1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
145 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
146 left undeleted; the attempt to re-create it then failed - resulting in
147 the usual "SIGHUP tp have daemon reload configuration" to not work.
148 This affected any platform not supporting "abstract" Unix-domain
149 sockets (i.e. not Linux).
151 JH/42 Bug 2692: Harden against a peer which reneges on a 452 "too many
152 recipients" response to RCPT in a later response, with a 250. The
153 previous coding assumed this would not happen, and under PIPELINING
154 would result in both lost and duplicate recipients for a message.
156 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
157 Previously the weighting was incorrectly applied. Similar fix for socks
158 proxies. Found and fixed by Heiko Schlichting.
160 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
161 not handle sub-lists included using the +namedlist syntax. While
162 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
163 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
166 HS/01 Enforce absolute PID file path name.
168 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
170 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
172 PP/02 Bug 2643: Correct TLS DH constants.
173 A missing NUL termination in our code-generation tool had led to some
174 incorrect Diffie-Hellman constants in the Exim source.
175 Reported by kylon94, code-gen tool fix by Simon Arlott.
177 PP/03 Impose security length checks on various command-line options.
178 Fixes CVE-2020-SPRSS reported by Qualys.
180 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
181 better. Reported by Qualys.
183 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
184 providing a particularly obnoxious sender full name.
187 pp/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
194 JH/01 Avoid costly startup code when not strictly needed. This reduces time
195 for some exim process initialisations. It does mean that the logging
196 of TLS configuration problems is only done for the daemon startup.
198 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
200 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
201 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
204 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
205 DKIM signing is being done. Previously a transport_filter would always
206 disable CHUNKING, falling back to traditional DATA.
208 JH/05 Regard command-line receipients as tainted.
210 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
212 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
213 PAM library frees one of the arguments given to it, despite the
214 documentation. Therefore a plain malloc must be used.
216 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
217 on-stack buffers were used, resulting in a taint trap when DSN information
218 copied from a received message was written into the buffer.
220 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
221 the ordering of its ARC headers. This caused a crash.
223 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
224 a new record was being constructed with information from the peer, a trap
227 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
228 installation would get error messages from DMARC verify, when it hit the
229 nonexistent file indicated by the default. Distros wanting DMARC enabled
230 should both provide the file and set the option.
231 Also enforce no DMARC verification for command-line sourced messages.
233 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
234 could, depending on the platform, hang at the STARTTLS response.
236 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
237 message on a connection. Previously if one message had ARC headers and
238 the following one did not, a crash could result when adding an
239 Authentication-Results: header.
241 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
242 Exim main code and Exim-related utities. The introduction of taint
243 tracking also did many adjustments to string handling. Since then, eximon
244 frequently terminated with an assert failure.
246 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
247 check for 452 responses. This slightly helps the inefficieny of doing
248 a large alias-expansion into a recipient-limited target. The max_rcpt
249 transport option still applies (and at the current default, will override
250 the new feature). The check is done for either cause of synch, and forces
251 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
252 connection. The new facility is not tunable at this time.
254 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
255 library live data was being used, so the results became garbage. Make
256 copies while it is still usable.
258 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
259 on delivery deferred (==) and failed (**) lines (if a delivery was
260 attemtped). Previously it was only on completion (=>) lines.
262 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
263 for the expansion of the server_scram_iter and server_scram_salt options.
265 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
266 are now specifically given a NO_DATA response without hitting the system
267 resolver. The library goes on to do the now-standard TXT lookup.
268 Use of dnsdb lookups is not affected.
270 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
271 only retrieve the errormessage once. Previously two calls to dlerror()
272 were used, and the second one (for mainlog/paniclog) retrieved null
275 JH/20 Taint checking: disallow use of tainted data for
276 - the appendfile transport file and directory options
277 - the pipe transport command
278 - the autoreply transport file, log and once options
279 - file names used by the redirect router (including filter files)
281 - paths used by single-key lookups
282 Previously this was permitted.
284 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
285 adjusted the size of a major service buffer; this failed because the
286 buffer was in use at the time. Change to a compile-time increase in the
287 buffer size, when this authenticator is compiled into exim.
289 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
290 previous fast-mode was untenable in the face of glibs using mmap to
291 support larger malloc requests.
293 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
294 New values supported, if defined on system where compiled:
295 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
296 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
298 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
299 running a limited number of proceses in parallel, a benefit is gained. The
300 amount varies with the platform hardware and load. The use of the option
301 queue_run_in_order means we cannot do this, as ordering becomes
304 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
305 had introduced a string-copy (for ensuring NUL-termination) which was not
306 appropriate for that case, which can include embedded NUL bytes in the
307 block of data. Investigation showed the copy to actually be needless, the
308 data being length-specified.
310 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
311 done during a receiving connection, and both used TLS, global info was
312 used rather than per-connection info for tracking the state of data
313 queued for transmission. This could result in a connection hang.
315 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
316 Previously, when delivering serveral messages down a single connection
317 only the first would provide a SIZE. This was due to the size information
318 not being properly tracked.
320 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
321 TAI (at 37 seconds currently), pretend to be in UTC for time-related
322 expansion and logging. Previously, spurious values such as a future
323 minute could be seen.
325 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
326 it could crash from a null-deref. This could also affect the
327 ${addresses: } operator and ${readsock } item.
329 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
330 message following a mime one, the variable was not reset.
332 JH/30 When an pipelined-connect fails at the first response, assume incorrect
333 cached capability (perhaps the peer reneged?) and immediately retry in
336 JH/31 Fix spurious detection of timeout while writing to transport filter.
338 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
339 an attempt to copy the string was made before checking it.
341 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
342 taint of the lookup key was maintained; we now regard the presence in the
343 filesystem as sufficient validation.
345 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
346 argument is supplied.
348 JH/35 The dsearch lookup now requires that the directory is an absolute path.
349 Previously this was not checked, and nonempty relative paths made an
350 access under Exim's current working directory.
352 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
353 Previously no event was raised.
355 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
356 parameter supplied by the sender MAIL FROM command. Previously it was
357 ignored, and only the check_spool_space option value for the required
360 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
361 the size of the signing public-key. Previously it was instead giving
362 the size of the signature hash.
364 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
365 the default. See the (new) dkim_verify_min_keysizes option.
367 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
368 and an ACL use a lookup for checking either the local_part or domain,
369 stale data could be accessed. Ensure that variable references are
370 dropped between messages.
372 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
373 by the client was not checked as pointing within response data before
374 being used. A malicious client could thus cause an out-of-bounds read and
375 possibly gain authentication. Fix by adding the check.
377 JH/42 Internationalisation: change the default for downconversion in the smtp
378 transport to be "if needed". Previously it was "as previously set" for
379 the message, which usually meant "if needed" for message-submission but
380 "no" for everything else. However, MTAs have been seen using SMTPUTF8
381 even when the envelope addresses did not need it, resulting in forwarding
382 failures to non-supporting MTAs. A downconvert in such cases will be
383 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
384 transport. The change does mean that addresses needing conversion will
385 be converted when previously a delivery failure would occur.
387 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
388 response was received it would be used unchecked in a fail-DSN, violating
389 standards on line-length limits. Truncate if needed.
391 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
392 doesn't work. (Additionally add a new main config option to configure the
399 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
400 side, to decode a TLS 1.3 packet capture.
402 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
403 Previously the default library behaviour applied, sending two, each in
406 JH/03 Debug output for ACL now gives the config file name and line number for
409 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
411 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
413 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
414 buffer overrun for (non-chunking) other transports.
416 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
417 TLS1.3, means that a server rejecting a client certificate is not visible
418 to the client until the first read of encrypted data (typically the
419 response to EHLO). Add detection for that case and treat it as a failed
420 TLS connection attempt, so that the normal retry-in-clear can work (if
421 suitably configured).
423 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
424 and/or domain. Found and fixed by Jason Betts.
426 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
427 configuration). If a CNAME target was not a wellformed name pattern, a
430 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
431 the OS reports them interleaved with other addresses.
433 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
434 used both for input and for a verify callout, both encrypted, SMTP
435 responses being sent by the server could be lost. This resulted in
436 dropped connections and sometimes bounces generated by a peer sending
439 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
440 string. Previously it was possible to cause undefined behaviour in a
441 library routine (usually a crash). Found by "zerons".
443 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
446 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
447 API was removed, so update to use the newer ones.
449 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
450 any timeout set, is taking a long time. Previously we would hang on to a
451 rotated logfile "forever" if the input was arriving with long gaps
452 (a previous attempt to fix addressed lack, for a long time, of initial
455 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
456 shared (NFS) environment. The length of the tempfile name is now
457 4 + 16 ("hdr.$message_exim_id") which might break on file
458 systems which restrict the file name length to lower values.
461 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
462 shared (NFS) environment.
464 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
465 did for all versions <4.90). Notably -M, -m, --invert, -I may be
468 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
469 on some platforms for bit 31.
471 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
472 to changes apparently associated with TLS1.3 handling some of the APIs
473 previously used were either nonfunctional or inappropriate. Strings
474 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
475 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
476 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
477 This affects log line X= elements, the $tls_{in,out}_cipher variables,
478 and the use of specific cipher names in the encrypted= ACL condition.
480 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
482 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
483 verification result was not updated unless hosts_require_ocsp applied.
485 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
486 queue_list_requires_admin set to false, non-admin users were denied the
489 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
490 directory-of-certs mode. Previously they were advertised despite the
493 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
494 A single TCP connection by a client will now hold a TLS connection open
495 for multiple message deliveries, by default. Previously the default was to
498 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
499 default. If built with the facility, DANE will be used. The facility
500 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
502 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
503 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
504 must be defined and you must still, unless you define DISABLE_TLS, manage
505 the the include-dir and library-file requirements that go with that
506 choice. Non-TLS builds are still supported.
508 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
511 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
512 default. If the platform supports and has the facility enabled, it will
513 be requested on all coneections.
515 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
516 controlled by the build-time option SUPPORT_PIPE_CONNECT.
518 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
520 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
521 success-DSN messages. Previously the From: header was always the default
522 one for these; the option was ignored.
524 JH/28 Fix the timeout on smtp response to apply to the whole response.
525 Previously it was reset for every read, so a teergrubing peer sending
526 single bytes within the time limit could extend the connection for a
527 long time. Credit to Qualsys Security Advisory Team for the discovery.
529 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
530 delivery address, which leaked information of the results of local
531 forwarding. Change to the original envelope recipient address, per
534 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
535 requested. Previously not bounce was generated and a log entry of
536 error ignored was made.
538 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
540 JH/32 Introduce a general tainting mechanism for values read from the input
541 channel, and values derived from them. Refuse to expand any tainted
542 values, to catch one form of exploit.
544 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
545 was unused and the unexpanded text used for the test. Found and
546 fixed by Ruben Jenster.
548 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
549 an attempt to use a TLS library read routine dereffed a nul pointer,
552 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
553 them in our smtp response.
555 JH/36 Have the general router option retry_use_local_part default to true when
556 any of the restrictive preconditions are set (to anything). Previously it
557 was only for check_local user. The change removes one item of manual
558 configuration which is required for proper retries when a remote router
559 handles a subset of addresses for a domain.
561 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
562 link count into consideration.
564 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
565 caused the extension of big_buffer, the following lines were ignored.
567 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
568 accordance with RFC 2308. Previously there was no expiry, so a longlived
569 receive process (eg. due to ACL delays) versus a short SOA value could
572 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
574 JH/39 Promote DMARC support to mainline.
576 JH/40 Bug 2452: Add a References: header to DSNs.
578 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
579 parameters. The relevant library call is documented as "Deprecated: This
580 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
581 3.6.0, DH parameters are negotiated following RFC7919."
583 HS/06 Change the default of dnssec_request_domains to "*"
585 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
586 carried on and emitted a BDAT command, even when PIPELINING was not
589 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
590 buffer was used for the filename, resulting in a trap when tainted
591 arguments (eg. $domain) were used.
593 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
594 recommended to avoid a possible server-load attack. The feature can be
595 re-enabled via the openssl_options main cofiguration option.
597 JH/45 local_scan API: documented the current smtp_printf() call. This changed
598 for version 4.90 - adding a "more data" boolean to the arguments.
599 Bumped the ABI version number also, this having been missed previously;
600 release versions 4.90 to 4.92.3 inclusive were effectively broken in
601 respect of usage of smtp_printf() by either local_scan code or libraries
602 accessed via the ${dlfunc } expansion item. Both will need coding
603 adjustment for any calls to smtp_printf() to match the new function
604 signature; a FALSE value for the new argument is always safe.
606 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
607 the file-offset (which the Linux syscall does, and exim expects); this
608 resulted in an indefinite loop.
610 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
611 to do ARC verification. The Authentication-Results: header line added
612 by the configuration then had no ARC item.
618 JH/01 Remove code calling the customisable local_scan function, unless a new
619 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
621 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
622 non-signal-safe functions being used.
624 JH/03 Bug 2269: When presented with a received message having a stupidly large
625 number of DKIM-Signature headers, disable DKIM verification to avoid
626 a resource-consumption attack. The limit is set at twenty.
628 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
629 report of oldest_pass in ${authres } in consequence, and separate out
630 some descriptions of reasons for verification fail.
632 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
633 files in the spool were present and unlocked. A queue-runner could spot
634 them, resulting in a duplicate delivery. Fix that by doing the unlock
635 after the unlink. Investigation by Tim Stewart. Take the opportunity to
636 add more error-checking on spoolfile handling while that code is being
639 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
640 No known attacks, no CVE, this is defensive hardening.
642 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
643 a queue-runner could start a delivery while other operations were ongoing.
644 Cutthrough delivery was a common victim, resulting in duplicate delivery.
645 Found and investigated by Tim Stewart. Fix by using the open message data
646 file handle rather than opening another, and not locally closing it (which
647 releases a lock) for that case, while creating the temporary .eml format
648 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
650 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
651 $sender_verify_failure/$recipient_verify_failure to "random".
653 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
656 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
657 Previously this would segfault.
659 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
662 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
663 like zero, since the resolver should be doing this for us, But we need one
664 as a CNAME but no MX presence gets the CNAME returned; we need to check
665 that doesn't point to an MX to declare it "no MX returned" rather than
666 "error, loop". A new main option is added so the older capability of
667 following some limited number of chain links is maintained.
669 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
671 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
672 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
673 now-deprecated earlier definitions used only specified the range up to TLS
674 1.2 (in the older-version library docs).
676 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
678 JH/15 Rework TLS client-side context management. Stop using a global, and
679 explicitly pass a context around. This enables future use of TLS for
680 connections to service-daemons (eg. malware scanning) while a client smtp
681 connection is using TLS; with cutthrough connections this is quite likely.
683 JH/16 Fix ARC verification to do AS checks in reverse order.
685 JH/17 Support a "tls" option on the ${readsocket } expansion item.
687 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
688 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
689 Previously the "utf8" would be re-prepended for every additional message.
691 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
692 Previously thery were accepted, resulting in issues when attempting to
693 forward messages to a non-supporting MTA.
695 PP/02 Let -n work with printing macros too, not just options.
697 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
698 one parent address was copied, and bogus data was used at delivery-logging
699 time. Either a crash (after delivery) or bogus log data could result.
700 Discovery and analysis by Tim Stewart.
702 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
703 Previously if the string ended mid-character, we did not insert the
704 promised '?' replacement.
706 PP/04 Documentation: current string operators work on bytes, not codepoints.
708 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
709 should pack well giving a smaller memory footprint so better caching and
710 therefore performance. Group the declarations where this can't be done so
711 that the byte-sized flag variables are not interspersed among pointer
712 variables, giving a better chance of good packing by the compiler.
714 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
715 non-null, to avoid issues with sites running BATV. Previously reports were
716 sent with an empty envelope sender so looked like bounces.
718 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
719 The ignore_error flag wasn't being returned from the filter subprocess so
720 was not set for later routers. Investigation and fix by Matthias Kurz.
722 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
723 and a msg:complete for the whole, when a message is manually removed using
724 -Mrm. Developement by Matthias Kurz, hacked on by JH.
726 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
727 a "Gnu special" function, asprintf() in the DB utility binary builds; I
728 hope that is portable enough.
730 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
731 requiring a known-CA anchor certificate; make it now rely entirely on the
732 TLSA as an anchor. Checking the name on the leaf cert against the name
733 on the A-record for the host is still done for TA (but not for EE mode).
735 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
736 included in delivery lines for non-proxied connections, when compiled with
737 SUPPORT_SOCKS and running with proxy logging enabled.
739 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
740 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
741 move the existing event to fire before the normal logging of message
742 failure so that custom logging is bracketed by normal logging.
744 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
745 msg:fail:internal event. Developement by Matthias Kurz.
747 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
748 far too small for todays use of crypto signatures stored there. Go all
749 the way to the max DNS message size of 64kB, even though this might be
750 overmuch for IOT constrained device use.
752 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
753 copy a string over itself. The library routine is documented as not
754 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
756 JH/32 For main options check_spool_space and check_inode_space, where the
757 platform supports 64b integers, support more than the previous 2^31 kB
758 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
759 the previous G, M, k.
761 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
762 $authenticated_fail_id variable on authentication failure. Previously
765 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
766 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
767 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
768 GNUTLS_SEC_PARAM_MEDIUM.
770 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
771 side. Previously we would continue as if no SNI had been received.
773 JH/36 Harden the handling of string-lists. When a list consisted of a sole
774 "<" character, which should be a list-separator specification, we walked
775 off past the nul-terimation.
777 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
778 causes) even when the retry time is not yet met. Previously they were
779 not, meaning that when (say) an account was over-quota and temp-rejecting,
780 and multiple senders' messages were queued, only one sender would get
781 notified on each configured delay_warning cycle.
783 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
785 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
786 cipher-suites, an error can be left on the stack even for a succeeding
787 accept; this results in impossible error messages when a later operation
790 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
791 return error codes indicating retry. Under TLS1.3 this becomes required.
793 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
794 it only wrote the new authenticators, resulting in a lack of tracking of
795 peer changes of ESMTP extensions until the next cache flush.
797 JH/41 Fix the loop reading a message header line to check for integer overflow,
798 and more-often against header_maxsize. Previously a crafted message could
799 induce a crash of the recive process; now the message is cleanly rejected.
801 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
802 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
808 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
809 When redis_servers is set to a list of > 1 element, and the Redis servers
810 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
811 case of MOVED into a DEFER case instead, thus moving the query onto the
812 next server in the list. For a cluster of N elements, all N servers must
813 be defined in redis_servers.
815 GF/02 Catch and remove uninitialized value warning in exiqsumm
816 Check for existence of @ARGV before looking at $ARGV[0]
818 JH/01 Replace the store_release() internal interface with store_newblock(),
819 which internalises the check required to safely use the old one, plus
820 the allocate and data copy operations duplicated in both (!) of the
821 extant use locations.
823 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
824 modifier. This matches the restriction on the commandline.
826 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
827 Previously only the last row was returned.
829 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
830 we assumed that tags in the header were well-formed, and parsed the
831 element content after inspecting only the first char of the tag.
832 Assumptions at that stage could crash the receive process on malformed
835 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
836 While running the DKIM ACL we operate on the Permanent memory pool so that
837 variables created with "set" persist to the DATA ACL. Also (at any time)
838 DNS lookups that fail create cache records using the Permanent pool. But
839 expansions release any allocations made on the current pool - so a dnsdb
840 lookup expansion done in the DKIM ACL releases the memory used for the
841 DNS negative-cache, and bad things result. Solution is to switch to the
842 Main pool for expansions.
843 While we're in that code, add checks on the DNS cache during store_reset,
844 active in the testsuite.
845 Problem spotted, and debugging aided, by Wolfgang Breyha.
847 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
848 When none of the hosts presented to a transport match an already-open
849 connection, close it and proceed with the list. Previously we would
850 queue the message. Spotted by Lena with Yahoo, probably involving
853 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
854 Previously a spurious "250 OK id=" response was appended to the proper
857 JH/08 The "support for" informational output now, which built with Content
858 Scanning support, has a line for the malware scanner interfaces compiled
859 in. Interface can be individually included or not at build time.
861 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
862 by the template makefile "src/EDITME". The "STREAM" support for an older
863 ClamAV interface method is removed.
865 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
866 rows affected is given instead).
868 JH/11 The runtime Berkeley DB library version is now additionally output by
869 "exim -d -bV". Previously only the compile-time version was shown.
871 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
872 SMTP connection. Previously, when one had more receipients than the
873 first, an abortive onward connection was made. Move to full support for
874 multiple onward connections in sequence, handling cutthrough connection
875 for all multi-message initiating connections.
877 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
878 routers. Previously, a multi-recipient message would fail to match the
879 onward-connection opened for the first recipient, and cause its closure.
881 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
882 a timeout on read on a GnuTLS initiating connection, resulting in the
883 initiating connection being dropped. This mattered most when the callout
884 was marked defer_ok. Fix to keep the two timeout-detection methods
887 JH/15 Relax results from ACL control request to enable cutthrough, in
888 unsupported situations, from error to silently (except under debug)
889 ignoring. This covers use with PRDR, frozen messages, queue-only and
892 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
894 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
895 metadata, resulting in a crash in free().
897 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
898 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
899 Broken also in d185889f4, with init system revamp.
901 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
902 Previously we abruptly closed the connection after reading a malware-
903 found indication; now we go on to read the "scan ok" response line,
906 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
907 ACL. Previously, a crash would result.
909 JH/19 Speed up macro lookups during configuration file read, by skipping non-
910 macro text after a replacement (previously it was only once per line) and
911 by skipping builtin macros when searching for an uppercase lead character.
913 JH/20 DANE support moved from Experimental to mainline. The Makefile control
914 for the build is renamed.
916 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
917 was allocated for every new TLS startup, meaning one per message. Fix
918 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
920 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
921 reported the original. Fix to report (as far as possible) the ACL
922 result replacing the original.
924 JH/23 Fix memory leak during multi-message connections using STARTTLS under
925 OpenSSL. Certificate information is loaded for every new TLS startup,
926 and the resources needed to be freed.
928 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
930 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
933 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
934 DATA response info to the (existing) per-recipient response info for
935 the "C=" log element. It can have useful tracking info from the
936 destination system. Patch from Simon Arlott.
938 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
939 length value. Previously this would segfault.
941 HS/02 Support Avast multiline protoocol, this allows passing flags to
942 newer versions of the scanner.
944 JH/28 Ensure that variables possibly set during message acceptance are marked
945 dead before release of memory in the daemon loop. This stops complaints
946 about them when the debug_store option is enabled. Discovered specifically
947 for sender_rate_period, but applies to a whole set of variables.
948 Do the same for the queue-runner and queue-list loops, for variables set
949 from spool message files. Do the same for the SMTP per-message loop, for
950 certain variables indirectly set in ACL operations.
952 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
953 as a multi-recipient message from a mailinglist manager). The coding had
954 an arbitrary cutoff number of characters while checking for more input;
955 enforced by writing a NUL into the buffer. This corrupted long / fast
956 input. The problem was exposed more widely when more pipelineing of SMTP
957 responses was introduced, and one Exim system was feeding another.
958 The symptom is log complaints of SMTP syntax error (NUL chars) on the
959 receiving system, and refused recipients seen by the sending system
960 (propating to people being dropped from mailing lists).
961 Discovered and pinpointed by David Carter.
963 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
964 replaced by the ${authresults } expansion.
966 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
968 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
969 allows proper process termination in container environments.
971 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
972 Previously the "final dot" had a newline after it; ensure it is CR,LF.
974 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
975 and "err_perm", deprecated since 4.83 when the RFC-defined words
976 "temperror" and "permerror" were introduced.
978 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
979 transport-filters or DKIM-signing. The restriction was lost in the
980 consolidation of verify-callout and delivery SMTP handling.
981 Extend the restriction to also cover ARC-signing.
983 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
984 in defer=pass mode supply a 450 to the initiator. Previously the message
987 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
988 tls_require_ciphers is used as before.
990 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
991 "pass_unscanned". Only tmpfails from the scanner are written to
992 the paniclog, as they may require admin intervention (permission
993 denied, license issues). Other scanner errors (like decompression
994 bombs) do not cause a paniclog entry.
996 JH/36 Fix reinitialisation of DKIM logging variable between messages.
997 Previously it was possible to log spurious information in receive log
1000 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1001 triggered odd behaviour from Outlook Express clients.
1003 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1006 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1007 since the IETF WG has not yet settled on that versus the original
1008 "bare" representation.
1010 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1011 Previously the millisecond value corrupted the output.
1012 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1013 corrupted the output.
1019 JH/01 Rework error string handling in TLS interface so that the caller in
1020 more cases is responsible for logging. This permits library-sourced
1021 string to be attached to addresses during delivery, and collapses
1022 pairs of long lines into single ones.
1024 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1025 during configuration. Wildcards are allowed and expanded.
1027 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1028 This permits better logging.
1030 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1031 do not close it down and have the child transport start it up again on
1032 the passed-on TCP connection. Instead, proxy the child (and any
1033 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1034 affected: the continued delivery log lines do not have any DNSSEC, TLS
1035 Certificate or OCSP information. TLS cipher information is still logged.
1037 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1038 identical IP addresses on different listening ports. Will also affect
1041 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1042 add noisy ifdef guards to special-case this sillyness.
1043 Patch from Bernd Kuhls.
1045 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1046 than 255 are no longer allowed.
1048 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1049 Disable the session-cache too, which might reduce our load. Since we
1050 currrectly use a new context for every connection, both as server and
1051 client, there is no benefit for these.
1052 GnuTLS appears to not support tickets server-side by default (we don't
1053 call gnutls_session_ticket_enable_server()) but client side is enabled
1054 by default on recent versions (3.1.3 +) unless the PFS priority string
1057 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1058 <https://reproducible-builds.org/specs/source-date-epoch/>.
1060 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1061 the check for any unsuccessful recipients did not notice the limit, and
1062 erroneously found still-pending ones.
1064 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1065 MSG_MORE. Only in-clear (not on TLS connections).
1067 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1068 transport-filter is involved we can buffer the headers in memory for
1069 creating the signature, and read the spool data file once for the
1070 signature and again for transmission.
1072 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1073 4.77 as the kernel support then wasn't solid, having issues in 64bit
1074 mode. Now, it's been long enough. Add support for FreeBSD also.
1076 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1077 case where the routing stage had gathered several addresses to send to
1078 a host before calling the transport for the first, we previously failed
1079 to close down TLS in the old transport process before passing the TCP
1080 connection to the new process. The new one sent a STARTTLS command
1081 which naturally failed, giving a failed delivery and bloating the retry
1082 database. Investigation and fix prototype from Wolfgang Breyha.
1084 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1085 false-negatives in the check that the sender had not preempted a response
1086 or prompt from Exim (running as a server), due to that code's lack of
1087 awareness of the SMTP input buffering.
1089 PP/04 Add commandline_checks_require_admin option.
1090 Exim drops privileges sanely, various checks such as -be aren't a
1091 security problem, as long as you trust local users with access to their
1092 own account. When invoked by services which pass untrusted data to
1093 Exim, this might be an issue. Set this option in main configuration
1094 AND make fixes to the calling application, such as using `--` to stop
1097 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1098 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1099 responses to those, into a single TLS record each way (this usually means
1100 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1103 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1104 affects you only if you're dancing at the edge of the param size limits.
1105 If you are, and this message makes sense to you, then: raise the
1106 configured limit or use OpenSSL 1.1. Nothing we can do for older
1109 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1110 cmdline element to get the documented default one. Previously it was
1113 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1114 are unsuable for non-message-specific reasons - eg. connection timeout,
1115 banner-time rejection.
1117 JH/16 Fix logging of delivery remote port, when specified by router, under
1120 PP/06 Repair manualroute's ability to take options in any order, even if one
1121 is the name of a transport.
1124 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1126 JH/17 Change the list-building routines interface to use the expanding-string
1127 triplet model, for better allocation and copying behaviour.
1129 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1130 Previously it was constructed the first time a possibly-matching string
1131 was met in the configuration file input during startup; now it is done
1134 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1135 compatible one, to avoid the (poorly documented) possibility of a config
1136 file in the working directory redirecting the DB files, possibly correpting
1137 some existing file. CVE-2017-10140 assigned for BDB.
1139 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1140 cache-hot. Previously, although the result was properly cached, the
1141 initial verify call returned a defer.
1143 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1144 the main verify for receipient in uncached-mode.
1146 JH/22 Retire historical build files to an "unsupported" subdir. These are
1147 defined as "ones for which we have no current evidence of testing".
1149 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1150 if present. Previously it was ignored.
1152 JH/24 Start using specified-initialisers in C structure init coding. This is
1153 a C99 feature (it's 2017, so now considered safe).
1155 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1156 if was a fixed-sized field and bitmask ops via macros; it is now more
1159 PP/07 GitHub PR 56: Apply MariaDB build fix.
1160 Patch provided by Jaroslav Škarvada.
1162 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1163 during Coverity cleanups [4.87 JH/47]
1164 Diagnosis and fix provided by Michael Fischer v. Mollard.
1166 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1167 the right size to place the terminating semicolon on its own folded
1168 line, the header hash was calculated to an incorrect value thanks to
1169 the (relaxed) space the fold became.
1171 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1172 and confused the parent.
1174 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1175 which could crash as a result. This could lead to undeliverable messages.
1177 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1180 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1181 matters on fast-turnover and PID-randomising systems, which were getting
1182 out-of-order delivery.
1184 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1185 a possibly-overlapping copy. The symptom was that "Remote host closed
1186 connection in response to HELO" was logged instead of the actual 4xx
1189 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1190 Previously only that bufferd was discarded, resulting in SYMTP command
1193 JH/32 DKIM: when a message has multiple signatures matching an identity given
1194 in dkim_verify_signers, run the dkim acl once for each. Previously only
1195 one run was done. Bug 2189.
1197 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1198 "panic the current process" to "deliberately defer". The panic log is
1199 still written with the problem list name; the mail and reject logs now
1200 get a temp-reject line for the message that was being handled, saying
1201 something like "domains check lookup or other defer". The SMTP 451
1202 message is still "Temporary local problem".
1204 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1205 A crafted sequence of BDAT commands could result in in-use memory beeing
1206 freed. CVE-2017-16943.
1208 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1209 from SMTP input. Previously it was always done; now only done for DATA
1210 and not BDAT commands. CVE-2017-16944.
1212 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1213 to the message (such as an overlong header line). Previously this was
1214 not done and we did not exit BDAT mode. Followon from the previous item
1215 though a different problem.
1221 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1222 than -2003 did; needs libidn2 in addition to libidn.
1224 JH/02 The path option on a pipe transport is now expanded before use.
1226 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1227 Patch provided by "Björn", documentation fix added too.
1229 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1230 missing a wire-to-host endian conversion.
1232 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1233 close after a BDAT command line could be taken as a following command,
1234 giving a synch failure. Fix by only checking for synch immediately
1235 before acknowledging the chunk.
1237 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1238 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1239 macro. Patches provided by Josh Soref.
1241 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1242 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1243 by IANA as a service extension. However, John Klensin suggests that we
1246 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1247 the dkim code may be unix-mode line endings rather than smtp wire-format
1248 CRLF, so prepend a CR to any bare LF.
1250 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1251 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1253 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1254 the first were themselves being wrongly included in the feed into dkim
1255 processing; with most chunk sizes in use this resulted in an incorrect
1256 body hash calculated value.
1258 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1259 DKIM signature block, for verification. Although advised against by
1260 standards it is specifically not ruled illegal.
1262 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1264 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1265 missing a body hash (the bh= tag).
1267 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1268 It seems that HAProxy sends the Proxy Protocol information in clear and
1269 only then does a TLS startup, so do the same.
1271 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1272 TCP connections (such as for Spamd) unless the daemon successfully set
1273 Fast Open mode on its listening sockets. This fixes breakage seen on
1274 too-old kernels or those not configured for Fast Open, at the cost of
1275 requiring both directions being enabled for TFO, and TFO never being used
1276 by non-daemon-related Exim processes.
1278 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1279 endings, at least on the first header line. Try to canonify any that get
1280 past that check, despite the cost.
1282 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1283 now limited to an arbitrary five deep, while parsing addresses with the
1284 strip_excess_angle_brackets option enabled.
1286 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1287 instead leave the unprompted TLS handshake in socket buffer for the
1288 TLS library to consume.
1290 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1292 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1294 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1295 region in one operation, for speed, and this leaves assigned pointers
1296 dangling. Add checks run only under the testsuite which checks all
1297 variables at a store-reset and panics on a dangling pointer; add code
1298 explicitly nulling out all the variables discovered. Fixes one known
1299 bug: a transport crash, where a dangling pointer for $sending_ip_address
1300 originally assigned in a verify callout, is re-used.
1302 PP/06 Drop '.' from @INC in various Perl scripts.
1304 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1306 PP/08 Reduce a number of compilation warnings under clang; building with
1307 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1308 should be warning-free.
1310 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1312 HS/01 Fix portability problems introduced by PP/08 for platforms where
1313 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1315 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1316 chunk. This allows us to accept broken chunked messages. We need a more
1317 general solution here.
1319 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1320 already-broken messages in the queue.
1322 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1324 JH/19 Fix reference counting bug in routing-generated-address tracking.
1330 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1331 supports it and a size is available (ie. the sending peer gave us one).
1333 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1334 years of being deprecated). The replacements are the ACLs
1335 acl_smtp_mime and acl_not_smtp_mime.
1337 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1338 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1339 or even an in-clear connection were permitted. Now, if the host lookup
1340 was dnssec and dane was requested then the host is only used if the
1341 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1342 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1343 if one fails this test.
1344 This means that a poorly-configured remote DNS will make it incommunicado;
1345 but it protects against a DNS-interception attack on it.
1347 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1348 non-noisy when a race steals the message being considered.
1350 JH/05 If main configuration option tls_certificate is unset, generate a
1351 self-signed certificate for inbound TLS connections.
1353 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1354 in rewrites and routers.
1356 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1357 and logged a warning sing 4.83; now they are a configuration file error.
1359 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1360 (lacking @domain). Apply the same qualification processing as RCPT.
1362 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1364 JH/10 Support ${sha256:} applied to a string (as well as the previous
1367 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1368 a cutthrough deliver is pending, as we always want to make a connection.
1369 This also avoids re-routing the message when later placing the cutthrough
1370 connection after a verify cache hit.
1371 Do not update it with the verify result either.
1373 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1374 when routing results in more than one destination address.
1376 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1377 signing (which inhibits the cutthrough capability). Previously only
1378 the presence of an option was tested; now an expansion evaluating as
1379 empty is permissible (obviously it should depend only on data available
1380 when the cutthrough connection is made).
1382 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1383 the relevant preceding SMTP command did not note the pipelining mode.
1385 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1386 Previously they were not counted.
1388 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1389 as one having no matching records. Previously we deferred the message
1390 that needed the lookup.
1392 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1393 distinguished as "(=".
1395 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1396 for missing MX records. Previously it only worked for missing A records.
1398 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1400 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1401 after the data-go-ahead and data-ack. Patch from Jason Betts.
1403 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1404 even for a "none" policy. Patch from Tony Meyer.
1406 JH/22 Fix continued use of a connection for further deliveries. If a port was
1407 specified by a router, it must also match for the delivery to be
1410 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1411 When one of the recipients of a message was unsuitable for the connection
1412 (has no matching addresses), we lost track of needing to mark it
1413 deferred. As a result mail would be lost.
1415 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1417 JH/25 Decoding ACL controls is now done using a binary search; the source code
1418 takes up less space and should be simpler to maintain. Merge the ACL
1419 condition decode tables also, with similar effect.
1421 JH/26 Fix problem with one_time used on a redirect router which returned the
1422 parent address unchanged. A retry would see the parent address marked as
1423 delivered, so not attempt the (identical) child. As a result mail would
1426 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1427 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1428 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1431 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1432 Main config options check_{log,spool}_{inodes,space} are now
1433 100 inodes, 10MB unless set otherwise in the configuration.
1435 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1436 Previously it only applied to the main-section connection policy
1439 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1441 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1442 by me. Added RFC7919 DH primes as an alternative.
1444 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1445 are not in the system include path.
1447 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1448 GnuTLS, when a session startup failed (eg because the client disconnected)
1449 Exim did stdio operations after fclose. This was exposed by a recent
1450 change which nulled out the file handle after the fclose.
1452 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1453 signed directly by the cert-signing cert, rather than an intermediate
1454 OCSP-signing cert. This is the model used by LetsEncrypt.
1456 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1458 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1459 an incoming connection.
1461 HS/02 Bug 1802: Do not half-close the connection after sending a request
1464 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1465 fallback to "prime256v1".
1467 JH/34 SECURITY: Use proper copy of DATA command in error message.
1468 Could leak key material. Remotely exploitable. CVE-2016-9963.
1474 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1475 and 3.4.4 - once the server is enabled to respond to an OCSP request
1476 it does even when not requested, resulting in a stapling non-aware
1477 client dropping the TLS connection.
1479 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1480 support variable-length bit vectors. No functional change.
1482 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1483 The I= interface field on outgoing lines is now after the H= remote
1484 host field, same as incoming lines. There is a separate
1485 outgoing_interface log selector which allows you to disable the
1488 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1489 If not running log_selector +smtp_connection the mainlog would be held
1490 open indefinitely after a "too many connections" event, including to a
1491 deleted file after a log rotate. Leave the per net connection logging
1492 leaving it open for efficiency as that will be quickly detected by the
1493 check on the next write.
1495 HS/01 Bug 1671: Fix post transport crash.
1496 Processing the wait-<transport> messages could crash the delivery
1497 process if the message IDs didn't exist for some reason. When
1498 using 'split_spool_directory=yes' the construction of the spool
1499 file name failed already, exposing the same netto behaviour.
1501 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1502 mime_regex ACL conditions.
1504 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1505 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1506 to HELO, local diagnostic string.
1508 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1509 log line to debug. Even when configured with a tls authenticator many
1510 client connections are expected to not authenticate in this way, so
1511 an authenticate fail is not an error.
1513 HS/02 Add the Exim version string to the process info. This way exiwhat
1514 gives some more detail about the running daemon.
1516 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1517 matter for fast-change records such as DNSBLs.
1519 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1520 retry record, even if constant. There may be multiple transports with
1521 different interface settings and the retry behaviour needs to be kept
1524 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1526 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1528 JH/10 Bug 840: fix log_defer_output option of pipe transport
1530 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1531 hosts_randomize. This matters a lot when combined with mua_wrapper.
1533 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1534 ${quote_pgsql:<string>} operator.
1536 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1538 JH/14 Change header file name for freeradius-client. Relevant if compiling
1539 with Radius support; from the Gentoo tree and checked under Fedora.
1541 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1543 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1544 option transport_filter, meaning no filtering.
1546 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1548 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1549 defaults to "*" (all hosts). The variable is now available when not built
1550 with TLS, default unset, mainly to enable keeping the testsuite sane.
1551 If a server certificate is not supplied (via tls_certificate) an error is
1552 logged, and clients will find TLS connections fail on startup. Presumably
1553 they will retry in-clear.
1554 Packagers of Exim are strongly encouraged to create a server certificate
1555 at installation time.
1557 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1558 with the $config_file variable.
1560 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1561 in transport context, after the attempt, and per-recipient. The latter type
1562 is per host attempted. The event data is the error message, and the errno
1563 information encodes the lookup type (A vs. MX) used for the (first) host,
1564 and the trailing two digits of the smtp 4xx response.
1566 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1567 to write to mainlog (or rejectlog, paniclog) in the window between file
1568 creation and permissions/ownership being changed. Particularly affects
1569 installations where exicyclog is run as root, rather than exim user;
1570 result is that the running daemon panics and dies.
1572 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1574 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1575 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1576 "pri" and "weight". Note that the previous implicit priority given by the
1577 list order is no longer honoured.
1579 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1580 for DKIM processing.
1582 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1583 by defining SUPPORT_SOCKS.
1585 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1586 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1587 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1588 variables are renamed to proxy_{local,external}_{address,port}.
1590 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1591 for a build by defining SUPPORT_I18N
1593 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1594 of the query string, and make ${quote_redis:} do that quoting.
1596 JH/29 Move Events support from Experimental to mainline, enabled by default
1597 and removable for a build by defining DISABLE_EVENT.
1599 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1601 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1602 cached by the daemon.
1604 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1605 by defining LOOKUP_REDIS. The libhiredis library is required.
1607 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1608 keys are given for lookup.
1610 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1611 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1612 only supported when built with TLS support. The PolarSSL SHA routines
1613 are still used when the TLS library is too old for convenient support.
1615 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1616 openssl_options), for security. OpenSSL forces this from version 1.1.0
1617 server-side so match that on older versions.
1619 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1620 allocation for $value could be released as the expansion processing
1621 concluded, but leaving the global pointer active for it.
1623 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1624 and to use the domains and local_parts ACL conditions.
1626 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1627 incorrectly not doubled on cutthrough transmission, hence seen as a
1628 body-termination at the receiving system - resulting in truncated mails.
1629 Commonly the sender saw a TCP-level error, and retransmitted the message
1630 via the normal store-and-forward channel. This could result in duplicates
1631 received - but deduplicating mailstores were liable to retain only the
1632 initial truncated version.
1634 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1636 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1638 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1639 we're in there, support oversigning also; bug 1309.
1641 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1643 HS/04 Add support for keep_environment and add_environment options.
1645 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1646 either intentional arithmetic overflow during PRNG, or testing config-
1649 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1650 delivery resulted in actual delivery. Cancel cutthrough before DATA
1653 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1654 rejects a recipient: pass the reject to the originator.
1656 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1657 Many were false-positives and ignorable, but it's worth fixing the
1660 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1661 for the new environment-manipulation done at startup. Move the routines
1662 from being local to tls.c to being global via the os.c file.
1664 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1665 an extract embedded as result-arg for a map, the first arg for extract
1666 is unavailable so we cannot tell if this is a numbered or keyed
1667 extraction. Accept either.
1673 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1676 JH/02 The smtp transport option "multi_domain" is now expanded.
1678 JH/03 The smtp transport now requests PRDR by default, if the server offers
1681 JH/04 Certificate name checking on server certificates, when exim is a client,
1682 is now done by default. The transport option tls_verify_cert_hostnames
1683 can be used to disable this per-host. The build option
1684 EXPERIMENTAL_CERTNAMES is withdrawn.
1686 JH/05 The value of the tls_verify_certificates smtp transport and main options
1687 default to the word "system" to access the system default CA bundle.
1688 For GnuTLS, only version 3.0.20 or later.
1690 JH/06 Verification of the server certificate for a TLS connection is now tried
1691 (but not required) by default. The verification status is now logged by
1692 default, for both outbound TLS and client-certificate supplying inbound
1695 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1698 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1699 Status Notification (bounce) messages are now MIME format per RFC 3464.
1700 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1701 under the control of the dsn_advertise_hosts option, and routers may
1702 have a dsn_lasthop option.
1704 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1705 default, modifiable by a malware= option. The list separator for
1706 the options can now be changed in the usual way. Bug 68.
1708 JH/10 The smtp_receive_timeout main option is now expanded before use.
1710 JH/11 The incoming_interface log option now also enables logging of the
1711 local interface on delivery outgoing connections.
1713 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1714 if the interface and destination host and port all match.
1716 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1719 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1720 Patch from Andrew Lewis.
1722 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1723 now supports optional time-restrictions, weighting, and priority
1724 modifiers per server. Patch originally by <rommer@active.by>.
1726 JH/16 The spamd_address main option now supports a mixed list of local
1727 and remote servers. Remote servers can be IPv6 addresses, and
1728 specify a port-range.
1730 JH/17 Bug 68: The spamd_address main option now supports an optional
1731 timeout value per server.
1733 JH/18 Bug 1581: Router and transport options headers_add/remove can
1734 now have the list separator specified.
1736 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1739 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1742 JH/21 Support for the A6 type of dns record is withdrawn.
1744 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1745 rather than the verbs used.
1747 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1748 from 255 to 1024 chars.
1750 JH/24 Verification callouts now attempt to use TLS by default.
1752 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1753 are generic router options now. The defaults didn't change.
1755 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1756 Original patch from Alexander Shikoff, worked over by JH.
1758 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1759 files if ZCAT_COMMAND is not executable.
1761 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1763 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1765 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1766 Normally benign, it bites when the pair was led to by a CNAME;
1767 modern usage is to not canonicalize the domain to a CNAME target
1768 (and we were inconsistent anyway for A-only vs AAAA+A).
1770 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1772 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1773 when evaluating $sender_host_dnssec.
1775 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1776 $sender_helo_dnssec variable.
1778 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1780 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1782 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1784 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1785 documented as working, but never had. Support all but $spam_report.
1787 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1788 added for tls authenticator.
1790 HS/03 Add perl_taintmode main config option
1796 TL/01 When running the test suite, the README says that variables such as
1797 no_msglog_check are global and can be placed anywhere in a specific
1798 test's script, however it was observed that placement needed to be near
1799 the beginning for it to behave that way. Changed the runtest perl
1800 script to read through the entire script once to detect and set these
1801 variables, reset to the beginning of the script, and then run through
1802 the script parsing/test process like normal.
1804 TL/02 The BSD's have an arc4random API. One of the functions to induce
1805 adding randomness was arc4random_stir(), but it has been removed in
1806 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1807 function when detected.
1809 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1810 cause callback expansion.
1812 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1813 syntax errors in an expansion can be treated as a string instead of
1814 logging or causing an error, due to the internal use of bool_lax
1815 instead of bool when processing it.
1817 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1818 server certificates when making smtp deliveries.
1820 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1822 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1824 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1826 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1827 Merged patch from Sebastian Wiedenroth.
1829 JH/05 Fix results-pipe from transport process. Several recipients, combined
1830 with certificate use, exposed issues where response data items split
1831 over buffer boundaries were not parsed properly. This eventually
1832 resulted in duplicates being sent. This issue only became common enough
1833 to notice due to the introduction of connection certificate information,
1834 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1836 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1837 size buffer was used, resulting in syntax errors when an expansion
1840 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1841 version 3.3.6 or later.
1843 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1844 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1845 both become "event_action", the variables become $event_name, $event_data
1846 and $event_defer_errno. There is a new variable $verify_mode, usable in
1847 routers, transports and related events. The tls:cert event is now also
1848 raised for inbound connections, if the main configuration event_action
1851 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1852 early OCSP support, but no stapling (appears to be less than 1.0.0).
1854 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1855 server certificate names available under the smtp transport option
1856 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1859 JH/10 Time-related extraction expansions from certificates now use the main
1860 option "timezone" setting for output formatting, and are consistent
1861 between OpenSSL and GnuTLS compilations. Bug 1541.
1863 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1864 encoded parameter in the incoming message. Bug 1558.
1866 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1867 include certificate info, eximon was claiming there were spoolfile
1870 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1872 JH/14 Log delivery-related information more consistently, using the sequence
1873 "H=<name> [<ip>]" wherever possible.
1875 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1876 are problematic for Debian distribution, omit them from the release
1879 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1881 JH/16 Fix string representation of time values on 64bit time_t architectures.
1884 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1885 output list separator was used.
1890 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1891 checkers that were complaining about end of non-void function with no
1894 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1895 This was a regression introduced in 4.83 by another bugfix.
1897 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1899 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1900 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1906 TF/01 Correctly close the server side of TLS when forking for delivery.
1908 When a message was received over SMTP with TLS, Exim failed to clear up
1909 the incoming connection properly after forking off the child process to
1910 deliver the message. In some situations the subsequent outgoing
1911 delivery connection happened to have the same fd number as the incoming
1912 connection previously had. Exim would try to use TLS and fail, logging
1913 a "Bad file descriptor" error.
1915 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1916 utilities have not been installed.
1918 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1919 temporary space as the ACL may create new global variables.
1921 TL/01 LDAP support uses per connection or global context settings, depending
1922 upon the detected version of the libraries at build time.
1924 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1925 to extract and use the src ip:port in logging and expansions as if it
1926 were a direct connection from the outside internet. PPv2 support was
1927 updated based on HAProxy spec change in May 2014.
1929 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1931 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1932 Properly escape header and check for NULL return.
1934 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1937 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1939 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1940 characters in header names, implemented as a verify condition.
1941 Contributed by Michael Fischer v. Mollard.
1943 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1944 results permerror and temperror. Previous values are deprecated but
1945 still accepted. In a future release, err_perm and err_temp will be
1946 completely removed, which will be a backward incompatibility if the
1947 ACL tests for either of these two old results. Patch contributed by
1948 user bes-internal on the mailing list.
1950 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1952 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1953 selectors, in both main and reject logs.
1955 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1958 JH/07 Add malware type "sock" for talking to simple daemon.
1960 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1962 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1963 routers/transports under cutthrough routing.
1965 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1966 numbers. Touch up "bool" conditional to keep the same definition.
1968 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1970 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1972 JH/12 Expand items in router/transport headers_add or headers_remove lists
1973 individually rather than the list as a whole. Bug 1452.
1975 Required for reasonable handling of multiple headers_ options when
1976 they may be empty; requires that headers_remove items with embedded
1977 colons must have them doubled (or the list-separator changed).
1979 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1980 view the policy declared in the DMARC record. Currently, $dmarc_status
1981 is a combined value of both the record presence and the result of the
1984 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1986 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1987 dnslookup router and the smtp transport (applying to the forward
1990 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1991 of ldap servers used for a specific lookup. Patch provided by Heiko
1994 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1995 New variable $lookup_dnssec_authenticated for observability.
1997 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1998 Patch submitted by Lars Timman.
2000 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2002 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2003 Requires trusted mode and valid format message id, aborts otherwise.
2004 Patch contributed by Heiko Schlichting.
2006 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2007 certextract with support for various fields. Bug 1358.
2009 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2010 is requested by default, modifiable by smtp transport option
2013 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2014 operate on certificate variables to give certificate fingerprints
2015 Also new ${sha256:cert_variable}.
2017 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2019 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2020 Christian Aistleitner.
2022 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2024 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2025 file. Patch from Wolfgang Breyha.
2027 JH/25 Expand the coverage of the delivery $host and $host_address to
2028 client authenticators run in verify callout. Bug 1476.
2030 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2031 align with daemon_smtp_ports. Bug 72.
2033 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2034 support and error reporting did not work properly.
2036 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2037 and is readable. Patch from Andrew Colin Kissa.
2039 TL/14 Enhance documentation of ${run expansion and how it parses the
2040 commandline after expansion, particularly in the case when an
2041 unquoted variable expansion results in an empty value.
2043 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2045 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2046 by renaming away from it. Addresses GH issue 15, reported by
2049 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2051 TL/15 SECURITY: prevent double expansion in math comparison functions
2052 (can expand unsanitized data). Not remotely exploitable.
2059 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2061 PP/02 Make -n do something, by making it not do something.
2062 When combined with -bP, the name of an option is not output.
2064 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2067 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2068 $sender_host_name and config options to manage this, and basic check
2071 PP/05 DSCP support for outbound connections and control modifier for inbound.
2073 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2074 (Only plugin which currently uses this is kerberos4, which nobody should
2075 be using, but we should make it available and other future plugins might
2076 conceivably use it, even though it would break NAT; stuff *should* be
2077 using channel bindings instead).
2079 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2080 name; added for Sendmail compatibility; requires admin caller.
2081 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2082 just ignore it); requires trusted caller.
2083 Also parse but ignore: -Ac -Am -X<logfile>
2086 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2088 TL/02 Add +smtp_confirmation as a default logging option.
2090 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2091 Patch by Magnus Holmgren from 2007-02-20.
2093 TL/04 Bugzilla 1281 - Spec typo.
2094 Bugzilla 1283 - Spec typo.
2095 Bugzilla 1290 - Spec grammar fixes.
2097 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2099 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2101 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2102 mailing list by by Dmitry Isaikin.
2104 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2106 JH/02 Support "G" suffix to numbers in ${if comparisons.
2108 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2110 NM/01 Bugzilla 1197 - Spec typo
2111 Bugzilla 1196 - Spec examples corrections
2113 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2115 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2116 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2119 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2120 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2122 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2123 "acl {{name}{arg}...}", and optional args on acl condition
2126 JH/05 Permit multiple router/transport headers_add/remove lines.
2128 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2130 JH/07 Avoid using a waiting database for a single-message-only transport.
2131 Performance patch from Paul Fisher. Bugzilla 1262.
2133 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2136 JH/09 Add $headers_added variable, with content from use of ACL modifier
2137 add_header (but not yet added to the message). Bugzilla 199.
2139 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2140 Pulled from Bugzilla 817 by Wolfgang Breyha.
2142 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2144 (nb: this is the same fix as in Exim 4.80.1)
2146 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2149 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2152 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2154 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2155 advertise SMTP AUTH mechanism to us, instead of a generic
2156 protocol violation error. Also, make Exim more robust to bad
2157 data from the Dovecot auth socket.
2159 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2161 When a queue runner is handling a message, Exim first routes the
2162 recipient addresses, during which it prunes them based on the retry
2163 hints database. After that it attempts to deliver the message to
2164 any remaining recipients. It then updates the hints database using
2167 So if a recipient address works intermittently, it can get repeatedly
2168 deferred at routing time. The retry hints record remains fresh so the
2169 address never reaches the final cutoff time.
2171 This is a fairly common occurrence when a user is bumping up against
2172 their storage quota. Exim had some logic in its local delivery code
2173 to deal with this. However it did not apply to per-recipient defers
2174 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2176 This change adds a proper retry rule check during routing so that the
2177 final cutoff time is checked against the message's age. We only do
2178 this check if there is an address retry record and there is not a
2179 domain retry record; this implies that previous attempts to handle
2180 the address had the retry_use_local_parts option turned on. We use
2181 this as an approximation for the destination being like a local
2182 delivery, as in LMTP.
2184 I suspect this new check makes the old local delivery cutoff check
2185 redundant, but I have not verified this so I left the code in place.
2187 TF/02 Correct gecos expansion when From: is a prefix of the username.
2189 Test 0254 submits a message to Exim with the header
2193 When I ran the test suite under the user fanf2, Exim expanded
2194 the header to contain my full name, whereas it should have added
2195 a Resent-Sender: header. It erroneously treats any prefix of the
2196 username as equal to the username.
2198 This change corrects that bug.
2200 GF/01 DCC debug and logging tidyup
2201 Error conditions log to paniclog rather than rejectlog.
2202 Debug lines prefixed by "DCC: " to remove any ambiguity.
2204 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2206 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2207 Bug spotted by Jeremy Harris; was flawed since initial commit.
2208 Would have resulted in OCSP responses post-SNI triggering an Exim
2209 NULL dereference and crash.
2211 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2213 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2214 Bug detection, analysis and fix by Samuel Thibault.
2215 Bugzilla 1331, Debian bug #698092.
2217 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2219 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2220 Server implementation by Todd Lyons, client by JH.
2221 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2222 config variable "prdr_enable" controls whether the server
2223 advertises the facility. If the client requests PRDR a new
2224 acl_data_smtp_prdr ACL is called once for each recipient, after
2225 the body content is received and before the acl_smtp_data ACL.
2226 The client is controlled by both of: a hosts_try_prdr option
2227 on the smtp transport, and the server advertisement.
2228 Default client logging of deliveries and rejections involving
2229 PRDR are flagged with the string "PRDR".
2231 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2232 fclose(). Diagnosis by Todd Lyons.
2234 PP/17 Update configure.default to handle IPv6 localhost better.
2235 Patch by Alain Williams (plus minor tweaks).
2238 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2239 This is now consistent with GnuTLS, and is now documented: the
2240 previous undocumented portable approach to treating the option as
2241 unset was to force an expansion failure. That still works, and
2242 an empty string is now equivalent.
2244 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2245 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2246 not performing validation itself.
2248 PP/20 Added force_command boolean option to pipe transport.
2249 Patch from Nick Koston, of cPanel Inc.
2251 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2254 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2256 PP/21 Fix eximon continuous updating with timestamped log-files.
2257 Broken in a format-string cleanup in 4.80, missed when I repaired the
2258 other false fix of the same issue.
2259 Report and fix from Heiko Schlichting.
2262 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2263 Report from Prashanth Katuri.
2265 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2266 It's SecureTransport, so affects any MacOS clients which use the
2267 system-integrated TLS libraries, including email clients.
2269 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2270 using a MIME ACL for non-SMTP local injection.
2271 Report and assistance in diagnosis by Warren Baker.
2273 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2275 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2277 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2278 last id that failed so it may be referenced in subsequent ACL's.
2280 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2283 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2284 ldap library initialization, allowing self-signed CA's to be
2285 used. Also properly sets require_cert option later in code by
2286 using NULL (global ldap config) instead of ldap handle (per
2287 session). Bug diagnosis and testing by alxgomz.
2289 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2290 the src/util/ subdirectory.
2292 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2293 renamed to Transport Post Delivery Action by Jeremy Harris, as
2296 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2297 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2298 redis_servers = needs to be configured which will be used by the redis
2299 lookup. Patch from Warren Baker, of The Packet Hub.
2301 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2303 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2304 hostname or reverse DNS when processing a host list. Used suggestions
2305 from multiple comments on this bug.
2307 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2309 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2310 Missed a few lines, added it to make the runtest require no keyboard
2313 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2314 contains upper case chars. Make router use caseful_local_part.
2316 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2317 support when GnuTLS has been built with p11-kit.
2323 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2325 This, or similar/improved, will also be change PP/11 of 4.82.
2331 PP/01 Handle short writes when writing local log-files.
2332 In practice, only affects FreeBSD (8 onwards).
2333 Bugzilla 1053, with thanks to Dmitry Isaikin.
2335 NM/01 Bugzilla 949 - Documentation tweak
2337 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2340 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2342 PP/02 Implemented gsasl authenticator.
2344 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2346 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2347 `pkg-config foo` for cflags/libs.
2349 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2350 with rest of GSASL and with heimdal_gssapi.
2352 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2353 `pkg-config foo` for cflags/libs for the TLS implementation.
2355 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2356 properties get this fed in as external SSF. A number of robustness
2357 and debugging improvements to the cyrus_sasl authenticator.
2359 PP/08 cyrus_sasl server now expands the server_realm option.
2361 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2362 Patch by Jeremy Harris.
2364 PP/10 Added dbmjz lookup type.
2366 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2368 PP/12 MAIL args handles TAB as well as SP, for better interop with
2369 non-compliant senders.
2370 Analysis and variant patch by Todd Lyons.
2372 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2373 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2374 Patch from Dirk Mueller <dmueller@suse.com>
2376 PP/13 tls_peerdn now print-escaped for spool files.
2377 Observed some $tls_peerdn in wild which contained \n, which resulted
2378 in spool file corruption.
2380 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2381 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2382 or write after TLS renegotiation, which otherwise led to messages
2385 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2386 as a tracking header (ie: a signed header comes before the signature).
2387 Patch from Wolfgang Breyha.
2389 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2390 comma-sep list; embedded commas doubled.
2392 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2394 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2396 Report and patch from Dmitry Banschikov.
2398 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2399 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2400 failures. We appear to now support TLS1.1+ with Exim.
2402 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2403 lets Exim select keys and certificates based upon TLS SNI from client.
2404 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2405 before an outbound SMTP session. New log_selector, +tls_sni.
2407 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2408 NULL dereference. Report and patch from Alun Jones.
2410 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2411 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2412 Not seeing resolver debug output on NetBSD, but suspect this is a
2413 resolver implementation change.
2415 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2416 Left warnings. Added "eximon gdb" invocation mode.
2418 PP/21 Defaulting "accept_8bitmime" to true, not false.
2420 PP/22 Added -bw for inetd wait mode support.
2422 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2423 locate the relevant includes and libraries. Made this the default.
2425 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2426 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2428 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2429 This may cause build issues on older platforms.
2431 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2432 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2433 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2434 Added SNI support via GnuTLS too.
2435 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2437 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2439 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2440 Applied second patch from Janne, implementing suggestion to default
2441 multiple-strings-in-record handling to match SPF spec.
2443 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2445 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2446 read-only, out of scope).
2447 Patch from Wolfgang Breyha, report from Stuart Northfield.
2449 PP/29 Fix three issues highlighted by clang analyser static analysis.
2450 Only crash-plausible issue would require the Cambridge-specific
2451 iplookup router and a misconfiguration.
2452 Report from Marcin Mirosław.
2454 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2456 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2457 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2458 As part of this, removing so much warning spew let me fix some minor
2459 real issues in debug logging.
2461 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2462 assignment on my part. Fixed.
2464 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2465 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2466 Janne Snabb (who went above and beyond: thank you).
2468 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2469 string otherwise requires a connection and a bunch more work and it's
2470 relatively easy to get wrong. Should also expose TLS library linkage
2473 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2474 64-bit ${eval} (JH/03).
2476 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2477 GNU libc to support some of the 64-bit stuff, should not lead to
2478 conflicts. Defined before os.h is pulled in, so if a given platform
2479 needs to override this, it can.
2481 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2482 protection layer was required, which is not implemented.
2483 Bugzilla 1254, patch from Wolfgang Breyha.
2485 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2486 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2487 tls_dhparam take prime identifiers. Also unbreak combination of
2488 OpenSSL+DH_params+TLSSNI.
2490 PP/39 Disable SSLv2 by default in OpenSSL support.
2496 PP/01 Solaris build fix for Oracle's LDAP libraries.
2497 Bugzilla 1109, patch from Stephen Usher.
2499 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2501 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2504 TF/02 Fix a couple more cases where we did not log the error message
2505 when unlink() failed. See also change 4.74-TF/03.
2507 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2508 lock up or crash if it happened to be inside a call to libc when it
2509 got a SIGUSR1 from exiwhat.
2511 The SIGUSR1 handler appends the current process status to the process
2512 log which is later printed by exiwhat. It used to use the general
2513 purpose logging code to do this, but several functions it calls are
2514 not safe for signals.
2516 The new output code in the SIGUSR1 handler is specific to the process
2517 log, and simple enough that it's easy to inspect for signal safety.
2518 Removing some special cases also simplifies the general logging code.
2519 Removing the spurious timestamps from the process log simplifies
2522 TF/04 Improved ratelimit ACL condition.
2524 The /noupdate option has been deprecated in favour of /readonly which
2525 has clearer semantics. The /leaky, /strict, and /readonly update modes
2526 are mutually exclusive. The update mode is no longer included in the
2527 database key; it just determines when the database is updated. (This
2528 means that when you upgrade Exim will forget old rate measurements.)
2530 Exim now checks that the per_* options are used with an update mode that
2531 makes sense for the current ACL. For example, when Exim is processing a
2532 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2533 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2534 must specify per_mail/readonly. If you omit the update mode it defaults to
2535 /leaky where that makes sense (as before) or /readonly where required.
2537 The /noupdate option is now undocumented but still supported for
2538 backwards compatibility. It is equivalent to /readonly except that in
2539 ACLs where /readonly is required you may specify /leaky/noupdate or
2540 /strict/noupdate which are treated the same as /readonly.
2542 A useful new feature is the /count= option. This is a generalization
2543 of the per_byte option, so that you can measure the throughput of other
2544 aggregate values. For example, the per_byte option is now equivalent
2545 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2547 The per_rcpt option has been generalized using the /count= mechanism
2548 (though it's more complicated than the per_byte equivalence). When it is
2549 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2550 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2551 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2552 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2553 Note that using per_rcpt with a non-readonly update mode in more than
2554 one ACL will cause the recipients to be double-counted. (The per_mail
2555 and per_byte options don't have this problem.)
2557 The handling of very low rates has changed slightly. If the computed rate
2558 is less than the event's count (usually one) then this event is the first
2559 after a long gap. In this case the rate is set to the same as this event's
2560 count, so that the first message of a spam run is counted properly.
2562 The major new feature is a mechanism for counting the rate of unique
2563 events. The new per_addr option counts the number of different
2564 recipients that someone has sent messages to in the last time period. It
2565 behaves like per_rcpt if all the recipient addresses are different, but
2566 duplicate recipient addresses do not increase the measured rate. Like
2567 the /count= option this is a general mechanism, so the per_addr option
2568 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2569 example, measure the rate that a client uses different sender addresses
2570 with the options per_mail/unique=$sender_address. There are further
2571 details in the main documentation.
2573 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2575 TF/06 Removed a few PCRE remnants.
2577 TF/07 Automatically extract Exim's version number from tags in the git
2578 repository when doing development or release builds.
2580 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2581 Bugzilla 879. Patch from Paul Fisher.
2583 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2584 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2587 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2589 PP/05 Variable $av_failed, true if the AV scanner deferred.
2590 Bugzilla 1078. Patch from John Horne.
2592 PP/06 Stop make process more reliably on build failure.
2593 Bugzilla 1087. Patch from Heiko Schlittermann.
2595 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2596 Bugzilla 1089. Patch from Heiko Schlittermann.
2598 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2599 Bugzilla 1131. Patch from Holger Weiß.
2601 PP/09 Handle IPv6 addresses with SPF.
2602 Bugzilla 860. Patch from Wolfgang Breyha.
2604 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2606 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2609 PP/11 match_* no longer expand right-hand-side by default.
2610 New compile-time build option, EXPAND_LISTMATCH_RHS.
2611 New expansion conditions, "inlist", "inlisti".
2613 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2615 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2617 PP/14 fix log_write() format string regression from TF/03.
2618 Bugzilla 1152. Patch from Dmitry Isaikin.
2624 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2626 PP/02 Harmonised TLS library version reporting; only show if debugging.
2627 Layout now matches that introduced for other libraries in 4.74 PP/03.
2629 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2631 PP/04 New "dns_use_edns0" global option.
2633 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2636 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2637 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2639 TK/01 Updated PolarSSL code to 0.14.2.
2640 Bugzilla 1097. Patch from Andreas Metzler.
2642 PP/07 Catch divide-by-zero in ${eval:...}.
2643 Fixes bugzilla 1102.
2645 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2648 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2649 format-string attack -- SECURITY: remote arbitrary code execution.
2651 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2652 time unintentionally subject to list matching rules, letting the header
2653 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2654 arbitrary string expansion). This allowed for information disclosure.
2656 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2657 INT_MIN/-1 -- value coerced to INT_MAX.
2663 NM/01 Workaround for PCRE version dependency in version reporting
2666 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2667 This fixes portability to compilers other than gcc, notably
2668 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2670 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2671 makefiles for portability to HP-UX and POSIX correctness.
2673 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2674 Also via indented variable definition in the Makefile.
2675 (Debugging by Oliver Heesakkers).
2677 PP/02 Restore caching of spamd results with expanded spamd_address.
2678 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2680 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2681 Improves build reliability. Fix from: Frank Elsner
2683 NM/02 Fix wide character breakage in the rfc2047 coding
2684 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2686 NM/03 Allow underscore in dnslist lookups
2687 Fixes bug 1026. Patch from Graeme Fowler
2689 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2690 Code patches from Adam Ciarcinski of NetBSD.
2692 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2695 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2696 is logged, to avoid truncation. Patch from John Horne.
2698 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2699 Patch from Jakob Hirsch.
2701 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2702 SQL string expansion failure details.
2703 Patch from Andrey Oktyabrski.
2705 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2706 Patch from Simon Arlott.
2708 PP/09 New lookups functionality failed to compile on old gcc which rejects
2709 extern declarations in function scope.
2710 Patch from Oliver Fleischmann
2712 PP/10 Use sig_atomic_t for flags set from signal handlers.
2713 Check getgroups() return and improve debugging.
2714 Fixed developed for diagnosis in bug 927 (which turned out to be
2717 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2718 Patch from Mark Zealey.
2720 PP/12 Bugzilla 1056: Improved spamd server selection.
2721 Patch from Mark Zealey.
2723 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2724 Based on patch from Heiko Schlittermann.
2726 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2727 Patch from Uwe Doering, sign-off by Michael Haardt.
2729 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2730 variable declaration deep within a block. Bug and patch from
2733 PP/15 lookups-Makefile IRIX compatibility coercion.
2735 PP/16 Make DISABLE_DKIM build knob functional.
2737 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2738 Patch by Simon Arlott
2740 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2741 variable argument macros. Our copy now differs from upstream.
2747 TF/01 Failure to get a lock on a hints database can have serious
2748 consequences so log it to the panic log.
2750 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2751 controlled using the smtp_confirmation log selector.
2753 TF/03 Include the error message when we fail to unlink a spool file.
2755 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2756 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2757 for maintaining out-of-tree patches for some time.
2759 PP/01 Bugzilla 139: Documentation and portability issues.
2760 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2761 Handle per-OS dynamic-module compilation flags.
2763 PP/02 Let /dev/null have normal permissions.
2764 The 4.73 fixes were a little too stringent and complained about the
2765 permissions on /dev/null. Exempt it from some checks.
2766 Reported by Andreas M. Kirchwitz.
2768 PP/03 Report version information for many libraries, including
2769 Exim version information for dynamically loaded libraries. Created
2770 version.h, now support a version extension string for distributors
2771 who patch heavily. Dynamic module ABI change.
2773 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2774 privilege escalation vulnerability whereby the Exim run-time user
2775 can cause root to append content of the attacker's choosing to
2778 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2781 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2782 If dropping privileges for untrusted macros, we disabled normal logging
2783 on the basis that it would fail; for the Exim run-time user, this is not
2784 the case, and it resulted in successful deliveries going unlogged.
2785 Fixed. Reported by Andreas Metzler.
2791 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2792 only prepend for the Resent-* case. Fixes regression introduced in
2793 Exim 4.70 by NM/22 for Bugzilla 607.
2795 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2796 increasing numbers of administrators be bitten by this.
2798 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2800 PP/03 Bugzilla 994: added openssl_options main configuration option.
2802 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2804 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2806 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2808 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2809 without a peer certificate, leading to a segfault because of an
2810 assumption that peers always have certificates. Be a little more
2811 paranoid. Problem reported by Martin Tscholak.
2813 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2814 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2815 NB: ClamAV planning to remove STREAM in "middle of 2010".
2816 CL also introduces -bmalware, various -d+acl logging additions and
2817 more caution in buffer sizes.
2819 PP/09 Implemented reverse_ip expansion operator.
2821 PP/10 Bugzilla 937: provide a "debug" ACL control.
2823 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2825 PP/12 Bugzilla 973: Implement --version.
2827 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2829 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2831 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2833 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2834 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2835 ignore trailing whitespace.
2837 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2839 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2842 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2843 Notification from Dr Andrew Aitchison.
2845 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2846 ExtendedDetectionInfo response format.
2847 Notification from John Horne.
2849 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2852 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2853 XSL and documented dependency on system catalogs, with examples of how
2856 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2859 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2860 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2861 configuration file which is writeable by the Exim user or group.
2863 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2864 of configuration files to cover files specified with the -C option if
2865 they are going to be used with root privileges, not just the default
2868 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2869 option (effectively making it always true).
2871 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2872 files to be used while preserving root privileges.
2874 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2875 that rogue child processes cannot use them.
2877 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2878 run-time user, instead of root.
2880 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2881 Exim run-time user without dropping privileges.
2883 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2884 result string, instead of calling string_vformat() twice with the same
2887 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2888 for other users. Others should always drop root privileges if they use
2889 -C on the command line, even for a whitelisted configure file.
2891 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2893 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2899 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2900 $data_path, and $header_path variables; fixed documentation bugs and
2903 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2904 exipick to access non-standard spools, including the "frozen" queue
2907 NM/01 Bugzilla 965: Support mysql stored procedures.
2908 Patch from Alain Williams
2910 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2912 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2913 Patch from Andreas Metzler
2915 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2916 Patch from Kirill Miazine
2918 NM/05 Bugzilla 671: Added umask to procmail example.
2920 JJ/03 installed exipick 20100323.0, fixing doc bug
2922 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2923 directory. Notification and patch from Dan Rosenberg.
2925 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2927 TK/02 Improve log output when DKIM signing operation fails.
2929 MH/01 Treat the transport option dkim_domain as a colon separated
2930 list, not as a single string, and sign the message with each element,
2931 omitting multiple occurences of the same signer.
2933 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2934 Bugzilla 985, 986. Patch by Simon Arlott
2936 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2937 Patch by Simon Arlott
2939 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2940 MBX locking. Notification from Dan Rosenberg.
2946 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2948 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2950 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2952 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2954 NM/04 Bugzilla 915: Flag broken perl installation during build.
2960 TK/01 Added patch by Johannes Berg that expands the main option
2961 "spamd_address" if it starts with a dollar sign.
2963 TK/02 Write list of recipients to X-Envelope-Sender header when building
2964 the mbox-format spool file for content scanning (suggested by Jakob
2967 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2968 (http://www.dcc-servers.net/) support via dccifd. Activated by
2969 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2971 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2972 by Mark Daniel Reidel <mr@df.eu>.
2974 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2975 When building exim an external PCRE library is now needed -
2976 PCRE is a system library on the majority of modern systems.
2977 See entry on PCRE_LIBS in EDITME file.
2979 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2980 conversation. Added nologin parameter to request.
2981 Patch contributed by Kirill Miazine.
2983 TF/01 Do not log submission mode rewrites if they do not change the address.
2985 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2987 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2988 log files in place. Contributed by Roberto Lima.
2990 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2992 TF/03 Bugzilla 615: When checking the local_parts router precondition
2993 after a local_part_suffix or local_part_prefix option, Exim now
2994 does not use the address's named list lookup cache, since this
2995 contains cached lookups for the whole local part.
2997 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2998 Robert Millan. Documentation is in experimental-spec.txt.
3000 TF/04 Bugzilla 668: Fix parallel build (make -j).
3002 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3004 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3005 Patch provided by Jan Srzednicki.
3007 TF/05 Leading white space used to be stripped from $spam_report which
3008 wrecked the formatting. Now it is preserved.
3010 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3011 that they are available at delivery time.
3013 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3015 TF/08 TLS error reporting now respects the incoming_interface and
3016 incoming_port log selectors.
3018 TF/09 Produce a more useful error message if an SMTP transport's hosts
3019 setting expands to an empty string.
3021 NM/06 Bugzilla 744: EXPN did not work under TLS.
3022 Patch provided by Phil Pennock.
3024 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3025 Patch provided by Richard Godbee.
3027 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3028 acl_smtp_notquit, added index entry.
3030 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3031 Patch provided by Eugene Bujak.
3033 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3034 accept(). Patch provided by Maxim Dounin.
3036 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3037 Patch provided by Phil Pennock.
3039 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3041 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3042 Patch provided by Brad "anomie" Jorsch.
3044 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3045 Patch provided by Dean Brooks.
3047 TK/05 Add native DKIM support (does not depend on external libraries).
3049 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3050 Patch provided by Graeme Fowler.
3052 NM/16 Bugzilla 851: Documentation example syntax fix.
3054 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3056 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3059 NM/19 Bugzilla 745: TLS version reporting.
3060 Patch provided by Phil Pennock.
3062 NM/20 Bugzilla 167: bool: condition support.
3063 Patch provided by Phil Pennock.
3065 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3066 clients. Patch provided by Phil Pennock.
3068 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3069 Patch provided by Brad "anomie" Jorsch.
3071 NM/23 Bugzilla 687: Fix misparses in eximstats.
3072 Patch provided by Heiko Schlittermann.
3074 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3075 Patch provided by Heiko Schlittermann.
3077 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3078 plus update to original patch.
3080 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3082 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3083 Patch provided by David Brownlee.
3085 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3087 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3089 NM/30 Bugzilla 888: TLS documentation bugfixes.
3091 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3093 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3094 Unlike the original bugzilla I have changed all shell scripts in src tree.
3096 NM/33 Bugzilla 898: Transport filter timeout fix.
3097 Patch by Todd Rinaldo.
3099 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3100 Patch by Serge Demonchaux.
3102 NM/35 Bugzilla 39: Base64 decode bug fixes.
3103 Patch by Jakob Hirsch.
3105 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3107 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3109 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3111 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3117 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3118 ALT-N's libdkim that I have put here:
3119 http://duncanthrax.net/exim-experimental/
3121 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3122 were called 'true' and it seems that C99 defines that as a reserved
3123 keyword to be used with 'bool' variable types. That means you could
3124 not include C99-style headers which use bools without triggering
3125 build errors in sieve.c.
3127 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3128 as mailq or other aliases. Changed the --help handling significantly
3129 to do whats expected. exim_usage() emits usage/help information.
3131 SC/01 Added the -bylocaldomain option to eximstats.
3133 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3135 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3137 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3143 PH/01 Another patch from the Sieve maintainer.
3145 PH/02 When an IPv6 address is converted to a string for single-key lookup
3146 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3147 dots are used instead of colons so that keys in lsearch files need not
3148 contain colons. This was done some time before quoting was made available
3149 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3150 (notated using the quote facility) so as to distinguish them from IPv4
3151 keys. This meant that lookups for IP addresses in host lists did not work
3152 for iplsearch lookups.
3154 This has been fixed by arranging for IPv6 addresses to be expressed with
3155 colons if the lookup type is iplsearch. This is not incompatible, because
3156 previously such lookups could never work.
3158 The situation is now rather anomalous, since one *can* have colons in
3159 ordinary lsearch keys. However, making the change in all cases is
3160 incompatible and would probably break a number of configurations.
3162 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3165 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3166 conversion specification without a maximum field width, thereby enabling
3167 a rogue spamd server to cause a buffer overflow. While nobody in their
3168 right mind would setup Exim to query an untrusted spamd server, an
3169 attacker that gains access to a server running spamd could potentially
3170 exploit this vulnerability to run arbitrary code as the Exim user.
3172 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3173 $primary_hostname instead of what libspf2 thinks the hosts name is.
3175 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3176 a directory entry by the name of the lookup key. Previously, if a
3177 symlink pointed to a non-existing file or a file in a directory that
3178 Exim lacked permissions to read, a lookup for a key matching that
3179 symlink would fail. Now it is enough that a matching directory entry
3180 exists, symlink or not. (Bugzilla 503.)
3182 PH/03 The body_linecount and body_zerocount variables are now exported in the
3185 PH/04 Added the $dnslist_matched variable.
3187 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3188 This means they are set thereafter only if the connection becomes
3191 PH/06 Added the client_condition to authenticators so that some can be skipped
3192 by clients under certain conditions.
3194 PH/07 The error message for a badly-placed control=no_multiline_responses left
3195 "_responses" off the end of the name.
3197 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3199 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3200 (without spaces) instead of just copying the configuration text.
3202 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3204 PH/11 Added $max_received_linelength.
3206 PH/12 Added +ignore_defer and +include_defer to host lists.
3208 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3209 way in which PCRE > 7.0 is built.
3211 PH/14 Implemented queue_only_load_latch.
3213 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3214 MAIL command. The effect was to mangle the value on 64-bit systems.
3216 PH/16 Another patch from the Sieve maintainer.
3218 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3220 PH/18 If a system quota error occurred while trying to create the file for
3221 a maildir delivery, the message "Mailbox is full" was not appended to the
3222 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3223 only to a quota excession during the actual writing of the file.
3225 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3226 characters?) which causes problems in log lines. The DN values are now
3227 passed through string_printing() before being added to log lines.
3229 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3230 and InterBase are left for another time.)
3232 PH/21 Added message_body_newlines option.
3234 PH/22 Guard against possible overflow in moan_check_errorcopy().
3236 PH/23 POSIX allows open() to be a macro; guard against that.
3238 PH/24 If the recipient of an error message contained an @ in the local part
3239 (suitably quoted, of course), incorrect values were put in $domain and
3240 $local_part during the evaluation of errors_copy.
3246 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3247 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3250 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3251 issue a MAIL command.
3253 PH/02 In an ACL statement such as
3255 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3257 if a client was not listed at all, or was listed with a value other than
3258 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3259 the condition was not true (as it should be), so access was not denied.
3260 The bug was that the ! inversion was incorrectly passed on to the second
3261 item. This has been fixed.
3263 PH/03 Added additional dnslists conditions == and =& which are different from
3264 = and & when the dns lookup returns more than one IP address.
3266 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3267 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3269 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3270 FSYNC, which compiles an option called disable_fsync that allows for
3271 bypassing fsync(). The documentation is heavily laced with warnings.
3273 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3275 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3276 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3277 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3278 including adding "make clean"; (3) Added -fPIC when compiling the test
3279 dynamically loaded module, to get rid of a warning.
3281 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3282 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3283 = 0s. The bug is otherwise harmless.
3285 PH/07 There was a bug in the dovecot authenticator such that the value of
3286 $auth1 could be overwritten, and so not correctly preserved, after a
3287 successful authentication. This usually meant that the value preserved by
3288 the server_setid option was incorrect.
3290 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3292 PH/09 Installed PCRE release 7.0.
3294 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3295 run for batched SMTP input. It is now run at the start of every message
3296 in the batch. While fixing this I discovered that the process information
3297 (output by running exiwhat) was not always getting set for -bs and -bS
3298 input. This is fixed, and it now also says "batched" for BSMTP.
3300 PH/11 Added control=no_pipelining.
3302 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3303 patch, slightly modified), and move the expansion of helo_data till after
3304 the connection is made in the smtp transport (so it can use these
3307 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3309 PH/14 Added log_selector = +pid.
3311 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3313 PH/16 Add ${if forany and ${if forall.
3315 PH/17 Added dsn_from option to vary the From: line in DSNs.
3317 PH/18 Flush SMTP output before performing a callout, unless control =
3318 no_callout_flush is set.
3320 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3321 was true (the default) a successful delivery failed to delete the retry
3322 item, thus causing premature timeout of the address. The bug is now
3325 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3327 PH/21 Long custom messages for fakedefer and fakereject are now split up
3328 into multiline responses in the same way that messages for "deny" and
3329 other ACL rejections are.
3331 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3332 with slight modification.
3334 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3335 draft, changing the syntax and factoring some duplicate code".
3337 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3338 for deliveries of the second and subsequent messages over the same SMTP
3341 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3342 ${reduce, with only minor "tidies".
3344 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3346 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3347 expansion side effects.
3349 PH/27 When a message times out after an over-quota error from an Exim-imposed
3350 quota, the bounce message says "mailbox is full". This message was not
3351 being given when it was a system quota that was exceeded. It now should
3354 MH/03 Made $recipients available in local_scan(). local_scan() already has
3355 better access to the recipient list through recipients_list[], but
3356 $recipients can be useful in postmaster-provided expansion strings.
3358 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3359 in the case of a MAIL command with additional options following the
3360 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3361 were accidentally chopped off.
3363 PH/29 SMTP synchronization checks are implemented when a command is read -
3364 there is a check that no more input is waiting when there shouldn't be
3365 any. However, for some commands, a delay in an ACL can mean that it is
3366 some time before the response is written. In this time, more input might
3367 arrive, invalidly. So now there are extra checks after an ACL has run for
3368 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3369 pipelining has not been advertised.
3371 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3373 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3374 returned in the SMTP response when smtp_return_error_details was false.
3375 This has been fixed.
3377 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3378 instead of the C I/O that was originally supplied, because problems were
3379 reported on Solaris.
3381 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3382 Exim which did not show up earlier: it was assuming that a call to
3383 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3384 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3385 was a macro that became an assignment, so it seemed to work. This has
3386 changed to a proper function call with a void return, hence the compile
3387 error. Exim's code has been fixed.
3389 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3392 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3394 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3396 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3397 because it turns out to be too big or there is a timeout) there is no
3398 "Completed" line in the log. When some messages of this type were
3399 selected by exigrep, they were listed as "not completed". Others were
3400 picked up by some special patterns. I have improved the selection
3401 criteria to be more general.
3403 PH/38 The host_find_failed option in the manualroute router can now be set
3404 to "ignore", to completely ignore a host whose IP address cannot be
3405 found. If all hosts are ignored, the behaviour is controlled by the new
3406 host_all_ignored option.
3408 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3409 homing or because of multiple MX records with /mx) generated more than
3410 one IP address, and the following item turned out to be the local host,
3411 all the secondary addresses of the first item were incorrectly removed
3412 from the list, along with the local host and any following hosts (which
3413 is what is supposed to happen).
3415 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3416 whoever called Exim into the -H file. In the case of the daemon it was
3417 behaving confusingly. When first started, it used values for whoever
3418 started the daemon, but after a SIGHUP it used the Exim user (because it
3419 calls itself on a restart). I have changed the code so that it now always
3422 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3423 message are rejected with the same error (e.g. no authentication or bad
3424 sender address), and a DATA command is nevertheless sent (as can happen
3425 with PIPELINING or a stupid MUA), the error message that was given to the
3426 RCPT commands is included in the rejection of the DATA command. This is
3427 intended to be helpful for MUAs that show only the final error to their
3430 PH/42 Another patch from the Sieve maintainer.
3432 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3433 Eximstats - Fixed some broken HTML links and added missing column headers
3435 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3436 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3438 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3440 PH/43 Yet another patch from the Sieve maintainer.
3442 PH/44 I found a way to check for a TCP/IP connection going away before sending
3443 the response to the final '.' that terminates a message, but only in the
3444 case where the client has not sent further data following the '.'
3445 (unfortunately, this is allowed). However, in many cases there won't be
3446 any further data because there won't be any more messages to send. A call
3447 to select() can be used: if it shows that the input is "ready", there is
3448 either input waiting, or the socket has been closed. An attempt to read
3449 the next input character can distinguish the two cases. Previously, Exim
3450 would have sent an OK response which the client would never have see.
3451 This could lead to message repetition. This fix should cure that, at
3452 least in a lot of common cases.
3454 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3455 advertised in response to EHLO.
3461 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3462 fixed by 4.65/MH/01 (is this a record?) are fixed:
3464 (i) An empty string was always treated as zero by the numeric comparison
3465 operators. This behaviour has been restored.
3467 (ii) It is documented that the numeric comparison operators always treat
3468 their arguments as decimal numbers. This was broken in that numbers
3469 starting with 0 were being interpreted as octal.
3471 While fixing these problems I realized that there was another issue that
3472 hadn't been noticed. Values of message_size_limit (both the global option
3473 and the transport option) were treated as octal if they started with 0.
3474 The documentation was vague. These values are now always treated as
3475 decimal, and I will make that clear in the documentation.
3481 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3482 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3485 MH/01 Don't check that the operands of numeric comparison operators are
3486 integers when their expansion is in "skipping" mode (fixes bug
3487 introduced by 4.64-PH/07).
3489 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3490 child addresses, Exim now panics and dies. Previously, because the count
3491 is held in a short int, deliveries were likely to be lost. As such a
3492 large number of recipients for a single message is ridiculous
3493 (performance will be very, very poor), I have chosen to impose a limit
3494 rather than extend the field.
3500 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3501 leftover -K file (the existence of which was triggered by #402).
3502 While we were at it, introduced process PID as part of the -K
3503 filename. This should rule out race conditions when creating
3506 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3507 processing considerably. Previous code took too long for large mails,
3508 triggering a timeout which in turn triggers #401.
3510 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3511 in the DK code in transports.c. sendfile() is not really portable,
3512 hence the _LINUX specificness.
3514 TF/01 In the add_headers option to the mail command in an Exim filter,
3515 there was a bug that Exim would claim a syntax error in any
3516 header after the first one which had an odd number of characters
3519 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3520 callout verification, Exim cached a "reject" for the entire domain. This
3521 is correct for most verifications, but it is not correct for a recipient
3522 verification with use_sender or use_postmaster set, because in that case
3523 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3524 case of MAIL FROM:<> rejection from other early rejections (e.g.
3525 rejection of HELO). When verifying a recipient using a non-null MAIL
3526 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3527 Whatever the result of the callout, the value of the domain cache is
3528 left unchanged (for any other kind of callout, getting as far as trying
3529 RCPT means that the domain itself is ok).
3531 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3534 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3535 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3536 session, the daemon ignores ECONNECT errors and logs others; it now
3537 ignores EPIPE as well.
3539 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3540 (quoted-printable decoding).
3542 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3543 later the small subsequent patch to fix an introduced bug.
3545 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3547 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3549 PH/08 An error is now given if message_size_limit is specified negative.
3551 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3552 to be given (somewhat) arbitrary names.
3554 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3557 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3558 miscellaneous code fixes
3560 PH/10 Added the log_reject_target ACL modifier to specify where to log
3563 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3564 hostname. This is wrong, because it relates to the incoming message (and
3565 probably the interface on which it is arriving) and not to the outgoing
3566 callout (which could be using a different interface). This has been
3567 changed to use the value of the helo_data option from the smtp transport
3568 instead - this is what is used when a message is actually being sent. If
3569 there is no remote transport (possible with a router that sets up host
3570 addresses), $smtp_active_hostname is used.
3572 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3573 tweaks were necessary in order to get it to work (see also 21 below):
3574 (a) The code assumed that strncpy() returns a negative number on buffer
3575 overflow, which isn't the case. Replaced with Exim's string_format()
3577 (b) There were several signed/unsigned issues. I just did the minimum
3578 hacking in of casts. There is scope for a larger refactoring.
3579 (c) The code used strcasecmp() which is not a standard C function.
3580 Replaced with Exim's strcmpic() function.
3581 (d) The code set only $1; it now sets $auth1 as well.
3582 (e) A simple test gave the error "authentication client didn't specify
3583 service in request". It would seem that Dovecot has changed its
3584 interface. Fortunately there's a specification; I followed it and
3585 changed what the client sends and it appears to be working now.
3587 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3590 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3591 address A is aliased to B and C, where B exists and C does not. Without
3592 -v the output is "A verified" because verification stops after a
3593 successful redirection if more than one address is generated. However,
3594 with -v the child addresses are also verified. Exim was outputting "A
3595 failed to verify" and then showing the successful verification for C,
3596 with its parentage. It now outputs "B failed to verify", showing B's
3597 parentage before showing the successful verification of C.
3599 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3600 look up a TXT record in a specific list after matching in a combined
3603 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3604 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3605 they consult the DNS. I had assumed they would set it the way they
3606 wanted; and indeed my experiments on Linux seem to show that in some
3607 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3608 To be on the safe side, however, I have now made the interface to
3609 host_find_byname() similar to host_find_bydns(), with an argument
3610 containing the DNS resolver options. The host_find_byname() function now
3611 sets these options at its start, just as host_find_bydns() does. The smtp
3612 transport options dns_qualify_single and dns_search_parents are passed to
3613 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3614 of host_find_byname() use the default settings of RES_DEFNAMES
3615 (qualify_single) but not RES_DNSRCH (search_parents).
3617 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3618 spool_read_header() do less string testing, by means of a preliminary
3619 switch on the second character of optional "-foo" lines. (This is
3620 overdue, caused by the large number of possibilities that now exist.
3621 Originally there were few.) While I was there, I also converted the
3622 str(n)cmp tests so they don't re-test the leading "-" and the first
3623 character, in the hope this might squeeze out yet more improvement.
3625 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3626 flag allowing group syntax was set by the header_syntax check but not
3627 turned off, possible causing trouble later; (2) The flag was not being
3628 set at all for the header_verify test, causing "group"-style headers to
3629 be rejected. I have now set it in this case, and also caused header_
3630 verify to ignore an empty address taken from a group. While doing this, I
3631 came across some other cases where the code for allowing group syntax
3632 while scanning a header line wasn't quite right (mostly, not resetting
3633 the flag correctly in the right place). These bugs could have caused
3634 trouble for malformed header lines. I hope it is now all correct.
3636 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3637 with the "reply" argument non-NULL. The code, however (which originally
3638 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3639 but it didn't always do it. This confused somebody who was copying the
3640 code for some other use. I have removed all the tests.
3642 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3643 feature that was used to support insecure browsers during the U.S. crypto
3644 embargo. It requires special client support, and Exim is probably the
3645 only MTA that supported it -- and would never use it because real RSA is
3646 always available. This code has been removed, because it had the bad
3647 effect of slowing Exim down by computing (never used) parameters for the
3648 RSA_EXPORT functionality.
3650 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3651 authenticator to fail if there's a tab character in the incoming data
3652 (there should never be unless someone is messing about, as it's supposed
3653 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3654 if the connection is using TLS or if the remote IP is the same as the
3655 local IP, and the "valid-client-cert option" if a client certificate has
3658 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3659 authenticators. This can be used for authorization after authentication
3660 succeeds. (In the case of plaintext, it servers for both authentication
3663 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3664 if any retry times were supplied.
3666 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3667 connection, where there is no client IP address to check. In this
3668 situation, the verify now always succeeds.
3670 PH/25 Applied John Jetmore's -Mset patch.
3672 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3674 PH/27 In a string expansion for a processed (not raw) header when multiple
3675 headers of the same name were present, leading whitespace was being
3676 removed from all of them, but trailing whitespace was being removed only
3677 from the last one. Now trailing whitespace is removed from each header
3678 before concatenation. Completely empty headers in a concatenation (as
3679 before) are ignored.
3681 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3682 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3684 PH/29 [Removed. This was a change that I later backed out, and forgot to
3685 correct the ChangeLog entry (that I had efficiently created) before
3686 committing the later change.]
3688 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3689 address errors (4xx response to RCPT) over the same connection as other
3690 messages routed to the same hosts. Such deliveries are always "forced",
3691 so retry times are not inspected. This resulted in far too many retries
3692 for the affected addresses. The effect occurred only when there were more
3693 hosts than the hosts_max_try setting in the smtp transport when it had
3694 the 4xx errors. Those hosts that it had tried were not added to the list
3695 of hosts for which the message was waiting, so if all were tried, there
3696 was no problem. Two fixes have been applied:
3698 (i) If there are any address or message errors in an SMTP delivery, none
3699 of the hosts (tried or untried) are now added to the list of hosts
3700 for which the message is waiting, so the message should not be a
3701 candidate for sending over the same connection that was used for a
3702 successful delivery of some other message. This seems entirely
3703 reasonable: after all the message is NOT "waiting for some host".
3704 This is so "obvious" that I'm not sure why it wasn't done
3705 previously. Hope I haven't missed anything, but it can't do any
3706 harm, as the worst effect is to miss an optimization.
3708 (ii) If, despite (i), such a delivery is accidentally attempted, the
3709 routing retry time is respected, so at least it doesn't keep
3710 hammering the server.
3712 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3713 in ${readsocket because some servers need this prod.
3715 PH/32 Added some extra debug output when updating a wait-xxx database.
3717 PH/33 The hint "could be header name not terminated by colon", which has been
3718 given for certain expansion errors for a long time, was not being given
3719 for the ${if def:h_colon_omitted{... case.
3721 PH/34 The spec says: "With one important exception, whenever a domain list is
3722 being scanned, $domain contains the subject domain." There was at least
3723 one case where this was not true.
3725 PH/35 The error "getsockname() failed: connection reset by peer" was being
3726 written to the panic log as well as the main log, but it isn't really
3727 panic-worthy as it just means the connection died rather early on. I have
3728 removed the panic log writing for the ECONNRESET error when getsockname()
3731 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3732 runs only) independently of the message's sender address. This meant
3733 that, if the 4xx error was in fact related to the sender, a different
3734 message to the same recipient with a different sender could confuse
3735 things. In particular, this can happen when sending to a greylisting
3736 server, but other circumstances could also provoke similar problems.
3737 I have changed the default so that the retry time for these errors is now
3738 based a combination of the sender and recipient addresses. This change
3739 can be overridden by setting address_retry_include_sender=false in the
3742 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3743 remote server are returned as part of bounce messages. This was not
3744 happening for LMTP over a pipe (the lmtp transport), but now it is the
3745 same for both kinds of LMTP.
3747 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3748 in header lines that were in fact CNAMEs. This is no longer the case.
3750 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3751 and queue runs started by the daemon processed all messages. This has
3752 been fixed so that -R and -S can now usefully be given with -q<time>.
3754 PH/40 Import PCRE release 6.7 (fixes some bugs).
3756 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3758 PH/42 Give an error if -q is specified more than once.
3760 PH/43 Renamed the variables $interface_address and $interface_port as
3761 $received_ip_address and $received_port, to make it clear that these
3762 values apply to message reception, and not to the outgoing interface when
3763 a message is delivered. (The old names remain recognized, of course.)
3765 PH/44 There was no timeout on the connect() call when using a Unix domain
3766 socket in the ${readsocket expansion. There now is.
3768 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3769 be meaningful with "accept".
3771 SC/01 Eximstats V1.43
3772 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3774 SC/02 Eximstats V1.44
3775 Use a glob alias rather than an array ref in the generated
3776 parser. This improves both readability and performance.
3778 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3779 Collect SpamAssassin and rejection statistics.
3780 Don't display local sender or destination tables unless
3781 there is data to show.
3782 Added average volumes into the top table text output.
3784 SC/04 Eximstats V1.46
3785 Collect data on the number of addresses (recipients)
3786 as well as the number of messages.
3788 SC/05 Eximstats V1.47
3789 Added 'Message too big' to the list of mail rejection
3790 reasons (thanks to Marco Gaiarin).
3792 SC/06 Eximstats V1.48
3793 Mainlog lines which have GMT offsets and are too short to
3794 have a flag are now skipped.
3796 SC/07 Eximstats V1.49 (Alain Williams)
3797 Added the -emptyok flag.
3799 SC/08 Eximstats V1.50
3800 Fixes for obtaining the IP address from reject messages.
3802 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3803 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3804 whitespace changes from 4.64-PH/27
3806 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3809 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3810 are found, allow negative numbers in numeric criteria)
3812 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3814 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3815 to match changes made in 4.64-PH/43
3817 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3819 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3820 "domain" is actually an IP address, and force a failure. This locks out
3821 those revolvers/nameservers that support "A-for-A" lookups, in
3822 contravention of the specifications.
3824 PH/48 When a host name was looked up from an IP address, and the subsequent
3825 forward lookup of the name timed out, the host name was left in
3826 $sender_host_name, contrary to the specification.
3828 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3829 restricted to single-key lookups, Exim was not diagnosing an error if
3830 * or *@ was used with a query-style lookup.
3832 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3834 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3835 long ago, but noone interested enough thought of it. Let's just say that
3836 the "1.1" means that there are some new functions that weren't there at
3837 some point in the past.
3839 PH/51 Error processing for expansion failure of helo_data from an smtp
3840 transport during callout processing was broken.
3842 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3843 tested/used via the -bh/-bhc/-bs options.
3845 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3846 bug, fixed in subsequent PCRE releases).
3848 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3849 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3851 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3857 SC/01 Use a glob alias rather than an array ref in eximstats generated
3858 parser. This improves both readability and performance.
3860 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3861 Don't display local sender or destination tables in eximstats unless
3862 there is data to show.
3863 Added average volumes into the eximstats top table text output.
3865 SC/03 Collect data on the number of addresses (recipients) as well
3866 as the number of messages in eximstats.
3868 TF/01 Correct an error in the documentation for the redirect router. Exim
3869 does (usually) call initgroups() when daemonizing.
3871 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3872 with consistent privilege compared to when running as a daemon.
3874 TF/03 Note in the spec that $authenticated_id is not set for local
3875 submissions from trusted users.
3877 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3878 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3880 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3881 by adding some example configuration directives to the default
3882 configuration file. A little bit of work is required to uncomment the
3883 directives and define how usernames and passwords are checked, but
3884 there is now a framework to start from.
3886 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3887 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3888 without this. I don't know how relevant this is to other LDAP libraries.
3890 PH/02 Add the verb name to the "unknown ACL verb" error.
3892 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3894 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3896 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3897 directory not expanded when it should be if an expanded home directory
3898 was set for the address (which is overridden by the transport).
3900 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3903 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3904 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3905 because it is too late at that time, and has no effect.
3907 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3908 security issue with \' (bugzilla #107). I could not use the
3909 PQescapeStringConn() function, because it needs a PGconn value as one of
3912 PH/08 When testing addresses using -bt, indicate those final addresses that
3913 are duplicates that would not cause an additional delivery. At least one
3914 person was confused, thinking that -bt output corresponded to deliveries.
3915 (Suppressing duplicates isn't a good idea as you lose the information
3916 about possibly different redirections that led to the duplicates.)
3918 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3919 systems where poll() doesn't work, in particular OS X.
3921 PH/10 Added more information to debugging output for retry time not reached.
3923 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3924 operations in malware.c.
3926 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3929 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3930 syslog_duplication set false, log lines that would normally be written
3931 both the the main log and to the reject log were not written to syslog at
3934 PH/14 In the default configuration, change the use of "message" in ACL warn
3935 statements to "add_header".
3937 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3938 not followed by a command (e.g. "seen endif").
3940 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3941 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3944 PH/17 Added extra conditions to the default value of delay_warning_condition
3948 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3949 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3950 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3953 The Auto-Submitted: and various List- headers are standardised, whereas I
3954 don't think Precedence: ever was.
3956 PH/18 Refactored debugging code in route_finduser() to show more information,
3957 in particular, the error code if getpwnam() issues one.
3959 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3960 This is apparently needed in addition to the PH/07 change above to avoid
3961 any possible encoding problems.
3963 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3964 but not after initializing Perl.
3966 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3967 output them only if debugging. By default they are written stderr,
3968 apparently, which is not desirable.
3970 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3973 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3976 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3978 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3979 authenticated or an ident call has been made. Suppress the default
3980 values for $authenticated_id and $authenticated_sender (but permit -oMai
3981 and -oMas) when testing with -bh.
3983 PH/24 Re-jigged the order of the tests in the default configuration so that the
3984 tests for valid domains and recipients precede the DNS black list and CSA
3985 tests, on the grounds that those ones are more expensive.
3987 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3988 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3989 This bug exists in every version of Exim that I still have, right back to
3992 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3993 However, an attempt to turn on case-sensitivity in a regex key by
3994 including (?-i) didn't work because the subject string was already
3995 lowercased, and the effects were non-intuitive. It turns out that a
3996 one-line patch can be used to allow (?-i) to work as expected.
4002 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4003 other effects) broke the use of negated acl sub-conditions.
4005 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4008 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4009 "Deny" causes Exim to reject the incoming connection with a 554 error.
4010 Unfortunately, if there is a major crisis, such as a disk failure,
4011 tcp-wrappers gives "deny", whereas what one would like would be some
4012 kind of temporary error. A kludge has been added to help with this.
4013 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4014 554 error is used if errno is still zero or contains ENOENT (which occurs
4015 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4018 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4020 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4021 errors. Otherwise a message that provokes a temporary error (when other
4022 messages do not) can cause a whole host to time out.
4024 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4025 addresses were routed directly to files or pipes from a redirect router.
4026 File deliveries just didn't batch; pipe deliveries might have suffered
4029 PH/06 A failure to get a lock for a hints database would erroneously always say
4030 "Failed to get write lock", even when it was really a read lock.
4032 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4033 of 0600. This has been changed to use the value of the lockfile_mode
4034 option (which defaults to 0600).
4036 PH/08 Applied small patch from the Sieve maintainer.
4038 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4039 folder from quota calculations, a direct delivery into this folder messed
4040 up the contents of the maildirsize file. This was because the regex was
4041 used only to exclude .Trash (or whatever) when the size of the mailbox
4042 was calculated. There was no check that a delivery was happening into an
4043 excluded directory. This bug has been fixed by ignoring all quota
4044 processing for deliveries into excluded directories.
4046 PH/10 Added the maildirfolder_create_regex option to appendfile.
4052 PH/01 The code for finding all the local interface addresses on a FreeBSD
4053 system running IPv6 was broken. This may well have applied to all BSD
4054 systems, as well as to others that have similar system calls. The broken
4055 code found IPv4 interfaces correctly, but gave incorrect values for the
4056 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4057 that it would not match correctly against @[] and not recognize the IPv6
4060 PH/02 The ipliteral router was not recognizing addresses of the form user@
4061 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4063 PH/03 Added disable_ipv6.
4065 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4066 decoded form, because it is most often used to construct To: headers
4067 lines in autoreplies, and the decoded form may well be syntactically
4068 invalid. However, $reply_address has leading white space removed, and all
4069 newlines turned into spaces so that the autoreply transport does not
4072 PH/05 If group was specified without a user on a router, and no group or user
4073 was specified on a transport, the group from the router was ignored.
4075 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4076 for visible compile-time settings that can be used to change these
4077 numbers, for those that want even more. Backwards compatibility with old
4078 spool files has been maintained. However, going back to a previous Exim
4079 release will lost any variables that are in spool files.
4081 PH/07 Two small changes when running in the test harness: increase delay when
4082 passing a TCP/IP connection to a new process, in case the original
4083 process has to generate a bounce, and remove special handling of
4084 127.0.0.2 (sic), which is no longer necessary.
4086 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4087 be the same on different OS.
4089 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4092 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4093 whether --show-vars was specified or not
4095 JJ/02 exipick: Added support for new ACL variable spool format introduced
4098 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4099 syntactically invalid From: or Reply-to: line, and a filter used this to
4100 generate an autoreply, and therefore failed to obtain an address for the
4101 autoreply, Exim could try to deliver to a non-existent relative file
4102 name, causing unrelated and misleading errors. What now happens is that
4103 it logs this as a hard delivery error, but does not attempt to create a
4106 PH/11 The exinext utility has a -C option for testing purposes, but although
4107 the given file was scanned by exinext itself; it wasn't being passed on
4108 when Exim was called.
4110 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4111 an end-of-file indication when reading a command response.
4113 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4114 compiled. In many other places in Exim, IPv6 addresses are always
4115 recognized, so I have changed this. It also means that IPv4 domain
4116 literals of the form [IPV4:n.n.n.n] are now always recognized.
4118 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4119 used if the router is not running as root, for example, when verifying at
4120 ACL time, or when using -bh. The debugging output from this situation was
4121 non-existent - all you got was a failure to exec. I have made two
4124 (a) Failures to set uid/gid, the current directory, or a process leader
4125 in a subprocess such as that created by queryprogram now generate
4126 suitable debugging output when -d is set.
4128 (b) The queryprogram router detects when it is not running as root,
4129 outputs suitable debugging information if -d is set, and then runs
4130 the subprocess without attempting to change uid/gid.
4132 PH/15 Minor change to Makefile for building test_host (undocumented testing
4135 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4136 additional section of a DNS packet that returns MX or SRV records.
4137 Instead, it always explicitly searches for A/AAAA records. This avoids
4138 major problems that occur when a DNS server includes only records of one
4139 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4140 fixed another bug: if SRV records were looked up and the corresponding
4141 address records were *not* found in the additional section, the port
4142 values from the SRV records were lost.
4144 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4145 using the correct key (the original address) when searching the retry
4146 rules in order to find which one to use for generating the retry hint.
4148 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4149 adding the default one. Similarly, if it contains a Reply-To: header, the
4150 errors_reply_to option, if set, is not used.
4152 PH/19 When calculating a retry time, Exim used to measure the "time since
4153 failure" by looking at the "first failed" field in the retry record. Now
4154 it does not use this if it is later than than the arrival time of the
4155 message. Instead it uses the arrival time. This makes for better
4156 behaviour in cases where some deliveries succeed, thus re-setting the
4157 "first failed" field. An example is a quota failure for a huge message
4158 when small messages continue to be delivered. Without this change, the
4159 "time since failure" will always be short, possible causing more frequent
4160 delivery attempts for the huge message than are intended.
4161 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4163 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4164 $1, $2, $3) because the numerical variables can be reset during some
4165 expansion items (e.g. "match"), thereby losing the authentication data.
4167 PH/21 Make -bV show the size of off_t variables so that the test suite can
4168 decide whether to run tests for quotas > 2G.
4170 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4171 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4172 filecount value is greater than 2G or if a quota value is greater than 2G
4173 on a system where the size of off_t is not greater than 4, a panic error
4176 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4177 never match. The debug and -bh output now contains an explicit error
4178 message indicating a malformed IPv4 address or mask.
4180 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4181 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4182 PH/23 above applies.
4184 PH/25 Do not write to syslog when running in the test harness. The only
4185 occasion when this arises is a failure to open the main or panic logs
4186 (for which there is an explicit test).
4188 PH/26 Added the /no_tell option to "control=freeze".
4190 PH/27 If a host name lookup failed very early in a connection, for example, if
4191 the IP address matched host_lookup and the reverse lookup yielded a name
4192 that did not have a forward lookup, an error message of the form "no IP
4193 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4194 could be logged. Now it outputs the IP address instead of "NULL".
4196 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4197 allows the sender and the authenticated sender to be set when
4198 submitting a message from within Exim. Since child_open_exim() is
4199 documented for local_scan(), the new function should be too.
4201 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4202 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4203 results in an empty string is now treated as unset.
4205 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4207 PH/31 Added sender_verify_fail logging option.
4209 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4210 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4211 tidied the source and removed it altogether.
4213 PH/33 When a queue run was abandoned because the load average was too high, a
4214 log line was always written; now it is written only if the queue_run log
4215 selector is set. In addition, the log line for abandonment now contains
4216 information about the queue run such as the pid. This is always present
4217 in "start" and "stop" lines but was omitted from the "abandon" line.
4219 PH/34 Omit spaces between a header name and the colon in the error message that
4220 is given when verify = headers_syntax fails (if there are lots of them,
4221 the message gets confusing).
4223 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4224 names, as there are now some PTR records that contain slashes. This check
4225 is only to protect against broken name servers that fall over on strange
4226 characters, so the fact that it applies to all lookups doesn't matter.
4228 PH/36 Now that the new test suite is complete, we can remove some of the
4229 special code in Exim that was needed for the old test suite. For example,
4230 sorting DNS records because real resolvers return them in an arbitrary
4231 order. The new test suite's fake resolver always returns records in the
4234 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4235 bounces) except when queue_only is set, to avoid logging races between
4236 the different processes.
4238 PH/38 Panic-die if .include specifies a non-absolute path.
4240 PH/39 A tweak to the "H" retry rule from its user.
4242 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4243 a label. They prevented compilation on older perls.
4245 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4246 a warning to be raised on newish perls.
4248 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4249 on queue. Changes to match documented behaviour of showing count of
4250 messages matching specified criteria.
4252 PH/40 Changed the default ident timeout from 30s to 5s.
4254 PH/41 Added support for the use of login_cap features, on those BSD systems
4255 that have them, for controlling the resources used by pipe deliveries.
4257 PH/42 The content-scanning code uses fopen() to create files in which to put
4258 message data. Previously it was not paying any attention to the mode of
4259 the files. Exim runs with umask(0) because the rest of the code creates
4260 files with open(), and sets the required mode explicitly. Thus, these
4261 files were ending up world-writeable. This was not a big issue, because,
4262 being within the spool directory, they were not world-accessible. I have
4263 created a function called modefopen, which takes an additional mode
4264 argument. It sets umask(777), creates the file, chmods it to the required
4265 mode, then resets the umask. All the relevant calls to fopen() in the
4266 content scanning code have been changed to use this function.
4268 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4269 to 24 hours. This avoids potential overflow problems when processing G
4270 and H retry rules. I suspect nobody ever tinkers with this value.
4272 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4274 PH/45 When the plaintext authenticator is running as a client, the server's
4275 challenges are checked to ensure they are valid base64 strings. By
4276 default, the authentication attempt is cancelled if an invalid string is
4277 received. Setting client_ignore_invalid_base64 true ignores these errors.
4278 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4279 they are received. Thus, the responses can be made to depend on the
4280 challenges. If an invalid string is ignored, an empty string is placed in
4283 PH/46 Messages that are created by the autoreply transport now contains a
4284 References: header, in accordance with RFCs 2822 and 3834.
4286 PH/47 Added authenticated_sender_force to the smtp transport.
4288 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4290 PH/49 Installed latest patch from the Sieve maintainer.
4292 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4293 was also set, the appendfile transport was unnecessarily scanning a
4294 directory of message files (e.g. for maildir delivery) to find the count
4295 of files (along with the size), even though it did not need this
4296 information. It now does the scan only if it needs to find either the
4297 size of the count of files.
4299 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4301 PH/52 Two bugs concerned with error handling when the smtp transport is
4304 (i) Exim was not creating retry information for temporary errors given
4305 for individual recipients after the DATA command when the smtp transport
4306 was used in LMTP mode. This meant that they could be retried too
4307 frequently, and not timed out correctly.
4309 (ii) Exim was setting the flag that allows error details to be returned
4310 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4311 recipients that were returned after the DATA command.
4313 PH/53 This is related to PH/52, but is more general: for any failing address,
4314 when detailed error information was permitted to be returned to the
4315 sender, but the error was temporary, then after the final timeout, only
4316 "retry timeout exceeded" was returned. Now it returns the full error as
4317 well as "retry timeout exceeded".
4319 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4320 do this, and (what is worse) MTAs that accept it.
4322 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4323 will now be deprecated.
4325 PH/56 New os.c-cygwin from the Cygwin maintainer.
4327 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4328 formats (previously only available in exim formats via -bpr, -bpru,
4329 and -bpra. Now also available in native and exiqgrep formats)
4331 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4332 with very large, slow to parse queues
4334 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4336 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4338 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4339 responses to authentication challenges, though it was showing the
4340 challenges; (ii) I've removed the CR characters from the debug output for
4343 PH/58 Allow for the insertion of a newline as well as a space when a string
4344 is turned into more than one encoded-word during RFC 2047 encoding. The
4345 Sieve code now uses this.
4347 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4348 data_4xx, lost_connection, tls_required.
4350 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4351 message was being sent as an SMTP response.
4353 PH/61 Add -l and -k options to exicyclog.
4355 PH/62 When verifying, if an address was redirected to one new address, so that
4356 verification continued, and the new address failed or deferred after
4357 having set something in $address_data, the value of $address_data was not
4358 passed back to the ACL. This was different to the case when no
4359 redirection occurred. The value is now passed back in both cases.
4361 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4362 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4363 use login_cap.h, so on its own it isn't the distinguishing feature. The
4364 new name refers directly to the setclassresources() function.
4366 PH/65 Added configuration files for NetBSD3.
4368 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4370 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4371 is preferred over IPv4.
4373 PH/68 The bounce_return_message and bounce_return_body options were not being
4374 honoured for bounces generated during the reception of non-SMTP messages.
4375 In particular, this applied to messages rejected by the ACL. This bug has
4376 been fixed. However, if bounce_return_message is true and bounce_return_
4377 body is false, the headers that are returned for a non-SMTP message
4378 include only those that have been read before the error was detected.
4379 (In the case of an ACL rejection, they have all been read.)
4381 PH/69 The HTML version of the specification is now built in a directory called
4382 spec_html instead of spec.html, because the latter looks like a path with
4383 a MIME-type, and this confuses some software.
4385 PH/70 Catch two compiler warnings in sieve.c.
4387 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4388 function verify_get_ident() calls ip_connect() to connect a socket, but
4389 if the "connect()" function timed out, ip_connect() used to close the
4390 socket. However, verify_get_ident() also closes the socket later, and in
4391 between Exim writes to the log, which may get opened at this point. When
4392 the socket was closed in ip_connect(), the log could get the same file
4393 descriptor number as the socket. This naturally causes chaos. The fix is
4394 not to close the socket in ip_connect(); the socket should be closed by
4395 the function that creates it. There was only one place in the code where
4396 this was missing, in the iplookup router, which I don't think anybody now
4397 uses, but I've fixed it anyway.
4399 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4400 well as to direct DNS lookups. Otherwise the handling of names in host
4401 lists is inconsistent and therefore confusing.
4407 PH/01 Two changes to the default runtime configuration:
4409 (1) Move the checks for relay_from_hosts and authenticated clients from
4410 after to before the (commented out) DNS black list checks.
4412 (2) Add control=submission to the relay_from_hosts and authenticated
4413 clients checks, on the grounds that messages accepted by these
4414 statements are most likely to be submissions.
4416 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4418 (1) Generate an error if the third argument for the ${prvs expansion is
4421 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4424 (3) Reset the variables that are obtained from the first argument of
4425 ${prvscheck and used in the second argument before leaving the code,
4426 because their memory is reclaimed, so using them afterwards may do
4429 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4430 one (it's much easier than Tom thought :-).
4432 (5) Because of (4), we can now allow for the use of $prvscheck_result
4433 inside the third argument.
4435 PH/03 For some reason, the default setting of PATH when running a command from
4436 a pipe transport was just "/usr/bin". I have changed it to
4439 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4440 anything to be listed in the output from -bV.
4442 PH/05 When a filter generated an autoreply, the entire To: header line was
4443 quoted in the delivery log line, like this:
4445 => >A.N.Other <ano@some.domain> <original@ddress> ...
4447 This has been changed so that it extracts the operative address. There
4448 may be more than one such address. If so, they are comma-separated, like
4451 => >ano@some.domain,ona@other.domain <original@ddress> ...
4453 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4454 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4455 not being looked up in the rDNS to get a host name, Exim was showing the
4456 IP address twice in Received: lines, even though the IP addresses were
4457 identical. For example:
4459 Received: from [1.2.3.4] (helo=[1.2.3.4])
4461 However, if the real host name was known, it was omitting the HELO data
4462 if it matched the actual IP address. This has been tidied up so that it
4463 doesn't show the same IP address twice.
4465 PH/07 When both +timestamp and +memory debugging was on, the value given by
4466 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4467 called by the debug printing, thereby overwriting the timestamp buffer.
4468 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4470 PH/08 When the original message was included in an autoreply transport, it
4471 always said "this is a copy of the message, including all the headers",
4472 even if body_only or headers_only was set. It now gives an appropriate
4475 PH/09 Applied a patch from the Sieve maintainer which:
4477 o fixes some comments
4478 o adds the (disabled) notify extension core
4479 o adds some debug output for the result of if/elsif tests
4480 o points to the current vacation draft in the documentation
4481 and documents the missing references header update
4485 o fixes a bug in processing the envelope test (when testing
4486 multiple envelope elements, the last element determined the
4489 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4490 Electronic Mail") by including:
4492 Auto-submitted: auto-generated
4494 in the messages that it generates (bounce messages and others, such as
4495 warnings). In the case of bounce messages for non-SMTP messages, there was
4496 also a typo: it was using "Auto_submitted" (underscore instead of
4497 hyphen). Since every message generated by Exim is necessarily in response
4498 to another message, thes have all been changed to:
4500 Auto-Submitted: auto-replied
4502 in accordance with these statements in the RFC:
4504 The auto-replied keyword:
4506 - SHOULD be used on messages sent in direct response to another
4507 message by an automatic process,
4509 - MUST NOT be used on manually-generated messages,
4511 - MAY be used on Delivery Status Notifications (DSNs) and Message
4512 Disposition Notifications (MDNs),
4514 - MUST NOT be used on messages generated by automatic or periodic
4515 processes, except for messages which are automatic responses to
4518 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4519 to the default Received: header definition.
4521 PH/12 Added log selector acl_warn_skipped (default on).
4523 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4524 variables because (a) they are in the wrong storage pool and (b) even if
4525 they were copied, it wouldn't work properly because of the caching.
4527 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4528 checking when decoding. Apparently there are clients that generate
4529 overlong encoded strings. Why am I not surprised?
4531 PH/15 If the first argument of "${if match_address" was not empty, but did not
4532 contain an "@" character, Exim crashed. Now it writes a panic log message
4533 and treats the condition as false.
4535 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4537 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4538 extension "envelope-auth". The code is finished and in agreement with
4539 other implementations, but there is no documentation so far and in fact,
4540 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4541 not changing the active code.
4543 Print executed "if" and "elsif" statements when debugging is used. This
4544 helps a great deal to understand what a filter does.
4546 Document more things not specified clearly in RFC3028. I had all this
4547 sorted out, when out of a sudden new issues came to my mind. Oops."
4549 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4552 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4553 canonical form (as documented). However, after a host name lookup from
4554 the IP address, check_host() was doing a simple string comparison with
4555 addresses acquired from the DNS when checking that the found name did
4556 have the original IP as one of its addresses. Since any found IPv6
4557 addresses are likely to be in abbreviated form, the comparison could
4558 fail. Luckily, there already exists a function for doing the comparison
4559 by converting both addresses to binary, so now that is used instead of
4560 the text comparison.
4562 PH/20 There was another similar case to PH/19, when a complete host name was
4563 given in a host list; looking up its IP address could give an abbreviated
4564 form, whereas the current host's name might or might not be abbreviated.
4565 The same fix has been applied.
4571 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4572 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4575 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4576 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4578 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4580 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4581 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4582 Also, for "Auto-Submitted", treat anything other than "no" as
4583 non-personal, in accordance with RFC 3834. (Previously it treated
4584 anything starting "auto-" as non-personal.)
4586 TF/01 The control=submission/name=... option had a problem with syntax
4587 errors if the name included a slash character. The /name= option
4588 now slurps the rest of the string, so it can include any characters
4589 but it must come last in the list of options (after /sender_retain
4592 PH/05 Some modifications to the interface to the fake nameserver for the new
4600 TK/01 Added the "success_on_redirect" address verification option. See
4601 NewStuff for rationale and an example.
4603 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4605 PH/02 Patch to exigrep to allow it to work on syslog lines.
4607 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4608 fread() to skip over the body file's header line, because in Cygwin the
4609 header line is locked and is inaccessible.
4611 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4612 co-exist for some time) to make it clear that it is the Exim ID that is
4613 referenced, not the Message-ID: header line.
4615 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4616 string_format() function, because snprintf() does not exist on all
4619 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4620 ${stat: expansion item.
4622 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4623 protocol synchronization error", to keep the pedants happy.
4625 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4626 well as for IRIX systems, when gcc is being used. See the host.c source
4629 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4631 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4634 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4635 transport and to the smtp transport in LMTP mode.
4637 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4639 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4640 run a filter in a subprocess. This could lead to confusion in subsequent
4641 lookups in the parent process. There should also be a search_tidyup() at
4642 the end of the subprocess.
4644 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4645 only if the host matched helo_try_verify_hosts, which caused the
4646 verification to occur when the EHLO/HELO command was issued. The ACL just
4647 tested the remembered result. Now, if a previous verification attempt has
4648 not happened, "verify = helo" does it there and then.
4650 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4652 TK/03 Fix log output including CR from clamd.
4654 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4655 exist provoked a memory error which could cause a segfault.
4657 PH/15 Installed PCRE 6.2
4659 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4661 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4662 of the problem. Specifically, suggested +O2 rather than +O1 for the
4665 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4667 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4668 record, the port was not correctly logged when the outgoing_port log
4669 selector was set (it logged the transort's default port).
4671 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4672 fallback_hosts, and "hosts" in the smtp transport.
4674 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4675 host errors such as "Connection refused".
4677 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4678 authentication with radiusclient 0.4.9:
4680 - Error returned from rc_read_config was caught wrongly
4681 - Username/password not passed on to radius server due to wrong length.
4683 The presumption is that some radiusclient API changes for 4.51/PH/17
4684 were not taken care of correctly. The code is still untested by me (my
4685 Linux distribution still has 0.3.2 of radiusclient), but it was
4686 contributed by a Radius user.
4688 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4689 expanding the "port" option of the smtp transport.
4691 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4692 while reading a MIME header. Thanks to Tom Hughes for a patch.
4694 PH/24 Include config.h inside local_scan.h so that configuration settings are
4697 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4698 that in an ACL for RCPT (for example), you can examine exactly what was
4701 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4702 commands, but it was not correctly comparing the address with the actual
4703 client host address. Thus, it would show the EHLO address in Received:
4704 header lines when this was not necessary.
4706 PH/27 Added the % operator to ${eval:}.
4708 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4709 it should be ignoring failures (because with -C, for example, it has lost
4710 privilege). It wasn't ignoring creation failures other than "already
4713 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4716 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4717 because an input line was too long, either on its own, or by virtue of
4718 too many continuations, the temporary file was not being removed, and the
4719 return code was incorrect.
4721 PH/31 Missing "BOOL" in function definition in filtertest.c.
4723 PH/32 Applied Sieve patches from the maintainer.
4725 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4727 PH/33 Added "verify = not_blind".
4729 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4730 Local/Makefile (with some defaults set). These are used in built scripts
4731 such as exicyclog, but they have never been used in the exim_install
4732 script (though there are many overriding facilities there). I have
4733 arranged that the exim_install script now takes note of these two
4736 PH/35 Installed configuration files for Dragonfly.
4738 PH/36 When a locally submitted message by a trusted user did not contain a
4739 From: header, and the sender address was obtained from -f or from an SMTP
4740 MAIL command, and the trusted user did not use -F to supply a sender
4741 name, $originator_name was incorrectly used when constructing a From:
4742 header. Furthermore, $originator_name was used for submission mode
4743 messages from external hosts without From: headers in a similar way,
4744 which is clearly wrong.
4746 PH/37 Added control=suppress_local_fixups.
4748 PH/38 When log_selector = +received_sender was set, and the addition of the
4749 sender made the log line's construction buffer exactly full, or one byte
4750 less than full, an overflow happened when the terminating "\n" was
4753 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4754 when the result of a list match is failure because a DNS lookup failed.
4756 PH/40 RM_COMMAND is now used in the building process.
4758 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4759 the "build-* directories that it finds.
4761 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4762 address in a domain literal was a prefix of an interface address.
4764 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4765 when verifying a sender address, unless rewrite_headers is false.
4767 PH/44 Wrote a long comment about why errors_to addresses are verified as
4768 recipients, not senders.
4770 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4771 the ratelimit ACL was added.
4773 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4775 PH/46 Added extra information about PostgreSQL errors to the error string.
4777 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4778 avoiding the need to install special zones in a real server. This is
4779 backwards compatible; if it can't find the fake resolver, it drops back.
4780 Thus, both old and new test suites can be run.
4782 TF/02 Added util/ratelimit.pl
4784 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4785 clock is set back in time.
4787 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4788 Candler <B.Candler@pobox.com>.
4790 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4791 OK for addresses that are the result of redirections.
4793 PH/48 A number of further additions for the benefit of the new test suite,
4794 including a fake gethostbyname() that interfaces to the fake DNS resolver
4797 TF/06 The fix for widen_domains has also been applied to qualify_single and
4798 search_parents which are the other dnslookup options that can cause
4801 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4804 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4806 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4807 service" if it encountered an MX record with an empty target hostname.
4808 The message is now "an MX or SRV record indicated no SMTP service".
4810 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4811 if the DNS of the sending site is misconfigured. This is quite a
4812 common situation. This change restores the behaviour of treating a
4813 helo verification defer as a failure.
4815 PH/51 If self=fail was set on a router, the bounce message did not include the
4816 actual error message.
4822 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4824 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4825 command itself ended in error, the underlying message about the transport
4826 filter timeout was being overwritten with the pipe command error. Now the
4827 underlying error message should be appended to the second error message.
4829 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4831 PH/02 Reduce the amount of output that "make" produces by default. Full output
4832 can still be requested.
4834 PH/03 The warning log line about a condition test deferring for a "warn" verb
4835 was being output only once per connection, rather than after each
4836 occurrence (because it was using the same function as for successful
4837 "warn" verbs). This seems wrong, so I have changed it.
4839 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4840 it should not have, which might have caused a crash in the right
4841 circumstances, but probably never did.
4843 PH/04 Installed a modified version of Tony Finch's patch to make submission
4844 mode fix the return path as well as the Sender: header line, and to
4845 add a /name= option so that you can make the user's friendly name appear
4848 TF/03 Added the control = fakedefer ACL modifier.
4850 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4851 Mark Lowes for thorough testing.
4853 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4855 TK/03 Merged latest SRS patch from Miles Wilton.
4857 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4858 with the definition in sysexits.h (which is #included earlier).
4859 Fortunately, Exim does not actually use EX_OK. The code used to try to
4860 preserve the sysexits.h value, by assuming that macro definitions were
4861 scanned for macro replacements. I have been disabused of this notion,
4862 so now the code just undefines EX_OK before #including unistd.h.
4864 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4865 in the smtp transport. When a block could not be written in a single
4866 write() function, the timeout was being re-applied to each part-write.
4867 This seems wrong - if the receiver was accepting one byte at a time it
4868 would take for ever. The timeout is now adjusted when this happens. It
4869 doesn't have to be particularly precise.
4871 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4872 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4874 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4875 without a domain if the check to <postmaster@domain> fails.
4877 SC/01 Eximstats: added -xls and the ability to specify output files
4878 (patch written by Frank Heydlauf).
4880 SC/02 Eximstats: use FileHandles for outputting results.
4882 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4884 SC/04 Eximstats: fixed display of large numbers with -nvr option
4886 SC/05 Eximstats: fixed merging of reports with empty tables.
4888 SC/06 Eximstats: added the -include_original_destination flag
4890 SC/07 Eximstats: removed tabs and trailing whitespace.
4892 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4894 TK/06 MBOX spool code: Add real "From " MBOX separator line
4895 so the .eml file is really in mbox format (even though
4896 most programs do not really care). Patch from Alex Miller.
4898 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4899 The latter is generated from $received_to and is only set if the
4900 message has one envelope recipient. SA can use these headers,
4901 obviously out-of-the-box. Patch from Alex Miller.
4903 PH/08 The ${def test on a variable was returning false if the variable's
4904 value was "0", contrary to what the specification has always said!
4905 The result should be true unless the variable is empty.
4907 PH/09 The syntax error of a character other than { following "${if
4908 def:variable_name" (after optional whitespace) was not being diagnosed.
4909 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4910 accidental colon was present, for example, could give incorrect results.
4912 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4913 result is used (not including appendfile, where other changes are about
4916 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4917 This involved changing a lot of size variables from int to off_t. It
4918 should work with maildirs and everything.
4920 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4921 spamd dying while we are connected to it.
4923 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4926 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4927 function for BDB 4.3.
4929 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4931 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4932 the sources for PCRE have been re-arranged and more files are now
4935 PH/15 The code I had for printing potentially long long variables in PH/11
4936 above was not the best (it lost precision). The length of off_t variables
4937 is now inspected at build time, and an appropriate printing format (%ld
4938 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4939 to be "long long int" or "long int". This is needed for the internal
4940 formatting function string_vformat().
4942 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4943 the configuration file to be ":syslog", then the script "guesses" where
4944 the logs files are, rather than using the compiled in default. In our
4945 case the guess is not the same as the compiled default, so the script
4946 suddenly stopped working when I started to use syslog. The patch checks
4947 to see if log_file_path is "". If so, it attempts to read it from exim
4948 with no configuration file to get the compiled in version, before it
4949 falls back to the previous guessing code."
4951 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4952 implementing BATV in an Exim configuration. See NewStuff for the gory
4955 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4956 Makefile that are specific to HP-UX.
4958 PH/18 If the "use_postmaster" option was set for a recipient callout together
4959 with the "random" option, the postmaster address was used as the MAIL
4960 FROM address for the random test, but not for the subsequent recipient
4961 test. It is now used for both.
4963 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4964 patch removes a few documentation additions to RFC 3028, because the
4965 latest draft now contains them. It adds the new en;ascii-case comparator
4966 and a new error check for 8bit text in MIME parts. Comparator and
4967 require names are now matched exactly. I enabled the subaddress
4968 extension, but it is not well tested yet (read: it works for me)."
4970 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4971 rework some of the code of TK/09 above to avoid the hardwired use of
4972 "%lld" and "long long". Replaced the call to snprintf() with a call to
4975 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4976 records point to non-existent hosts", "retry timeout exceeded", and
4977 "retry time not reached for any host after a long failure period".
4979 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4980 experimental DomainKeys support:
4982 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4983 (2) On an error such as an illegally used "control", the wrong name for
4984 the control was given.
4986 These problems did NOT occur unless DomainKeys support was compiled.
4988 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4990 PH/24 Added ${if match_ip condition.
4992 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4993 databases so that it will be absolutely obvious if a crash occurs in the
4994 DB library. This is a regular occurrence (often caused by mis-matched
4997 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4998 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4999 picked up on a user's system that detects such things. There doesn't seem
5000 to be a gcc warning option for this - only an attribute that has to be
5001 put on the function's prototype. It seems that in Fedora Core 4 they have
5002 set this on a number of new functions. No doubt there will be more in due
5005 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5006 specify a transport. However, if an address that was verified by such a
5007 router was the subject of a callout, Exim crashed because it tried to
5008 read the rcpt_include_affixes from the non-existent transport. Now it
5009 just assumes that the setting of that option is false. This bug was
5010 introduced by 4.51/PH/31.
5012 PH/28 Changed -d+all to exclude +memory, because that information is very
5013 rarely of interest, but it makes the output a lot bigger. People tend to
5014 do -d+all out of habit.
5016 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5017 code in os-type was giving problems when libc.so lives in lib64, like on
5020 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5021 aren't the modern standard, and it seems that some systems' include files
5022 don't always have them. Exim was already checking for some of the newer
5023 ones like T_AAAA, and defining it itself. I've added checks for all the
5024 record types that Exim uses.
5026 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5027 not automatically generating a new one, as it is supposed to. This
5028 prevented TLS from working. If the file did exist, but contained invalid
5029 data, a new version was generated, as expected. It was only the case of a
5030 non-existent file that was broken.
5032 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5033 with a change in libdomainkeys > 0.64.
5035 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5036 from DNS. If the selector record carries the flag, it now has
5037 precedence over the domain-wide flag.
5039 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5041 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5042 the use of a port name that isn't defined in /etc/services) occurred, the
5043 message was deferred as in a normal delivery, and thus remained on the
5044 spool, instead of being failed because of the mua_wrapper setting. This
5045 is now fixed, and I tidied up some of the mua_wrapper messages at the
5048 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5049 the messages in a hash of arrays rather than using individual hashes.
5050 This is a bit cleaner and results in dramatic memory savings, albeit
5051 at a slight CPU cost.
5053 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5054 as requested by Marc Sherman.
5056 SC/10 Eximstats: added histograms for user specified patterns as requested
5059 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5061 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5062 fopen() in the content-scanning modules that did not already have it.
5068 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5069 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5071 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5073 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5075 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5076 to test Sieve filters that use "vacation".
5078 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5079 that changes the way the GnuTLS parameters are stored in the cache file.
5080 The new format can be generated externally. For backward compatibility,
5081 if the data in the cache doesn't make sense, Exim assumes it has read an
5082 old-format file, and it generates new data and writes a new file. This
5083 means that you can't go back to an older release without removing the
5086 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5087 work if there are any delivery delays because "one_time" forces the
5088 parent to be marked "delivered", so its unseen clone is never tried
5089 again. For this reason, Exim now forbids the simultaneous setting of
5092 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5093 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5094 entry if you want to know the details. The fix, however, neglected to
5095 consider the case when local delivery batching is involved. The test for
5096 "previously delivered" was not happening when checking to see if an
5097 address could be batched with a previous (undelivered) one; under
5098 certain circumstances this could lead to multiple deliveries to the same
5101 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5102 in its include files, and this causes problems building Exim.
5104 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5105 header_syntax) but Exim was just ignoring anything given after a slash.
5106 In particular, this caused confusion with an attempt to use "verify =
5107 reverse_host_lookup/defer_ok". An error is now given when options are
5108 supplied for verify items that do not have them. (Maybe reverse_host_
5109 lookup should have a defer_ok option, but that's a different point.)
5111 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5112 defined by RFC 821) to 2048, because there were problems with some AUTH
5113 commands, and RFC 1869 says the size should be increased for extended
5114 SMTP commands that take arguments.
5116 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5119 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5120 "unknown" error; now it says that the functionality isn't in the binary.
5122 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5123 an address' error message when a string expansion fails (syntax or
5124 whatever). Otherwise the password may appear in the log. Following change
5125 PH/42 below, there is no longer a chance of it appearing in a bounce
5128 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5130 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5131 of its addresses was ever tried. (Bugzilla bug #2.)
5133 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5134 the result incorrectly in the debug output. (It correctly added a newline
5135 to what was transported.)
5137 TF/01 Added $received_time.
5139 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5140 commented out examples of how to interface to a virus scanner and to
5141 SpamAssassin. Also added commented examples of av_scanner and
5142 spamd_address settings.
5144 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5145 and controls are allowed in which ACLs. There were a couple of minor
5146 errors. Some of the entries in the conditions table (which is a table of
5147 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5148 negation of where the condition IS allowed.
5150 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5152 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5153 header file does not have a version number, so I've had to invent a new
5154 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5155 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5156 radiusclient), but it was contributed by a Radius user.
5158 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5159 files or directories, for OpenSSL.
5161 PH/19 When an Exim process that is running as root has to create an Exim log
5162 file, it does so in a subprocess that runs as exim:exim so as to get the
5163 ownership right at creation (otherwise, other Exim processes might see
5164 the file with the wrong ownership). There was no test for failure of this
5165 fork() call, which would lead to the process getting stuck as it waited
5166 for a non-existent subprocess. Forks do occasionally fail when resources
5167 run out. I reviewed all the other calls to fork(); they all seem to check
5170 PH/20 When checking for unexpected SMTP input at connect time (before writing
5171 the banner), Exim was not dealing correctly with a non-positive return
5172 from the read() function. If the client had disconnected by this time,
5173 the result was a log entry for a synchronization error with an empty
5174 string after "input=" when read() returned zero. If read() returned -1
5175 (an event I could not check), uninitialized data bytes were printed.
5176 There were reports of junk text (parts of files, etc) appearing after
5179 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5181 PH/22 Added support for macro redefinition, and (re)definition in between
5182 driver and ACL definitions.
5184 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5185 forgetting to use the resulting value; it was using the unexpanded value.
5187 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5188 hadn't been configured. The fix is from Juergen Kreileder, who
5189 understands it better than I do:
5191 "Here's what I see happening with three configured cyrus_sasl
5192 authenticators configured (plain, login, cram-md5):
5194 On startup auth_cyrus_sasl_init() gets called for each of these.
5195 This means three calls to sasl_listmech() without a specified mech_list.
5196 => SASL tests which mechs of all available mechs actually work
5197 => three warnings about OTP not working
5198 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5200 With the patch, sasl_listmech() also gets called three times. But now
5201 SASL's mech_list option is set to the server_mech specified in the the
5202 authenticator. Or in other words, the answer from sasl_listmech()
5203 gets limited to just the mech you're testing for (which is different
5205 => the return list contains just 'plain' or 'login', 'cram-md5' or
5206 nothing depending on the value of ob->server_mech.
5208 I've just tested the patch: Authentication still works fine,
5209 unavailable mechs specified in the exim configuration are still
5210 caught, and the auth.log warnings about OTP are gone."
5212 PH/25 When debugging is enabled, the contents of the command line are added
5213 to the debugging output, even when log_selector=+arguments is not
5216 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5217 answer is "GNU", and only if the return is "GNU/something" is the answer
5220 PH/27 $acl_verify_message is now set immediately after the failure of a
5221 verification in an ACL, and so is available in subsequent modifiers. In
5222 particular, the message can be preserved by coding like this:
5224 warn !verify = sender
5225 set acl_m0 = $acl_verify_message
5227 Previously, $acl_verify_message was set only while expanding "message"
5228 and "log_message" when a very denied access.
5230 PH/28 Modified OS/os.c-Linux with
5232 -#ifndef OS_LOAD_AVERAGE
5233 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5235 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5236 nomenclature these days.)
5238 PH/29 Installed patch from the Sieve maintainer that adds the options
5239 sieve_useraddress and sieve_subaddress to the redirect router.
5241 PH/30 In these circumstances:
5242 . Two addresses routed to the same list of hosts;
5243 . First host does not offer TLS;
5244 . First host accepts first address;
5245 . First host gives temporary error to second address;
5246 . Second host offers TLS and a TLS session is established;
5247 . Second host accepts second address.
5248 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5249 and peerdn, if requested) that were in fact used only for the second
5252 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5253 attention to any local part prefix or suffix that was matched by the
5254 router that accepted the address. It now behaves in the same way as it
5255 does for delivery: the affixes are removed from the local part unless
5256 rcpt_include_affixes is set on the transport.
5258 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5259 timeout during the DATA phase of an incoming message.
5261 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5262 applied a patch sanctioned by the Sieve maintainer.
5264 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5265 the uid or gid is negative. A case of a negative gid caused this to be
5266 noticed. The fix allows for either to be negative.
5268 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5269 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5272 PH/36 The change PH/12 above was broken. Fixed it.
5274 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5275 the grounds that routing the same address twice would always produce the
5276 same answer. This might have been true once, but it is certainly no
5277 longer true now. Routing a child address may depend on the previous
5278 routing that produced that child. Some complicated redirection strategies
5279 went wrong when messages had multiple recipients, and made Exim's
5280 behaviour dependent on the order in which the addresses were given.
5282 I have moved the duplicate checking until after the routing is complete.
5283 Exim scans the addresses that are assigned to local and remote
5284 transports, and removes any duplicates. This means that more work will be
5285 done, as duplicates will always all be routed, but duplicates are
5286 presumably rare, so I don't expect this is of any significance.
5288 For deliveries to pipes, files, and autoreplies, the duplicate checking
5289 still happens during the routing process, since they are not going to be
5292 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5293 It corrects a timeout issue with spamd. This is Ian's comment: "The
5294 background is that sometimes spamd either never reads data from a
5295 connection it has accepted, or it never writes response data. The exiscan
5296 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5297 blindly assumes that writes won't block so it may never time out."
5299 PH/39 Allow G after quota size as well as K and M.
5301 PH/40 The value set for $authenticated_id in an authenticator may not contain
5302 binary zeroes or newlines because the value is written to log lines and
5303 to spool files. There was no check on this. Now the value is run through
5304 the string_printing() function so that such characters are converted to
5305 printable escape sequences.
5307 PH/41 $message_linecount is a new variable that contains the total number of
5308 lines in the message. Compare $body_linecount, which is the count for the
5311 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5312 bounce and delay warning messages, except in certain special cases, which
5315 (a) An SMTP error message from a remote host;
5316 (b) A message specified in a :fail: redirection;
5317 (c) A message specified in a "fail" command in a system filter;
5318 (d) A message specified in a FAIL return from the queryprogram router;
5319 (e) A message specified by the cannot_route_message router option.
5321 In these cases only, Exim does include the error details in bounce and
5322 warning messages. There are also a few cases where bland messages such
5323 as "unrouteable address" or "local delivery error" are given.
5325 PH/43 $value is now also set for the "else" part of a ${run expansion.
5327 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5328 being worked on, but at least Exim now implements the latest version to
5331 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5332 process to complete was treated as a delivery failure, a timeout while
5333 writing the message to the pipe was logged, but erroneously treated as a
5334 successful delivery. Such timeouts include transport filter timeouts. For
5335 consistency with the overall process timeout, these timeouts are now
5336 treated as errors, giving rise to delivery failures by default. However,
5337 there is now a new Boolean option for the pipe transport called
5338 timeout_defer, which, if set TRUE, converts the failures into defers for
5339 both kinds of timeout. A transport filter timeout is now identified in
5342 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5343 systems where "make" and "gmake" are different, calling "gmake" at top
5344 level broke things. I've arranged for the value of $(MAKE) to be passed
5345 from the Makefile to this script so that it can call the same version of
5349 A note about Exim versions 4.44 and 4.50
5350 ----------------------------------------
5352 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5353 changes of various kinds. As a consequence, a big documentation update was
5354 needed. This delayed the release for rather longer than seemed good, especially
5355 in the light of a couple of (minor) security issues. Therefore, the changes
5356 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5357 release. So 4.44 and 4.50 are in effect two different branches that both start
5360 I have left the 4.50 change log unchanged; it contains all the changes since
5361 4.43. The change log for 4.44 is below; many of its items are identical to
5362 those for 4.50. This seems to be the most sensible way to preserve the
5363 historical information.
5369 1. Minor wording change to the doc/README.SIEVE file.
5371 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5372 computation of the current number of files was incorrect.
5374 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5375 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5378 4. Give more explanation in the error message when the command for a transport
5379 filter fails to execute.
5381 5. There are several places where Exim runs a non-Exim command in a
5382 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5383 was being done only for the command run by the queryprogram router. It is
5384 now done for all such subprocesses. The other cases are: ${run, transport
5385 filters, and the commands run by the lmtp and pipe transports.
5387 6. Added CONFIGURE_GROUP build-time option.
5389 7. Some older OS have a limit of 256 on the maximum number of file
5390 descriptors. Exim was using setrlimit() to set 1000 as a large value
5391 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5392 systems. I've change it so that if it can't get 1000, it tries for 256.
5394 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5395 was an oversight, and furthermore, ever since the addition of extra
5396 controls (e.g. 4.43/32), the checks on when to allow different forms of
5397 "control" were broken. There should now be diagnostics for all cases when a
5398 control that does not make sense is encountered.
5400 9. Added the /retain_sender option to "control=submission".
5402 10. $recipients is now available in the predata ACL (oversight).
5404 11. Tidy the search cache before the fork to do a delivery from a message
5405 received from the command line. Otherwise the child will trigger a lookup
5406 failure and thereby defer the delivery if it tries to use (for example) a
5407 cached ldap connection that the parent has called unbind on.
5409 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5410 of $address_data from the recipient verification was clobbered by the
5411 sender verification.
5413 13. The value of address_data from a sender verification is now available in
5414 $sender_address_data in subsequent conditions in the ACL statement.
5416 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5418 15. Added a new option "connect=<time>" to callout options, to set a different
5421 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5422 was its contents. (It was OK if the option was not defined at all.)
5424 17. A "Completed" log line is now written for messages that are removed from
5425 the spool by the -Mrm option.
5427 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5428 information about exactly what failed.
5430 19. Added -dd to debug only the daemon process.
5432 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5433 handles timeouts, both on the server side and network timeouts. Renamed the
5434 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5436 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5437 It is now set to "smtps".
5439 22. $host_address is now set to the target address during the checking of
5440 ignore_target_hosts.
5442 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5443 being passed; this would have caused $sender_host_name to have been used if
5444 matching the list had actually called for a host name (not very likely,
5445 since this list is usually IP addresses). A host name is now passed as
5448 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5449 code that specifies a non-restarting handler (typically sigaction() in
5450 modern systems) in an attempt to fix a rare and obscure crash bug.
5452 25. Narrowed the window for a race in the daemon that could cause it to ignore
5453 SIGCHLD signals. This is not a major problem, because they are used only to
5454 wake it up if nothing else does.
5456 26. A malformed maildirsize file could cause Exim to calculate negative values
5457 for the mailbox size or file count. Odd effects could occur as a result.
5458 The maildirsize information is now recalculated if the size or filecount
5461 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5462 support for a long time. Removed HAVE_SYS_VFS_H.
5464 28. Installed the latest version of exipick from John Jetmore.
5466 29. In an address list, if the pattern was not a regular expression, an empty
5467 subject address (from a bounce message) matched only if the pattern was an
5468 empty string. Non-empty patterns were not even tested. This was the wrong
5469 because it is perfectly reasonable to use an empty address as part of a
5470 database query. An empty address is now tested by patterns that are
5471 lookups. However, all the other forms of pattern expect the subject to
5472 contain a local part and a domain, and therefore, for them, an empty
5473 address still always fails if the pattern is not itself empty.
5475 30. Exim went into a mad DNS loop when attempting to do a callout where the
5476 host was specified on an smtp transport, and looking it up yielded more
5477 than one IP address.
5479 31. Re-factored the code for checking spool and log partition space into a
5480 function that finds that data and another that does the check. The former
5481 is then used to implement four new variables: $spool_space, $log_space,
5482 $spool_inodes, and $log_inodes.
5484 32. The RFC2047 encoding function was originally intended for short strings
5485 such as real names; it was not keeping to the 75-character limit for
5486 encoded words that the RFC imposes. It now respects the limit, and
5487 generates multiple encoded words if necessary. To be on the safe side, I
5488 have increased the buffer size for the ${rfc2047: expansion operator from
5491 33. It is now permitted to omit both strings after an "if" condition; if the
5492 condition is true, the result is "true". As before, when the second string
5493 is omitted, a false condition yields an empty string. This makes it less
5494 cumbersome to write custom ACL and router conditions.
5496 34. Failure to deliver a bounce message always caused it to be frozen, even if
5497 there was an errors_to setting on the router. The errors_to setting is now
5500 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5501 canonical form (fully expanded) before being placed in
5502 $sender_host_address.
5504 36. The table in the code that translates DNS record types into text (T_A to
5505 "A" for instance) was missing entries for NS and CNAME. It is just possible
5506 that this could have caused confusion if both these types were looked up
5507 for the same domain, because the text type is used as part of Exim's
5508 per-process caching. But the chance of anyone hitting this buglet seems
5511 37. The dnsdb lookup has been extended in a number of ways.
5513 (1) There is a new type, "zns", which walks up the domain tree until it
5514 finds some nameserver records. It should be used with care.
5516 (2) There is a new type, "mxh", which is like "mx" except that it returns
5517 just the host names, not the priorities.
5519 (3) It is now possible to give a list of domains (or IP addresses) to be
5520 looked up. The behaviour when one of the lookups defers can be
5521 controlled by a keyword.
5523 (4) It is now possible to specify the separator character for use when
5524 multiple records are returned.
5526 38. The dnslists ACL condition has been extended: it is now possible to supply
5527 a list of IP addresses and/or domains to be looked up in a particular DNS
5530 39. Added log_selector=+queue_time_overall.
5532 40. When running the queue in the test harness, wait just a tad after forking a
5533 delivery process, to get repeatability of debugging output.
5535 41. Include certificate and key file names in error message when GnuTLS fails
5536 to set them up, because the GnuTLS error message doesn't include the name
5537 of the failing file when there is a problem reading it.
5539 42. Allow both -bf and -bF in the same test run.
5541 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5543 44. The "Exiscan patch" is now merged into the mainline Exim source.
5545 45. Sometimes the final signoff response after QUIT could fail to get
5546 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5547 < 0 before doing a fflush(). This bug looks as though it goes back to the
5548 introduction of TLS in release 3.20, but "sometimes" must have been rare
5549 because the tests only now provoked it.
5551 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5552 (this can affect the format of dates).
5554 47. exim_tidydb, when checking for the continued existence of a message for
5555 which it has found a message-specific retry record, was not finding
5556 messages that were in split spool directories. Consequently, it was
5557 deleting retry records that should have stayed in existence.
5559 48. Steve fixed some bugs in eximstats.
5561 49. The SPA authentication driver was not abandoning authentication and moving
5562 on to the next authenticator when an expansion was forced to fail,
5563 contradicting the general specification for all authenticators. Instead it
5564 was generating a temporary error. It now behaves as specified.
5566 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5567 (the order specifies the preference for clients). The order is now AES256,
5568 AES128, 3DES, ARCFOUR128.
5570 51. Small patch to Sieve code - explicitly set From: when generating an
5573 52. Exim crashed if a remote delivery caused a very long error message to be
5574 recorded - for instance if somebody sent an entire SpamAssassin report back
5575 as a large number of 550 error lines. This bug was coincidentally fixed by
5576 increasing the size of one of Exim's internal buffers (big_buffer) that
5577 happened as part of the Exiscan merge. However, to be on the safe side, I
5578 have made the code more robust (and fixed the comments that describe what
5581 53. Now that there can be additional text after "Completed" in log lines (if
5582 the queue_time_overall log selector is set), a one-byte patch to exigrep
5583 was needed to allow it to recognize "Completed" as not the last thing in
5586 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5587 patch that reportedly fixes this has been added. I am not expert enough to
5588 create a test for it. This is what the patch creator wrote:
5590 "I found a little strange behaviour of ldap code when working with
5591 Windows 2003 AD Domain, where users was placed in more than one
5592 Organization Units. When I tried to give exim partial DN, the exit code
5593 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5594 But simultaneously result of request was absolutely normal ldap result,
5595 so I produce this patch..."
5597 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5598 REFERENCE, so I have modified the code to exclude the patch when that macro
5601 55. Some experimental protocols are using DNS PTR records for new purposes. The
5602 keys for these records are domain names, not reversed IP addresses. The
5603 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5604 leaves it alone. Component reversal etc. now happens only for IP addresses.
5607 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5609 57. Double the size of the debug message buffer (to 2048) so that more of very
5610 long debug lines gets shown.
5612 58. The exicyclog utility now does better if the number of log files to keep
5613 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5615 59. Two changes related to the smtp_active_hostname option:
5617 (1) $smtp_active_hostname is now available as a variable.
5618 (2) The default for smtp_banner uses $smtp_active_hostname instead
5619 of $primary_hostname.
5621 60. The host_aton() function is supposed to be passed a string that is known
5622 to be a valid IP address. However, in the case of IPv6 addresses, it was
5623 not checking this. This is a hostage to fortune. Exim now panics and dies
5624 if the condition is not met. A case was found where this could be provoked
5625 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5626 components; fortuitously, this particular loophole had already been fixed
5627 by change 4.50/55 above.
5629 If there are any other similar loopholes, the new check in host_aton()
5630 itself should stop them being exploited. The report I received stated that
5631 data on the command line could provoke the exploit when Exim was running as
5632 exim, but did not say which command line option was involved. All I could
5633 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5634 running as the user.
5637 61. There was a buffer overflow vulnerability in the SPA authentication code
5638 (which came originally from the Samba project). I have added a test to the
5639 spa_base64_to_bits() function which I hope fixes it.
5642 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5643 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5645 63. The daemon start-up calls getloadavg() while still root for those OS that
5646 need the first call to be done as root, but it missed one case: when
5647 deliver_queue_load_max is set with deliver_drop_privilege. This is
5648 necessary for the benefit of the queue runner, because there is no re-exec
5649 when deliver_drop_privilege is set.
5651 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5652 This has been fixed.
5654 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5655 was in use, was not putting the data itself into the right store pool;
5656 consequently, it could be overwritten for a subsequent message in the same
5657 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5660 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5662 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5663 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5664 the function were treating the return as a boolean value, which happened to
5665 work because 0=false and not-0=true, but is not correct code.
5667 68. The host_aton() function was not handling scoped IPv6 addresses (those
5668 with, for example, "%eth0" on the end) correctly.
5670 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5671 negated items (that is, ~something) in unsigned ints. Some compilers
5672 apparently mutter when there is no cast.
5674 70. If an address verification called from an ACL failed, and did not produce a
5675 user-specific message (i.e. there was only a "system" message), nothing was
5676 put in $acl_verify_message. In this situation, it now puts the system
5679 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5680 session; change 4.31/43 added the unwanted input to the log line - except
5681 that it did not do this in the start of session case. It now does.
5683 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5684 This is wrong and can cause the other end to generate a synchronization
5685 error if it is another Exim or anything else that does the synchronization
5686 check. A QUIT command is no longer sent after a timeout.
5688 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5689 during host lookups.
5691 74. The defer_ok option of callout verification was not working if it was used
5692 when verifying addresses in header lines, that is, for this case:
5694 verify = header_sender/callout=defer_ok
5696 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5697 those file descriptors could be used for SMTP connections. If anything
5698 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5699 could be sent to the SMTP client, causing chaos. The daemon now opens
5700 stdin, stdout, and stderr to /dev/null when it puts itself into the
5703 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5704 log by default. The user can override this with suitable Perl magic.
5706 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5707 the log message when discard triggers, was not working for the DATA ACL or
5708 for the non-SMTP ACL.
5710 78. Error message wording change in sieve.c.
5712 79. If smtp_accept_max_per_host was set, the number of connections could be
5713 restricted to fewer than expected, because the daemon was trying to set up
5714 a new connection before checking whether the processes handling previous
5715 connections had finished. The check for completed processes is now done
5716 earlier. On busy systems, this bug wouldn't be noticed because something
5717 else would have woken the daemon, and it would have reaped the completed
5720 80. If a message was submitted locally by a user whose login name contained one
5721 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5722 It caused a spool format error. I have fixed the spool reading code. A
5723 related problem was that the "from" clause in the Received: line became
5724 illegal because of the space(s). It is now covered by ${quote_local_part.
5726 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5729 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5731 83. Patch from Sieve maintainer to fix unterminated string problem in
5732 "vacation" handling.
5734 84. Some minor changes to the Linux configuration files to help with other
5735 OS variants using glibc.
5737 85. One more patch for Sieve to update vacation handling to latest spec.
5740 ----------------------------------------------------
5741 See the note above about the 4.44 and 4.50 releases.
5742 ----------------------------------------------------
5748 1. Change 4.43/35 introduced a bug that caused file counts to be
5749 incorrectly computed when quota_filecount was set in an appendfile
5752 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5753 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5756 3. Give more explanation in the error message when the command for a transport
5757 filter fails to execute.
5759 4. There are several places where Exim runs a non-Exim command in a
5760 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5761 was being done only for the command run by the queryprogram router. It is
5762 now done for all such subprocesses. The other cases are: ${run, transport
5763 filters, and the commands run by the lmtp and pipe transports.
5765 5. Some older OS have a limit of 256 on the maximum number of file
5766 descriptors. Exim was using setrlimit() to set 1000 as a large value
5767 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5768 systems. I've change it so that if it can't get 1000, it tries for 256.
5770 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5771 was an oversight, and furthermore, ever since the addition of extra
5772 controls (e.g. 4.43/32), the checks on when to allow different forms of
5773 "control" were broken. There should now be diagnostics for all cases when a
5774 control that does not make sense is encountered.
5776 7. $recipients is now available in the predata ACL (oversight).
5778 8. Tidy the search cache before the fork to do a delivery from a message
5779 received from the command line. Otherwise the child will trigger a lookup
5780 failure and thereby defer the delivery if it tries to use (for example) a
5781 cached ldap connection that the parent has called unbind on.
5783 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5784 of $address_data from the recipient verification was clobbered by the
5785 sender verification.
5787 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5788 was its contents. (It was OK if the option was not defined at all.)
5790 11. A "Completed" log line is now written for messages that are removed from
5791 the spool by the -Mrm option.
5793 12. $host_address is now set to the target address during the checking of
5794 ignore_target_hosts.
5796 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5797 being passed; this would have caused $sender_host_name to have been used if
5798 matching the list had actually called for a host name (not very likely,
5799 since this list is usually IP addresses). A host name is now passed as
5802 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5803 code that specifies a non-restarting handler (typically sigaction() in
5804 modern systems) in an attempt to fix a rare and obscure crash bug.
5806 15. Narrowed the window for a race in the daemon that could cause it to ignore
5807 SIGCHLD signals. This is not a major problem, because they are used only to
5808 wake it up if nothing else does.
5810 16. A malformed maildirsize file could cause Exim to calculate negative values
5811 for the mailbox size or file count. Odd effects could occur as a result.
5812 The maildirsize information is now recalculated if the size or filecount
5815 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5816 support for a long time. Removed HAVE_SYS_VFS_H.
5818 18. Updated exipick to current release from John Jetmore.
5820 19. Allow an empty sender to be matched against a lookup in an address list.
5821 Previously the only cases considered were a regular expression, or an
5824 20. Exim went into a mad DNS lookup loop when doing a callout where the
5825 host was specified on the transport, if the DNS lookup yielded more than
5828 21. The RFC2047 encoding function was originally intended for short strings
5829 such as real names; it was not keeping to the 75-character limit for
5830 encoded words that the RFC imposes. It now respects the limit, and
5831 generates multiple encoded words if necessary. To be on the safe side, I
5832 have increased the buffer size for the ${rfc2047: expansion operator from
5835 22. Failure to deliver a bounce message always caused it to be frozen, even if
5836 there was an errors_to setting on the router. The errors_to setting is now
5839 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5840 canonical form (fully expanded) before being placed in
5841 $sender_host_address.
5843 24. Updated eximstats to version 1.33
5845 25. Include certificate and key file names in error message when GnuTLS fails
5846 to set them up, because the GnuTLS error message doesn't include the name
5847 of the failing file when there is a problem reading it.
5849 26. Expand error message when OpenSSL has problems setting up cert/key files.
5852 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5853 (this can affect the format of dates).
5855 28. exim_tidydb, when checking for the continued existence of a message for
5856 which it has found a message-specific retry record, was not finding
5857 messages that were in split spool directories. Consequently, it was
5858 deleting retry records that should have stayed in existence.
5860 29. eximstats updated to version 1.35
5861 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5862 1.35 - bugfix such that pie charts by volume are generated correctly
5864 30. The SPA authentication driver was not abandoning authentication and moving
5865 on to the next authenticator when an expansion was forced to fail,
5866 contradicting the general specification for all authenticators. Instead it
5867 was generating a temporary error. It now behaves as specified.
5869 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5870 (the order specifies the preference for clients). The order is now AES256,
5871 AES128, 3DES, ARCFOUR128.
5873 31. Small patch to Sieve code - explicitly set From: when generating an
5876 32. Exim crashed if a remote delivery caused a very long error message to be
5877 recorded - for instance if somebody sent an entire SpamAssassin report back
5878 as a large number of 550 error lines. This bug was coincidentally fixed by
5879 increasing the size of one of Exim's internal buffers (big_buffer) that
5880 happened as part of the Exiscan merge. However, to be on the safe side, I
5881 have made the code more robust (and fixed the comments that describe what
5884 33. Some experimental protocols are using DNS PTR records for new purposes. The
5885 keys for these records are domain names, not reversed IP addresses. The
5886 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5887 leaves it alone. Component reversal etc. now happens only for IP addresses.
5890 34. The host_aton() function is supposed to be passed a string that is known
5891 to be a valid IP address. However, in the case of IPv6 addresses, it was
5892 not checking this. This is a hostage to fortune. Exim now panics and dies
5893 if the condition is not met. A case was found where this could be provoked
5894 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5895 components; fortuitously, this particular loophole had already been fixed
5896 by change 4.50/55 or 4.44/33 above.
5898 If there are any other similar loopholes, the new check in host_aton()
5899 itself should stop them being exploited. The report I received stated that
5900 data on the command line could provoke the exploit when Exim was running as
5901 exim, but did not say which command line option was involved. All I could
5902 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5903 running as the user.
5906 35. There was a buffer overflow vulnerability in the SPA authentication code
5907 (which came originally from the Samba project). I have added a test to the
5908 spa_base64_to_bits() function which I hope fixes it.
5911 36. The daemon start-up calls getloadavg() while still root for those OS that
5912 need the first call to be done as root, but it missed one case: when
5913 deliver_queue_load_max is set with deliver_drop_privilege. This is
5914 necessary for the benefit of the queue runner, because there is no re-exec
5915 when deliver_drop_privilege is set.
5917 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5918 was in use, was not putting the data itself into the right store pool;
5919 consequently, it could be overwritten for a subsequent message in the same
5920 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5923 38. Sometimes the final signoff response after QUIT could fail to get
5924 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5925 < 0 before doing a fflush(). This bug looks as though it goes back to the
5926 introduction of TLS in release 3.20, but "sometimes" must have been rare
5927 because the tests only now provoked it.
5933 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5934 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5935 Now it returns an integer. A number of calls were still expecting a T/F
5936 return. Fortuitously, in all cases, the tests worked in OK situations,
5937 which is the norm. However, things would have gone wrong on any write
5938 failures on the smtp file descriptor. This function is used when sending
5939 messages over SMTP and also when doing verify callouts.
5941 2. When Exim is called to do synchronous delivery of a locally submitted
5942 message (the -odf or -odi options), it no longer closes stderr before doing
5945 3. Implemented the mua_wrapper option.
5947 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5949 5. Implemented the functions header_remove(), header_testname(),
5950 header_add_at_position(), and receive_remove_recipient(), and exported them
5953 6. If an ACL "warn" statement specified the addition of headers, Exim already
5954 inserted X-ACL-Warn: at the start if there was no header name. However, it
5955 was not making this test for the second and subsequent header lines if
5956 there were newlines in the string. This meant that an invalid header could
5957 be inserted if Exim was badly configured.
5959 7. Allow an ACL "warn" statement to add header lines at the start or after all
5960 the Received: headers, as well as at the end.
5962 8. Added the rcpt_4xx retry error code.
5964 9. Added postmaster_mailfrom=xxx to callout verification option.
5966 10. Added mailfrom=xxxx to the callout verification option, for verify=
5969 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5970 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5972 12. Inserted some casts to stop certain compilers warning when using pointer
5973 differences as field lengths or precisions in printf-type calls (mostly
5974 affecting debugging statements).
5976 13. Added optional readline() support for -be (dynamically loaded).
5978 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5979 same clock tick as a message's arrival, so that its received time was the
5980 same as the "first fail" time on the retry record, and that message
5981 remained on the queue past the ultimate address timeout, every queue runner
5982 would try a delivery (because it was past the ultimate address timeout) but
5983 after another failure, the ultimate address timeout, which should have then
5984 bounced the address, did not kick in. This was a "< instead of <=" error;
5985 in most cases the first failure would have been in the next clock tick
5986 after the received time, and all would be well.
5988 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5989 being recognized when the domain list was tested by the match_domain
5990 condition in an expansion string.
5992 16. Added the ${str2b64: operator.
5994 17. Exim was always calling setrlimit() to set a large limit for the number of
5995 processes, without checking whether the existing limit was already
5996 adequate. (It did check for the limit on file descriptors.) Furthermore,
5997 errors from getrlimit() and setrlimit() were being ignored. Now they are
5998 logged to the main and panic logs, but Exim does carry on, to try to do its
5999 job under whatever limits there are.
6001 18. Imported PCRE 5.0.
6003 19. Trivial typo in log message " temporarily refused connection" (the leading
6006 20. If the log selector return_path_on_delivery was set and an address was
6007 redirected to /dev/null, the delivery process crashed because it assumed
6008 that a return path would always be set for a "successful" delivery. In this
6009 case, the whole delivery is bypassed as an optimization, and therefore no
6012 21. Internal re-arrangement: the function for sending a challenge and reading
6013 a response while authentication was assuming a zero-terminated challenge
6014 string. It's now changed to take a pointer and a length, to allow for
6015 binary data in such strings.
6017 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6019 23. Exim was not respecting finduser_retries when seeking the login of the
6020 uid under which it was called; it was always trying 10 times. (The default
6021 setting of finduser_retries is zero.) Also, it was sleeping after the final
6022 failure, which is pointless.
6024 24. Implemented tls_on_connect_ports.
6026 25. Implemented acl_smtp_predata.
6028 26. If the domain in control=submission is set empty, Exim assumes that the
6029 authenticated id is a complete email address when it generates From: or
6030 Sender: header lines.
6032 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6033 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6034 chown and chgrp in /bin and hostname in /usr/bin.
6036 28. Exim was keeping the "process log" file open after each use, just as it
6037 does for the main log. This opens the possibility of it remaining open for
6038 long periods when the USR1 signal hits a daemon. Occasional processlog
6039 errors were reported, that could have been caused by this. Anyway, it seems
6040 much more sensible not to leave this file open at all, so that is what now
6043 29. The long-running daemon process does not normally write to the log once it
6044 has entered its main loop, and it closes the log before doing so. This is
6045 so that log files can straightforwardly be renamed and moved. However,
6046 there are a couple of unusual error situations where the daemon does write
6047 log entries, and I had neglected to close the log afterwards.
6049 30. The text of an SMTP error response that was received during a remote
6050 delivery was being truncated at 512 bytes. This is too short for some of
6051 the long messages that one sometimes sees. I've increased the limit to
6054 31. It is now possible to make retry rules that apply only when a message has a
6055 specific sender, in particular, an empty sender.
6057 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6058 it possible to be selective about when SMTP synchronization is enforced.
6060 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6062 32. Implemented hosts_connection_nolog.
6064 33. Added an ACL for QUIT.
6066 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6069 35. Added mailbox_size and mailbox_filecount to appendfile.
6071 36. Added control = no_multiline_responses to ACLs.
6073 37. There was a bug in the logic of the code that waits for the clock to tick
6074 in the case where the clock went backwards by a substantial amount such
6075 that the microsecond fraction of "now" was more than the microsecond
6076 fraction of "then" (but the whole seconds number was less).
6078 38. Added support for the libradius Radius client library this is found on
6079 FreeBSD (previously only the radiusclient library was supported).
6085 1. When certain lookups returned multiple values in the form name=value, the
6086 quoting of the values was not always being done properly. Specifically:
6087 (a) If the value started with a double quote, but contained no whitespace,
6089 (b) If the value contained whitespace other than a space character (i.e.
6090 tabs or newlines or carriage returns) it was not quoted.
6091 This fix has been applied to the mysql and pgsql lookups by writing a
6092 separate quoting function and calling it from the lookup code. The fix
6093 should probably also be applied to nisplus, ibase and oracle lookups, but
6094 since I cannot test any of those, I have not disturbed their existing code.
6096 2. A hit in the callout cache for a specific address caused a log line with no
6097 reason for rejecting RCPT. Now it says "Previous (cached) callout
6098 verification failure".
6100 3. There was an off-by-one bug in the queryprogram router. An over-long
6101 return line was truncated at 256 instead of 255 characters, thereby
6102 overflowing its buffer with the terminating zero. As well as fixing this, I
6103 have increased the buffer size to 1024 (and made a note to document this).
6105 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6106 when Exim is waiting for an SMTP response from a remote server, Exim
6107 restarts its select() call on the socket, thereby resetting its timeout.
6108 This is not a problem when such interrupts are rare. Somebody set up a cron
6109 job to run exiwhat every 2 minutes, which is less than the normal select()
6110 timeout (5 or 10 minutes). This meant that the select() timeout never
6111 kicked in because it was always reset. I have fixed this by comparing the
6112 time when an interrupt arrives with the time at the start of the first call
6113 to select(). If more time than the timeout has elapsed, the interrupt is
6114 treated as a timeout.
6116 5. Some internal re-factoring in preparation for the addition of Sieve
6117 extensions (by MH). In particular, the "personal" test is moved to a
6118 separate function, and given an option for scanning Cc: and Bcc: (which is
6119 not set for Exim filters).
6121 6. When Exim created an email address using the login of the caller as the
6122 local part (e.g. when creating a From: or Sender: header line), it was not
6123 quoting the local part when it contained special characters such as @.
6125 7. Installed new OpenBSD configuration files.
6127 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6128 try to make them clearer.
6130 9. Callout options, other than the timeout value, were being ignored when
6131 verifying sender addresses in header lines. For example, when using
6133 verify = header_sender/callout=no_cache
6135 the cache was (incorrectly) being used.
6137 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6138 only the Cygwin environment.
6140 11. When return_path_on_delivery was set as a log selector, if different remote
6141 addresses in the same message used different return paths and parallel
6142 remote delivery occurred, the wrong values would sometimes be logged.
6143 (Whenever a remote delivery process finished, the return path value from
6144 the most recently started remote delivery process was logged.)
6146 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6147 lines when AUTH and/or TLS are in use. This is the "received protocol"
6148 field. Exim used to use "asmtp" for authenticated SMTP, without any
6149 indication (in the protocol name) for TLS use. Now it follows the RFC and
6150 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6151 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6152 names appear in log lines as well as in Received: header lines.
6154 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6155 extensions, and comparison tests, and to fix some bugs.
6157 14. Changes to the "personal" filter test:
6159 (1) The test was buggy in that it was just doing the equivalent of
6160 "contains" tests on header lines. For example, if a user's address was
6161 anne@some.where, the "personal" test would incorrectly be true for
6163 To: susanne@some.where
6165 This test is now done by extracting each address from the header in turn,
6166 and checking the entire address. Other tests that are part of "personal"
6167 are now done using regular expressions (for example, to check local parts
6168 of addresses in From: header lines).
6170 (2) The list of non-personal local parts in From: addresses has been
6171 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6172 taken from the Sieve specification recommendations.
6174 (3) If the message contains any header line starting with "List-" it is
6175 treated as non-personal.
6177 (4) The test for "circular" in the Subject: header line has been removed
6178 because it now seems ill-conceived.
6180 15. Minor typos in src/EDITME comments corrected.
6182 16. Installed latest exipick from John Jetmore.
6184 17. If headers_add on a router specified a text string that was too long for
6185 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6186 of string_sprintf() is now avoided.
6188 18. $message_body_size was not set (it was always zero) when running the DATA
6189 ACL and the local_scan() function.
6191 19. For the "mail" command in an Exim filter, no default was being set for
6192 the once_repeat time, causing a random time value to be used if "once" was
6193 specified. (If the value happened to be <= 0, no repeat happened.) The
6194 default is now 0s, meaning "never repeat". The "vacation" command was OK
6195 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6196 (I found it when inspecting the code).
6198 20. There is now an overall timeout for performing a callout verification. It
6199 defaults to 4 times the callout timeout, which applies to individual SMTP
6200 commands during the callout. The overall timeout applies when there is more
6201 than one host that can be tried. The timeout is checked before trying the
6202 next host. This prevents very long delays if there are a large number of
6203 hosts and all are timing out (e.g. when the network connections are timing
6204 out). The value of the overall timeout can be changed by specifying an
6205 additional sub-option for "callout", called "maxwait". For example:
6207 verify = sender/callout=5s,maxwait=20s
6209 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6210 to the end before writing, but this should make it even safer).
6212 22. Exim was forgetting that it had advertised PIPELINING for the second and
6213 subsequent messages on an SMTP connection. It was also not resetting its
6214 memory on STARTTLS and an internal HELO.
6216 23. When Exim logs an SMTP synchronization error within a session, it now
6217 records whether PIPELINING has been advertised or not.
6219 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6220 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6221 rather than long int.
6223 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6225 26. Added the never_mail option to autoreply.
6231 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6232 crash if the getsockname() call failed; this can happen if a connection is
6233 closed very soon after it is established. The problem was simply in the
6234 order in which certain operations were done, causing Exim to try to write
6235 to the SMTP stream before it had set up the file descriptor. The bug has
6236 been fixed by making things happen in the correct order.
6242 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6243 before the connection was closed, thus losing the rejection response.
6245 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6246 some early Solaris releases, but causes trouble in current releases where
6247 socklen_t is defined.
6249 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6252 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6255 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6256 that releasing the top of it at the end releases what was used for sub-
6257 expansions (unless the block got too big). However, discard this block if
6258 the first thing is a variable or header, so that we can use its block when
6259 it is dynamic (useful for very large $message_headers, for example).
6261 6. Lookups now cache *every* query, not just the most recent. A new, separate
6262 store pool is used for this. It can be recovered when all lookup caches are
6263 flushed. Lookups now release memory at the end of their result strings.
6264 This has involved some general refactoring of the lookup sources.
6266 7. Some code has been added to the store_xxx() functions to reduce the amount
6267 of flapping under certain conditions.
6269 8. log_incoming_interface used to affect only the <= reception log lines. Now
6270 it causes the local interface and port to be added to several more SMTP log
6271 lines, for example "SMTP connection from", and rejection lines.
6273 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6275 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6277 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6278 could be overwritten at the end of the current message (or the start of a
6279 new message if it was set in a HELO ACL). The value is now preserved for
6280 the duration of the SMTP connection.
6282 12. If a transport had a headers_rewrite setting, and a matching header line
6283 contained an unqualified address, that address was qualified, even if it
6284 did not match any rewriting rules. The underlying bug was that the values
6285 of the flags that permit the existence of unqualified sender and recipient
6286 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6287 non-local messages, and by -bnq for local messages) were not being
6288 preserved with the message after it was received.
6290 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6291 "next input=" as part of the text comprising the host identity instead of
6292 the correct text. The code was using the same buffer for two different
6293 strings. However, depending on which order the printing function evaluated
6294 its arguments, the bug did not always show up. Under Linux, for example, my
6295 test suite worked just fine.
6297 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6298 doesn't work with some older versions of Perl. It has been changed to "my",
6299 which in any case is probably the better facility to use.
6301 15. A really picky compiler found some instances of statements for creating
6302 error messages that either had too many or two few arguments for the format
6305 16. The size of the buffer for calls to the DNS resolver has been increased
6306 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6307 for addresses that have a lot of PTR records. This alleviates a problem; it
6308 does not fully solve it.
6310 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6311 buffer now truncates the list and logs the incident, which is the same
6312 action as happens when Exim is looking up a host name and its aliases.
6313 Previously in this situation something unpredictable would happen;
6314 sometimes it was "internal error: store_reset failed".
6316 18. If a server dropped the connection unexpectedly when an Exim client was
6317 using GnuTLS and trying to read a response, the client delivery process
6318 crashed while trying to generate an error log message.
6320 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6321 string, for example:
6323 warn message = H1: something\nH2: something
6325 the text was added as a single header line from Exim's point of view
6326 though it ended up OK in the delivered message. However, searching for the
6327 second and subsequent header lines using $h_h2: did not work. This has been
6328 fixed. Similarly, if a system filter added multiple headers in this way,
6329 the routers could not see them.
6331 20. Expanded the error message when iplsearch is called with an invalid key to
6332 suggest using net-iplsearch in a host list.
6334 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6335 ACLs are no longer actually imposed (and a message to that effect is
6338 22. If a "gecos" field in a passwd entry contained escaped characters, in
6339 particular, if it contained a \" sequence, Exim got it wrong when building
6340 a From: or a Sender: header from that name. A second bug also caused
6341 incorrect handling when an unquoted " was present following a character
6342 that needed quoting.
6344 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6345 was not being matched caselessly.
6347 24. Arranged for all hyphens in the exim.8 source to be escaped with
6350 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6351 properly. Recipient callout cache records were still being keyed to include
6352 the sender, even when use_sender was set false. This led to far more
6353 callouts that were necessary. The sender is no longer included in the key
6354 when use_sender is false.
6356 26. Added "control = submission" modifier to ACLs.
6358 27. Added the ${base62d: operator to decode base 62 numbers.
6360 28. dnsdb lookups can now access SRV records.
6362 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6363 the configuration file.
6365 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6366 (-v) mode. This makes the output for a verbose queue run more intelligible.
6368 31. Added a use_postmaster feature to recipient callouts.
6370 32. Added the $body_zerocount variable, containing the number of binary zero
6371 bytes in the message body.
6373 33. The time of last modification of the "new" subdirectory is now used as the
6374 "mailbox time last read" when there is a quota error for a maildir
6377 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6379 35. Added +ignore_unknown as a special item in host lists.
6381 36. Code for decoding IPv6 addresses in host lists is now included, even if
6382 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6383 address was recognized as an IP address, but was then not correctly decoded
6384 into binary, causing unexpected and incorrect effects when compared with
6391 1. Very minor rewording of debugging text in manualroute to say "list of
6392 hosts" instead of "hostlist".
6394 2. If verify=header_syntax was set, and a header line with an unqualified
6395 address (no domain) and a large number of spaces between the end of the
6396 name and the colon was received, the reception process suffered a buffer
6397 overflow, and (when I tested it) crashed. This was caused by some obsolete
6398 code that should have been removed. The fix is to remove it!
6400 3. When running in the test harness, delay a bit after writing a bounce
6401 message to get a bit more predictability in the log output.
6403 4. Added a call to search_tidyup() just before forking a reception process. In
6404 theory, someone could use a lookup in the expansion of smtp_accept_max_
6405 per_host which, without the tidyup, could leave open a database connection.
6407 5. Added the variables $recipient_data and $sender_data which get set from a
6408 lookup success in an ACL "recipients" or "senders" condition, or a router
6409 "senders" option, similar to $domain_data and $local_part_data.
6411 6. Moved the writing of debug_print from before to after the "senders" test
6414 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6415 problems for message scanning, either using a data ACL, or using
6416 local_scan() because the Received: header was not generated till after they
6417 were called (in order to set the time as the time of reception completion).
6418 I have revised the way this works. The header is now generated after the
6419 body is received, but before the ACL or local_scan() are called. After they
6420 are run, the timestamp in the header is updated.
6426 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6427 before starting a queue runner without re-exec. This happened only when
6428 deliver_drop_privilege was set or when the Exim user was set to root. The
6429 effect of the bug was that timeouts during subsequent deliveries caused
6430 crashes instead of being properly handled. The handler is now left at its
6431 default (and expected) setting.
6433 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6434 message, again when deliver_drop_privilege is set or Exim is run as root.
6435 The bug described in (1) was not present in this case, but the tidying up
6436 of the other signals was missing. I have made the two cases consistent.
6438 3. The ignore_target_hosts setting on a manualroute router was being ignored
6439 for hosts that were looked up using the /MX notation.
6441 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6444 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6445 operated on the sender address. After changing the $sender_address to <>
6446 for the sender address verify, Exim was re-instated it as the original
6447 (before rewriting) address, but remembering that it had rewritten it, so it
6448 wasn't rewriting it again. This bug also had the effect of breaking the
6449 sender address verification caching when the sender address was rewritten.
6451 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6452 This has been changed so that if the ip literal address matches
6453 ignore_target_hosts, the router declines.
6455 7. Added expansion conditions match_domain, match_address, and match_local_
6456 part (NOT match_host).
6458 8. The placeholder for the Received: header didn't have a length field set.
6460 9. Added code to Exim itself and to exim_lock to test for a specific race
6461 condition that could lead to file corruption when using MBX delivery. The
6462 issue is with the lockfile that is created in /tmp. If this file is removed
6463 after a process has opened it but before that process has acquired a lock,
6464 there is the potential for a second process to recreate the file and also
6465 acquire a lock. This could lead to two Exim processes writing to the file
6466 at the same time. The added code performs the same test as UW imapd; it
6467 checks after acquiring the lock that its file descriptor still refers to
6468 the same named file.
6470 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6471 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6474 11. Added the smtp_active_hostname option. If used, this will typically be made
6475 to depend on the incoming interface address. Because $interface_address is
6476 not set up until the daemon has forked a reception process, error responses
6477 that can happen earlier (such as "too many connections") no longer contain
6480 12. If an expansion in a condition on a "warn" statement fails because a lookup
6481 defers, the "warn" statement is abandoned, and the next ACL statement is
6482 processed. Previously this caused the whole ACL to be aborted.
6484 13. Added the iplsearch lookup type.
6486 14. Added ident_timeout as a log selector.
6488 15. Added tls_certificate_verified as a log selector.
6490 16. Added a global option tls_require_ciphers (compare the smtp transport
6491 option of the same name). This controls incoming TLS connections.
6493 17. I finally figured out how to make tls_require_ciphers do a similar thing
6494 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6495 before starting the TLS session.
6497 18. Tabs are now shown as \t in -bP output.
6499 19. If the log selector return_path_on_delivery was set, Exim crashed when
6500 bouncing a message because it had too many Received: header lines.
6502 20. If two routers both had headers_remove settings, and the first one included
6503 a superfluous trailing colon, the final name in the first list and the
6504 first name in the second list were incorrectly joined into one item (with a
6505 colon in the middle).
6511 1. Added -C and -D options to the exinext utility, mainly to make it easier
6512 to include in the automated testing, but these could be helpful when
6513 multiple configurations are in use.
6515 2. The exinext utility was not formatting the output nicely when there was
6516 an alternate port involved in the retry record key, nor when there was a
6517 message id as well (for retries that were specific to a specific message
6518 and a specific host). It was also confused by IPv6 addresses, because of
6519 the additional colons they contain. I have fixed the IPv4 problem, and
6520 patched it up to do a reasonable job for IPv6.
6522 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6523 delivery, the log line now contains "pipelined" if PIPELINING was used.
6525 4. An SMTP transport process used to panic and die if the bind() call to set
6526 an explicit outgoing interface failed. This has been changed; it is now
6527 treated in the same way as a connect() failure.
6529 5. A reference to $sender_host_name in the part of a conditional expansion
6530 that was being skipped was still causing a DNS lookup. This no longer
6533 6. The def: expansion condition was not recognizing references to header lines
6534 that used bh_ and bheader_.
6536 7. Added the _cache feature to named lists.
6538 8. The code for checking quota_filecount in the appendfile transport was
6539 allowing one more file than it should have been.
6541 9. For compatibility with Sendmail, the command line option
6549 and sets the incoming protocol and host name (for trusted callers). The
6550 host name and its colon can be omitted when only the protocol is to be set.
6551 Note the Exim already has two private options, -pd and -ps, that refer to
6552 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6553 "s", but I don't think that's a major issue.
6555 10. A number of refactoring changes to the code, none of which should affect
6558 (a) The number of logging options was getting close to filling up the
6559 32-bit word that was used as a bit map. I have split them into two classes:
6560 those that are passed in the argument to log_write(), and those that are
6561 only ever tested independently outside of that function. These are now in
6562 separate 32-bit words, so there is plenty of room for expansion again.
6563 There is no change in the user interface or the logging behaviour.
6565 (b) When building, for example, log lines, the code previously used a
6566 macro that called string_cat() twice, in order to add two strings. This is
6567 not really sufficiently general. Furthermore, there was one instance where
6568 it was actually wrong because one of the argument was used twice, and in
6569 one call a function was used. (As it happened, calling the function twice
6570 did not affect the overall behaviour.) The macro has been replaced by a
6571 function that can join an arbitrary number of extra strings onto a growing
6574 (c) The code for expansion conditions now uses a table and a binary chop
6575 instead of a serial search (which was left over from when there were very
6576 few conditions). Also, it now recognizes conditions like "pam" even when
6577 the relevant support is not compiled in: a suitably worded error message is
6578 given if an attempt is made to use such a condition.
6580 11. Added ${time_interval:xxxxx}.
6582 12. A bug was causing one of the ddress fields not to be passed back correctly
6583 from remote delivery subprocesses. The field in question was not being
6584 subsequently used, so this caused to problems in practice.
6586 13. Added new log selectors queue_time and deliver_time.
6588 14. Might have fixed a bug in maildirsizefile handling that threw up
6589 "unexpected character" debug warnings, and recalculated the data
6590 unnecessarily. In any case, I expanded the warning message to give more
6593 15. Added the message "Restricted characters in address" to the statements in
6594 the default ACL that block characters like @ and % in local parts.
6596 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6597 Three changes have been made:
6599 (a) There was a serious bug; a negative response to MAIL caused the whole
6600 recipient domain to be cached as invalid, thereby blocking all messages
6601 to all local parts at the same domain, from all senders. This bug has
6602 been fixed. The domain is no longer cached after a negative response to
6603 MAIL if the sender used is not empty.
6605 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6608 (c) A new callout option, "use_sender" has been added for people who want
6609 the modified behaviour.
6615 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6618 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6619 indeed breaks things for older releases.
6621 3. Added additional logging to the case where there is a problem reading data
6622 from a filter that is running in a subprocess using a pipe, in order to
6623 try to track down a specific problem.
6625 4. Testing facility fudge: when running in the test harness and attempting
6626 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6627 getting "No route to host". Convert this to a timeout.
6629 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6632 6. Some OS don't have socklen_t but use size_t instead. This affects the
6633 fifth argument of getsockopt() amongst other things. This is now
6634 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6635 can be set for individual OS. I have set it for SunOS5, OSF1, and
6636 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6637 some earlier ones do not.
6639 7. Change 4.30/15 was not doing the test caselessly.
6641 8. The standard form for an IPv6 address literal was being rejected by address
6642 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6643 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6644 this, as well as the form without the "IPv6" on the front (but only when
6645 address literals are enabled, of course).
6647 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6649 10. Exim crashed if a message with an empty sender address specified by -f
6650 encountered a router with an errors_to setting. This could be provoked only
6651 by a command such as
6655 where an empty string was supplied; "<>" did not hit this bug.
6657 11. Installed PCRE release 4.5.
6659 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6660 remained set. It is now erased.
6662 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6663 times from message ids (which are base 36 rather than the normal 62).
6665 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6666 were being counted as actual protocol errors, and logged if the log
6667 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6668 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6669 and DATA following a set of rejected RCPTs do not count as protocol errors.
6670 In other words, Exim assumes they were pipelined, though this may not
6671 actually be the case. Of course, in all cases the client gets an
6672 appropriate error code.
6674 15. If a lookup fails in an ACL condition, a message about the failure may
6675 be available; it is used if testing the ACL cannot continue, because most
6676 such messages specify what the cause of the deferral is. However, some
6677 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6678 that caused an old message to be retained and used if a later statement
6679 caused a defer, replacing the real cause of the deferral.
6681 16. If an IP address had so many PTR records that the DNS lookup buffer
6682 was not large enough to hold them, Exim could crash while trying to process
6683 the truncated data. It now detects and logs this case.
6685 17. Further to 4.21/58, another change has been made: if (and only if) the
6686 first line of a message (the first header line) ends with CRLF, a bare LF
6687 in a subsequent header line has a space inserted after it, so as not to
6688 terminate the header.
6690 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6691 unnecessarily, used atoi() instead of strtol(), and didn't check the
6692 termination when getting file sizes from file names by regex.
6694 19. Completely re-implemented the support for maildirsize files, in the light
6695 of a number of problems with the previous contributed implementation
6696 (4.30/29). In particular:
6698 . If the quota is zero, the maildirsize file is maintained, but no quota is
6701 . If the maildir directory does not exist, it is created before any attempt
6702 to write a maildirsize file.
6704 . The quota value in the file is just a cache; if the quota is changed in
6705 the transport, the new value overrides.
6707 . A regular expression is available for excluding directories from the
6710 20. The autoreply transport checks the characters in options that define the
6711 message's headers; it allows continued headers, but it was checking with
6712 isspace() after an embedded newline instead of explicitly looking for a
6715 21. If all the "regular" hosts to which an address was routed had passed their
6716 expiry times, and had not reached their retry times, the address was
6717 bounced, even if fallback hosts were defined. Now Exim should go on to try
6720 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6721 equivalent code in the SMTP transport. Some hosts send humungous responses
6722 to HELO/EHLO, more than 1024 it seems.
6724 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6725 gives compiler warnings in some environments. I've now done it "properly",
6728 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6729 (because of problems with the built-in one) was declared to return uschar *
6730 instead of char *, causing compiler failure.
6732 25. Fixed a file descriptor leak when processing alias/forward files.
6734 26. Fixed a minor format string issue in dbfn.c.
6736 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6738 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6739 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6742 29. When a sender address is verified, it is cached, to save repeating the test
6743 when there is more than one recipient in a message. However, when the
6744 verification involves a callout, it is possible for different callout
6745 options to be set for different recipients. It is too complicated to keep
6746 track of this in the cache, so now Exim always runs a verification when a
6747 callout is required, relying on the callout cache for the optimization.
6748 The overhead is duplication of the address routing, but this should not be
6751 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6752 to be verified with callout=postmaster, and the main callout worked but the
6753 postmaster check failed, the verification correctly failed. However, if a
6754 subsequent RCPT command asked for sender verification *without* the
6755 postmaster check, incorrect caching caused this verification also to fail,
6758 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6759 it was not caching the DNS options (qualify_single, search_parents) that
6760 were used when the lookup failed. A subsequent lookup with different
6761 options therefore always gave the same answer, though there were cases
6762 where it should not have. (Example: a "domains = !$mx_any" option on a
6763 dnslookup router: the "domains" option is always processed without any
6764 widening, but the router might have qualify_single set.) Now Exim uses the
6765 cached value only when the same options are set.
6767 32. Added John Jetmore's "exipick" utility to the distribution.
6769 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6770 than a timeout (e.g. a certificate is required, and is not provided), an
6771 Exim server now closes the connection immediately. Previously it waited for
6772 the client to close - but if the client is SSL, it seems that they each
6773 wait for each other, leading to a delay before one of them times out.
6775 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6776 maintained 0.8.x compatibility because I don't think many are using it, and
6777 it is clearly obsolete.
6779 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6782 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6783 client certificate was expired. A simple patch fixes this, though I don't
6784 understand the full logic of why the verify callback is called multiple
6787 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6788 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6789 which causes problems with some clients (such as the Certicom SSL Plus
6790 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6791 disables the coutermeasure allowing Eudora to connect."
6793 38. Exim was not checking that a write() to a log file succeeded. This could
6794 lead to Bad Things if a log got too big, in particular if it hit a file
6795 size limit. Exim now panics and dies if it cannot write to a log file, just
6796 as it does if it cannot open a log file.
6798 39. Modified OS/Makefile-Linux so that it now contains
6800 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6802 The two -D definitions ensure that Exim is compiled with large file
6803 support, which makes it possible to handle log files that are bigger than
6806 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6807 instance) a domain was checked against a named list that involved a lookup,
6808 causing $domain_data to be set, then another domain was checked against the
6809 same list, then the first domain was re-checked, the value of $domain_data
6810 after the final check could be wrong. In particular, if the second check
6811 failed, it could be set empty. This bug probably also applied to
6814 41. The strip_trailing_dot option was not being applied to the address given
6815 with the -f command-line option.
6817 42. The code for reading a message's header from the spool was incrementing
6818 $received_count, but never initializing it. This meant that the value was
6819 incorrect (doubled) while delivering a message in the same process in which
6820 it was received. In the most common configuration of Exim, this never
6821 happens - a fresh exec is done - but it can happen when
6822 deliver_drop_privilege is set.
6824 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6825 it now includes up to 150 characters of the unexpected data in the log
6828 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6829 and building data strings. The size of both of these buffers was 10 000
6830 bytes - far larger than anybody would *ever* want, thought I. Needless to
6831 say, somebody hit the limit. I have increased the maximum line length to
6832 20 000 and the maximum data length of concatenated lines to 100 000. I have
6833 also fixed two bugs, because there was no checking on these buffers. Tsk,
6834 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6835 buffer is too small.
6837 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6838 lsearch lookups. Now it does.
6840 46. When parsing a route_list item in a manualroute router, a fixed-length
6841 buffer was used for the list of hosts. I made this 1024 bytes long,
6842 thinking that nobody would ever have a list of hosts that long. Wrong.
6843 Somebody had a whole pile of complicated expansion conditions, and the
6844 string was silently truncated, leading to an expansion error. It turns out
6845 that it is easier to change to an unlimited length (owing to other changes
6846 that have happened since this code was originally written) than to build
6847 structure for giving a limitation error. The length of the item that
6848 expands into the list of hosts is now unlimited.
6850 47. The lsearch lookup could not handle data where the length of text line was
6851 more than 4095 characters. Such lines were truncated, leading to shortened
6852 data being returned. It should now handle lines of any length.
6854 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6855 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6858 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6859 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6860 no longer happens in comment lines. A list of replacements is now placed
6861 at the head of all of the source files, except those whose only change is
6862 to replace PERL_COMMAND in the very first #! line.
6864 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6865 messages on the queue, with a bottom-up merge sort, using code contributed
6866 by Michael Haardt. This should make operations like -bp somewhat faster on
6867 large queues. It won't affect queue runners, except when queue_run_in_order
6870 51. Installed eximstats 1.31 in the distribution.
6872 52. Added support for SRV lookups to the dnslookup router.
6874 53. If an ACL referred to $message_body or $message_body_end, the value was not
6875 reset for any messages that followed in the same SMTP session.
6877 54. The store-handling optimization for building very long strings was not
6878 differentiating between the different store pools. I don't think this
6879 actually made any difference in practice, but I've tidied it.
6881 55. While running the routers to verify a sender address, $sender_address
6882 was still set to the sender address. This is wrong, because when routing to
6883 send a bounce to the sender, it would be empty. Therefore, I have changed
6884 it so that, while verifying a sender address, $sender_address is set to <>.
6885 (There is no change to what happens when verifying a recipient address.)
6887 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6888 target A or AAAA records (if not already returned) without resetting the
6889 qualify_single or search_parents options of the DNS resolver. These are
6890 inappropriate in this case because the targets of MX and SRV records must
6891 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6892 target that, when qualified, matched something in the local domain. These
6893 two options are now turned off when doing these lookups.
6895 57. It seems that at least some releases of Reiserfs (which does not have the
6896 concept of a fixed number of inodes) returns zero and not -1 for the
6897 number of available inodes. This interacted badly with check_spool_inodes,
6898 which assumed that -1 was the "no such thing" setting. What I have done is
6899 to check that the total number of inodes is greater than zero before doing
6900 the test of how many are available.
6902 58. When a "warn" ACL statement has a log_message modifier, the message is
6903 remembered, and not repeated. This is to avoid a lot of repetition when a
6904 message has many recipients that cause the same warning to be written.
6905 However, Exim was preserving the list of already written lines for an
6906 entire SMTP session, which doesn't seem right. The memory is now reset if a
6907 new message is started.
6909 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6910 debugging output unless log_rewrite was also set.
6912 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6913 of (int)(handle) when we know that handle contains (void *)(-1).
6915 61. The Exim daemon panic-logs an error return when it closes the incoming
6916 connection. However "connection reset by peer" seems to be common, and
6917 isn't really an error worthy of noting specially, so that particular error
6920 62. When Exim is trying to find all the local interfaces, it used to panic and
6921 die if the ioctl to get the interface flags failed. However, it seems that
6922 on at least one OS (Solaris 9) it is possible to have an interface that is
6923 included in the list of interfaces, but for which you get a failure error
6924 for this call. This happens when the interface is not "plumbed" into a
6925 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6926 failure of the "get flags" call assumes that the interface is down.
6928 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6929 makes life easier for people who are doing arithmetic on fields extracted
6930 from dates, where you often get leading zeros that should not be
6931 interpreted as octal.
6933 64. Added qualify_domain to the redirect router, to override the global
6936 65. If a pathologically long header line contained very many addresses (the
6937 report of this problem mentioned 10 000) and each of them was rewritten,
6938 Exim could use up a very large amount of memory. (It kept on making new
6939 copies of the header line as it rewrote, and never released the old ones.)
6940 At the expense of a bit more processing, the header rewriting function has
6941 been changed so that it no longer eats memory in this way.
6943 66. The generation of the Received: header has been moved from the time that a
6944 message starts to be received, to the time that it finishes. The timestamp
6945 in the Received: header should now be very close to that of the <= log
6946 line. There are two side-effects of this change:
6948 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6949 logged header lines no longer include the local Received: line, because
6950 it has not yet been created. The same applies to a copy of the message
6951 that is returned to a non-SMTP sender when a message is rejected.
6953 (b) When a filter file is tested using -bf, no additional Received: header
6954 is added to the test message. After some thought, I decided that this
6957 This change does not affect the value of $received_for. It is still set
6958 after address rewriting, but before local_scan() is called.
6960 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6962 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6963 gave an unhelpful panic error message, and a defer error. I have managed to
6964 change this behaviour so that it now rejects any supplied certificate,
6965 which seems right, as the list of acceptable certificates is empty.
6967 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6968 gave an unhelpful defer error. I have not managed to make this reject any
6969 supplied certificates, but the error message it gives is "no certificate
6970 supplied", which is not helpful.
6972 70. exigrep's output now also includes lines that are not associated with any
6973 message, but which match the given pattern. Implemented by a patch from
6974 Martin Sluka, which also tidied up the Perl a bit.
6976 71. Recipient callout verification, like sender verification, was using <> in
6977 the MAIL FROM command. This isn't really the right thing, since the actual
6978 sender may affect whether the remote host accepts the recipient or not. I
6979 have changed it to use the actual sender in the callout; this means that
6980 the cache record is now keyed on a recipient/sender pair, not just the
6981 recipient address. There doesn't seem to be a real danger of callout loops,
6982 since a callout by the remote host to check the sender would use <>.
6983 [SEE ABOVE: changed after hitting problems.]
6985 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6986 temporary errors. However, in the case of such a code being given after
6987 the end of a data transmission (i.e. after ".") Exim was failing to write
6988 a retry record for the message. (Yes, there was some broken host that was
6989 actually sending 8xx at this point.)
6991 73. An unknown lookup type in a host list could cause Exim to panic-die when
6992 the list was checked. (An example that provoked this was putting <; in the
6993 middle of a list instead of at the start.) If this happened during a DATA
6994 ACL check, a -D file could be left lying around. This kind of configuration
6995 error no longer causes Exim to die; instead it causes a defer error. The
6996 incident is still logged to the main and panic logs.
6998 74. Buglet left over from Exim 3 conversion. The message "too many messages
6999 in one connection" was written to the rejectlog but not the mainlog, except
7000 when address rewriting (yes!) was being logged.
7002 75. Added write_rejectlog option.
7004 76. When a system filter was run not as root (that is, when system_filter_user
7005 was set), the values of the $n variables were not being returned to the
7006 main process; thus, they were not subsequently available in the $sn
7009 77. Added +return_path_on_delivery log selector.
7011 78. A connection timeout was being treated differently from recipients deferred
7012 when testing hosts_max_try with a message that was older than the host's
7013 retry timeout. (The host should not be counted, thus allowing all hosts to
7014 be tried at least once before bouncing.) This may have been the cause of an
7015 occasionally reported bug whereby a message would remain on the queue
7016 longer than the retry timeout, but would be bounced if a delivery was
7017 forced. I say "may" because I never totally pinned down the problem;
7018 setting up timeout/retry tests is difficult. See also the next item.
7020 79. The ultimate address timeout was not being applied to errors that involved
7021 a combination of host plus message (for example, a timeout on a MAIL
7022 command). When an address resolved to a number of possible hosts, and they
7023 were not all tried for each delivery (e.g. because of hosts_max_try), a
7024 message could remain on the queue longer than the retry timeout.
7026 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7029 81. Fixed an obscure SMTP outgoing bug which required at least the following
7030 conditions: (a) there was another message waiting for the same server;
7031 (b) the server returned 5xx to all RCPT commands in the first message so
7032 that the message was not completed; (c) the server dropped the connection
7033 or gave a negative response to the RSET that Exim sends to abort the
7034 transaction. The observed case was a dropped connection after DATA that had
7035 been sent in pipelining mode. That is, the server had advertised PIPELINING
7036 but was not implementing it correctly. The effect of the bug was incorrect
7037 behaviour, such as trying another host, and this could lead to a crash.
7043 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7044 and daemon.c were passed as pointers to ints; they should have been
7045 pointers to socklen_t variables (which are typically unsigned ints).
7047 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7050 3. Fixed a really odd bug that affected only the testing scheme; patching a
7051 certain fixed string in the binary changed the value of another string that
7052 happened to be identical to the end of the original first string.
7054 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7055 name", it returns that address as the IP address. On some operating
7056 systems (e.g. Solaris), it also passes back the IP address string as the
7057 "host name". However, on others (e.g. Linux), it passes back an empty
7058 string. Exim wasn't checking for this, and was changing the host name to an
7059 empty string, assuming it had been canonicalized.
7061 5. Although rare, it is permitted to have more than one PTR record for a given
7062 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7063 all the names associated with an address, because they do in Solaris.
7064 However, it seems that they do not in Linux for data that comes from the
7065 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7066 I found this out when I moved to a new Linux workstation and tried to run
7067 the Exim test suite.
7069 To get round this problem I have changed the code so that it now does its
7070 own call to the DNS to look up PTR records when searching for a host name.
7071 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7072 addresses that are only in /etc/hosts are still found.
7074 This behaviour is, however, controlled by an option called host_lookup_
7075 order, which defaults to "bydns:byaddr". If people want to use the other
7076 order, or indeed, just use one or the other means of lookup, they can
7077 specify it in this variable.
7079 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7080 some operating systems, this comes back from gethostbyaddr() as an empty
7081 string, and this is what Exim used to test for. However, it seems that in
7082 other systems, "." is yielded. Exim now tests for this case too.
7084 7. The values of check_spool_space and check_log_space are now held internally
7085 as a number of kilobytes instead of an absolute number of bytes. If a
7086 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7087 kilobyte. This means that much larger values can be stored.
7089 8. Exim monitor: an attempt to get the action menu when not actually pointing
7090 at a message produces an empty menu entitled "No message selected". This
7091 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7092 no entries in it ("Shell widget menu has zero width and/or height"). So I
7093 have added a single, blank menu entry in this case.
7095 9. Added ${quote_local_part.
7097 10. MIME decoding is now applied to the contents of Subject: header lines when
7100 11. Now that a reference to $sender_host_address automatically causes a reverse
7101 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7102 host lookup before query-style lookups in lists that might use this
7103 variable. This has therefore been abolished, and the "net-" prefix is no
7104 longer necessary for query-style lookups.
7106 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7107 have been a typo for LFLAGS, so it has been changed.
7109 13. The install script calls Exim with "-C /dev/null" in order to find the
7110 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7111 to be output. However, since Exim outputs its version number before the
7112 error, it didn't break the script. It just looked ugly. I fixed this by
7113 always allowing "-C /dev/null" if the caller is root.
7115 14. Ignore overlarge ACL variable number when reading spool file - insurance
7116 against a later release with more variables having written the file.
7118 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7119 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7120 this, as well as the form without the "IPv6" on the front.
7122 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7123 OS/Makefile-Darwin file.
7125 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7126 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7128 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7129 variable. However, some OS use ints for inodes. I've added cast to long int
7130 to get rid of the compiler warning.
7132 19. I had forgotten to lock out "/../" in configuration file names when
7133 ALT_CONFIG_PREFIX was set.
7135 20. Routers used for verification do not need to specify transports. However,
7136 if such a router generated a host list, and callout was configured, Exim
7137 crashed, because it could not find a port number from the (non-existent)
7138 transport. It now assumes port 25 in this circumstance.
7140 21. Added the -t option to exigrep.
7142 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7143 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7144 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7145 them is set, LOOKUP_LSEARCH is forced.
7147 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7149 24. Added sender and host information to the "rejected by local_scan()" log
7150 line; previously there was no indication of these.
7152 25. Added .include_if_exists.
7154 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7155 Linux. It turns out that not all file systems support this. Apparently some
7156 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7157 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7158 not supported on the file descriptor, is now ignored when Exim is trying to
7159 sync a directory. This applies only to Linux.
7161 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7163 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7164 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7167 29. Added support for maildirsize files from supplied patch (modified a bit).
7169 30. The use of :fail: followed by an empty string could lead Exim to respond to
7170 sender verification failures with (e.g.):
7172 550 Verification failed for <xxx>
7173 550 Sender verify failed
7175 where the first response line was missing the '-' that indicates it is not
7176 the final line of the response.
7178 31. The loop for finding the name of the user that called Exim had a hardwired
7179 limit of 10; it now uses the value of finduser_retries, which is used for
7180 all other user lookups.
7182 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7185 33. Exim was neglecting to zero errno before one call of strtol() when
7186 expanding a string and expecting an integer value. On some systems this
7187 resulted in spurious "integer overflow" errors. Also, it was casting the
7188 result into an int without checking.
7190 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7191 did not work. The code looks as if it has *never* worked, though it appears
7192 to have been documented since at least release 1.62. I have made it work.
7194 35. The "timeout_DNS" error in retry rules, also documented since at least
7195 1.62, also never worked. As it isn't clear exactly what this means, and
7196 clearly it isn't a major issue, I have abolished the feature by treating it
7197 as "timeout", and writing a warning to the main and panic logs.
7199 36. The display of retry rules for -brt wasn't always showing the error code
7202 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7203 timeout_connect_A, timeout_connect_MX.
7205 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7206 to the empty sender.
7208 39. The daemon was not analysing the content of -oX till after it had closed
7209 stderr and disconnected from the controlling terminal. This meant that any
7210 syntax errors were only noted on the panic log, and the return code from
7211 the command was 0. By re-arranging the code a little, I've made the
7212 decoding happen first, so such errors now appear on stderr, and the return
7213 code is 1. However, the actual setting up of the sockets still happens in
7214 the disconnected process, so errors there are still only recorded on the
7217 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7218 connections (as happens on some IP stacks) was logged at start up time as
7219 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7220 it from "IPv6 and IPv4", which means that two separate sockets are being
7223 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7224 says whether AF_INET or AF_INET6 was passed as an argument.
7226 42. Exiwhat output was messed up when time zones were included in log
7229 43. Exiwhat now gives more information about the daemon's listening ports,
7230 and whether -tls-on-connect was used.
7232 44. The "port" option of the smtp transport is now expanded.
7234 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7235 silently ignored. Now an error message is written to the main and panic
7238 46. There's a new ACL modifier called "logwrite" which writes to a log file
7239 as soon as it is encountered.
7241 47. Added $local_user_uid and $local_user_gid at routing time.
7243 48. Exim crashed when trying to verify a sender address that was being
7246 49. Exim was recognizing only a space character after ".include". It now also
7247 recognizes a tab character.
7249 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7250 extracting the relevant information from the specification. The man page no
7251 longer contains scrambled data for the -d option, and I've added a section
7252 at the front about calling Exim under different names.
7254 51. Added "extra_headers" argument to the "mail" command in filter files.
7256 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7259 53. Installed eximstats 1.29.
7261 54. Added transport_filter_timeout as a generic transport option.
7263 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7264 Cc: header lines. This was required by RFC 822, but it not required by RFC
7267 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7268 incoming messages that did not have them. Now it does so only if the
7269 message originates locally, that is, if there is no associated remote host
7270 address. When Resent- header lines are present, this applies to the Resent-
7271 lines rather than the non-Resent- lines.
7273 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7274 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7276 58. Messages for configuration errors now include the name of the main
7277 configuration file - useful now that there may be more than one file in a
7278 list (.included file names were always shown).
7280 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7281 for those rare installations that do not start the daemon as root or run it
7282 setuid root. I've cut out the call to initgroups() if the daemon is not
7285 60. The Exim user and group can now be bound into the binary as text strings
7286 that are looked up at the start of Exim's processing.
7288 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7290 62. Added $mailstore_basename variable.
7292 63. Installed patch to sieve.c from Michael Haardt.
7294 64. When Exim failed to open the panic log after failing to open the main log,
7295 the original message it was trying to log was written to stderr and debug
7296 output, but if they were not available (the usual case in production), it
7297 was lost. Now it is written to syslog before the two lines that record the
7298 failures to open the logs.
7300 65. Users' Exim filters run in subprocesses under the user's uid. It is
7301 possible for a "deliver" command or an alias in a "personal" command to
7302 provoke an address rewrite. If logging of address rewriting is configured,
7303 this fails because the process is not running as root or exim. There may be
7304 a better way of dealing with this, but for the moment (because 4.30 needs
7305 to be released), I have disabled address rewrite logging when running a
7306 filter in a non-root, non-exim process.
7312 1. The buildconfig auxiliary program wasn't quoting the value set for
7313 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7314 not defined. This bug was masked in 4.22 by the effect that was fixed in
7317 2. Some messages that were rejected after a message id was allocated were
7318 shown as "incomplete" by exigrep. It no longer does this for messages that
7319 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7321 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7322 have allow_domain_literals set, the ID did not get logged in the <= line.
7323 Domain literals are now always recognized in Message-ID: header lines.
7325 4. The first argument for a ${extract expansion item is the key name or field
7326 number. Leading and trailing spaces in this item were not being ignored,
7327 causing some misleading effects.
7329 5. When deliver_drop_privilege was set, single queue runner processes started
7330 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7331 same command in the process it spins off) were not dropping privilege.
7333 6. When the daemon running as "exim" started a queue runner, it always
7334 re-executed Exim in the spun-off process. This is a waste of effort when
7335 deliver_drop_privilege is set. The new process now just calls the
7336 queue-runner function directly.
7342 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7345 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7346 ignored. Though the use of -oP was forcing the writing of a pid file, it
7347 was always written to the default place.
7349 3. If the message "no IP address found for host xxxx" is generated during
7350 incoming verification, it is now followed by identification of the incoming
7351 connection (so you can more easily find what provoked it).
7353 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7355 5. Added some features to "harden" Exim a bit more against certain attacks:
7357 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7358 be put in Local/Makefile. This is like the never_users runtime option,
7359 but it cannot be overridden. The default setting is "root".
7361 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7362 prefix string with which any file named in a -C command line option
7365 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7366 is retained for -C and -D only if the caller of Exim is root. Without
7367 it, the exim user may also use -C and -D and retain privilege.
7369 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7370 command line option is disabled.
7372 6. Macro names set by the -D option must start with an upper case letter, just
7373 like macro names defined in the configuration file.
7375 7. Added "dereference=" facility to LDAP.
7377 8. Two instances of the typo "uknown" in the source files are fixed.
7379 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7380 the Configure-Makefile script screwed up while processing it.
7382 10. Incorporated PCRE 4.4.
7384 11. The SMTP synchronization check was not operating right at the start of an
7385 SMTP session. For example, it could not catch a HELO sent before the client
7386 waited for the greeting. There is now a check for outstanding input at the
7387 point when the greeting is written. Because of the duplex, asynchronous
7388 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7389 way, but not yet received, when the check is performed.
7391 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7392 on TCP/IP sockets, because this apparently causes some broken clients to
7395 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7396 unchanged) from the Cygwin maintainer.
7398 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7399 was supported instead of testing for mbx. Effectively a typo.
7401 15. The spa authenticator server code was not checking that the input it
7402 received was valid base64.
7404 16. The debug output line for the "set" modifier in ACLs was not showing the
7405 name of the variable that was being set.
7407 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7409 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7410 lookup on its own. Something else was needed to trigger the lookup. For
7411 example, a match in host_lookup or the need for a host name in a host list.
7412 Now, if $sender_host_name is referenced and the host name has not yet been
7413 looked up, a lookup is performed. If the lookup fails, the variable remains
7414 empty, and $host_lookup_failed is set to "1".
7416 19. Added "eqi" as a case-independent comparison operator.
7418 20. The saslauthd authentication condition could segfault if neither service
7419 nor realm was specified.
7421 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7422 error message that was logged was misleading, and incoming SMTP
7423 connections were dropped. The message is now more accurate, and temporary
7424 errors are given to SMTP connections.
7426 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7427 (except RSET) with a 421 error, until QUIT is received. However, it was
7428 failing to send a response to QUIT.
7430 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7431 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7432 ACL is now run after the helo_try_verify_hosts code.
7434 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7435 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7436 case-independent, so other case variants are also recognized). Apparently
7437 some systems use these upper case variants.
7439 25. If more than two messages were waiting for the same host, and a transport
7440 filter was specified for the transport, Exim sent two messages over the
7441 same TCP/IP connection, and then failed with "socket operation on non-
7442 socket" when it tried to send the third.
7444 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7446 27. The extern definition of crypt16() in expand.c was not being excluded when
7447 the OS had its own crypt16() function.
7449 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7450 as a preferred synonym for return_size_limit, both as an option and as an
7453 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7455 30. Changed the default configuration ACL to relax the local part checking rule
7456 for addresses that are not in any local domains. For these addresses,
7457 slashes and pipe symbols are allowed within local parts, but the sequence
7458 /../ is explicitly forbidden.
7460 31. SPA server authentication was not clearing the challenge buffer before
7463 32. log_message in a "warn" ACL statement was writing to the reject log as
7464 well as to the main log, which contradicts the documentation and doesn't
7465 seem right (because no rejection is happening). So I have stopped it.
7467 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7468 However, I am unable to do any testing of this.
7470 34. Fixed an infelicity in the appendfile transport. When checking directories
7471 for a mailbox, to see if any needed to be created, it was accidentally
7472 using path names with one or more superfluous leading slashes; tracing
7473 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7475 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7476 contents are added to the log line that is written for every discarded
7477 recipient. (Previously a log_message setting was ignored.)
7479 36. The ${quote: operator now quotes the string if it is empty.
7481 37. The install script runs exim in order to find its version number. If for
7482 some reason other than non-existence or emptiness, which it checks, it
7483 could not run './exim', it was installing it with an empty version number,
7484 i.e. as "exim-". This error state is now caught, and the installation is
7487 38. An argument was missing from the function that creates an error message
7488 when Exim fails to connect to the socket for saslauthd authentication.
7489 This could cause Exim to crash, or give a corrupted message.
7491 39. Added isip, isip4, and isip6 to ${if conditions.
7493 40. The ACL variables $acl_xx are now saved with the message, and can be
7494 accessed later in routers, transports, and filters.
7496 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7497 strings in the file are not string-expanded.
7499 42. If a MAIL command specified a SIZE value that was too large to fit into an
7500 int variable, the check against message_size_limit failed. Such values are
7501 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7502 day this will have to be increased, but I don't think I want to be around
7503 when emails are that large.
7510 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7511 iconv() is not standard in FreeBSD.
7513 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7514 IPv6 enabled. The observed symptom was a segmentation fault on return from
7515 the function os_common_find_running_interfaces() in src/os.c.
7517 3. In the check_special_case() function in daemon.c I had used "errno" as an
7518 argument name, which causes warnings on some systems. This was basically a
7519 typo, since it was named "eno" in the comments!
7521 4. The code that waits for the clock to tick (at a resolution of some fraction
7522 of a second) so as to ensure message-id uniqueness was always waiting for
7523 at least one whole tick, when it could have waited for less. [This is
7524 almost certainly not relevant at current processor speeds, where it is
7525 unlikely to ever wait at all. But we try to future-proof.]
7527 5. The function that sleeps for a time interval that includes fractions of a
7528 second contained a race. It did not block SIGALRM between setting the
7529 timer, and suspending (a couple of lines later). If the interval was short
7530 and the sigsuspend() was delayed until after it had expired, the suspension
7531 never ended. On busy systems this could lead to processes getting stuck for
7534 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7535 process, before it forks any delivery processes. The open lookup caching
7536 mechanism meant that the open file or database connection was passed into
7537 the delivery process. The problem was that delivery processes always tidy
7538 up cached lookup data. This could cause a problem for the next delivery
7539 process started by the queue runner, because the external queue runner
7540 process does not know about the closure. So the next delivery process
7541 still has data in the lookup cache. In the case of a file lookup, there was
7542 no problem because closing a file descriptor in a subprocess doesn't affect
7543 the parent. However, if the lookup was caching a connection to a database,
7544 the connection was closed, and the second delivery process was likely to
7545 see errors such as "PGSQL: query failed: server closed the connection
7546 unexpectedly". The problem has been fixed by closing all cached lookups
7547 in a queue runner before running a delivery process.
7549 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7550 seem to have the "const" qualifier which it has on other OS. I've
7553 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7554 *of the same type* (client or server) with the same public name that an
7555 error should be diagnosed.
7557 9. When Exim looked up a host name for an IP address, but failed to find the
7558 original IP address when looking up the host name (a safety check), it
7559 output the message "<ip address> does not match any IP for NULL", which was
7560 confusing, to say the least. The bug was that the host name should have
7561 appeared instead of "NULL".
7563 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7564 user that is built into the binary, and the -C or -D options is used, root
7565 privilege is dropped before the configuration file is read. In addition,
7566 logging is switched to stderr instead of the normal log files. If the
7567 configuration then re-defines the Exim user, the unprivileged environment
7568 is probably not what is expected, so Exim logs a panic warning message (but
7571 However, if deliver_drop_privilege is set, the unprivileged state may well
7572 be exactly what is intended, so the warning has been cut out in that case,
7573 and Exim is allowed to try to write to its normal log files.
7579 1. smtp_return_error_details was not giving details for temporary sender
7580 or receiver verification errors.
7582 2. Diagnose a configuration error if two authenticators have the same public
7585 3. Exim used not to create the message log file for a message until the first
7586 delivery attempt. This could be confusing when incoming messages were held
7587 for policy or load reasons. The message log file is now created at the time
7588 the message is received, and an initial "Received" line is written to it.
7590 4. The automatically generated man page for command line options had a minor
7591 bug that caused no ill effects; however, a more serious problem was that
7592 the procedure for building the man page automatically didn't always
7593 operate. Consequently, release 4.20 contains an out-of-date version. This
7594 shouldn't happen again.
7596 5. When building Exim with embedded Perl support, the script that builds the
7597 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7598 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7600 6. The freeze_tell option was not being used for messages that were frozen on
7601 arrival, either by an ACL or by local_scan().
7603 7. Added the smtp_incomplete_transaction log selector.
7605 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7606 was accepting AUTH without a new EHLO.
7608 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7609 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7612 10. Logging of TCP/IP connections (when configured) now happens in the main
7613 daemon process instead of the child process, so that the TCP/IP connection
7614 count is more accurate (but it can never be perfect).
7616 11. The use of "drop" in a nested ACL was not being handled correctly in the
7617 outer ACL. Now, if condition failure induced by the nested "drop" causes
7618 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7619 or "require"), the connection is dropped.
7621 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7622 that yield "discard" can now be used with an "accept" or a "discard" verb,
7623 but an error is generated for any others (because I can't see a useful way
7624 to define what should happen).
7626 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7627 are now processed in the same way as lines in the Exim configuration file.
7628 In particular, continuation lines are supported.
7630 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7632 15. Added -ti meaning -t -i.
7634 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7635 domains, and warn by logging if others are found.
7637 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7638 returned from GIFCONF when Exim is trying to find the list of interfaces on
7639 a host. The code in os.c has been modified to copy each ifreq to an aligned
7640 structure in all cases.
7642 Also, in some cases, the returned ifreq's were being copied to a 'struct
7643 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7644 means the last couple of bytes of an IPv6 address could be chopped if the
7645 ifreq contained only a normal sockaddr (14 bytes storage).
7647 18. Named domain lists were not supported in the hosts_treat_as_local option.
7648 An entry such as +xxxx was not recognized, and was treated as a literal
7651 19. Ensure that header lines added by a DATA ACL are included in the reject log
7652 if the ACL subsequently rejects the message.
7654 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7655 MD5 (which is deprecated).
7657 21. When testing a filter file using -bf, Exim was writing a message when it
7658 took the sender from a "From " line in the message, but it was not doing so
7659 when it took $return_path from a Return-Path: header line. It now does.
7661 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7662 with a valid header line field name (a series of printing characters
7663 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7665 23. Changed "disc" in the source to "disk" to conform to the documentation and
7666 the book and for uniformity.
7668 24. Ignore Sendmail's -Ooption=value command line item.
7670 25. When execve() failed while trying to run a command in a pipe transport,
7671 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7672 could be confused with a return value of 69 from the command itself. This
7673 has been changed to 127, the value the shell returns if it is asked to run
7674 a non-existent command. The wording for the related log line suggests a
7675 non-existent command as the problem.
7677 26. If received_header_text expands to an empty string, do not add a Received:
7678 header line to the message. (Well, it adds a token one on the spool, but
7679 marks it "old" so that it doesn't get used or transmitted.)
7681 27. Installed eximstats 1.28 (addition of -nt option).
7683 28. There was no check for failure on the call to getsockname() in the daemon
7684 code. This can fail if there is a shortage of resources on the system, with
7685 ENOMEM, for example. A temporary error is now given on failure.
7687 29. Contrary to the C standard, it seems that in some environments, the
7688 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7689 program. Exim now does this explicitly; it affects the formatting of
7690 timestamps using strftime().
7692 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7693 complaints. I've now initialized all the variables, to avoid this.
7695 32. Header lines added by a system filter were not being "seen" during
7696 transport-time rewrites.
7698 33. The info_callback() function passed to OpenSSL is set up with type void
7699 (*)(SSL *, int, int), as described somewhere. However, when calling the
7700 function (actually a macro) that sets it up, the type void(*)() is
7701 expected. I've put in a cast to prevent warnings from picky compilers.
7703 34. If a DNS black list lookup found a CNAME record, but there were no A
7704 records associated with the domain it pointed at, Exim crashed.
7706 35. If a DNS black list lookup returned more than one A record, Exim ignored
7707 all but the first. It now scans all returned addresses if a particular IP
7708 value is being sought. In this situation, the contents of the
7709 $dnslist_value variable are a list of all the addresses, separated by a
7712 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7713 to accept a host name and all its aliases if the forward lookup for any of
7714 them yielded the IP address of the incoming connection. Now it accepts only
7715 those names whose forward lookup yields the correct IP address. Any other
7716 names are discarded. This closes a loophole whereby a rogue DNS
7717 administrator could create reverse DNS records to break through a
7718 wildcarded host restriction in an ACL.
7720 37. If a user filter or a system filter that ran in a subprocess used any of
7721 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7722 the wrong values were passed to the pipe command ($thisaddress had the
7723 value of $0, $0 had the value of $1, etc). This bug was introduced by
7724 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7726 38. Improved the line breaking for long SMTP error messages from ACLs.
7727 Previously, if there was no break point between 40 and 75 characters, Exim
7728 left the rest of the message alone. Two changes have been made: (a) I've
7729 reduced the minimum length to 35 characters; (b) if it can't find a break
7730 point between 35 and 75 characters, it looks ahead and uses the first one
7731 that it finds. This may give the occasional overlong line, but at least the
7732 remaining text gets split now.
7734 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7735 file descriptors might be low, and that setting 1000 would always raise it.
7736 It turns out that in some environments, the limit is already over 1000 and
7737 that lowering it causes trouble. So now Exim takes care not to decrease it.
7739 40. When delivering a message, the value of $return_path is set to $sender_
7740 address at the start of routing (routers may change the value). By an
7741 oversight, this default was not being set up when an address was tested by
7742 -bt or -bv, which affected the outcome if any router or filter referred to
7745 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7746 log only when "message" or "log_message" are set. However, if one of the
7747 conditions was an address verification, or a call to a nested ACL, the
7748 messages generated by the underlying test were being passed through. This
7749 no longer happens. The underlying message is available in $acl_verify_
7750 message for both "message" and "log_message" expansions, so it can be
7751 passed through if needed.
7753 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7754 new expansion $bh_ to give the encoded byte string without charset
7755 translation. Translation happens only if iconv() is available; HAVE_ICONV
7756 indicates this at build time. HEADERS_CHARSET gives the charset to
7757 translate to; headers_charset can change it in the configuration, and
7758 "headers charset" can change it in an individual filter file.
7760 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7761 that creates RFC 2047 encoded "words" labels them as that charset instead
7762 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7763 expansion operator; (ii) when Exim creates a From: line for a local
7764 message; (iii) when a header line is rewritten to include a "phrase" part.
7766 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7767 buggy, causing it to skip the first lines of messages whose message ID
7768 ended in 'D'. This would not have bitten before Exim release 4.14, because
7769 message IDs were unlikely to end in 'D' before then. The effect was to have
7770 incorrect size information for certain domains.
7772 45. #include "config.h" was missing at the start of the crypt16.c module. This
7773 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7776 46. If there was a timeout during a "random" callout check, Exim treated it as
7777 a failure of the random address, and carried on sending RSET and the real
7778 address. If the delay was just some slowness somewhere, the response to the
7779 original RCPT would be taken as a response to RSET and so on, causing
7780 mayhem of various kinds.
7782 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7783 when I implemented it. It didn't allow for the fact that some option values
7784 may legitimately be negative (e.g. size_addition), and it didn't even do
7785 the right test for positive values.
7787 48. Domain names in DNS records are case-independent. Exim always looks them up
7788 in lower case. Some resolvers return domain names in exactly the case they
7789 appear in the zone file, that is, they may contain uppercase letters. Not
7790 all resolvers do this - some return always lower case. Exim was treating a
7791 change of case by a resolver as a change of domain, similar to a widening
7792 of a domain abbreviation. This triggered its re-routing code and so it was
7793 trying to route what was effectively the same domain again. This normally
7794 caused routing to fail (because the router wouldn't handle the domain
7795 twice). Now Exim checks for this case specially, and just changes the
7796 casing of the domain that it ultimately uses when it transmits the message
7799 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7802 50. If a filter generated a file delivery with a non-absolute name (possible if
7803 no home directory exists for the router), the forbid_file option was not
7806 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7807 the existing equality matching.
7809 52. Exim was using ints instead of ino_t variables in some places where it was
7810 dealing with inode numbers.
7812 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7813 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7814 called TMPDIR, and if it finds it is different, it changes its value.
7816 54. The smtp_printf() function is now made available to local_scan() so
7817 additional output lines can be written before returning. There is also an
7818 smtp_fflush() function to enable the detection of a dropped connection.
7819 The variables smtp_input and smtp_batched_input are exported to
7822 55. Changed the default runtime configuration: the message "Unknown user"
7823 has been removed from the ACL, and instead placed on the localuser router,
7824 using the cannot_route_message feature. This means that any verification
7825 failures that generate their own messages won't get overridden. Similarly,
7826 the "Unrouteable address" message that was in the ACL for unverifiable
7827 relay addresses has also been removed.
7829 56. Added hosts_avoid_esmtp to the smtp transport.
7831 57. The exicyclog script was not checking for the esoteric option
7832 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7833 will work only if exicyclog is run under the appropriate euid.
7835 58. Following a discussion on the list, the rules by which Exim recognises line
7836 endings on incoming messages have been changed. The -dropcr and drop_cr
7837 options are now no-ops, retained only for backwards compatibility. The
7838 following line terminators are recognized: LF CRLF CR. However, special
7839 processing applies to CR:
7841 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7842 nor a local message in the state where . is a terminator.
7844 (ii) If a bare CR is encountered in a header line, an extra space is added
7845 after the line terminator so as not to end the header. The reasoning
7846 behind this is that bare CRs in header lines are most likely either
7847 to be mistakes, or people trying to play silly games.
7849 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7850 was being incorrectly given as 18 bytes larger than it should have been.
7851 This is a VOB (very old bug).
7853 60. This may never have affected anything current, but just in case it has:
7854 When the local host is found other than at the start of a list of hosts,
7855 the local host, those with the same MX, and any that follow, are discarded.
7856 When the list in question was part of a longer list of hosts, the following
7857 hosts (not currently being processed) were also being discarded. This no
7858 longer happens. I'm not sure if this situation could ever has previously
7861 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7864 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7865 header. This is something that is recommended in a new Internet Draft, and
7866 is something that is documented as being done by Sendmail. There are two
7867 possible values. For messages generated by the autoreply transport, Exim
7870 Auto-Submitted: auto-replied
7872 whereas for all other generated messages (e.g. bounces) it adds
7874 Auto-Submitted: auto-generated
7876 63. The "personal" condition in filters now includes a test for the
7877 Auto-Submitted: header. If it contains the string "auto-" the message it
7878 not considered personal.
7880 64. Added rcpt_include_affixes as a generic transport option.
7882 65. Added queue_only_override (default true).
7884 66. Added the syslog_duplication option.
7886 67. If what should have been the first header line of a message consisted of
7887 a space followed by a colon, Exim was mis-interpreting it as a header line.
7888 It isn't of course - it is syntactically invalid and should therefore be
7889 treated as the start of the message body. The misbehaviour could have
7890 caused a number of strange effects, including loss of data in subsequent
7891 header lines, and spool format errors.
7893 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7894 client host had authenticated. This control can now be exercised by an ACL
7895 for more flexibility.
7897 69. By default, callouts do not happen when testing with -bh. There is now a
7898 variant, -bhc, which does actually run the callout code, including
7899 consulting and updating the callout cache.
7901 70. Added support for saslauthd authentication, courtesy of Alexander
7904 71. If statvfs() failed on the spool or log directories while checking their
7905 size for availability, Exim confusingly gave the error "space shortage".
7906 Furthermore, in debugging mode it crashed with a floating point exception.
7907 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7908 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7909 serious problem, Exim now writes to the main and panic logs when this
7910 happens, with details of the failure. It then refuses to accept the
7911 incoming message, giving the message "spool directory problem" or "log
7912 directory problem" with a 421 code for SMTP messages.
7914 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7915 0, 1, and 2 exist, because some OS complain for execs without them (see
7916 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7917 descriptors. However, the code omitted to check that the open succeeded,
7918 causing mysterious errors if for some reason the permissions on /dev/null
7919 got screwed. Now Exim writes a message to the main and panic logs, and
7920 bombs out if it can't open /dev/null.
7922 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7923 interact so that it is all more flexible. It is supposed to remain
7924 backwards compatible. Also added extra_local_interfaces.
7926 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7927 to bomb out with an assertion failure - to the client this appears as a
7928 connection drop. This problem occurs in the part of the code that was taken
7929 from the Samba project. Fortunately, the assertion is in a very simple
7930 function, so I have fixed this by reproducing the function inline in the
7931 one place where it is called, and arranging for authentication to fail
7932 instead of killing the process with assert().
7934 75. The SPA client code was not working when the server requested OEM rather
7935 than Unicode encoding.
7937 76. Added code to make require_files with a specific uid setting more usable in
7938 the case where statting the file as root fails - usually a non-root-mounted
7939 NFS file system. When this happens and the failure is EACCES, Exim now
7940 forks a subprocess and does the per-uid checking as the relevant uid.
7942 77. Added process_log_path.
7944 78. If log_file_path was not explicitly set, a setting of check_log_space or
7945 check_log_inodes was ignored.
7947 79. If a space check for the spool or log partitions fails, the incident is now
7948 logged. Of course, in the latter case the data may get lost...
7950 80. Added the %p formatting code to string_format() so that it can be used to
7951 print addresses in debug_print(). Adjusted all the address printing in the
7952 debugging in store.c to use %p rather than %d.
7954 81. There was a concern that a line of code in smtp_in.c could overflow a
7955 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7956 initially expressed, the concern was not well-founded, because trailing
7957 spaces are removed early. However, if the trailing spaces were followed by
7958 a NULL, they did not get removed, so the overflow was possible. Two fixes
7961 (a) I re-wrote the offending code in a cleaner fashion.
7962 (b) If an incoming SMTP command contains a NULL character, it is rejected
7965 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7966 runs initgroups(), so that if the Exim user is in any additional groups,
7967 they will be used during message reception.
7973 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob