done as root, but it missed one case: when
5640 deliver_queue_load_max is set with deliver_drop_privilege. This is
5641 necessary for the benefit of the queue runner, because there is no re-exec
5642 when deliver_drop_privilege is set.
5644 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5645 This has been fixed.
5647 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5648 was in use, was not putting the data itself into the right store pool;
5649 consequently, it could be overwritten for a subsequent message in the same
5650 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5653 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5655 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5656 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5657 the function were treating the return as a boolean value, which happened to
5658 work because 0=false and not-0=true, but is not correct code.
5660 68. The host_aton() function was not handling scoped IPv6 addresses (those
5661 with, for example, "%eth0" on the end) correctly.
5663 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5664 negated items (that is, ~something) in unsigned ints. Some compilers
5665 apparently mutter when there is no cast.
5667 70. If an address verification called from an ACL failed, and did not produce a
5668 user-specific message (i.e. there was only a "system" message), nothing was
5669 put in $acl_verify_message. In this situation, it now puts the system
5672 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5673 session; change 4.31/43 added the unwanted input to the log line - except
5674 that it did not do this in the start of session case. It now does.
5676 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5677 This is wrong and can cause the other end to generate a synchronization
5678 error if it is another Exim or anything else that does the synchronization
5679 check. A QUIT command is no longer sent after a timeout.
5681 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5682 during host lookups.
5684 74. The defer_ok option of callout verification was not working if it was used
5685 when verifying addresses in header lines, that is, for this case:
5687 verify = header_sender/callout=defer_ok
5689 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5690 those file descriptors could be used for SMTP connections. If anything
5691 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5692 could be sent to the SMTP client, causing chaos. The daemon now opens
5693 stdin, stdout, and stderr to /dev/null when it puts itself into the
5696 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5697 log by default. The user can override this with suitable Perl magic.
5699 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5700 the log message when discard triggers, was not working for the DATA ACL or
5701 for the non-SMTP ACL.
5703 78. Error message wording change in sieve.c.
5705 79. If smtp_accept_max_per_host was set, the number of connections could be
5706 restricted to fewer than expected, because the daemon was trying to set up
5707 a new connection before checking whether the processes handling previous
5708 connections had finished. The check for completed processes is now done
5709 earlier. On busy systems, this bug wouldn't be noticed because something
5710 else would have woken the daemon, and it would have reaped the completed
5713 80. If a message was submitted locally by a user whose login name contained one
5714 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5715 It caused a spool format error. I have fixed the spool reading code. A
5716 related problem was that the "from" clause in the Received: line became
5717 illegal because of the space(s). It is now covered by ${quote_local_part.
5719 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5722 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5724 83. Patch from Sieve maintainer to fix unterminated string problem in
5725 "vacation" handling.
5727 84. Some minor changes to the Linux configuration files to help with other
5728 OS variants using glibc.
5730 85. One more patch for Sieve to update vacation handling to latest spec.
5733 ----------------------------------------------------
5734 See the note above about the 4.44 and 4.50 releases.
5735 ----------------------------------------------------
5741 1. Change 4.43/35 introduced a bug that caused file counts to be
5742 incorrectly computed when quota_filecount was set in an appendfile
5745 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5746 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5749 3. Give more explanation in the error message when the command for a transport
5750 filter fails to execute.
5752 4. There are several places where Exim runs a non-Exim command in a
5753 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5754 was being done only for the command run by the queryprogram router. It is
5755 now done for all such subprocesses. The other cases are: ${run, transport
5756 filters, and the commands run by the lmtp and pipe transports.
5758 5. Some older OS have a limit of 256 on the maximum number of file
5759 descriptors. Exim was using setrlimit() to set 1000 as a large value
5760 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5761 systems. I've change it so that if it can't get 1000, it tries for 256.
5763 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5764 was an oversight, and furthermore, ever since the addition of extra
5765 controls (e.g. 4.43/32), the checks on when to allow different forms of
5766 "control" were broken. There should now be diagnostics for all cases when a
5767 control that does not make sense is encountered.
5769 7. $recipients is now available in the predata ACL (oversight).
5771 8. Tidy the search cache before the fork to do a delivery from a message
5772 received from the command line. Otherwise the child will trigger a lookup
5773 failure and thereby defer the delivery if it tries to use (for example) a
5774 cached ldap connection that the parent has called unbind on.
5776 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5777 of $address_data from the recipient verification was clobbered by the
5778 sender verification.
5780 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5781 was its contents. (It was OK if the option was not defined at all.)
5783 11. A "Completed" log line is now written for messages that are removed from
5784 the spool by the -Mrm option.
5786 12. $host_address is now set to the target address during the checking of
5787 ignore_target_hosts.
5789 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5790 being passed; this would have caused $sender_host_name to have been used if
5791 matching the list had actually called for a host name (not very likely,
5792 since this list is usually IP addresses). A host name is now passed as
5795 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5796 code that specifies a non-restarting handler (typically sigaction() in
5797 modern systems) in an attempt to fix a rare and obscure crash bug.
5799 15. Narrowed the window for a race in the daemon that could cause it to ignore
5800 SIGCHLD signals. This is not a major problem, because they are used only to
5801 wake it up if nothing else does.
5803 16. A malformed maildirsize file could cause Exim to calculate negative values
5804 for the mailbox size or file count. Odd effects could occur as a result.
5805 The maildirsize information is now recalculated if the size or filecount
5808 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5809 support for a long time. Removed HAVE_SYS_VFS_H.
5811 18. Updated exipick to current release from John Jetmore.
5813 19. Allow an empty sender to be matched against a lookup in an address list.
5814 Previously the only cases considered were a regular expression, or an
5817 20. Exim went into a mad DNS lookup loop when doing a callout where the
5818 host was specified on the transport, if the DNS lookup yielded more than
5821 21. The RFC2047 encoding function was originally intended for short strings
5822 such as real names; it was not keeping to the 75-character limit for
5823 encoded words that the RFC imposes. It now respects the limit, and
5824 generates multiple encoded words if necessary. To be on the safe side, I
5825 have increased the buffer size for the ${rfc2047: expansion operator from
5828 22. Failure to deliver a bounce message always caused it to be frozen, even if
5829 there was an errors_to setting on the router. The errors_to setting is now
5832 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5833 canonical form (fully expanded) before being placed in
5834 $sender_host_address.
5836 24. Updated eximstats to version 1.33
5838 25. Include certificate and key file names in error message when GnuTLS fails
5839 to set them up, because the GnuTLS error message doesn't include the name
5840 of the failing file when there is a problem reading it.
5842 26. Expand error message when OpenSSL has problems setting up cert/key files.
5845 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5846 (this can affect the format of dates).
5848 28. exim_tidydb, when checking for the continued existence of a message for
5849 which it has found a message-specific retry record, was not finding
5850 messages that were in split spool directories. Consequently, it was
5851 deleting retry records that should have stayed in existence.
5853 29. eximstats updated to version 1.35
5854 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5855 1.35 - bugfix such that pie charts by volume are generated correctly
5857 30. The SPA authentication driver was not abandoning authentication and moving
5858 on to the next authenticator when an expansion was forced to fail,
5859 contradicting the general specification for all authenticators. Instead it
5860 was generating a temporary error. It now behaves as specified.
5862 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5863 (the order specifies the preference for clients). The order is now AES256,
5864 AES128, 3DES, ARCFOUR128.
5866 31. Small patch to Sieve code - explicitly set From: when generating an
5869 32. Exim crashed if a remote delivery caused a very long error message to be
5870 recorded - for instance if somebody sent an entire SpamAssassin report back
5871 as a large number of 550 error lines. This bug was coincidentally fixed by
5872 increasing the size of one of Exim's internal buffers (big_buffer) that
5873 happened as part of the Exiscan merge. However, to be on the safe side, I
5874 have made the code more robust (and fixed the comments that describe what
5877 33. Some experimental protocols are using DNS PTR records for new purposes. The
5878 keys for these records are domain names, not reversed IP addresses. The
5879 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5880 leaves it alone. Component reversal etc. now happens only for IP addresses.
5883 34. The host_aton() function is supposed to be passed a string that is known
5884 to be a valid IP address. However, in the case of IPv6 addresses, it was
5885 not checking this. This is a hostage to fortune. Exim now panics and dies
5886 if the condition is not met. A case was found where this could be provoked
5887 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5888 components; fortuitously, this particular loophole had already been fixed
5889 by change 4.50/55 or 4.44/33 above.
5891 If there are any other similar loopholes, the new check in host_aton()
5892 itself should stop them being exploited. The report I received stated that
5893 data on the command line could provoke the exploit when Exim was running as
5894 exim, but did not say which command line option was involved. All I could
5895 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5896 running as the user.
5899 35. There was a buffer overflow vulnerability in the SPA authentication code
5900 (which came originally from the Samba project). I have added a test to the
5901 spa_base64_to_bits() function which I hope fixes it.
5904 36. The daemon start-up calls getloadavg() while still root for those OS that
5905 need the first call to be done as root, but it missed one case: when
5906 deliver_queue_load_max is set with deliver_drop_privilege. This is
5907 necessary for the benefit of the queue runner, because there is no re-exec
5908 when deliver_drop_privilege is set.
5910 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5911 was in use, was not putting the data itself into the right store pool;
5912 consequently, it could be overwritten for a subsequent message in the same
5913 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5916 38. Sometimes the final signoff response after QUIT could fail to get
5917 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5918 < 0 before doing a fflush(). This bug looks as though it goes back to the
5919 introduction of TLS in release 3.20, but "sometimes" must have been rare
5920 because the tests only now provoked it.
5926 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5927 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5928 Now it returns an integer. A number of calls were still expecting a T/F
5929 return. Fortuitously, in all cases, the tests worked in OK situations,
5930 which is the norm. However, things would have gone wrong on any write
5931 failures on the smtp file descriptor. This function is used when sending
5932 messages over SMTP and also when doing verify callouts.
5934 2. When Exim is called to do synchronous delivery of a locally submitted
5935 message (the -odf or -odi options), it no longer closes stderr before doing
5938 3. Implemented the mua_wrapper option.
5940 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5942 5. Implemented the functions header_remove(), header_testname(),
5943 header_add_at_position(), and receive_remove_recipient(), and exported them
5946 6. If an ACL "warn" statement specified the addition of headers, Exim already
5947 inserted X-ACL-Warn: at the start if there was no header name. However, it
5948 was not making this test for the second and subsequent header lines if
5949 there were newlines in the string. This meant that an invalid header could
5950 be inserted if Exim was badly configured.
5952 7. Allow an ACL "warn" statement to add header lines at the start or after all
5953 the Received: headers, as well as at the end.
5955 8. Added the rcpt_4xx retry error code.
5957 9. Added postmaster_mailfrom=xxx to callout verification option.
5959 10. Added mailfrom=xxxx to the callout verification option, for verify=
5962 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5963 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5965 12. Inserted some casts to stop certain compilers warning when using pointer
5966 differences as field lengths or precisions in printf-type calls (mostly
5967 affecting debugging statements).
5969 13. Added optional readline() support for -be (dynamically loaded).
5971 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5972 same clock tick as a message's arrival, so that its received time was the
5973 same as the "first fail" time on the retry record, and that message
5974 remained on the queue past the ultimate address timeout, every queue runner
5975 would try a delivery (because it was past the ultimate address timeout) but
5976 after another failure, the ultimate address timeout, which should have then
5977 bounced the address, did not kick in. This was a "< instead of <=" error;
5978 in most cases the first failure would have been in the next clock tick
5979 after the received time, and all would be well.
5981 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5982 being recognized when the domain list was tested by the match_domain
5983 condition in an expansion string.
5985 16. Added the ${str2b64: operator.
5987 17. Exim was always calling setrlimit() to set a large limit for the number of
5988 processes, without checking whether the existing limit was already
5989 adequate. (It did check for the limit on file descriptors.) Furthermore,
5990 errors from getrlimit() and setrlimit() were being ignored. Now they are
5991 logged to the main and panic logs, but Exim does carry on, to try to do its
5992 job under whatever limits there are.
5994 18. Imported PCRE 5.0.
5996 19. Trivial typo in log message " temporarily refused connection" (the leading
5999 20. If the log selector return_path_on_delivery was set and an address was
6000 redirected to /dev/null, the delivery process crashed because it assumed
6001 that a return path would always be set for a "successful" delivery. In this
6002 case, the whole delivery is bypassed as an optimization, and therefore no
6005 21. Internal re-arrangement: the function for sending a challenge and reading
6006 a response while authentication was assuming a zero-terminated challenge
6007 string. It's now changed to take a pointer and a length, to allow for
6008 binary data in such strings.
6010 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6012 23. Exim was not respecting finduser_retries when seeking the login of the
6013 uid under which it was called; it was always trying 10 times. (The default
6014 setting of finduser_retries is zero.) Also, it was sleeping after the final
6015 failure, which is pointless.
6017 24. Implemented tls_on_connect_ports.
6019 25. Implemented acl_smtp_predata.
6021 26. If the domain in control=submission is set empty, Exim assumes that the
6022 authenticated id is a complete email address when it generates From: or
6023 Sender: header lines.
6025 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6026 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6027 chown and chgrp in /bin and hostname in /usr/bin.
6029 28. Exim was keeping the "process log" file open after each use, just as it
6030 does for the main log. This opens the possibility of it remaining open for
6031 long periods when the USR1 signal hits a daemon. Occasional processlog
6032 errors were reported, that could have been caused by this. Anyway, it seems
6033 much more sensible not to leave this file open at all, so that is what now
6036 29. The long-running daemon process does not normally write to the log once it
6037 has entered its main loop, and it closes the log before doing so. This is
6038 so that log files can straightforwardly be renamed and moved. However,
6039 there are a couple of unusual error situations where the daemon does write
6040 log entries, and I had neglected to close the log afterwards.
6042 30. The text of an SMTP error response that was received during a remote
6043 delivery was being truncated at 512 bytes. This is too short for some of
6044 the long messages that one sometimes sees. I've increased the limit to
6047 31. It is now possible to make retry rules that apply only when a message has a
6048 specific sender, in particular, an empty sender.
6050 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6051 it possible to be selective about when SMTP synchronization is enforced.
6053 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6055 32. Implemented hosts_connection_nolog.
6057 33. Added an ACL for QUIT.
6059 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6062 35. Added mailbox_size and mailbox_filecount to appendfile.
6064 36. Added control = no_multiline_responses to ACLs.
6066 37. There was a bug in the logic of the code that waits for the clock to tick
6067 in the case where the clock went backwards by a substantial amount such
6068 that the microsecond fraction of "now" was more than the microsecond
6069 fraction of "then" (but the whole seconds number was less).
6071 38. Added support for the libradius Radius client library this is found on
6072 FreeBSD (previously only the radiusclient library was supported).
6078 1. When certain lookups returned multiple values in the form name=value, the
6079 quoting of the values was not always being done properly. Specifically:
6080 (a) If the value started with a double quote, but contained no whitespace,
6082 (b) If the value contained whitespace other than a space character (i.e.
6083 tabs or newlines or carriage returns) it was not quoted.
6084 This fix has been applied to the mysql and pgsql lookups by writing a
6085 separate quoting function and calling it from the lookup code. The fix
6086 should probably also be applied to nisplus, ibase and oracle lookups, but
6087 since I cannot test any of those, I have not disturbed their existing code.
6089 2. A hit in the callout cache for a specific address caused a log line with no
6090 reason for rejecting RCPT. Now it says "Previous (cached) callout
6091 verification failure".
6093 3. There was an off-by-one bug in the queryprogram router. An over-long
6094 return line was truncated at 256 instead of 255 characters, thereby
6095 overflowing its buffer with the terminating zero. As well as fixing this, I
6096 have increased the buffer size to 1024 (and made a note to document this).
6098 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6099 when Exim is waiting for an SMTP response from a remote server, Exim
6100 restarts its select() call on the socket, thereby resetting its timeout.
6101 This is not a problem when such interrupts are rare. Somebody set up a cron
6102 job to run exiwhat every 2 minutes, which is less than the normal select()
6103 timeout (5 or 10 minutes). This meant that the select() timeout never
6104 kicked in because it was always reset. I have fixed this by comparing the
6105 time when an interrupt arrives with the time at the start of the first call
6106 to select(). If more time than the timeout has elapsed, the interrupt is
6107 treated as a timeout.
6109 5. Some internal re-factoring in preparation for the addition of Sieve
6110 extensions (by MH). In particular, the "personal" test is moved to a
6111 separate function, and given an option for scanning Cc: and Bcc: (which is
6112 not set for Exim filters).
6114 6. When Exim created an email address using the login of the caller as the
6115 local part (e.g. when creating a From: or Sender: header line), it was not
6116 quoting the local part when it contained special characters such as @.
6118 7. Installed new OpenBSD configuration files.
6120 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6121 try to make them clearer.
6123 9. Callout options, other than the timeout value, were being ignored when
6124 verifying sender addresses in header lines. For example, when using
6126 verify = header_sender/callout=no_cache
6128 the cache was (incorrectly) being used.
6130 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6131 only the Cygwin environment.
6133 11. When return_path_on_delivery was set as a log selector, if different remote
6134 addresses in the same message used different return paths and parallel
6135 remote delivery occurred, the wrong values would sometimes be logged.
6136 (Whenever a remote delivery process finished, the return path value from
6137 the most recently started remote delivery process was logged.)
6139 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6140 lines when AUTH and/or TLS are in use. This is the "received protocol"
6141 field. Exim used to use "asmtp" for authenticated SMTP, without any
6142 indication (in the protocol name) for TLS use. Now it follows the RFC and
6143 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6144 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6145 names appear in log lines as well as in Received: header lines.
6147 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6148 extensions, and comparison tests, and to fix some bugs.
6150 14. Changes to the "personal" filter test:
6152 (1) The test was buggy in that it was just doing the equivalent of
6153 "contains" tests on header lines. For example, if a user's address was
6154 anne@some.where, the "personal" test would incorrectly be true for
6156 To: susanne@some.where
6158 This test is now done by extracting each address from the header in turn,
6159 and checking the entire address. Other tests that are part of "personal"
6160 are now done using regular expressions (for example, to check local parts
6161 of addresses in From: header lines).
6163 (2) The list of non-personal local parts in From: addresses has been
6164 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6165 taken from the Sieve specification recommendations.
6167 (3) If the message contains any header line starting with "List-" it is
6168 treated as non-personal.
6170 (4) The test for "circular" in the Subject: header line has been removed
6171 because it now seems ill-conceived.
6173 15. Minor typos in src/EDITME comments corrected.
6175 16. Installed latest exipick from John Jetmore.
6177 17. If headers_add on a router specified a text string that was too long for
6178 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6179 of string_sprintf() is now avoided.
6181 18. $message_body_size was not set (it was always zero) when running the DATA
6182 ACL and the local_scan() function.
6184 19. For the "mail" command in an Exim filter, no default was being set for
6185 the once_repeat time, causing a random time value to be used if "once" was
6186 specified. (If the value happened to be <= 0, no repeat happened.) The
6187 default is now 0s, meaning "never repeat". The "vacation" command was OK
6188 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6189 (I found it when inspecting the code).
6191 20. There is now an overall timeout for performing a callout verification. It
6192 defaults to 4 times the callout timeout, which applies to individual SMTP
6193 commands during the callout. The overall timeout applies when there is more
6194 than one host that can be tried. The timeout is checked before trying the
6195 next host. This prevents very long delays if there are a large number of
6196 hosts and all are timing out (e.g. when the network connections are timing
6197 out). The value of the overall timeout can be changed by specifying an
6198 additional sub-option for "callout", called "maxwait". For example:
6200 verify = sender/callout=5s,maxwait=20s
6202 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6203 to the end before writing, but this should make it even safer).
6205 22. Exim was forgetting that it had advertised PIPELINING for the second and
6206 subsequent messages on an SMTP connection. It was also not resetting its
6207 memory on STARTTLS and an internal HELO.
6209 23. When Exim logs an SMTP synchronization error within a session, it now
6210 records whether PIPELINING has been advertised or not.
6212 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6213 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6214 rather than long int.
6216 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6218 26. Added the never_mail option to autoreply.
6224 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6225 crash if the getsockname() call failed; this can happen if a connection is
6226 closed very soon after it is established. The problem was simply in the
6227 order in which certain operations were done, causing Exim to try to write
6228 to the SMTP stream before it had set up the file descriptor. The bug has
6229 been fixed by making things happen in the correct order.
6235 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6236 before the connection was closed, thus losing the rejection response.
6238 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6239 some early Solaris releases, but causes trouble in current releases where
6240 socklen_t is defined.
6242 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6245 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6248 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6249 that releasing the top of it at the end releases what was used for sub-
6250 expansions (unless the block got too big). However, discard this block if
6251 the first thing is a variable or header, so that we can use its block when
6252 it is dynamic (useful for very large $message_headers, for example).
6254 6. Lookups now cache *every* query, not just the most recent. A new, separate
6255 store pool is used for this. It can be recovered when all lookup caches are
6256 flushed. Lookups now release memory at the end of their result strings.
6257 This has involved some general refactoring of the lookup sources.
6259 7. Some code has been added to the store_xxx() functions to reduce the amount
6260 of flapping under certain conditions.
6262 8. log_incoming_interface used to affect only the <= reception log lines. Now
6263 it causes the local interface and port to be added to several more SMTP log
6264 lines, for example "SMTP connection from", and rejection lines.
6266 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6268 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6270 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6271 could be overwritten at the end of the current message (or the start of a
6272 new message if it was set in a HELO ACL). The value is now preserved for
6273 the duration of the SMTP connection.
6275 12. If a transport had a headers_rewrite setting, and a matching header line
6276 contained an unqualified address, that address was qualified, even if it
6277 did not match any rewriting rules. The underlying bug was that the values
6278 of the flags that permit the existence of unqualified sender and recipient
6279 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6280 non-local messages, and by -bnq for local messages) were not being
6281 preserved with the message after it was received.
6283 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6284 "next input=" as part of the text comprising the host identity instead of
6285 the correct text. The code was using the same buffer for two different
6286 strings. However, depending on which order the printing function evaluated
6287 its arguments, the bug did not always show up. Under Linux, for example, my
6288 test suite worked just fine.
6290 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6291 doesn't work with some older versions of Perl. It has been changed to "my",
6292 which in any case is probably the better facility to use.
6294 15. A really picky compiler found some instances of statements for creating
6295 error messages that either had too many or two few arguments for the format
6298 16. The size of the buffer for calls to the DNS resolver has been increased
6299 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6300 for addresses that have a lot of PTR records. This alleviates a problem; it
6301 does not fully solve it.
6303 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6304 buffer now truncates the list and logs the incident, which is the same
6305 action as happens when Exim is looking up a host name and its aliases.
6306 Previously in this situation something unpredictable would happen;
6307 sometimes it was "internal error: store_reset failed".
6309 18. If a server dropped the connection unexpectedly when an Exim client was
6310 using GnuTLS and trying to read a response, the client delivery process
6311 crashed while trying to generate an error log message.
6313 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6314 string, for example:
6316 warn message = H1: something\nH2: something
6318 the text was added as a single header line from Exim's point of view
6319 though it ended up OK in the delivered message. However, searching for the
6320 second and subsequent header lines using $h_h2: did not work. This has been
6321 fixed. Similarly, if a system filter added multiple headers in this way,
6322 the routers could not see them.
6324 20. Expanded the error message when iplsearch is called with an invalid key to
6325 suggest using net-iplsearch in a host list.
6327 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6328 ACLs are no longer actually imposed (and a message to that effect is
6331 22. If a "gecos" field in a passwd entry contained escaped characters, in
6332 particular, if it contained a \" sequence, Exim got it wrong when building
6333 a From: or a Sender: header from that name. A second bug also caused
6334 incorrect handling when an unquoted " was present following a character
6335 that needed quoting.
6337 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6338 was not being matched caselessly.
6340 24. Arranged for all hyphens in the exim.8 source to be escaped with
6343 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6344 properly. Recipient callout cache records were still being keyed to include
6345 the sender, even when use_sender was set false. This led to far more
6346 callouts that were necessary. The sender is no longer included in the key
6347 when use_sender is false.
6349 26. Added "control = submission" modifier to ACLs.
6351 27. Added the ${base62d: operator to decode base 62 numbers.
6353 28. dnsdb lookups can now access SRV records.
6355 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6356 the configuration file.
6358 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6359 (-v) mode. This makes the output for a verbose queue run more intelligible.
6361 31. Added a use_postmaster feature to recipient callouts.
6363 32. Added the $body_zerocount variable, containing the number of binary zero
6364 bytes in the message body.
6366 33. The time of last modification of the "new" subdirectory is now used as the
6367 "mailbox time last read" when there is a quota error for a maildir
6370 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6372 35. Added +ignore_unknown as a special item in host lists.
6374 36. Code for decoding IPv6 addresses in host lists is now included, even if
6375 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6376 address was recognized as an IP address, but was then not correctly decoded
6377 into binary, causing unexpected and incorrect effects when compared with
6384 1. Very minor rewording of debugging text in manualroute to say "list of
6385 hosts" instead of "hostlist".
6387 2. If verify=header_syntax was set, and a header line with an unqualified
6388 address (no domain) and a large number of spaces between the end of the
6389 name and the colon was received, the reception process suffered a buffer
6390 overflow, and (when I tested it) crashed. This was caused by some obsolete
6391 code that should have been removed. The fix is to remove it!
6393 3. When running in the test harness, delay a bit after writing a bounce
6394 message to get a bit more predictability in the log output.
6396 4. Added a call to search_tidyup() just before forking a reception process. In
6397 theory, someone could use a lookup in the expansion of smtp_accept_max_
6398 per_host which, without the tidyup, could leave open a database connection.
6400 5. Added the variables $recipient_data and $sender_data which get set from a
6401 lookup success in an ACL "recipients" or "senders" condition, or a router
6402 "senders" option, similar to $domain_data and $local_part_data.
6404 6. Moved the writing of debug_print from before to after the "senders" test
6407 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6408 problems for message scanning, either using a data ACL, or using
6409 local_scan() because the Received: header was not generated till after they
6410 were called (in order to set the time as the time of reception completion).
6411 I have revised the way this works. The header is now generated after the
6412 body is received, but before the ACL or local_scan() are called. After they
6413 are run, the timestamp in the header is updated.
6419 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6420 before starting a queue runner without re-exec. This happened only when
6421 deliver_drop_privilege was set or when the Exim user was set to root. The
6422 effect of the bug was that timeouts during subsequent deliveries caused
6423 crashes instead of being properly handled. The handler is now left at its
6424 default (and expected) setting.
6426 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6427 message, again when deliver_drop_privilege is set or Exim is run as root.
6428 The bug described in (1) was not present in this case, but the tidying up
6429 of the other signals was missing. I have made the two cases consistent.
6431 3. The ignore_target_hosts setting on a manualroute router was being ignored
6432 for hosts that were looked up using the /MX notation.
6434 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6437 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6438 operated on the sender address. After changing the $sender_address to <>
6439 for the sender address verify, Exim was re-instated it as the original
6440 (before rewriting) address, but remembering that it had rewritten it, so it
6441 wasn't rewriting it again. This bug also had the effect of breaking the
6442 sender address verification caching when the sender address was rewritten.
6444 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6445 This has been changed so that if the ip literal address matches
6446 ignore_target_hosts, the router declines.
6448 7. Added expansion conditions match_domain, match_address, and match_local_
6449 part (NOT match_host).
6451 8. The placeholder for the Received: header didn't have a length field set.
6453 9. Added code to Exim itself and to exim_lock to test for a specific race
6454 condition that could lead to file corruption when using MBX delivery. The
6455 issue is with the lockfile that is created in /tmp. If this file is removed
6456 after a process has opened it but before that process has acquired a lock,
6457 there is the potential for a second process to recreate the file and also
6458 acquire a lock. This could lead to two Exim processes writing to the file
6459 at the same time. The added code performs the same test as UW imapd; it
6460 checks after acquiring the lock that its file descriptor still refers to
6461 the same named file.
6463 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6464 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6467 11. Added the smtp_active_hostname option. If used, this will typically be made
6468 to depend on the incoming interface address. Because $interface_address is
6469 not set up until the daemon has forked a reception process, error responses
6470 that can happen earlier (such as "too many connections") no longer contain
6473 12. If an expansion in a condition on a "warn" statement fails because a lookup
6474 defers, the "warn" statement is abandoned, and the next ACL statement is
6475 processed. Previously this caused the whole ACL to be aborted.
6477 13. Added the iplsearch lookup type.
6479 14. Added ident_timeout as a log selector.
6481 15. Added tls_certificate_verified as a log selector.
6483 16. Added a global option tls_require_ciphers (compare the smtp transport
6484 option of the same name). This controls incoming TLS connections.
6486 17. I finally figured out how to make tls_require_ciphers do a similar thing
6487 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6488 before starting the TLS session.
6490 18. Tabs are now shown as \t in -bP output.
6492 19. If the log selector return_path_on_delivery was set, Exim crashed when
6493 bouncing a message because it had too many Received: header lines.
6495 20. If two routers both had headers_remove settings, and the first one included
6496 a superfluous trailing colon, the final name in the first list and the
6497 first name in the second list were incorrectly joined into one item (with a
6498 colon in the middle).
6504 1. Added -C and -D options to the exinext utility, mainly to make it easier
6505 to include in the automated testing, but these could be helpful when
6506 multiple configurations are in use.
6508 2. The exinext utility was not formatting the output nicely when there was
6509 an alternate port involved in the retry record key, nor when there was a
6510 message id as well (for retries that were specific to a specific message
6511 and a specific host). It was also confused by IPv6 addresses, because of
6512 the additional colons they contain. I have fixed the IPv4 problem, and
6513 patched it up to do a reasonable job for IPv6.
6515 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6516 delivery, the log line now contains "pipelined" if PIPELINING was used.
6518 4. An SMTP transport process used to panic and die if the bind() call to set
6519 an explicit outgoing interface failed. This has been changed; it is now
6520 treated in the same way as a connect() failure.
6522 5. A reference to $sender_host_name in the part of a conditional expansion
6523 that was being skipped was still causing a DNS lookup. This no longer
6526 6. The def: expansion condition was not recognizing references to header lines
6527 that used bh_ and bheader_.
6529 7. Added the _cache feature to named lists.
6531 8. The code for checking quota_filecount in the appendfile transport was
6532 allowing one more file than it should have been.
6534 9. For compatibility with Sendmail, the command line option
6542 and sets the incoming protocol and host name (for trusted callers). The
6543 host name and its colon can be omitted when only the protocol is to be set.
6544 Note the Exim already has two private options, -pd and -ps, that refer to
6545 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6546 "s", but I don't think that's a major issue.
6548 10. A number of refactoring changes to the code, none of which should affect
6551 (a) The number of logging options was getting close to filling up the
6552 32-bit word that was used as a bit map. I have split them into two classes:
6553 those that are passed in the argument to log_write(), and those that are
6554 only ever tested independently outside of that function. These are now in
6555 separate 32-bit words, so there is plenty of room for expansion again.
6556 There is no change in the user interface or the logging behaviour.
6558 (b) When building, for example, log lines, the code previously used a
6559 macro that called string_cat() twice, in order to add two strings. This is
6560 not really sufficiently general. Furthermore, there was one instance where
6561 it was actually wrong because one of the argument was used twice, and in
6562 one call a function was used. (As it happened, calling the function twice
6563 did not affect the overall behaviour.) The macro has been replaced by a
6564 function that can join an arbitrary number of extra strings onto a growing
6567 (c) The code for expansion conditions now uses a table and a binary chop
6568 instead of a serial search (which was left over from when there were very
6569 few conditions). Also, it now recognizes conditions like "pam" even when
6570 the relevant support is not compiled in: a suitably worded error message is
6571 given if an attempt is made to use such a condition.
6573 11. Added ${time_interval:xxxxx}.
6575 12. A bug was causing one of the ddress fields not to be passed back correctly
6576 from remote delivery subprocesses. The field in question was not being
6577 subsequently used, so this caused to problems in practice.
6579 13. Added new log selectors queue_time and deliver_time.
6581 14. Might have fixed a bug in maildirsizefile handling that threw up
6582 "unexpected character" debug warnings, and recalculated the data
6583 unnecessarily. In any case, I expanded the warning message to give more
6586 15. Added the message "Restricted characters in address" to the statements in
6587 the default ACL that block characters like @ and % in local parts.
6589 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6590 Three changes have been made:
6592 (a) There was a serious bug; a negative response to MAIL caused the whole
6593 recipient domain to be cached as invalid, thereby blocking all messages
6594 to all local parts at the same domain, from all senders. This bug has
6595 been fixed. The domain is no longer cached after a negative response to
6596 MAIL if the sender used is not empty.
6598 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6601 (c) A new callout option, "use_sender" has been added for people who want
6602 the modified behaviour.
6608 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6611 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6612 indeed breaks things for older releases.
6614 3. Added additional logging to the case where there is a problem reading data
6615 from a filter that is running in a subprocess using a pipe, in order to
6616 try to track down a specific problem.
6618 4. Testing facility fudge: when running in the test harness and attempting
6619 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6620 getting "No route to host". Convert this to a timeout.
6622 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6625 6. Some OS don't have socklen_t but use size_t instead. This affects the
6626 fifth argument of getsockopt() amongst other things. This is now
6627 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6628 can be set for individual OS. I have set it for SunOS5, OSF1, and
6629 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6630 some earlier ones do not.
6632 7. Change 4.30/15 was not doing the test caselessly.
6634 8. The standard form for an IPv6 address literal was being rejected by address
6635 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6636 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6637 this, as well as the form without the "IPv6" on the front (but only when
6638 address literals are enabled, of course).
6640 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6642 10. Exim crashed if a message with an empty sender address specified by -f
6643 encountered a router with an errors_to setting. This could be provoked only
6644 by a command such as
6648 where an empty string was supplied; "<>" did not hit this bug.
6650 11. Installed PCRE release 4.5.
6652 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6653 remained set. It is now erased.
6655 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6656 times from message ids (which are base 36 rather than the normal 62).
6658 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6659 were being counted as actual protocol errors, and logged if the log
6660 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6661 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6662 and DATA following a set of rejected RCPTs do not count as protocol errors.
6663 In other words, Exim assumes they were pipelined, though this may not
6664 actually be the case. Of course, in all cases the client gets an
6665 appropriate error code.
6667 15. If a lookup fails in an ACL condition, a message about the failure may
6668 be available; it is used if testing the ACL cannot continue, because most
6669 such messages specify what the cause of the deferral is. However, some
6670 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6671 that caused an old message to be retained and used if a later statement
6672 caused a defer, replacing the real cause of the deferral.
6674 16. If an IP address had so many PTR records that the DNS lookup buffer
6675 was not large enough to hold them, Exim could crash while trying to process
6676 the truncated data. It now detects and logs this case.
6678 17. Further to 4.21/58, another change has been made: if (and only if) the
6679 first line of a message (the first header line) ends with CRLF, a bare LF
6680 in a subsequent header line has a space inserted after it, so as not to
6681 terminate the header.
6683 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6684 unnecessarily, used atoi() instead of strtol(), and didn't check the
6685 termination when getting file sizes from file names by regex.
6687 19. Completely re-implemented the support for maildirsize files, in the light
6688 of a number of problems with the previous contributed implementation
6689 (4.30/29). In particular:
6691 . If the quota is zero, the maildirsize file is maintained, but no quota is
6694 . If the maildir directory does not exist, it is created before any attempt
6695 to write a maildirsize file.
6697 . The quota value in the file is just a cache; if the quota is changed in
6698 the transport, the new value overrides.
6700 . A regular expression is available for excluding directories from the
6703 20. The autoreply transport checks the characters in options that define the
6704 message's headers; it allows continued headers, but it was checking with
6705 isspace() after an embedded newline instead of explicitly looking for a
6708 21. If all the "regular" hosts to which an address was routed had passed their
6709 expiry times, and had not reached their retry times, the address was
6710 bounced, even if fallback hosts were defined. Now Exim should go on to try
6713 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6714 equivalent code in the SMTP transport. Some hosts send humungous responses
6715 to HELO/EHLO, more than 1024 it seems.
6717 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6718 gives compiler warnings in some environments. I've now done it "properly",
6721 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6722 (because of problems with the built-in one) was declared to return uschar *
6723 instead of char *, causing compiler failure.
6725 25. Fixed a file descriptor leak when processing alias/forward files.
6727 26. Fixed a minor format string issue in dbfn.c.
6729 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6731 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6732 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6735 29. When a sender address is verified, it is cached, to save repeating the test
6736 when there is more than one recipient in a message. However, when the
6737 verification involves a callout, it is possible for different callout
6738 options to be set for different recipients. It is too complicated to keep
6739 track of this in the cache, so now Exim always runs a verification when a
6740 callout is required, relying on the callout cache for the optimization.
6741 The overhead is duplication of the address routing, but this should not be
6744 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6745 to be verified with callout=postmaster, and the main callout worked but the
6746 postmaster check failed, the verification correctly failed. However, if a
6747 subsequent RCPT command asked for sender verification *without* the
6748 postmaster check, incorrect caching caused this verification also to fail,
6751 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6752 it was not caching the DNS options (qualify_single, search_parents) that
6753 were used when the lookup failed. A subsequent lookup with different
6754 options therefore always gave the same answer, though there were cases
6755 where it should not have. (Example: a "domains = !$mx_any" option on a
6756 dnslookup router: the "domains" option is always processed without any
6757 widening, but the router might have qualify_single set.) Now Exim uses the
6758 cached value only when the same options are set.
6760 32. Added John Jetmore's "exipick" utility to the distribution.
6762 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6763 than a timeout (e.g. a certificate is required, and is not provided), an
6764 Exim server now closes the connection immediately. Previously it waited for
6765 the client to close - but if the client is SSL, it seems that they each
6766 wait for each other, leading to a delay before one of them times out.
6768 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6769 maintained 0.8.x compatibility because I don't think many are using it, and
6770 it is clearly obsolete.
6772 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6775 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6776 client certificate was expired. A simple patch fixes this, though I don't
6777 understand the full logic of why the verify callback is called multiple
6780 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6781 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6782 which causes problems with some clients (such as the Certicom SSL Plus
6783 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6784 disables the coutermeasure allowing Eudora to connect."
6786 38. Exim was not checking that a write() to a log file succeeded. This could
6787 lead to Bad Things if a log got too big, in particular if it hit a file
6788 size limit. Exim now panics and dies if it cannot write to a log file, just
6789 as it does if it cannot open a log file.
6791 39. Modified OS/Makefile-Linux so that it now contains
6793 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6795 The two -D definitions ensure that Exim is compiled with large file
6796 support, which makes it possible to handle log files that are bigger than
6799 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6800 instance) a domain was checked against a named list that involved a lookup,
6801 causing $domain_data to be set, then another domain was checked against the
6802 same list, then the first domain was re-checked, the value of $domain_data
6803 after the final check could be wrong. In particular, if the second check
6804 failed, it could be set empty. This bug probably also applied to
6807 41. The strip_trailing_dot option was not being applied to the address given
6808 with the -f command-line option.
6810 42. The code for reading a message's header from the spool was incrementing
6811 $received_count, but never initializing it. This meant that the value was
6812 incorrect (doubled) while delivering a message in the same process in which
6813 it was received. In the most common configuration of Exim, this never
6814 happens - a fresh exec is done - but it can happen when
6815 deliver_drop_privilege is set.
6817 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6818 it now includes up to 150 characters of the unexpected data in the log
6821 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6822 and building data strings. The size of both of these buffers was 10 000
6823 bytes - far larger than anybody would *ever* want, thought I. Needless to
6824 say, somebody hit the limit. I have increased the maximum line length to
6825 20 000 and the maximum data length of concatenated lines to 100 000. I have
6826 also fixed two bugs, because there was no checking on these buffers. Tsk,
6827 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6828 buffer is too small.
6830 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6831 lsearch lookups. Now it does.
6833 46. When parsing a route_list item in a manualroute router, a fixed-length
6834 buffer was used for the list of hosts. I made this 1024 bytes long,
6835 thinking that nobody would ever have a list of hosts that long. Wrong.
6836 Somebody had a whole pile of complicated expansion conditions, and the
6837 string was silently truncated, leading to an expansion error. It turns out
6838 that it is easier to change to an unlimited length (owing to other changes
6839 that have happened since this code was originally written) than to build
6840 structure for giving a limitation error. The length of the item that
6841 expands into the list of hosts is now unlimited.
6843 47. The lsearch lookup could not handle data where the length of text line was
6844 more than 4095 characters. Such lines were truncated, leading to shortened
6845 data being returned. It should now handle lines of any length.
6847 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6848 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6851 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6852 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6853 no longer happens in comment lines. A list of replacements is now placed
6854 at the head of all of the source files, except those whose only change is
6855 to replace PERL_COMMAND in the very first #! line.
6857 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6858 messages on the queue, with a bottom-up merge sort, using code contributed
6859 by Michael Haardt. This should make operations like -bp somewhat faster on
6860 large queues. It won't affect queue runners, except when queue_run_in_order
6863 51. Installed eximstats 1.31 in the distribution.
6865 52. Added support for SRV lookups to the dnslookup router.
6867 53. If an ACL referred to $message_body or $message_body_end, the value was not
6868 reset for any messages that followed in the same SMTP session.
6870 54. The store-handling optimization for building very long strings was not
6871 differentiating between the different store pools. I don't think this
6872 actually made any difference in practice, but I've tidied it.
6874 55. While running the routers to verify a sender address, $sender_address
6875 was still set to the sender address. This is wrong, because when routing to
6876 send a bounce to the sender, it would be empty. Therefore, I have changed
6877 it so that, while verifying a sender address, $sender_address is set to <>.
6878 (There is no change to what happens when verifying a recipient address.)
6880 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6881 target A or AAAA records (if not already returned) without resetting the
6882 qualify_single or search_parents options of the DNS resolver. These are
6883 inappropriate in this case because the targets of MX and SRV records must
6884 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6885 target that, when qualified, matched something in the local domain. These
6886 two options are now turned off when doing these lookups.
6888 57. It seems that at least some releases of Reiserfs (which does not have the
6889 concept of a fixed number of inodes) returns zero and not -1 for the
6890 number of available inodes. This interacted badly with check_spool_inodes,
6891 which assumed that -1 was the "no such thing" setting. What I have done is
6892 to check that the total number of inodes is greater than zero before doing
6893 the test of how many are available.
6895 58. When a "warn" ACL statement has a log_message modifier, the message is
6896 remembered, and not repeated. This is to avoid a lot of repetition when a
6897 message has many recipients that cause the same warning to be written.
6898 However, Exim was preserving the list of already written lines for an
6899 entire SMTP session, which doesn't seem right. The memory is now reset if a
6900 new message is started.
6902 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6903 debugging output unless log_rewrite was also set.
6905 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6906 of (int)(handle) when we know that handle contains (void *)(-1).
6908 61. The Exim daemon panic-logs an error return when it closes the incoming
6909 connection. However "connection reset by peer" seems to be common, and
6910 isn't really an error worthy of noting specially, so that particular error
6913 62. When Exim is trying to find all the local interfaces, it used to panic and
6914 die if the ioctl to get the interface flags failed. However, it seems that
6915 on at least one OS (Solaris 9) it is possible to have an interface that is
6916 included in the list of interfaces, but for which you get a failure error
6917 for this call. This happens when the interface is not "plumbed" into a
6918 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6919 failure of the "get flags" call assumes that the interface is down.
6921 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6922 makes life easier for people who are doing arithmetic on fields extracted
6923 from dates, where you often get leading zeros that should not be
6924 interpreted as octal.
6926 64. Added qualify_domain to the redirect router, to override the global
6929 65. If a pathologically long header line contained very many addresses (the
6930 report of this problem mentioned 10 000) and each of them was rewritten,
6931 Exim could use up a very large amount of memory. (It kept on making new
6932 copies of the header line as it rewrote, and never released the old ones.)
6933 At the expense of a bit more processing, the header rewriting function has
6934 been changed so that it no longer eats memory in this way.
6936 66. The generation of the Received: header has been moved from the time that a
6937 message starts to be received, to the time that it finishes. The timestamp
6938 in the Received: header should now be very close to that of the <= log
6939 line. There are two side-effects of this change:
6941 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6942 logged header lines no longer include the local Received: line, because
6943 it has not yet been created. The same applies to a copy of the message
6944 that is returned to a non-SMTP sender when a message is rejected.
6946 (b) When a filter file is tested using -bf, no additional Received: header
6947 is added to the test message. After some thought, I decided that this
6950 This change does not affect the value of $received_for. It is still set
6951 after address rewriting, but before local_scan() is called.
6953 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6955 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6956 gave an unhelpful panic error message, and a defer error. I have managed to
6957 change this behaviour so that it now rejects any supplied certificate,
6958 which seems right, as the list of acceptable certificates is empty.
6960 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6961 gave an unhelpful defer error. I have not managed to make this reject any
6962 supplied certificates, but the error message it gives is "no certificate
6963 supplied", which is not helpful.
6965 70. exigrep's output now also includes lines that are not associated with any
6966 message, but which match the given pattern. Implemented by a patch from
6967 Martin Sluka, which also tidied up the Perl a bit.
6969 71. Recipient callout verification, like sender verification, was using <> in
6970 the MAIL FROM command. This isn't really the right thing, since the actual
6971 sender may affect whether the remote host accepts the recipient or not. I
6972 have changed it to use the actual sender in the callout; this means that
6973 the cache record is now keyed on a recipient/sender pair, not just the
6974 recipient address. There doesn't seem to be a real danger of callout loops,
6975 since a callout by the remote host to check the sender would use <>.
6976 [SEE ABOVE: changed after hitting problems.]
6978 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6979 temporary errors. However, in the case of such a code being given after
6980 the end of a data transmission (i.e. after ".") Exim was failing to write
6981 a retry record for the message. (Yes, there was some broken host that was
6982 actually sending 8xx at this point.)
6984 73. An unknown lookup type in a host list could cause Exim to panic-die when
6985 the list was checked. (An example that provoked this was putting <; in the
6986 middle of a list instead of at the start.) If this happened during a DATA
6987 ACL check, a -D file could be left lying around. This kind of configuration
6988 error no longer causes Exim to die; instead it causes a defer error. The
6989 incident is still logged to the main and panic logs.
6991 74. Buglet left over from Exim 3 conversion. The message "too many messages
6992 in one connection" was written to the rejectlog but not the mainlog, except
6993 when address rewriting (yes!) was being logged.
6995 75. Added write_rejectlog option.
6997 76. When a system filter was run not as root (that is, when system_filter_user
6998 was set), the values of the $n variables were not being returned to the
6999 main process; thus, they were not subsequently available in the $sn
7002 77. Added +return_path_on_delivery log selector.
7004 78. A connection timeout was being treated differently from recipients deferred
7005 when testing hosts_max_try with a message that was older than the host's
7006 retry timeout. (The host should not be counted, thus allowing all hosts to
7007 be tried at least once before bouncing.) This may have been the cause of an
7008 occasionally reported bug whereby a message would remain on the queue
7009 longer than the retry timeout, but would be bounced if a delivery was
7010 forced. I say "may" because I never totally pinned down the problem;
7011 setting up timeout/retry tests is difficult. See also the next item.
7013 79. The ultimate address timeout was not being applied to errors that involved
7014 a combination of host plus message (for example, a timeout on a MAIL
7015 command). When an address resolved to a number of possible hosts, and they
7016 were not all tried for each delivery (e.g. because of hosts_max_try), a
7017 message could remain on the queue longer than the retry timeout.
7019 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7022 81. Fixed an obscure SMTP outgoing bug which required at least the following
7023 conditions: (a) there was another message waiting for the same server;
7024 (b) the server returned 5xx to all RCPT commands in the first message so
7025 that the message was not completed; (c) the server dropped the connection
7026 or gave a negative response to the RSET that Exim sends to abort the
7027 transaction. The observed case was a dropped connection after DATA that had
7028 been sent in pipelining mode. That is, the server had advertised PIPELINING
7029 but was not implementing it correctly. The effect of the bug was incorrect
7030 behaviour, such as trying another host, and this could lead to a crash.
7036 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7037 and daemon.c were passed as pointers to ints; they should have been
7038 pointers to socklen_t variables (which are typically unsigned ints).
7040 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7043 3. Fixed a really odd bug that affected only the testing scheme; patching a
7044 certain fixed string in the binary changed the value of another string that
7045 happened to be identical to the end of the original first string.
7047 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7048 name", it returns that address as the IP address. On some operating
7049 systems (e.g. Solaris), it also passes back the IP address string as the
7050 "host name". However, on others (e.g. Linux), it passes back an empty
7051 string. Exim wasn't checking for this, and was changing the host name to an
7052 empty string, assuming it had been canonicalized.
7054 5. Although rare, it is permitted to have more than one PTR record for a given
7055 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7056 all the names associated with an address, because they do in Solaris.
7057 However, it seems that they do not in Linux for data that comes from the
7058 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7059 I found this out when I moved to a new Linux workstation and tried to run
7060 the Exim test suite.
7062 To get round this problem I have changed the code so that it now does its
7063 own call to the DNS to look up PTR records when searching for a host name.
7064 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7065 addresses that are only in /etc/hosts are still found.
7067 This behaviour is, however, controlled by an option called host_lookup_
7068 order, which defaults to "bydns:byaddr". If people want to use the other
7069 order, or indeed, just use one or the other means of lookup, they can
7070 specify it in this variable.
7072 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7073 some operating systems, this comes back from gethostbyaddr() as an empty
7074 string, and this is what Exim used to test for. However, it seems that in
7075 other systems, "." is yielded. Exim now tests for this case too.
7077 7. The values of check_spool_space and check_log_space are now held internally
7078 as a number of kilobytes instead of an absolute number of bytes. If a
7079 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7080 kilobyte. This means that much larger values can be stored.
7082 8. Exim monitor: an attempt to get the action menu when not actually pointing
7083 at a message produces an empty menu entitled "No message selected". This
7084 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7085 no entries in it ("Shell widget menu has zero width and/or height"). So I
7086 have added a single, blank menu entry in this case.
7088 9. Added ${quote_local_part.
7090 10. MIME decoding is now applied to the contents of Subject: header lines when
7093 11. Now that a reference to $sender_host_address automatically causes a reverse
7094 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7095 host lookup before query-style lookups in lists that might use this
7096 variable. This has therefore been abolished, and the "net-" prefix is no
7097 longer necessary for query-style lookups.
7099 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7100 have been a typo for LFLAGS, so it has been changed.
7102 13. The install script calls Exim with "-C /dev/null" in order to find the
7103 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7104 to be output. However, since Exim outputs its version number before the
7105 error, it didn't break the script. It just looked ugly. I fixed this by
7106 always allowing "-C /dev/null" if the caller is root.
7108 14. Ignore overlarge ACL variable number when reading spool file - insurance
7109 against a later release with more variables having written the file.
7111 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7112 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7113 this, as well as the form without the "IPv6" on the front.
7115 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7116 OS/Makefile-Darwin file.
7118 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7119 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7121 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7122 variable. However, some OS use ints for inodes. I've added cast to long int
7123 to get rid of the compiler warning.
7125 19. I had forgotten to lock out "/../" in configuration file names when
7126 ALT_CONFIG_PREFIX was set.
7128 20. Routers used for verification do not need to specify transports. However,
7129 if such a router generated a host list, and callout was configured, Exim
7130 crashed, because it could not find a port number from the (non-existent)
7131 transport. It now assumes port 25 in this circumstance.
7133 21. Added the -t option to exigrep.
7135 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7136 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7137 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7138 them is set, LOOKUP_LSEARCH is forced.
7140 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7142 24. Added sender and host information to the "rejected by local_scan()" log
7143 line; previously there was no indication of these.
7145 25. Added .include_if_exists.
7147 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7148 Linux. It turns out that not all file systems support this. Apparently some
7149 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7150 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7151 not supported on the file descriptor, is now ignored when Exim is trying to
7152 sync a directory. This applies only to Linux.
7154 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7156 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7157 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7160 29. Added support for maildirsize files from supplied patch (modified a bit).
7162 30. The use of :fail: followed by an empty string could lead Exim to respond to
7163 sender verification failures with (e.g.):
7165 550 Verification failed for <xxx>
7166 550 Sender verify failed
7168 where the first response line was missing the '-' that indicates it is not
7169 the final line of the response.
7171 31. The loop for finding the name of the user that called Exim had a hardwired
7172 limit of 10; it now uses the value of finduser_retries, which is used for
7173 all other user lookups.
7175 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7178 33. Exim was neglecting to zero errno before one call of strtol() when
7179 expanding a string and expecting an integer value. On some systems this
7180 resulted in spurious "integer overflow" errors. Also, it was casting the
7181 result into an int without checking.
7183 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7184 did not work. The code looks as if it has *never* worked, though it appears
7185 to have been documented since at least release 1.62. I have made it work.
7187 35. The "timeout_DNS" error in retry rules, also documented since at least
7188 1.62, also never worked. As it isn't clear exactly what this means, and
7189 clearly it isn't a major issue, I have abolished the feature by treating it
7190 as "timeout", and writing a warning to the main and panic logs.
7192 36. The display of retry rules for -brt wasn't always showing the error code
7195 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7196 timeout_connect_A, timeout_connect_MX.
7198 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7199 to the empty sender.
7201 39. The daemon was not analysing the content of -oX till after it had closed
7202 stderr and disconnected from the controlling terminal. This meant that any
7203 syntax errors were only noted on the panic log, and the return code from
7204 the command was 0. By re-arranging the code a little, I've made the
7205 decoding happen first, so such errors now appear on stderr, and the return
7206 code is 1. However, the actual setting up of the sockets still happens in
7207 the disconnected process, so errors there are still only recorded on the
7210 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7211 connections (as happens on some IP stacks) was logged at start up time as
7212 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7213 it from "IPv6 and IPv4", which means that two separate sockets are being
7216 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7217 says whether AF_INET or AF_INET6 was passed as an argument.
7219 42. Exiwhat output was messed up when time zones were included in log
7222 43. Exiwhat now gives more information about the daemon's listening ports,
7223 and whether -tls-on-connect was used.
7225 44. The "port" option of the smtp transport is now expanded.
7227 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7228 silently ignored. Now an error message is written to the main and panic
7231 46. There's a new ACL modifier called "logwrite" which writes to a log file
7232 as soon as it is encountered.
7234 47. Added $local_user_uid and $local_user_gid at routing time.
7236 48. Exim crashed when trying to verify a sender address that was being
7239 49. Exim was recognizing only a space character after ".include". It now also
7240 recognizes a tab character.
7242 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7243 extracting the relevant information from the specification. The man page no
7244 longer contains scrambled data for the -d option, and I've added a section
7245 at the front about calling Exim under different names.
7247 51. Added "extra_headers" argument to the "mail" command in filter files.
7249 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7252 53. Installed eximstats 1.29.
7254 54. Added transport_filter_timeout as a generic transport option.
7256 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7257 Cc: header lines. This was required by RFC 822, but it not required by RFC
7260 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7261 incoming messages that did not have them. Now it does so only if the
7262 message originates locally, that is, if there is no associated remote host
7263 address. When Resent- header lines are present, this applies to the Resent-
7264 lines rather than the non-Resent- lines.
7266 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7267 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7269 58. Messages for configuration errors now include the name of the main
7270 configuration file - useful now that there may be more than one file in a
7271 list (.included file names were always shown).
7273 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7274 for those rare installations that do not start the daemon as root or run it
7275 setuid root. I've cut out the call to initgroups() if the daemon is not
7278 60. The Exim user and group can now be bound into the binary as text strings
7279 that are looked up at the start of Exim's processing.
7281 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7283 62. Added $mailstore_basename variable.
7285 63. Installed patch to sieve.c from Michael Haardt.
7287 64. When Exim failed to open the panic log after failing to open the main log,
7288 the original message it was trying to log was written to stderr and debug
7289 output, but if they were not available (the usual case in production), it
7290 was lost. Now it is written to syslog before the two lines that record the
7291 failures to open the logs.
7293 65. Users' Exim filters run in subprocesses under the user's uid. It is
7294 possible for a "deliver" command or an alias in a "personal" command to
7295 provoke an address rewrite. If logging of address rewriting is configured,
7296 this fails because the process is not running as root or exim. There may be
7297 a better way of dealing with this, but for the moment (because 4.30 needs
7298 to be released), I have disabled address rewrite logging when running a
7299 filter in a non-root, non-exim process.
7305 1. The buildconfig auxiliary program wasn't quoting the value set for
7306 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7307 not defined. This bug was masked in 4.22 by the effect that was fixed in
7310 2. Some messages that were rejected after a message id was allocated were
7311 shown as "incomplete" by exigrep. It no longer does this for messages that
7312 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7314 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7315 have allow_domain_literals set, the ID did not get logged in the <= line.
7316 Domain literals are now always recognized in Message-ID: header lines.
7318 4. The first argument for a ${extract expansion item is the key name or field
7319 number. Leading and trailing spaces in this item were not being ignored,
7320 causing some misleading effects.
7322 5. When deliver_drop_privilege was set, single queue runner processes started
7323 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7324 same command in the process it spins off) were not dropping privilege.
7326 6. When the daemon running as "exim" started a queue runner, it always
7327 re-executed Exim in the spun-off process. This is a waste of effort when
7328 deliver_drop_privilege is set. The new process now just calls the
7329 queue-runner function directly.
7335 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7338 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7339 ignored. Though the use of -oP was forcing the writing of a pid file, it
7340 was always written to the default place.
7342 3. If the message "no IP address found for host xxxx" is generated during
7343 incoming verification, it is now followed by identification of the incoming
7344 connection (so you can more easily find what provoked it).
7346 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7348 5. Added some features to "harden" Exim a bit more against certain attacks:
7350 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7351 be put in Local/Makefile. This is like the never_users runtime option,
7352 but it cannot be overridden. The default setting is "root".
7354 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7355 prefix string with which any file named in a -C command line option
7358 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7359 is retained for -C and -D only if the caller of Exim is root. Without
7360 it, the exim user may also use -C and -D and retain privilege.
7362 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7363 command line option is disabled.
7365 6. Macro names set by the -D option must start with an upper case letter, just
7366 like macro names defined in the configuration file.
7368 7. Added "dereference=" facility to LDAP.
7370 8. Two instances of the typo "uknown" in the source files are fixed.
7372 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7373 the Configure-Makefile script screwed up while processing it.
7375 10. Incorporated PCRE 4.4.
7377 11. The SMTP synchronization check was not operating right at the start of an
7378 SMTP session. For example, it could not catch a HELO sent before the client
7379 waited for the greeting. There is now a check for outstanding input at the
7380 point when the greeting is written. Because of the duplex, asynchronous
7381 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7382 way, but not yet received, when the check is performed.
7384 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7385 on TCP/IP sockets, because this apparently causes some broken clients to
7388 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7389 unchanged) from the Cygwin maintainer.
7391 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7392 was supported instead of testing for mbx. Effectively a typo.
7394 15. The spa authenticator server code was not checking that the input it
7395 received was valid base64.
7397 16. The debug output line for the "set" modifier in ACLs was not showing the
7398 name of the variable that was being set.
7400 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7402 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7403 lookup on its own. Something else was needed to trigger the lookup. For
7404 example, a match in host_lookup or the need for a host name in a host list.
7405 Now, if $sender_host_name is referenced and the host name has not yet been
7406 looked up, a lookup is performed. If the lookup fails, the variable remains
7407 empty, and $host_lookup_failed is set to "1".
7409 19. Added "eqi" as a case-independent comparison operator.
7411 20. The saslauthd authentication condition could segfault if neither service
7412 nor realm was specified.
7414 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7415 error message that was logged was misleading, and incoming SMTP
7416 connections were dropped. The message is now more accurate, and temporary
7417 errors are given to SMTP connections.
7419 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7420 (except RSET) with a 421 error, until QUIT is received. However, it was
7421 failing to send a response to QUIT.
7423 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7424 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7425 ACL is now run after the helo_try_verify_hosts code.
7427 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7428 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7429 case-independent, so other case variants are also recognized). Apparently
7430 some systems use these upper case variants.
7432 25. If more than two messages were waiting for the same host, and a transport
7433 filter was specified for the transport, Exim sent two messages over the
7434 same TCP/IP connection, and then failed with "socket operation on non-
7435 socket" when it tried to send the third.
7437 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7439 27. The extern definition of crypt16() in expand.c was not being excluded when
7440 the OS had its own crypt16() function.
7442 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7443 as a preferred synonym for return_size_limit, both as an option and as an
7446 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7448 30. Changed the default configuration ACL to relax the local part checking rule
7449 for addresses that are not in any local domains. For these addresses,
7450 slashes and pipe symbols are allowed within local parts, but the sequence
7451 /../ is explicitly forbidden.
7453 31. SPA server authentication was not clearing the challenge buffer before
7456 32. log_message in a "warn" ACL statement was writing to the reject log as
7457 well as to the main log, which contradicts the documentation and doesn't
7458 seem right (because no rejection is happening). So I have stopped it.
7460 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7461 However, I am unable to do any testing of this.
7463 34. Fixed an infelicity in the appendfile transport. When checking directories
7464 for a mailbox, to see if any needed to be created, it was accidentally
7465 using path names with one or more superfluous leading slashes; tracing
7466 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7468 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7469 contents are added to the log line that is written for every discarded
7470 recipient. (Previously a log_message setting was ignored.)
7472 36. The ${quote: operator now quotes the string if it is empty.
7474 37. The install script runs exim in order to find its version number. If for
7475 some reason other than non-existence or emptiness, which it checks, it
7476 could not run './exim', it was installing it with an empty version number,
7477 i.e. as "exim-". This error state is now caught, and the installation is
7480 38. An argument was missing from the function that creates an error message
7481 when Exim fails to connect to the socket for saslauthd authentication.
7482 This could cause Exim to crash, or give a corrupted message.
7484 39. Added isip, isip4, and isip6 to ${if conditions.
7486 40. The ACL variables $acl_xx are now saved with the message, and can be
7487 accessed later in routers, transports, and filters.
7489 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7490 strings in the file are not string-expanded.
7492 42. If a MAIL command specified a SIZE value that was too large to fit into an
7493 int variable, the check against message_size_limit failed. Such values are
7494 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7495 day this will have to be increased, but I don't think I want to be around
7496 when emails are that large.
7503 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7504 iconv() is not standard in FreeBSD.
7506 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7507 IPv6 enabled. The observed symptom was a segmentation fault on return from
7508 the function os_common_find_running_interfaces() in src/os.c.
7510 3. In the check_special_case() function in daemon.c I had used "errno" as an
7511 argument name, which causes warnings on some systems. This was basically a
7512 typo, since it was named "eno" in the comments!
7514 4. The code that waits for the clock to tick (at a resolution of some fraction
7515 of a second) so as to ensure message-id uniqueness was always waiting for
7516 at least one whole tick, when it could have waited for less. [This is
7517 almost certainly not relevant at current processor speeds, where it is
7518 unlikely to ever wait at all. But we try to future-proof.]
7520 5. The function that sleeps for a time interval that includes fractions of a
7521 second contained a race. It did not block SIGALRM between setting the
7522 timer, and suspending (a couple of lines later). If the interval was short
7523 and the sigsuspend() was delayed until after it had expired, the suspension
7524 never ended. On busy systems this could lead to processes getting stuck for
7527 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7528 process, before it forks any delivery processes. The open lookup caching
7529 mechanism meant that the open file or database connection was passed into
7530 the delivery process. The problem was that delivery processes always tidy
7531 up cached lookup data. This could cause a problem for the next delivery
7532 process started by the queue runner, because the external queue runner
7533 process does not know about the closure. So the next delivery process
7534 still has data in the lookup cache. In the case of a file lookup, there was
7535 no problem because closing a file descriptor in a subprocess doesn't affect
7536 the parent. However, if the lookup was caching a connection to a database,
7537 the connection was closed, and the second delivery process was likely to
7538 see errors such as "PGSQL: query failed: server closed the connection
7539 unexpectedly". The problem has been fixed by closing all cached lookups
7540 in a queue runner before running a delivery process.
7542 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7543 seem to have the "const" qualifier which it has on other OS. I've
7546 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7547 *of the same type* (client or server) with the same public name that an
7548 error should be diagnosed.
7550 9. When Exim looked up a host name for an IP address, but failed to find the
7551 original IP address when looking up the host name (a safety check), it
7552 output the message "<ip address> does not match any IP for NULL", which was
7553 confusing, to say the least. The bug was that the host name should have
7554 appeared instead of "NULL".
7556 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7557 user that is built into the binary, and the -C or -D options is used, root
7558 privilege is dropped before the configuration file is read. In addition,
7559 logging is switched to stderr instead of the normal log files. If the
7560 configuration then re-defines the Exim user, the unprivileged environment
7561 is probably not what is expected, so Exim logs a panic warning message (but
7564 However, if deliver_drop_privilege is set, the unprivileged state may well
7565 be exactly what is intended, so the warning has been cut out in that case,
7566 and Exim is allowed to try to write to its normal log files.
7572 1. smtp_return_error_details was not giving details for temporary sender
7573 or receiver verification errors.
7575 2. Diagnose a configuration error if two authenticators have the same public
7578 3. Exim used not to create the message log file for a message until the first
7579 delivery attempt. This could be confusing when incoming messages were held
7580 for policy or load reasons. The message log file is now created at the time
7581 the message is received, and an initial "Received" line is written to it.
7583 4. The automatically generated man page for command line options had a minor
7584 bug that caused no ill effects; however, a more serious problem was that
7585 the procedure for building the man page automatically didn't always
7586 operate. Consequently, release 4.20 contains an out-of-date version. This
7587 shouldn't happen again.
7589 5. When building Exim with embedded Perl support, the script that builds the
7590 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7591 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7593 6. The freeze_tell option was not being used for messages that were frozen on
7594 arrival, either by an ACL or by local_scan().
7596 7. Added the smtp_incomplete_transaction log selector.
7598 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7599 was accepting AUTH without a new EHLO.
7601 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7602 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7605 10. Logging of TCP/IP connections (when configured) now happens in the main
7606 daemon process instead of the child process, so that the TCP/IP connection
7607 count is more accurate (but it can never be perfect).
7609 11. The use of "drop" in a nested ACL was not being handled correctly in the
7610 outer ACL. Now, if condition failure induced by the nested "drop" causes
7611 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7612 or "require"), the connection is dropped.
7614 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7615 that yield "discard" can now be used with an "accept" or a "discard" verb,
7616 but an error is generated for any others (because I can't see a useful way
7617 to define what should happen).
7619 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7620 are now processed in the same way as lines in the Exim configuration file.
7621 In particular, continuation lines are supported.
7623 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7625 15. Added -ti meaning -t -i.
7627 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7628 domains, and warn by logging if others are found.
7630 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7631 returned from GIFCONF when Exim is trying to find the list of interfaces on
7632 a host. The code in os.c has been modified to copy each ifreq to an aligned
7633 structure in all cases.
7635 Also, in some cases, the returned ifreq's were being copied to a 'struct
7636 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7637 means the last couple of bytes of an IPv6 address could be chopped if the
7638 ifreq contained only a normal sockaddr (14 bytes storage).
7640 18. Named domain lists were not supported in the hosts_treat_as_local option.
7641 An entry such as +xxxx was not recognized, and was treated as a literal
7644 19. Ensure that header lines added by a DATA ACL are included in the reject log
7645 if the ACL subsequently rejects the message.
7647 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7648 MD5 (which is deprecated).
7650 21. When testing a filter file using -bf, Exim was writing a message when it
7651 took the sender from a "From " line in the message, but it was not doing so
7652 when it took $return_path from a Return-Path: header line. It now does.
7654 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7655 with a valid header line field name (a series of printing characters
7656 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7658 23. Changed "disc" in the source to "disk" to conform to the documentation and
7659 the book and for uniformity.
7661 24. Ignore Sendmail's -Ooption=value command line item.
7663 25. When execve() failed while trying to run a command in a pipe transport,
7664 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7665 could be confused with a return value of 69 from the command itself. This
7666 has been changed to 127, the value the shell returns if it is asked to run
7667 a non-existent command. The wording for the related log line suggests a
7668 non-existent command as the problem.
7670 26. If received_header_text expands to an empty string, do not add a Received:
7671 header line to the message. (Well, it adds a token one on the spool, but
7672 marks it "old" so that it doesn't get used or transmitted.)
7674 27. Installed eximstats 1.28 (addition of -nt option).
7676 28. There was no check for failure on the call to getsockname() in the daemon
7677 code. This can fail if there is a shortage of resources on the system, with
7678 ENOMEM, for example. A temporary error is now given on failure.
7680 29. Contrary to the C standard, it seems that in some environments, the
7681 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7682 program. Exim now does this explicitly; it affects the formatting of
7683 timestamps using strftime().
7685 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7686 complaints. I've now initialized all the variables, to avoid this.
7688 32. Header lines added by a system filter were not being "seen" during
7689 transport-time rewrites.
7691 33. The info_callback() function passed to OpenSSL is set up with type void
7692 (*)(SSL *, int, int), as described somewhere. However, when calling the
7693 function (actually a macro) that sets it up, the type void(*)() is
7694 expected. I've put in a cast to prevent warnings from picky compilers.
7696 34. If a DNS black list lookup found a CNAME record, but there were no A
7697 records associated with the domain it pointed at, Exim crashed.
7699 35. If a DNS black list lookup returned more than one A record, Exim ignored
7700 all but the first. It now scans all returned addresses if a particular IP
7701 value is being sought. In this situation, the contents of the
7702 $dnslist_value variable are a list of all the addresses, separated by a
7705 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7706 to accept a host name and all its aliases if the forward lookup for any of
7707 them yielded the IP address of the incoming connection. Now it accepts only
7708 those names whose forward lookup yields the correct IP address. Any other
7709 names are discarded. This closes a loophole whereby a rogue DNS
7710 administrator could create reverse DNS records to break through a
7711 wildcarded host restriction in an ACL.
7713 37. If a user filter or a system filter that ran in a subprocess used any of
7714 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7715 the wrong values were passed to the pipe command ($thisaddress had the
7716 value of $0, $0 had the value of $1, etc). This bug was introduced by
7717 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7719 38. Improved the line breaking for long SMTP error messages from ACLs.
7720 Previously, if there was no break point between 40 and 75 characters, Exim
7721 left the rest of the message alone. Two changes have been made: (a) I've
7722 reduced the minimum length to 35 characters; (b) if it can't find a break
7723 point between 35 and 75 characters, it looks ahead and uses the first one
7724 that it finds. This may give the occasional overlong line, but at least the
7725 remaining text gets split now.
7727 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7728 file descriptors might be low, and that setting 1000 would always raise it.
7729 It turns out that in some environments, the limit is already over 1000 and
7730 that lowering it causes trouble. So now Exim takes care not to decrease it.
7732 40. When delivering a message, the value of $return_path is set to $sender_
7733 address at the start of routing (routers may change the value). By an
7734 oversight, this default was not being set up when an address was tested by
7735 -bt or -bv, which affected the outcome if any router or filter referred to
7738 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7739 log only when "message" or "log_message" are set. However, if one of the
7740 conditions was an address verification, or a call to a nested ACL, the
7741 messages generated by the underlying test were being passed through. This
7742 no longer happens. The underlying message is available in $acl_verify_
7743 message for both "message" and "log_message" expansions, so it can be
7744 passed through if needed.
7746 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7747 new expansion $bh_ to give the encoded byte string without charset
7748 translation. Translation happens only if iconv() is available; HAVE_ICONV
7749 indicates this at build time. HEADERS_CHARSET gives the charset to
7750 translate to; headers_charset can change it in the configuration, and
7751 "headers charset" can change it in an individual filter file.
7753 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7754 that creates RFC 2047 encoded "words" labels them as that charset instead
7755 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7756 expansion operator; (ii) when Exim creates a From: line for a local
7757 message; (iii) when a header line is rewritten to include a "phrase" part.
7759 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7760 buggy, causing it to skip the first lines of messages whose message ID
7761 ended in 'D'. This would not have bitten before Exim release 4.14, because
7762 message IDs were unlikely to end in 'D' before then. The effect was to have
7763 incorrect size information for certain domains.
7765 45. #include "config.h" was missing at the start of the crypt16.c module. This
7766 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7769 46. If there was a timeout during a "random" callout check, Exim treated it as
7770 a failure of the random address, and carried on sending RSET and the real
7771 address. If the delay was just some slowness somewhere, the response to the
7772 original RCPT would be taken as a response to RSET and so on, causing
7773 mayhem of various kinds.
7775 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7776 when I implemented it. It didn't allow for the fact that some option values
7777 may legitimately be negative (e.g. size_addition), and it didn't even do
7778 the right test for positive values.
7780 48. Domain names in DNS records are case-independent. Exim always looks them up
7781 in lower case. Some resolvers return domain names in exactly the case they
7782 appear in the zone file, that is, they may contain uppercase letters. Not
7783 all resolvers do this - some return always lower case. Exim was treating a
7784 change of case by a resolver as a change of domain, similar to a widening
7785 of a domain abbreviation. This triggered its re-routing code and so it was
7786 trying to route what was effectively the same domain again. This normally
7787 caused routing to fail (because the router wouldn't handle the domain
7788 twice). Now Exim checks for this case specially, and just changes the
7789 casing of the domain that it ultimately uses when it transmits the message
7792 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7795 50. If a filter generated a file delivery with a non-absolute name (possible if
7796 no home directory exists for the router), the forbid_file option was not
7799 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7800 the existing equality matching.
7802 52. Exim was using ints instead of ino_t variables in some places where it was
7803 dealing with inode numbers.
7805 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7806 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7807 called TMPDIR, and if it finds it is different, it changes its value.
7809 54. The smtp_printf() function is now made available to local_scan() so
7810 additional output lines can be written before returning. There is also an
7811 smtp_fflush() function to enable the detection of a dropped connection.
7812 The variables smtp_input and smtp_batched_input are exported to
7815 55. Changed the default runtime configuration: the message "Unknown user"
7816 has been removed from the ACL, and instead placed on the localuser router,
7817 using the cannot_route_message feature. This means that any verification
7818 failures that generate their own messages won't get overridden. Similarly,
7819 the "Unrouteable address" message that was in the ACL for unverifiable
7820 relay addresses has also been removed.
7822 56. Added hosts_avoid_esmtp to the smtp transport.
7824 57. The exicyclog script was not checking for the esoteric option
7825 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7826 will work only if exicyclog is run under the appropriate euid.
7828 58. Following a discussion on the list, the rules by which Exim recognises line
7829 endings on incoming messages have been changed. The -dropcr and drop_cr
7830 options are now no-ops, retained only for backwards compatibility. The
7831 following line terminators are recognized: LF CRLF CR. However, special
7832 processing applies to CR:
7834 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7835 nor a local message in the state where . is a terminator.
7837 (ii) If a bare CR is encountered in a header line, an extra space is added
7838 after the line terminator so as not to end the header. The reasoning
7839 behind this is that bare CRs in header lines are most likely either
7840 to be mistakes, or people trying to play silly games.
7842 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7843 was being incorrectly given as 18 bytes larger than it should have been.
7844 This is a VOB (very old bug).
7846 60. This may never have affected anything current, but just in case it has:
7847 When the local host is found other than at the start of a list of hosts,
7848 the local host, those with the same MX, and any that follow, are discarded.
7849 When the list in question was part of a longer list of hosts, the following
7850 hosts (not currently being processed) were also being discarded. This no
7851 longer happens. I'm not sure if this situation could ever has previously
7854 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7857 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7858 header. This is something that is recommended in a new Internet Draft, and
7859 is something that is documented as being done by Sendmail. There are two
7860 possible values. For messages generated by the autoreply transport, Exim
7863 Auto-Submitted: auto-replied
7865 whereas for all other generated messages (e.g. bounces) it adds
7867 Auto-Submitted: auto-generated
7869 63. The "personal" condition in filters now includes a test for the
7870 Auto-Submitted: header. If it contains the string "auto-" the message it
7871 not considered personal.
7873 64. Added rcpt_include_affixes as a generic transport option.
7875 65. Added queue_only_override (default true).
7877 66. Added the syslog_duplication option.
7879 67. If what should have been the first header line of a message consisted of
7880 a space followed by a colon, Exim was mis-interpreting it as a header line.
7881 It isn't of course - it is syntactically invalid and should therefore be
7882 treated as the start of the message body. The misbehaviour could have
7883 caused a number of strange effects, including loss of data in subsequent
7884 header lines, and spool format errors.
7886 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7887 client host had authenticated. This control can now be exercised by an ACL
7888 for more flexibility.
7890 69. By default, callouts do not happen when testing with -bh. There is now a
7891 variant, -bhc, which does actually run the callout code, including
7892 consulting and updating the callout cache.
7894 70. Added support for saslauthd authentication, courtesy of Alexander
7897 71. If statvfs() failed on the spool or log directories while checking their
7898 size for availability, Exim confusingly gave the error "space shortage".
7899 Furthermore, in debugging mode it crashed with a floating point exception.
7900 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7901 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7902 serious problem, Exim now writes to the main and panic logs when this
7903 happens, with details of the failure. It then refuses to accept the
7904 incoming message, giving the message "spool directory problem" or "log
7905 directory problem" with a 421 code for SMTP messages.
7907 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7908 0, 1, and 2 exist, because some OS complain for execs without them (see
7909 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7910 descriptors. However, the code omitted to check that the open succeeded,
7911 causing mysterious errors if for some reason the permissions on /dev/null
7912 got screwed. Now Exim writes a message to the main and panic logs, and
7913 bombs out if it can't open /dev/null.
7915 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7916 interact so that it is all more flexible. It is supposed to remain
7917 backwards compatible. Also added extra_local_interfaces.
7919 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7920 to bomb out with an assertion failure - to the client this appears as a
7921 connection drop. This problem occurs in the part of the code that was taken
7922 from the Samba project. Fortunately, the assertion is in a very simple
7923 function, so I have fixed this by reproducing the function inline in the
7924 one place where it is called, and arranging for authentication to fail
7925 instead of killing the process with assert().
7927 75. The SPA client code was not working when the server requested OEM rather
7928 than Unicode encoding.
7930 76. Added code to make require_files with a specific uid setting more usable in
7931 the case where statting the file as root fails - usually a non-root-mounted
7932 NFS file system. When this happens and the failure is EACCES, Exim now
7933 forks a subprocess and does the per-uid checking as the relevant uid.
7935 77. Added process_log_path.
7937 78. If log_file_path was not explicitly set, a setting of check_log_space or
7938 check_log_inodes was ignored.
7940 79. If a space check for the spool or log partitions fails, the incident is now
7941 logged. Of course, in the latter case the data may get lost...
7943 80. Added the %p formatting code to string_format() so that it can be used to
7944 print addresses in debug_print(). Adjusted all the address printing in the
7945 debugging in store.c to use %p rather than %d.
7947 81. There was a concern that a line of code in smtp_in.c could overflow a
7948 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7949 initially expressed, the concern was not well-founded, because trailing
7950 spaces are removed early. However, if the trailing spaces were followed by
7951 a NULL, they did not get removed, so the overflow was possible. Two fixes
7954 (a) I re-wrote the offending code in a cleaner fashion.
7955 (b) If an incoming SMTP command contains a NULL character, it is rejected
7958 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7959 runs initgroups(), so that if the Exim user is in any additional groups,
7960 they will be used during message reception.
7966 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob