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.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
163 JH/01 Avoid costly startup code when not strictly needed. This reduces time
164 for some exim process initialisations. It does mean that the logging
165 of TLS configuration problems is only done for the daemon startup.
167 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
169 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
170 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
173 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
174 DKIM signing is being done. Previously a transport_filter would always
175 disable CHUNKING, falling back to traditional DATA.
177 JH/05 Regard command-line recipients as tainted.
179 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
181 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
182 PAM library frees one of the arguments given to it, despite the
183 documentation. Therefore a plain malloc must be used.
185 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
186 on-stack buffers were used, resulting in a taint trap when DSN information
187 copied from a received message was written into the buffer.
189 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
190 the ordering of its ARC headers. This caused a crash.
192 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
193 a new record was being constructed with information from the peer, a trap
196 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
197 installation would get error messages from DMARC verify, when it hit the
198 nonexistent file indicated by the default. Distros wanting DMARC enabled
199 should both provide the file and set the option.
200 Also enforce no DMARC verification for command-line sourced messages.
202 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
203 could, depending on the platform, hang at the STARTTLS response.
205 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
206 message on a connection. Previously if one message had ARC headers and
207 the following one did not, a crash could result when adding an
208 Authentication-Results: header.
210 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
211 Exim main code and Exim-related utities. The introduction of taint
212 tracking also did many adjustments to string handling. Since then, eximon
213 frequently terminated with an assert failure.
215 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
216 check for 452 responses. This slightly helps the inefficieny of doing
217 a large alias-expansion into a recipient-limited target. The max_rcpt
218 transport option still applies (and at the current default, will override
219 the new feature). The check is done for either cause of synch, and forces
220 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
221 connection. The new facility is not tunable at this time.
223 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
224 library live data was being used, so the results became garbage. Make
225 copies while it is still usable.
227 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
228 on delivery deferred (==) and failed (**) lines (if a delivery was
229 attemtped). Previously it was only on completion (=>) lines.
231 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
232 for the expansion of the server_scram_iter and server_scram_salt options.
234 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
235 are now specifically given a NO_DATA response without hitting the system
236 resolver. The library goes on to do the now-standard TXT lookup.
237 Use of dnsdb lookups is not affected.
239 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
240 only retrieve the errormessage once. Previously two calls to dlerror()
241 were used, and the second one (for mainlog/paniclog) retrieved null
244 JH/20 Taint checking: disallow use of tainted data for
245 - the appendfile transport file and directory options
246 - the pipe transport command
247 - the autoreply transport file, log and once options
248 - file names used by the redirect router (including filter files)
250 - paths used by single-key lookups
251 Previously this was permitted.
253 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
254 adjusted the size of a major service buffer; this failed because the
255 buffer was in use at the time. Change to a compile-time increase in the
256 buffer size, when this authenticator is compiled into exim.
258 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
259 previous fast-mode was untenable in the face of glibs using mmap to
260 support larger malloc requests.
262 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
263 New values supported, if defined on system where compiled:
264 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
265 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
267 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
268 running a limited number of proceses in parallel, a benefit is gained. The
269 amount varies with the platform hardware and load. The use of the option
270 queue_run_in_order means we cannot do this, as ordering becomes
273 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
274 had introduced a string-copy (for ensuring NUL-termination) which was not
275 appropriate for that case, which can include embedded NUL bytes in the
276 block of data. Investigation showed the copy to actually be needless, the
277 data being length-specified.
279 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
280 done during a receiving connection, and both used TLS, global info was
281 used rather than per-connection info for tracking the state of data
282 queued for transmission. This could result in a connection hang.
284 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
285 Previously, when delivering serveral messages down a single connection
286 only the first would provide a SIZE. This was due to the size information
287 not being properly tracked.
289 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
290 TAI (at 37 seconds currently), pretend to be in UTC for time-related
291 expansion and logging. Previously, spurious values such as a future
292 minute could be seen.
294 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
295 it could crash from a null-deref. This could also affect the
296 ${addresses: } operator and ${readsock } item.
298 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
299 message following a mime one, the variable was not reset.
301 JH/30 When an pipelined-connect fails at the first response, assume incorrect
302 cached capability (perhaps the peer reneged?) and immediately retry in
305 JH/31 Fix spurious detection of timeout while writing to transport filter.
307 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
308 an attempt to copy the string was made before checking it.
310 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
311 taint of the lookup key was maintained; we now regard the presence in the
312 filesystem as sufficient validation.
314 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
315 argument is supplied.
317 JH/35 The dsearch lookup now requires that the directory is an absolute path.
318 Previously this was not checked, and nonempty relative paths made an
319 access under Exim's current working directory.
321 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
322 Previously no event was raised.
324 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
325 parameter supplied by the sender MAIL FROM command. Previously it was
326 ignored, and only the check_spool_space option value for the required
329 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
330 the size of the signing public-key. Previously it was instead giving
331 the size of the signature hash.
333 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
334 the default. See the (new) dkim_verify_min_keysizes option.
336 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
337 and an ACL use a lookup for checking either the local_part or domain,
338 stale data could be accessed. Ensure that variable references are
339 dropped between messages.
341 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
342 by the client was not checked as pointing within response data before
343 being used. A malicious client could thus cause an out-of-bounds read and
344 possibly gain authentication. Fix by adding the check.
346 JH/42 Internationalisation: change the default for downconversion in the smtp
347 transport to be "if needed". Previously it was "as previously set" for
348 the message, which usually meant "if needed" for message-submission but
349 "no" for everything else. However, MTAs have been seen using SMTPUTF8
350 even when the envelope addresses did not need it, resulting in forwarding
351 failures to non-supporting MTAs. A downconvert in such cases will be
352 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
353 transport. The change does mean that addresses needing conversion will
354 be converted when previously a delivery failure would occur.
356 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
357 response was received it would be used unchecked in a fail-DSN, violating
358 standards on line-length limits. Truncate if needed.
360 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
361 doesn't work. (Additionally add a new main config option to configure the
368 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
369 side, to decode a TLS 1.3 packet capture.
371 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
372 Previously the default library behaviour applied, sending two, each in
375 JH/03 Debug output for ACL now gives the config file name and line number for
378 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
380 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
382 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
383 buffer overrun for (non-chunking) other transports.
385 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
386 TLS1.3, means that a server rejecting a client certificate is not visible
387 to the client until the first read of encrypted data (typically the
388 response to EHLO). Add detection for that case and treat it as a failed
389 TLS connection attempt, so that the normal retry-in-clear can work (if
390 suitably configured).
392 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
393 and/or domain. Found and fixed by Jason Betts.
395 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
396 configuration). If a CNAME target was not a wellformed name pattern, a
399 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
400 the OS reports them interleaved with other addresses.
402 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
403 used both for input and for a verify callout, both encrypted, SMTP
404 responses being sent by the server could be lost. This resulted in
405 dropped connections and sometimes bounces generated by a peer sending
408 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
409 string. Previously it was possible to cause undefined behaviour in a
410 library routine (usually a crash). Found by "zerons".
412 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
415 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
416 API was removed, so update to use the newer ones.
418 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
419 any timeout set, is taking a long time. Previously we would hang on to a
420 rotated logfile "forever" if the input was arriving with long gaps
421 (a previous attempt to fix addressed lack, for a long time, of initial
424 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
425 shared (NFS) environment. The length of the tempfile name is now
426 4 + 16 ("hdr.$message_exim_id") which might break on file
427 systems which restrict the file name length to lower values.
430 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
431 shared (NFS) environment.
433 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
434 did for all versions <4.90). Notably -M, -m, --invert, -I may be
437 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
438 on some platforms for bit 31.
440 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
441 to changes apparently associated with TLS1.3 handling some of the APIs
442 previously used were either nonfunctional or inappropriate. Strings
443 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
444 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
445 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
446 This affects log line X= elements, the $tls_{in,out}_cipher variables,
447 and the use of specific cipher names in the encrypted= ACL condition.
449 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
451 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
452 verification result was not updated unless hosts_require_ocsp applied.
454 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
455 queue_list_requires_admin set to false, non-admin users were denied the
458 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
459 directory-of-certs mode. Previously they were advertised despite the
462 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
463 A single TCP connection by a client will now hold a TLS connection open
464 for multiple message deliveries, by default. Previously the default was to
467 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
468 default. If built with the facility, DANE will be used. The facility
469 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
471 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
472 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
473 must be defined and you must still, unless you define DISABLE_TLS, manage
474 the the include-dir and library-file requirements that go with that
475 choice. Non-TLS builds are still supported.
477 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
480 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
481 default. If the platform supports and has the facility enabled, it will
482 be requested on all coneections.
484 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
485 controlled by the build-time option SUPPORT_PIPE_CONNECT.
487 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
489 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
490 success-DSN messages. Previously the From: header was always the default
491 one for these; the option was ignored.
493 JH/28 Fix the timeout on smtp response to apply to the whole response.
494 Previously it was reset for every read, so a teergrubing peer sending
495 single bytes within the time limit could extend the connection for a
496 long time. Credit to Qualsys Security Advisory Team for the discovery.
498 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
499 delivery address, which leaked information of the results of local
500 forwarding. Change to the original envelope recipient address, per
503 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
504 requested. Previously not bounce was generated and a log entry of
505 error ignored was made.
507 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
509 JH/32 Introduce a general tainting mechanism for values read from the input
510 channel, and values derived from them. Refuse to expand any tainted
511 values, to catch one form of exploit.
513 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
514 was unused and the unexpanded text used for the test. Found and
515 fixed by Ruben Jenster.
517 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
518 an attempt to use a TLS library read routine dereffed a nul pointer,
521 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
522 them in our smtp response.
524 JH/36 Have the general router option retry_use_local_part default to true when
525 any of the restrictive preconditions are set (to anything). Previously it
526 was only for check_local user. The change removes one item of manual
527 configuration which is required for proper retries when a remote router
528 handles a subset of addresses for a domain.
530 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
531 link count into consideration.
533 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
534 caused the extension of big_buffer, the following lines were ignored.
536 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
537 accordance with RFC 2308. Previously there was no expiry, so a longlived
538 receive process (eg. due to ACL delays) versus a short SOA value could
541 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
543 JH/39 Promote DMARC support to mainline.
545 JH/40 Bug 2452: Add a References: header to DSNs.
547 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
548 parameters. The relevant library call is documented as "Deprecated: This
549 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
550 3.6.0, DH parameters are negotiated following RFC7919."
552 HS/06 Change the default of dnssec_request_domains to "*"
554 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
555 carried on and emitted a BDAT command, even when PIPELINING was not
558 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
559 buffer was used for the filename, resulting in a trap when tainted
560 arguments (eg. $domain) were used.
562 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
563 recommended to avoid a possible server-load attack. The feature can be
564 re-enabled via the openssl_options main cofiguration option.
566 JH/45 local_scan API: documented the current smtp_printf() call. This changed
567 for version 4.90 - adding a "more data" boolean to the arguments.
568 Bumped the ABI version number also, this having been missed previously;
569 release versions 4.90 to 4.92.3 inclusive were effectively broken in
570 respect of usage of smtp_printf() by either local_scan code or libraries
571 accessed via the ${dlfunc } expansion item. Both will need coding
572 adjustment for any calls to smtp_printf() to match the new function
573 signature; a FALSE value for the new argument is always safe.
575 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
576 the file-offset (which the Linux syscall does, and exim expects); this
577 resulted in an indefinite loop.
579 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
580 to do ARC verification. The Authentication-Results: header line added
581 by the configuration then had no ARC item.
587 JH/01 Remove code calling the customisable local_scan function, unless a new
588 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
590 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
591 non-signal-safe functions being used.
593 JH/03 Bug 2269: When presented with a received message having a stupidly large
594 number of DKIM-Signature headers, disable DKIM verification to avoid
595 a resource-consumption attack. The limit is set at twenty.
597 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
598 report of oldest_pass in ${authres } in consequence, and separate out
599 some descriptions of reasons for verification fail.
601 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
602 files in the spool were present and unlocked. A queue-runner could spot
603 them, resulting in a duplicate delivery. Fix that by doing the unlock
604 after the unlink. Investigation by Tim Stewart. Take the opportunity to
605 add more error-checking on spoolfile handling while that code is being
608 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
609 No known attacks, no CVE, this is defensive hardening.
611 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
612 a queue-runner could start a delivery while other operations were ongoing.
613 Cutthrough delivery was a common victim, resulting in duplicate delivery.
614 Found and investigated by Tim Stewart. Fix by using the open message data
615 file handle rather than opening another, and not locally closing it (which
616 releases a lock) for that case, while creating the temporary .eml format
617 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
619 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
620 $sender_verify_failure/$recipient_verify_failure to "random".
622 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
625 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
626 Previously this would segfault.
628 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
631 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
632 like zero, since the resolver should be doing this for us, But we need one
633 as a CNAME but no MX presence gets the CNAME returned; we need to check
634 that doesn't point to an MX to declare it "no MX returned" rather than
635 "error, loop". A new main option is added so the older capability of
636 following some limited number of chain links is maintained.
638 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
640 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
641 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
642 now-deprecated earlier definitions used only specified the range up to TLS
643 1.2 (in the older-version library docs).
645 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
647 JH/15 Rework TLS client-side context management. Stop using a global, and
648 explicitly pass a context around. This enables future use of TLS for
649 connections to service-daemons (eg. malware scanning) while a client smtp
650 connection is using TLS; with cutthrough connections this is quite likely.
652 JH/16 Fix ARC verification to do AS checks in reverse order.
654 JH/17 Support a "tls" option on the ${readsocket } expansion item.
656 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
657 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
658 Previously the "utf8" would be re-prepended for every additional message.
660 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
661 Previously thery were accepted, resulting in issues when attempting to
662 forward messages to a non-supporting MTA.
664 PP/02 Let -n work with printing macros too, not just options.
666 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
667 one parent address was copied, and bogus data was used at delivery-logging
668 time. Either a crash (after delivery) or bogus log data could result.
669 Discovery and analysis by Tim Stewart.
671 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
672 Previously if the string ended mid-character, we did not insert the
673 promised '?' replacement.
675 PP/04 Documentation: current string operators work on bytes, not codepoints.
677 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
678 should pack well giving a smaller memory footprint so better caching and
679 therefore performance. Group the declarations where this can't be done so
680 that the byte-sized flag variables are not interspersed among pointer
681 variables, giving a better chance of good packing by the compiler.
683 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
684 non-null, to avoid issues with sites running BATV. Previously reports were
685 sent with an empty envelope sender so looked like bounces.
687 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
688 The ignore_error flag wasn't being returned from the filter subprocess so
689 was not set for later routers. Investigation and fix by Matthias Kurz.
691 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
692 and a msg:complete for the whole, when a message is manually removed using
693 -Mrm. Developement by Matthias Kurz, hacked on by JH.
695 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
696 a "Gnu special" function, asprintf() in the DB utility binary builds; I
697 hope that is portable enough.
699 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
700 requiring a known-CA anchor certificate; make it now rely entirely on the
701 TLSA as an anchor. Checking the name on the leaf cert against the name
702 on the A-record for the host is still done for TA (but not for EE mode).
704 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
705 included in delivery lines for non-proxied connections, when compiled with
706 SUPPORT_SOCKS and running with proxy logging enabled.
708 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
709 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
710 move the existing event to fire before the normal logging of message
711 failure so that custom logging is bracketed by normal logging.
713 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
714 msg:fail:internal event. Developement by Matthias Kurz.
716 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
717 far too small for todays use of crypto signatures stored there. Go all
718 the way to the max DNS message size of 64kB, even though this might be
719 overmuch for IOT constrained device use.
721 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
722 copy a string over itself. The library routine is documented as not
723 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
725 JH/32 For main options check_spool_space and check_inode_space, where the
726 platform supports 64b integers, support more than the previous 2^31 kB
727 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
728 the previous G, M, k.
730 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
731 $authenticated_fail_id variable on authentication failure. Previously
734 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
735 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
736 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
737 GNUTLS_SEC_PARAM_MEDIUM.
739 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
740 side. Previously we would continue as if no SNI had been received.
742 JH/36 Harden the handling of string-lists. When a list consisted of a sole
743 "<" character, which should be a list-separator specification, we walked
744 off past the nul-terimation.
746 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
747 causes) even when the retry time is not yet met. Previously they were
748 not, meaning that when (say) an account was over-quota and temp-rejecting,
749 and multiple senders' messages were queued, only one sender would get
750 notified on each configured delay_warning cycle.
752 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
754 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
755 cipher-suites, an error can be left on the stack even for a succeeding
756 accept; this results in impossible error messages when a later operation
759 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
760 return error codes indicating retry. Under TLS1.3 this becomes required.
762 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
763 it only wrote the new authenticators, resulting in a lack of tracking of
764 peer changes of ESMTP extensions until the next cache flush.
766 JH/41 Fix the loop reading a message header line to check for integer overflow,
767 and more-often against header_maxsize. Previously a crafted message could
768 induce a crash of the recive process; now the message is cleanly rejected.
770 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
771 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
777 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
778 When redis_servers is set to a list of > 1 element, and the Redis servers
779 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
780 case of MOVED into a DEFER case instead, thus moving the query onto the
781 next server in the list. For a cluster of N elements, all N servers must
782 be defined in redis_servers.
784 GF/02 Catch and remove uninitialized value warning in exiqsumm
785 Check for existence of @ARGV before looking at $ARGV[0]
787 JH/01 Replace the store_release() internal interface with store_newblock(),
788 which internalises the check required to safely use the old one, plus
789 the allocate and data copy operations duplicated in both (!) of the
790 extant use locations.
792 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
793 modifier. This matches the restriction on the commandline.
795 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
796 Previously only the last row was returned.
798 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
799 we assumed that tags in the header were well-formed, and parsed the
800 element content after inspecting only the first char of the tag.
801 Assumptions at that stage could crash the receive process on malformed
804 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
805 While running the DKIM ACL we operate on the Permanent memory pool so that
806 variables created with "set" persist to the DATA ACL. Also (at any time)
807 DNS lookups that fail create cache records using the Permanent pool. But
808 expansions release any allocations made on the current pool - so a dnsdb
809 lookup expansion done in the DKIM ACL releases the memory used for the
810 DNS negative-cache, and bad things result. Solution is to switch to the
811 Main pool for expansions.
812 While we're in that code, add checks on the DNS cache during store_reset,
813 active in the testsuite.
814 Problem spotted, and debugging aided, by Wolfgang Breyha.
816 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
817 When none of the hosts presented to a transport match an already-open
818 connection, close it and proceed with the list. Previously we would
819 queue the message. Spotted by Lena with Yahoo, probably involving
822 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
823 Previously a spurious "250 OK id=" response was appended to the proper
826 JH/08 The "support for" informational output now, which built with Content
827 Scanning support, has a line for the malware scanner interfaces compiled
828 in. Interface can be individually included or not at build time.
830 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
831 by the template makefile "src/EDITME". The "STREAM" support for an older
832 ClamAV interface method is removed.
834 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
835 rows affected is given instead).
837 JH/11 The runtime Berkeley DB library version is now additionally output by
838 "exim -d -bV". Previously only the compile-time version was shown.
840 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
841 SMTP connection. Previously, when one had more recipients than the
842 first, an abortive onward connection was made. Move to full support for
843 multiple onward connections in sequence, handling cutthrough connection
844 for all multi-message initiating connections.
846 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
847 routers. Previously, a multi-recipient message would fail to match the
848 onward-connection opened for the first recipient, and cause its closure.
850 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
851 a timeout on read on a GnuTLS initiating connection, resulting in the
852 initiating connection being dropped. This mattered most when the callout
853 was marked defer_ok. Fix to keep the two timeout-detection methods
856 JH/15 Relax results from ACL control request to enable cutthrough, in
857 unsupported situations, from error to silently (except under debug)
858 ignoring. This covers use with PRDR, frozen messages, queue-only and
861 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
863 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
864 metadata, resulting in a crash in free().
866 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
867 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
868 Broken also in d185889f4, with init system revamp.
870 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
871 Previously we abruptly closed the connection after reading a malware-
872 found indication; now we go on to read the "scan ok" response line,
875 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
876 ACL. Previously, a crash would result.
878 JH/19 Speed up macro lookups during configuration file read, by skipping non-
879 macro text after a replacement (previously it was only once per line) and
880 by skipping builtin macros when searching for an uppercase lead character.
882 JH/20 DANE support moved from Experimental to mainline. The Makefile control
883 for the build is renamed.
885 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
886 was allocated for every new TLS startup, meaning one per message. Fix
887 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
889 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
890 reported the original. Fix to report (as far as possible) the ACL
891 result replacing the original.
893 JH/23 Fix memory leak during multi-message connections using STARTTLS under
894 OpenSSL. Certificate information is loaded for every new TLS startup,
895 and the resources needed to be freed.
897 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
899 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
902 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
903 DATA response info to the (existing) per-recipient response info for
904 the "C=" log element. It can have useful tracking info from the
905 destination system. Patch from Simon Arlott.
907 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
908 length value. Previously this would segfault.
910 HS/02 Support Avast multiline protoocol, this allows passing flags to
911 newer versions of the scanner.
913 JH/28 Ensure that variables possibly set during message acceptance are marked
914 dead before release of memory in the daemon loop. This stops complaints
915 about them when the debug_store option is enabled. Discovered specifically
916 for sender_rate_period, but applies to a whole set of variables.
917 Do the same for the queue-runner and queue-list loops, for variables set
918 from spool message files. Do the same for the SMTP per-message loop, for
919 certain variables indirectly set in ACL operations.
921 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
922 as a multi-recipient message from a mailinglist manager). The coding had
923 an arbitrary cutoff number of characters while checking for more input;
924 enforced by writing a NUL into the buffer. This corrupted long / fast
925 input. The problem was exposed more widely when more pipelineing of SMTP
926 responses was introduced, and one Exim system was feeding another.
927 The symptom is log complaints of SMTP syntax error (NUL chars) on the
928 receiving system, and refused recipients seen by the sending system
929 (propating to people being dropped from mailing lists).
930 Discovered and pinpointed by David Carter.
932 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
933 replaced by the ${authresults } expansion.
935 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
937 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
938 allows proper process termination in container environments.
940 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
941 Previously the "final dot" had a newline after it; ensure it is CR,LF.
943 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
944 and "err_perm", deprecated since 4.83 when the RFC-defined words
945 "temperror" and "permerror" were introduced.
947 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
948 transport-filters or DKIM-signing. The restriction was lost in the
949 consolidation of verify-callout and delivery SMTP handling.
950 Extend the restriction to also cover ARC-signing.
952 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
953 in defer=pass mode supply a 450 to the initiator. Previously the message
956 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
957 tls_require_ciphers is used as before.
959 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
960 "pass_unscanned". Only tmpfails from the scanner are written to
961 the paniclog, as they may require admin intervention (permission
962 denied, license issues). Other scanner errors (like decompression
963 bombs) do not cause a paniclog entry.
965 JH/36 Fix reinitialisation of DKIM logging variable between messages.
966 Previously it was possible to log spurious information in receive log
969 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
970 triggered odd behaviour from Outlook Express clients.
972 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
975 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
976 since the IETF WG has not yet settled on that versus the original
977 "bare" representation.
979 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
980 Previously the millisecond value corrupted the output.
981 Fix also for syslog_pid=no and log_selector +pid, for which the pid
982 corrupted the output.
988 JH/01 Rework error string handling in TLS interface so that the caller in
989 more cases is responsible for logging. This permits library-sourced
990 string to be attached to addresses during delivery, and collapses
991 pairs of long lines into single ones.
993 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
994 during configuration. Wildcards are allowed and expanded.
996 JH/02 Rework error string handling in DKIM to pass more info back to callers.
997 This permits better logging.
999 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1000 do not close it down and have the child transport start it up again on
1001 the passed-on TCP connection. Instead, proxy the child (and any
1002 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1003 affected: the continued delivery log lines do not have any DNSSEC, TLS
1004 Certificate or OCSP information. TLS cipher information is still logged.
1006 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1007 identical IP addresses on different listening ports. Will also affect
1010 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1011 add noisy ifdef guards to special-case this sillyness.
1012 Patch from Bernd Kuhls.
1014 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1015 than 255 are no longer allowed.
1017 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1018 Disable the session-cache too, which might reduce our load. Since we
1019 currrectly use a new context for every connection, both as server and
1020 client, there is no benefit for these.
1021 GnuTLS appears to not support tickets server-side by default (we don't
1022 call gnutls_session_ticket_enable_server()) but client side is enabled
1023 by default on recent versions (3.1.3 +) unless the PFS priority string
1026 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1027 <https://reproducible-builds.org/specs/source-date-epoch/>.
1029 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1030 the check for any unsuccessful recipients did not notice the limit, and
1031 erroneously found still-pending ones.
1033 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1034 MSG_MORE. Only in-clear (not on TLS connections).
1036 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1037 transport-filter is involved we can buffer the headers in memory for
1038 creating the signature, and read the spool data file once for the
1039 signature and again for transmission.
1041 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1042 4.77 as the kernel support then wasn't solid, having issues in 64bit
1043 mode. Now, it's been long enough. Add support for FreeBSD also.
1045 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1046 case where the routing stage had gathered several addresses to send to
1047 a host before calling the transport for the first, we previously failed
1048 to close down TLS in the old transport process before passing the TCP
1049 connection to the new process. The new one sent a STARTTLS command
1050 which naturally failed, giving a failed delivery and bloating the retry
1051 database. Investigation and fix prototype from Wolfgang Breyha.
1053 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1054 false-negatives in the check that the sender had not preempted a response
1055 or prompt from Exim (running as a server), due to that code's lack of
1056 awareness of the SMTP input buffering.
1058 PP/04 Add commandline_checks_require_admin option.
1059 Exim drops privileges sanely, various checks such as -be aren't a
1060 security problem, as long as you trust local users with access to their
1061 own account. When invoked by services which pass untrusted data to
1062 Exim, this might be an issue. Set this option in main configuration
1063 AND make fixes to the calling application, such as using `--` to stop
1066 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1067 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1068 responses to those, into a single TLS record each way (this usually means
1069 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1072 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1073 affects you only if you're dancing at the edge of the param size limits.
1074 If you are, and this message makes sense to you, then: raise the
1075 configured limit or use OpenSSL 1.1. Nothing we can do for older
1078 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1079 cmdline element to get the documented default one. Previously it was
1082 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1083 are unsuable for non-message-specific reasons - eg. connection timeout,
1084 banner-time rejection.
1086 JH/16 Fix logging of delivery remote port, when specified by router, under
1089 PP/06 Repair manualroute's ability to take options in any order, even if one
1090 is the name of a transport.
1093 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1095 JH/17 Change the list-building routines interface to use the expanding-string
1096 triplet model, for better allocation and copying behaviour.
1098 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1099 Previously it was constructed the first time a possibly-matching string
1100 was met in the configuration file input during startup; now it is done
1103 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1104 compatible one, to avoid the (poorly documented) possibility of a config
1105 file in the working directory redirecting the DB files, possibly correpting
1106 some existing file. CVE-2017-10140 assigned for BDB.
1108 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1109 cache-hot. Previously, although the result was properly cached, the
1110 initial verify call returned a defer.
1112 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1113 the main verify for receipient in uncached-mode.
1115 JH/22 Retire historical build files to an "unsupported" subdir. These are
1116 defined as "ones for which we have no current evidence of testing".
1118 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1119 if present. Previously it was ignored.
1121 JH/24 Start using specified-initialisers in C structure init coding. This is
1122 a C99 feature (it's 2017, so now considered safe).
1124 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1125 if was a fixed-sized field and bitmask ops via macros; it is now more
1128 PP/07 GitHub PR 56: Apply MariaDB build fix.
1129 Patch provided by Jaroslav Škarvada.
1131 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1132 during Coverity cleanups [4.87 JH/47]
1133 Diagnosis and fix provided by Michael Fischer v. Mollard.
1135 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1136 the right size to place the terminating semicolon on its own folded
1137 line, the header hash was calculated to an incorrect value thanks to
1138 the (relaxed) space the fold became.
1140 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1141 and confused the parent.
1143 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1144 which could crash as a result. This could lead to undeliverable messages.
1146 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1149 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1150 matters on fast-turnover and PID-randomising systems, which were getting
1151 out-of-order delivery.
1153 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1154 a possibly-overlapping copy. The symptom was that "Remote host closed
1155 connection in response to HELO" was logged instead of the actual 4xx
1158 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1159 Previously only that bufferd was discarded, resulting in SYMTP command
1162 JH/32 DKIM: when a message has multiple signatures matching an identity given
1163 in dkim_verify_signers, run the dkim acl once for each. Previously only
1164 one run was done. Bug 2189.
1166 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1167 "panic the current process" to "deliberately defer". The panic log is
1168 still written with the problem list name; the mail and reject logs now
1169 get a temp-reject line for the message that was being handled, saying
1170 something like "domains check lookup or other defer". The SMTP 451
1171 message is still "Temporary local problem".
1173 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1174 A crafted sequence of BDAT commands could result in in-use memory beeing
1175 freed. CVE-2017-16943.
1177 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1178 from SMTP input. Previously it was always done; now only done for DATA
1179 and not BDAT commands. CVE-2017-16944.
1181 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1182 to the message (such as an overlong header line). Previously this was
1183 not done and we did not exit BDAT mode. Followon from the previous item
1184 though a different problem.
1190 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1191 than -2003 did; needs libidn2 in addition to libidn.
1193 JH/02 The path option on a pipe transport is now expanded before use.
1195 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1196 Patch provided by "Björn", documentation fix added too.
1198 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1199 missing a wire-to-host endian conversion.
1201 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1202 close after a BDAT command line could be taken as a following command,
1203 giving a synch failure. Fix by only checking for synch immediately
1204 before acknowledging the chunk.
1206 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1207 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1208 macro. Patches provided by Josh Soref.
1210 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1211 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1212 by IANA as a service extension. However, John Klensin suggests that we
1215 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1216 the dkim code may be unix-mode line endings rather than smtp wire-format
1217 CRLF, so prepend a CR to any bare LF.
1219 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1220 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1222 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1223 the first were themselves being wrongly included in the feed into dkim
1224 processing; with most chunk sizes in use this resulted in an incorrect
1225 body hash calculated value.
1227 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1228 DKIM signature block, for verification. Although advised against by
1229 standards it is specifically not ruled illegal.
1231 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1233 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1234 missing a body hash (the bh= tag).
1236 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1237 It seems that HAProxy sends the Proxy Protocol information in clear and
1238 only then does a TLS startup, so do the same.
1240 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1241 TCP connections (such as for Spamd) unless the daemon successfully set
1242 Fast Open mode on its listening sockets. This fixes breakage seen on
1243 too-old kernels or those not configured for Fast Open, at the cost of
1244 requiring both directions being enabled for TFO, and TFO never being used
1245 by non-daemon-related Exim processes.
1247 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1248 endings, at least on the first header line. Try to canonify any that get
1249 past that check, despite the cost.
1251 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1252 now limited to an arbitrary five deep, while parsing addresses with the
1253 strip_excess_angle_brackets option enabled.
1255 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1256 instead leave the unprompted TLS handshake in socket buffer for the
1257 TLS library to consume.
1259 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1261 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1263 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1264 region in one operation, for speed, and this leaves assigned pointers
1265 dangling. Add checks run only under the testsuite which checks all
1266 variables at a store-reset and panics on a dangling pointer; add code
1267 explicitly nulling out all the variables discovered. Fixes one known
1268 bug: a transport crash, where a dangling pointer for $sending_ip_address
1269 originally assigned in a verify callout, is re-used.
1271 PP/06 Drop '.' from @INC in various Perl scripts.
1273 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1275 PP/08 Reduce a number of compilation warnings under clang; building with
1276 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1277 should be warning-free.
1279 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1281 HS/01 Fix portability problems introduced by PP/08 for platforms where
1282 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1284 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1285 chunk. This allows us to accept broken chunked messages. We need a more
1286 general solution here.
1288 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1289 already-broken messages in the queue.
1291 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1293 JH/19 Fix reference counting bug in routing-generated-address tracking.
1299 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1300 supports it and a size is available (ie. the sending peer gave us one).
1302 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1303 years of being deprecated). The replacements are the ACLs
1304 acl_smtp_mime and acl_not_smtp_mime.
1306 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1307 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1308 or even an in-clear connection were permitted. Now, if the host lookup
1309 was dnssec and dane was requested then the host is only used if the
1310 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1311 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1312 if one fails this test.
1313 This means that a poorly-configured remote DNS will make it incommunicado;
1314 but it protects against a DNS-interception attack on it.
1316 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1317 non-noisy when a race steals the message being considered.
1319 JH/05 If main configuration option tls_certificate is unset, generate a
1320 self-signed certificate for inbound TLS connections.
1322 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1323 in rewrites and routers.
1325 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1326 and logged a warning sing 4.83; now they are a configuration file error.
1328 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1329 (lacking @domain). Apply the same qualification processing as RCPT.
1331 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1333 JH/10 Support ${sha256:} applied to a string (as well as the previous
1336 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1337 a cutthrough deliver is pending, as we always want to make a connection.
1338 This also avoids re-routing the message when later placing the cutthrough
1339 connection after a verify cache hit.
1340 Do not update it with the verify result either.
1342 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1343 when routing results in more than one destination address.
1345 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1346 signing (which inhibits the cutthrough capability). Previously only
1347 the presence of an option was tested; now an expansion evaluating as
1348 empty is permissible (obviously it should depend only on data available
1349 when the cutthrough connection is made).
1351 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1352 the relevant preceding SMTP command did not note the pipelining mode.
1354 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1355 Previously they were not counted.
1357 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1358 as one having no matching records. Previously we deferred the message
1359 that needed the lookup.
1361 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1362 distinguished as "(=".
1364 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1365 for missing MX records. Previously it only worked for missing A records.
1367 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1369 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1370 after the data-go-ahead and data-ack. Patch from Jason Betts.
1372 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1373 even for a "none" policy. Patch from Tony Meyer.
1375 JH/22 Fix continued use of a connection for further deliveries. If a port was
1376 specified by a router, it must also match for the delivery to be
1379 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1380 When one of the recipients of a message was unsuitable for the connection
1381 (has no matching addresses), we lost track of needing to mark it
1382 deferred. As a result mail would be lost.
1384 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1386 JH/25 Decoding ACL controls is now done using a binary search; the source code
1387 takes up less space and should be simpler to maintain. Merge the ACL
1388 condition decode tables also, with similar effect.
1390 JH/26 Fix problem with one_time used on a redirect router which returned the
1391 parent address unchanged. A retry would see the parent address marked as
1392 delivered, so not attempt the (identical) child. As a result mail would
1395 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1396 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1397 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1400 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1401 Main config options check_{log,spool}_{inodes,space} are now
1402 100 inodes, 10MB unless set otherwise in the configuration.
1404 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1405 Previously it only applied to the main-section connection policy
1408 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1410 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1411 by me. Added RFC7919 DH primes as an alternative.
1413 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1414 are not in the system include path.
1416 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1417 GnuTLS, when a session startup failed (eg because the client disconnected)
1418 Exim did stdio operations after fclose. This was exposed by a recent
1419 change which nulled out the file handle after the fclose.
1421 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1422 signed directly by the cert-signing cert, rather than an intermediate
1423 OCSP-signing cert. This is the model used by LetsEncrypt.
1425 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1427 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1428 an incoming connection.
1430 HS/02 Bug 1802: Do not half-close the connection after sending a request
1433 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1434 fallback to "prime256v1".
1436 JH/34 SECURITY: Use proper copy of DATA command in error message.
1437 Could leak key material. Remotely exploitable. CVE-2016-9963.
1443 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1444 and 3.4.4 - once the server is enabled to respond to an OCSP request
1445 it does even when not requested, resulting in a stapling non-aware
1446 client dropping the TLS connection.
1448 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1449 support variable-length bit vectors. No functional change.
1451 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1452 The I= interface field on outgoing lines is now after the H= remote
1453 host field, same as incoming lines. There is a separate
1454 outgoing_interface log selector which allows you to disable the
1457 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1458 If not running log_selector +smtp_connection the mainlog would be held
1459 open indefinitely after a "too many connections" event, including to a
1460 deleted file after a log rotate. Leave the per net connection logging
1461 leaving it open for efficiency as that will be quickly detected by the
1462 check on the next write.
1464 HS/01 Bug 1671: Fix post transport crash.
1465 Processing the wait-<transport> messages could crash the delivery
1466 process if the message IDs didn't exist for some reason. When
1467 using 'split_spool_directory=yes' the construction of the spool
1468 file name failed already, exposing the same netto behaviour.
1470 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1471 mime_regex ACL conditions.
1473 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1474 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1475 to HELO, local diagnostic string.
1477 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1478 log line to debug. Even when configured with a tls authenticator many
1479 client connections are expected to not authenticate in this way, so
1480 an authenticate fail is not an error.
1482 HS/02 Add the Exim version string to the process info. This way exiwhat
1483 gives some more detail about the running daemon.
1485 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1486 matter for fast-change records such as DNSBLs.
1488 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1489 retry record, even if constant. There may be multiple transports with
1490 different interface settings and the retry behaviour needs to be kept
1493 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1495 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1497 JH/10 Bug 840: fix log_defer_output option of pipe transport
1499 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1500 hosts_randomize. This matters a lot when combined with mua_wrapper.
1502 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1503 ${quote_pgsql:<string>} operator.
1505 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1507 JH/14 Change header file name for freeradius-client. Relevant if compiling
1508 with Radius support; from the Gentoo tree and checked under Fedora.
1510 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1512 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1513 option transport_filter, meaning no filtering.
1515 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1517 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1518 defaults to "*" (all hosts). The variable is now available when not built
1519 with TLS, default unset, mainly to enable keeping the testsuite sane.
1520 If a server certificate is not supplied (via tls_certificate) an error is
1521 logged, and clients will find TLS connections fail on startup. Presumably
1522 they will retry in-clear.
1523 Packagers of Exim are strongly encouraged to create a server certificate
1524 at installation time.
1526 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1527 with the $config_file variable.
1529 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1530 in transport context, after the attempt, and per-recipient. The latter type
1531 is per host attempted. The event data is the error message, and the errno
1532 information encodes the lookup type (A vs. MX) used for the (first) host,
1533 and the trailing two digits of the smtp 4xx response.
1535 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1536 to write to mainlog (or rejectlog, paniclog) in the window between file
1537 creation and permissions/ownership being changed. Particularly affects
1538 installations where exicyclog is run as root, rather than exim user;
1539 result is that the running daemon panics and dies.
1541 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1543 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1544 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1545 "pri" and "weight". Note that the previous implicit priority given by the
1546 list order is no longer honoured.
1548 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1549 for DKIM processing.
1551 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1552 by defining SUPPORT_SOCKS.
1554 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1555 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1556 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1557 variables are renamed to proxy_{local,external}_{address,port}.
1559 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1560 for a build by defining SUPPORT_I18N
1562 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1563 of the query string, and make ${quote_redis:} do that quoting.
1565 JH/29 Move Events support from Experimental to mainline, enabled by default
1566 and removable for a build by defining DISABLE_EVENT.
1568 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1570 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1571 cached by the daemon.
1573 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1574 by defining LOOKUP_REDIS. The libhiredis library is required.
1576 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1577 keys are given for lookup.
1579 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1580 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1581 only supported when built with TLS support. The PolarSSL SHA routines
1582 are still used when the TLS library is too old for convenient support.
1584 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1585 openssl_options), for security. OpenSSL forces this from version 1.1.0
1586 server-side so match that on older versions.
1588 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1589 allocation for $value could be released as the expansion processing
1590 concluded, but leaving the global pointer active for it.
1592 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1593 and to use the domains and local_parts ACL conditions.
1595 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1596 incorrectly not doubled on cutthrough transmission, hence seen as a
1597 body-termination at the receiving system - resulting in truncated mails.
1598 Commonly the sender saw a TCP-level error, and retransmitted the message
1599 via the normal store-and-forward channel. This could result in duplicates
1600 received - but deduplicating mailstores were liable to retain only the
1601 initial truncated version.
1603 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1605 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1607 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1608 we're in there, support oversigning also; bug 1309.
1610 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1612 HS/04 Add support for keep_environment and add_environment options.
1614 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1615 either intentional arithmetic overflow during PRNG, or testing config-
1618 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1619 delivery resulted in actual delivery. Cancel cutthrough before DATA
1622 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1623 rejects a recipient: pass the reject to the originator.
1625 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1626 Many were false-positives and ignorable, but it's worth fixing the
1629 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1630 for the new environment-manipulation done at startup. Move the routines
1631 from being local to tls.c to being global via the os.c file.
1633 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1634 an extract embedded as result-arg for a map, the first arg for extract
1635 is unavailable so we cannot tell if this is a numbered or keyed
1636 extraction. Accept either.
1642 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1645 JH/02 The smtp transport option "multi_domain" is now expanded.
1647 JH/03 The smtp transport now requests PRDR by default, if the server offers
1650 JH/04 Certificate name checking on server certificates, when exim is a client,
1651 is now done by default. The transport option tls_verify_cert_hostnames
1652 can be used to disable this per-host. The build option
1653 EXPERIMENTAL_CERTNAMES is withdrawn.
1655 JH/05 The value of the tls_verify_certificates smtp transport and main options
1656 default to the word "system" to access the system default CA bundle.
1657 For GnuTLS, only version 3.0.20 or later.
1659 JH/06 Verification of the server certificate for a TLS connection is now tried
1660 (but not required) by default. The verification status is now logged by
1661 default, for both outbound TLS and client-certificate supplying inbound
1664 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1667 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1668 Status Notification (bounce) messages are now MIME format per RFC 3464.
1669 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1670 under the control of the dsn_advertise_hosts option, and routers may
1671 have a dsn_lasthop option.
1673 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1674 default, modifiable by a malware= option. The list separator for
1675 the options can now be changed in the usual way. Bug 68.
1677 JH/10 The smtp_receive_timeout main option is now expanded before use.
1679 JH/11 The incoming_interface log option now also enables logging of the
1680 local interface on delivery outgoing connections.
1682 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1683 if the interface and destination host and port all match.
1685 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1688 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1689 Patch from Andrew Lewis.
1691 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1692 now supports optional time-restrictions, weighting, and priority
1693 modifiers per server. Patch originally by <rommer@active.by>.
1695 JH/16 The spamd_address main option now supports a mixed list of local
1696 and remote servers. Remote servers can be IPv6 addresses, and
1697 specify a port-range.
1699 JH/17 Bug 68: The spamd_address main option now supports an optional
1700 timeout value per server.
1702 JH/18 Bug 1581: Router and transport options headers_add/remove can
1703 now have the list separator specified.
1705 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1708 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1711 JH/21 Support for the A6 type of dns record is withdrawn.
1713 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1714 rather than the verbs used.
1716 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1717 from 255 to 1024 chars.
1719 JH/24 Verification callouts now attempt to use TLS by default.
1721 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1722 are generic router options now. The defaults didn't change.
1724 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1725 Original patch from Alexander Shikoff, worked over by JH.
1727 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1728 files if ZCAT_COMMAND is not executable.
1730 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1732 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1734 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1735 Normally benign, it bites when the pair was led to by a CNAME;
1736 modern usage is to not canonicalize the domain to a CNAME target
1737 (and we were inconsistent anyway for A-only vs AAAA+A).
1739 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1741 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1742 when evaluating $sender_host_dnssec.
1744 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1745 $sender_helo_dnssec variable.
1747 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1749 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1751 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1753 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1754 documented as working, but never had. Support all but $spam_report.
1756 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1757 added for tls authenticator.
1759 HS/03 Add perl_taintmode main config option
1765 TL/01 When running the test suite, the README says that variables such as
1766 no_msglog_check are global and can be placed anywhere in a specific
1767 test's script, however it was observed that placement needed to be near
1768 the beginning for it to behave that way. Changed the runtest perl
1769 script to read through the entire script once to detect and set these
1770 variables, reset to the beginning of the script, and then run through
1771 the script parsing/test process like normal.
1773 TL/02 The BSD's have an arc4random API. One of the functions to induce
1774 adding randomness was arc4random_stir(), but it has been removed in
1775 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1776 function when detected.
1778 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1779 cause callback expansion.
1781 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1782 syntax errors in an expansion can be treated as a string instead of
1783 logging or causing an error, due to the internal use of bool_lax
1784 instead of bool when processing it.
1786 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1787 server certificates when making smtp deliveries.
1789 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1791 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1793 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1795 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1796 Merged patch from Sebastian Wiedenroth.
1798 JH/05 Fix results-pipe from transport process. Several recipients, combined
1799 with certificate use, exposed issues where response data items split
1800 over buffer boundaries were not parsed properly. This eventually
1801 resulted in duplicates being sent. This issue only became common enough
1802 to notice due to the introduction of connection certificate information,
1803 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1805 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1806 size buffer was used, resulting in syntax errors when an expansion
1809 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1810 version 3.3.6 or later.
1812 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1813 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1814 both become "event_action", the variables become $event_name, $event_data
1815 and $event_defer_errno. There is a new variable $verify_mode, usable in
1816 routers, transports and related events. The tls:cert event is now also
1817 raised for inbound connections, if the main configuration event_action
1820 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1821 early OCSP support, but no stapling (appears to be less than 1.0.0).
1823 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1824 server certificate names available under the smtp transport option
1825 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1828 JH/10 Time-related extraction expansions from certificates now use the main
1829 option "timezone" setting for output formatting, and are consistent
1830 between OpenSSL and GnuTLS compilations. Bug 1541.
1832 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1833 encoded parameter in the incoming message. Bug 1558.
1835 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1836 include certificate info, eximon was claiming there were spoolfile
1839 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1841 JH/14 Log delivery-related information more consistently, using the sequence
1842 "H=<name> [<ip>]" wherever possible.
1844 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1845 are problematic for Debian distribution, omit them from the release
1848 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1850 JH/16 Fix string representation of time values on 64bit time_t architectures.
1853 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1854 output list separator was used.
1859 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1860 checkers that were complaining about end of non-void function with no
1863 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1864 This was a regression introduced in 4.83 by another bugfix.
1866 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1868 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1869 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1875 TF/01 Correctly close the server side of TLS when forking for delivery.
1877 When a message was received over SMTP with TLS, Exim failed to clear up
1878 the incoming connection properly after forking off the child process to
1879 deliver the message. In some situations the subsequent outgoing
1880 delivery connection happened to have the same fd number as the incoming
1881 connection previously had. Exim would try to use TLS and fail, logging
1882 a "Bad file descriptor" error.
1884 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1885 utilities have not been installed.
1887 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1888 temporary space as the ACL may create new global variables.
1890 TL/01 LDAP support uses per connection or global context settings, depending
1891 upon the detected version of the libraries at build time.
1893 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1894 to extract and use the src ip:port in logging and expansions as if it
1895 were a direct connection from the outside internet. PPv2 support was
1896 updated based on HAProxy spec change in May 2014.
1898 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1900 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1901 Properly escape header and check for NULL return.
1903 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1906 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1908 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1909 characters in header names, implemented as a verify condition.
1910 Contributed by Michael Fischer v. Mollard.
1912 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1913 results permerror and temperror. Previous values are deprecated but
1914 still accepted. In a future release, err_perm and err_temp will be
1915 completely removed, which will be a backward incompatibility if the
1916 ACL tests for either of these two old results. Patch contributed by
1917 user bes-internal on the mailing list.
1919 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1921 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1922 selectors, in both main and reject logs.
1924 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1927 JH/07 Add malware type "sock" for talking to simple daemon.
1929 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1931 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1932 routers/transports under cutthrough routing.
1934 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1935 numbers. Touch up "bool" conditional to keep the same definition.
1937 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1939 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1941 JH/12 Expand items in router/transport headers_add or headers_remove lists
1942 individually rather than the list as a whole. Bug 1452.
1944 Required for reasonable handling of multiple headers_ options when
1945 they may be empty; requires that headers_remove items with embedded
1946 colons must have them doubled (or the list-separator changed).
1948 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1949 view the policy declared in the DMARC record. Currently, $dmarc_status
1950 is a combined value of both the record presence and the result of the
1953 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1955 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1956 dnslookup router and the smtp transport (applying to the forward
1959 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1960 of ldap servers used for a specific lookup. Patch provided by Heiko
1963 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1964 New variable $lookup_dnssec_authenticated for observability.
1966 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1967 Patch submitted by Lars Timman.
1969 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1971 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1972 Requires trusted mode and valid format message id, aborts otherwise.
1973 Patch contributed by Heiko Schlichting.
1975 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1976 certextract with support for various fields. Bug 1358.
1978 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1979 is requested by default, modifiable by smtp transport option
1982 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1983 operate on certificate variables to give certificate fingerprints
1984 Also new ${sha256:cert_variable}.
1986 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1988 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1989 Christian Aistleitner.
1991 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1993 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1994 file. Patch from Wolfgang Breyha.
1996 JH/25 Expand the coverage of the delivery $host and $host_address to
1997 client authenticators run in verify callout. Bug 1476.
1999 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2000 align with daemon_smtp_ports. Bug 72.
2002 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2003 support and error reporting did not work properly.
2005 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2006 and is readable. Patch from Andrew Colin Kissa.
2008 TL/14 Enhance documentation of ${run expansion and how it parses the
2009 commandline after expansion, particularly in the case when an
2010 unquoted variable expansion results in an empty value.
2012 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2014 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2015 by renaming away from it. Addresses GH issue 15, reported by
2018 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2020 TL/15 SECURITY: prevent double expansion in math comparison functions
2021 (can expand unsanitized data). Not remotely exploitable.
2028 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2030 PP/02 Make -n do something, by making it not do something.
2031 When combined with -bP, the name of an option is not output.
2033 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2036 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2037 $sender_host_name and config options to manage this, and basic check
2040 PP/05 DSCP support for outbound connections and control modifier for inbound.
2042 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2043 (Only plugin which currently uses this is kerberos4, which nobody should
2044 be using, but we should make it available and other future plugins might
2045 conceivably use it, even though it would break NAT; stuff *should* be
2046 using channel bindings instead).
2048 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2049 name; added for Sendmail compatibility; requires admin caller.
2050 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2051 just ignore it); requires trusted caller.
2052 Also parse but ignore: -Ac -Am -X<logfile>
2055 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2057 TL/02 Add +smtp_confirmation as a default logging option.
2059 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2060 Patch by Magnus Holmgren from 2007-02-20.
2062 TL/04 Bugzilla 1281 - Spec typo.
2063 Bugzilla 1283 - Spec typo.
2064 Bugzilla 1290 - Spec grammar fixes.
2066 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2068 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2070 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2071 mailing list by by Dmitry Isaikin.
2073 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2075 JH/02 Support "G" suffix to numbers in ${if comparisons.
2077 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2079 NM/01 Bugzilla 1197 - Spec typo
2080 Bugzilla 1196 - Spec examples corrections
2082 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2084 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2085 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2088 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2089 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2091 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2092 "acl {{name}{arg}...}", and optional args on acl condition
2095 JH/05 Permit multiple router/transport headers_add/remove lines.
2097 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2099 JH/07 Avoid using a waiting database for a single-message-only transport.
2100 Performance patch from Paul Fisher. Bugzilla 1262.
2102 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2105 JH/09 Add $headers_added variable, with content from use of ACL modifier
2106 add_header (but not yet added to the message). Bugzilla 199.
2108 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2109 Pulled from Bugzilla 817 by Wolfgang Breyha.
2111 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2113 (nb: this is the same fix as in Exim 4.80.1)
2115 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2118 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2121 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2123 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2124 advertise SMTP AUTH mechanism to us, instead of a generic
2125 protocol violation error. Also, make Exim more robust to bad
2126 data from the Dovecot auth socket.
2128 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2130 When a queue runner is handling a message, Exim first routes the
2131 recipient addresses, during which it prunes them based on the retry
2132 hints database. After that it attempts to deliver the message to
2133 any remaining recipients. It then updates the hints database using
2136 So if a recipient address works intermittently, it can get repeatedly
2137 deferred at routing time. The retry hints record remains fresh so the
2138 address never reaches the final cutoff time.
2140 This is a fairly common occurrence when a user is bumping up against
2141 their storage quota. Exim had some logic in its local delivery code
2142 to deal with this. However it did not apply to per-recipient defers
2143 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2145 This change adds a proper retry rule check during routing so that the
2146 final cutoff time is checked against the message's age. We only do
2147 this check if there is an address retry record and there is not a
2148 domain retry record; this implies that previous attempts to handle
2149 the address had the retry_use_local_parts option turned on. We use
2150 this as an approximation for the destination being like a local
2151 delivery, as in LMTP.
2153 I suspect this new check makes the old local delivery cutoff check
2154 redundant, but I have not verified this so I left the code in place.
2156 TF/02 Correct gecos expansion when From: is a prefix of the username.
2158 Test 0254 submits a message to Exim with the header
2162 When I ran the test suite under the user fanf2, Exim expanded
2163 the header to contain my full name, whereas it should have added
2164 a Resent-Sender: header. It erroneously treats any prefix of the
2165 username as equal to the username.
2167 This change corrects that bug.
2169 GF/01 DCC debug and logging tidyup
2170 Error conditions log to paniclog rather than rejectlog.
2171 Debug lines prefixed by "DCC: " to remove any ambiguity.
2173 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2175 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2176 Bug spotted by Jeremy Harris; was flawed since initial commit.
2177 Would have resulted in OCSP responses post-SNI triggering an Exim
2178 NULL dereference and crash.
2180 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2182 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2183 Bug detection, analysis and fix by Samuel Thibault.
2184 Bugzilla 1331, Debian bug #698092.
2186 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2188 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2189 Server implementation by Todd Lyons, client by JH.
2190 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2191 config variable "prdr_enable" controls whether the server
2192 advertises the facility. If the client requests PRDR a new
2193 acl_data_smtp_prdr ACL is called once for each recipient, after
2194 the body content is received and before the acl_smtp_data ACL.
2195 The client is controlled by both of: a hosts_try_prdr option
2196 on the smtp transport, and the server advertisement.
2197 Default client logging of deliveries and rejections involving
2198 PRDR are flagged with the string "PRDR".
2200 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2201 fclose(). Diagnosis by Todd Lyons.
2203 PP/17 Update configure.default to handle IPv6 localhost better.
2204 Patch by Alain Williams (plus minor tweaks).
2207 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2208 This is now consistent with GnuTLS, and is now documented: the
2209 previous undocumented portable approach to treating the option as
2210 unset was to force an expansion failure. That still works, and
2211 an empty string is now equivalent.
2213 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2214 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2215 not performing validation itself.
2217 PP/20 Added force_command boolean option to pipe transport.
2218 Patch from Nick Koston, of cPanel Inc.
2220 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2223 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2225 PP/21 Fix eximon continuous updating with timestamped log-files.
2226 Broken in a format-string cleanup in 4.80, missed when I repaired the
2227 other false fix of the same issue.
2228 Report and fix from Heiko Schlichting.
2231 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2232 Report from Prashanth Katuri.
2234 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2235 It's SecureTransport, so affects any MacOS clients which use the
2236 system-integrated TLS libraries, including email clients.
2238 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2239 using a MIME ACL for non-SMTP local injection.
2240 Report and assistance in diagnosis by Warren Baker.
2242 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2244 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2246 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2247 last id that failed so it may be referenced in subsequent ACL's.
2249 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2252 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2253 ldap library initialization, allowing self-signed CA's to be
2254 used. Also properly sets require_cert option later in code by
2255 using NULL (global ldap config) instead of ldap handle (per
2256 session). Bug diagnosis and testing by alxgomz.
2258 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2259 the src/util/ subdirectory.
2261 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2262 renamed to Transport Post Delivery Action by Jeremy Harris, as
2265 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2266 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2267 redis_servers = needs to be configured which will be used by the redis
2268 lookup. Patch from Warren Baker, of The Packet Hub.
2270 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2272 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2273 hostname or reverse DNS when processing a host list. Used suggestions
2274 from multiple comments on this bug.
2276 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2278 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2279 Missed a few lines, added it to make the runtest require no keyboard
2282 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2283 contains upper case chars. Make router use caseful_local_part.
2285 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2286 support when GnuTLS has been built with p11-kit.
2292 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2294 This, or similar/improved, will also be change PP/11 of 4.82.
2300 PP/01 Handle short writes when writing local log-files.
2301 In practice, only affects FreeBSD (8 onwards).
2302 Bugzilla 1053, with thanks to Dmitry Isaikin.
2304 NM/01 Bugzilla 949 - Documentation tweak
2306 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2309 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2311 PP/02 Implemented gsasl authenticator.
2313 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2315 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2316 `pkg-config foo` for cflags/libs.
2318 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2319 with rest of GSASL and with heimdal_gssapi.
2321 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2322 `pkg-config foo` for cflags/libs for the TLS implementation.
2324 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2325 properties get this fed in as external SSF. A number of robustness
2326 and debugging improvements to the cyrus_sasl authenticator.
2328 PP/08 cyrus_sasl server now expands the server_realm option.
2330 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2331 Patch by Jeremy Harris.
2333 PP/10 Added dbmjz lookup type.
2335 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2337 PP/12 MAIL args handles TAB as well as SP, for better interop with
2338 non-compliant senders.
2339 Analysis and variant patch by Todd Lyons.
2341 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2342 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2343 Patch from Dirk Mueller <dmueller@suse.com>
2345 PP/13 tls_peerdn now print-escaped for spool files.
2346 Observed some $tls_peerdn in wild which contained \n, which resulted
2347 in spool file corruption.
2349 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2350 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2351 or write after TLS renegotiation, which otherwise led to messages
2354 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2355 as a tracking header (ie: a signed header comes before the signature).
2356 Patch from Wolfgang Breyha.
2358 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2359 comma-sep list; embedded commas doubled.
2361 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2363 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2365 Report and patch from Dmitry Banschikov.
2367 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2368 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2369 failures. We appear to now support TLS1.1+ with Exim.
2371 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2372 lets Exim select keys and certificates based upon TLS SNI from client.
2373 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2374 before an outbound SMTP session. New log_selector, +tls_sni.
2376 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2377 NULL dereference. Report and patch from Alun Jones.
2379 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2380 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2381 Not seeing resolver debug output on NetBSD, but suspect this is a
2382 resolver implementation change.
2384 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2385 Left warnings. Added "eximon gdb" invocation mode.
2387 PP/21 Defaulting "accept_8bitmime" to true, not false.
2389 PP/22 Added -bw for inetd wait mode support.
2391 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2392 locate the relevant includes and libraries. Made this the default.
2394 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2395 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2397 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2398 This may cause build issues on older platforms.
2400 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2401 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2402 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2403 Added SNI support via GnuTLS too.
2404 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2406 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2408 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2409 Applied second patch from Janne, implementing suggestion to default
2410 multiple-strings-in-record handling to match SPF spec.
2412 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2414 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2415 read-only, out of scope).
2416 Patch from Wolfgang Breyha, report from Stuart Northfield.
2418 PP/29 Fix three issues highlighted by clang analyser static analysis.
2419 Only crash-plausible issue would require the Cambridge-specific
2420 iplookup router and a misconfiguration.
2421 Report from Marcin Mirosław.
2423 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2425 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2426 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2427 As part of this, removing so much warning spew let me fix some minor
2428 real issues in debug logging.
2430 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2431 assignment on my part. Fixed.
2433 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2434 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2435 Janne Snabb (who went above and beyond: thank you).
2437 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2438 string otherwise requires a connection and a bunch more work and it's
2439 relatively easy to get wrong. Should also expose TLS library linkage
2442 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2443 64-bit ${eval} (JH/03).
2445 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2446 GNU libc to support some of the 64-bit stuff, should not lead to
2447 conflicts. Defined before os.h is pulled in, so if a given platform
2448 needs to override this, it can.
2450 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2451 protection layer was required, which is not implemented.
2452 Bugzilla 1254, patch from Wolfgang Breyha.
2454 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2455 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2456 tls_dhparam take prime identifiers. Also unbreak combination of
2457 OpenSSL+DH_params+TLSSNI.
2459 PP/39 Disable SSLv2 by default in OpenSSL support.
2465 PP/01 Solaris build fix for Oracle's LDAP libraries.
2466 Bugzilla 1109, patch from Stephen Usher.
2468 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2470 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2473 TF/02 Fix a couple more cases where we did not log the error message
2474 when unlink() failed. See also change 4.74-TF/03.
2476 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2477 lock up or crash if it happened to be inside a call to libc when it
2478 got a SIGUSR1 from exiwhat.
2480 The SIGUSR1 handler appends the current process status to the process
2481 log which is later printed by exiwhat. It used to use the general
2482 purpose logging code to do this, but several functions it calls are
2483 not safe for signals.
2485 The new output code in the SIGUSR1 handler is specific to the process
2486 log, and simple enough that it's easy to inspect for signal safety.
2487 Removing some special cases also simplifies the general logging code.
2488 Removing the spurious timestamps from the process log simplifies
2491 TF/04 Improved ratelimit ACL condition.
2493 The /noupdate option has been deprecated in favour of /readonly which
2494 has clearer semantics. The /leaky, /strict, and /readonly update modes
2495 are mutually exclusive. The update mode is no longer included in the
2496 database key; it just determines when the database is updated. (This
2497 means that when you upgrade Exim will forget old rate measurements.)
2499 Exim now checks that the per_* options are used with an update mode that
2500 makes sense for the current ACL. For example, when Exim is processing a
2501 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2502 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2503 must specify per_mail/readonly. If you omit the update mode it defaults to
2504 /leaky where that makes sense (as before) or /readonly where required.
2506 The /noupdate option is now undocumented but still supported for
2507 backwards compatibility. It is equivalent to /readonly except that in
2508 ACLs where /readonly is required you may specify /leaky/noupdate or
2509 /strict/noupdate which are treated the same as /readonly.
2511 A useful new feature is the /count= option. This is a generalization
2512 of the per_byte option, so that you can measure the throughput of other
2513 aggregate values. For example, the per_byte option is now equivalent
2514 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2516 The per_rcpt option has been generalized using the /count= mechanism
2517 (though it's more complicated than the per_byte equivalence). When it is
2518 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2519 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2520 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2521 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2522 Note that using per_rcpt with a non-readonly update mode in more than
2523 one ACL will cause the recipients to be double-counted. (The per_mail
2524 and per_byte options don't have this problem.)
2526 The handling of very low rates has changed slightly. If the computed rate
2527 is less than the event's count (usually one) then this event is the first
2528 after a long gap. In this case the rate is set to the same as this event's
2529 count, so that the first message of a spam run is counted properly.
2531 The major new feature is a mechanism for counting the rate of unique
2532 events. The new per_addr option counts the number of different
2533 recipients that someone has sent messages to in the last time period. It
2534 behaves like per_rcpt if all the recipient addresses are different, but
2535 duplicate recipient addresses do not increase the measured rate. Like
2536 the /count= option this is a general mechanism, so the per_addr option
2537 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2538 example, measure the rate that a client uses different sender addresses
2539 with the options per_mail/unique=$sender_address. There are further
2540 details in the main documentation.
2542 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2544 TF/06 Removed a few PCRE remnants.
2546 TF/07 Automatically extract Exim's version number from tags in the git
2547 repository when doing development or release builds.
2549 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2550 Bugzilla 879. Patch from Paul Fisher.
2552 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2553 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2556 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2558 PP/05 Variable $av_failed, true if the AV scanner deferred.
2559 Bugzilla 1078. Patch from John Horne.
2561 PP/06 Stop make process more reliably on build failure.
2562 Bugzilla 1087. Patch from Heiko Schlittermann.
2564 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2565 Bugzilla 1089. Patch from Heiko Schlittermann.
2567 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2568 Bugzilla 1131. Patch from Holger Weiß.
2570 PP/09 Handle IPv6 addresses with SPF.
2571 Bugzilla 860. Patch from Wolfgang Breyha.
2573 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2575 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2578 PP/11 match_* no longer expand right-hand-side by default.
2579 New compile-time build option, EXPAND_LISTMATCH_RHS.
2580 New expansion conditions, "inlist", "inlisti".
2582 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2584 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2586 PP/14 fix log_write() format string regression from TF/03.
2587 Bugzilla 1152. Patch from Dmitry Isaikin.
2593 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2595 PP/02 Harmonised TLS library version reporting; only show if debugging.
2596 Layout now matches that introduced for other libraries in 4.74 PP/03.
2598 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2600 PP/04 New "dns_use_edns0" global option.
2602 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2605 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2606 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2608 TK/01 Updated PolarSSL code to 0.14.2.
2609 Bugzilla 1097. Patch from Andreas Metzler.
2611 PP/07 Catch divide-by-zero in ${eval:...}.
2612 Fixes bugzilla 1102.
2614 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2617 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2618 format-string attack -- SECURITY: remote arbitrary code execution.
2620 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2621 time unintentionally subject to list matching rules, letting the header
2622 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2623 arbitrary string expansion). This allowed for information disclosure.
2625 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2626 INT_MIN/-1 -- value coerced to INT_MAX.
2632 NM/01 Workaround for PCRE version dependency in version reporting
2635 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2636 This fixes portability to compilers other than gcc, notably
2637 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2639 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2640 makefiles for portability to HP-UX and POSIX correctness.
2642 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2643 Also via indented variable definition in the Makefile.
2644 (Debugging by Oliver Heesakkers).
2646 PP/02 Restore caching of spamd results with expanded spamd_address.
2647 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2649 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2650 Improves build reliability. Fix from: Frank Elsner
2652 NM/02 Fix wide character breakage in the rfc2047 coding
2653 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2655 NM/03 Allow underscore in dnslist lookups
2656 Fixes bug 1026. Patch from Graeme Fowler
2658 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2659 Code patches from Adam Ciarcinski of NetBSD.
2661 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2664 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2665 is logged, to avoid truncation. Patch from John Horne.
2667 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2668 Patch from Jakob Hirsch.
2670 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2671 SQL string expansion failure details.
2672 Patch from Andrey Oktyabrski.
2674 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2675 Patch from Simon Arlott.
2677 PP/09 New lookups functionality failed to compile on old gcc which rejects
2678 extern declarations in function scope.
2679 Patch from Oliver Fleischmann
2681 PP/10 Use sig_atomic_t for flags set from signal handlers.
2682 Check getgroups() return and improve debugging.
2683 Fixed developed for diagnosis in bug 927 (which turned out to be
2686 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2687 Patch from Mark Zealey.
2689 PP/12 Bugzilla 1056: Improved spamd server selection.
2690 Patch from Mark Zealey.
2692 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2693 Based on patch from Heiko Schlittermann.
2695 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2696 Patch from Uwe Doering, sign-off by Michael Haardt.
2698 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2699 variable declaration deep within a block. Bug and patch from
2702 PP/15 lookups-Makefile IRIX compatibility coercion.
2704 PP/16 Make DISABLE_DKIM build knob functional.
2706 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2707 Patch by Simon Arlott
2709 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2710 variable argument macros. Our copy now differs from upstream.
2716 TF/01 Failure to get a lock on a hints database can have serious
2717 consequences so log it to the panic log.
2719 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2720 controlled using the smtp_confirmation log selector.
2722 TF/03 Include the error message when we fail to unlink a spool file.
2724 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2725 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2726 for maintaining out-of-tree patches for some time.
2728 PP/01 Bugzilla 139: Documentation and portability issues.
2729 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2730 Handle per-OS dynamic-module compilation flags.
2732 PP/02 Let /dev/null have normal permissions.
2733 The 4.73 fixes were a little too stringent and complained about the
2734 permissions on /dev/null. Exempt it from some checks.
2735 Reported by Andreas M. Kirchwitz.
2737 PP/03 Report version information for many libraries, including
2738 Exim version information for dynamically loaded libraries. Created
2739 version.h, now support a version extension string for distributors
2740 who patch heavily. Dynamic module ABI change.
2742 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2743 privilege escalation vulnerability whereby the Exim run-time user
2744 can cause root to append content of the attacker's choosing to
2747 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2750 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2751 If dropping privileges for untrusted macros, we disabled normal logging
2752 on the basis that it would fail; for the Exim run-time user, this is not
2753 the case, and it resulted in successful deliveries going unlogged.
2754 Fixed. Reported by Andreas Metzler.
2760 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2761 only prepend for the Resent-* case. Fixes regression introduced in
2762 Exim 4.70 by NM/22 for Bugzilla 607.
2764 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2765 increasing numbers of administrators be bitten by this.
2767 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2769 PP/03 Bugzilla 994: added openssl_options main configuration option.
2771 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2773 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2775 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2777 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2778 without a peer certificate, leading to a segfault because of an
2779 assumption that peers always have certificates. Be a little more
2780 paranoid. Problem reported by Martin Tscholak.
2782 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2783 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2784 NB: ClamAV planning to remove STREAM in "middle of 2010".
2785 CL also introduces -bmalware, various -d+acl logging additions and
2786 more caution in buffer sizes.
2788 PP/09 Implemented reverse_ip expansion operator.
2790 PP/10 Bugzilla 937: provide a "debug" ACL control.
2792 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2794 PP/12 Bugzilla 973: Implement --version.
2796 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2798 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2800 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2802 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2803 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2804 ignore trailing whitespace.
2806 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2808 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2811 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2812 Notification from Dr Andrew Aitchison.
2814 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2815 ExtendedDetectionInfo response format.
2816 Notification from John Horne.
2818 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2821 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2822 XSL and documented dependency on system catalogs, with examples of how
2825 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2828 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2829 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2830 configuration file which is writeable by the Exim user or group.
2832 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2833 of configuration files to cover files specified with the -C option if
2834 they are going to be used with root privileges, not just the default
2837 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2838 option (effectively making it always true).
2840 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2841 files to be used while preserving root privileges.
2843 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2844 that rogue child processes cannot use them.
2846 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2847 run-time user, instead of root.
2849 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2850 Exim run-time user without dropping privileges.
2852 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2853 result string, instead of calling string_vformat() twice with the same
2856 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2857 for other users. Others should always drop root privileges if they use
2858 -C on the command line, even for a whitelisted configure file.
2860 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2862 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2868 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2869 $data_path, and $header_path variables; fixed documentation bugs and
2872 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2873 exipick to access non-standard spools, including the "frozen" queue
2876 NM/01 Bugzilla 965: Support mysql stored procedures.
2877 Patch from Alain Williams
2879 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2881 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2882 Patch from Andreas Metzler
2884 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2885 Patch from Kirill Miazine
2887 NM/05 Bugzilla 671: Added umask to procmail example.
2889 JJ/03 installed exipick 20100323.0, fixing doc bug
2891 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2892 directory. Notification and patch from Dan Rosenberg.
2894 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2896 TK/02 Improve log output when DKIM signing operation fails.
2898 MH/01 Treat the transport option dkim_domain as a colon separated
2899 list, not as a single string, and sign the message with each element,
2900 omitting multiple occurences of the same signer.
2902 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2903 Bugzilla 985, 986. Patch by Simon Arlott
2905 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2906 Patch by Simon Arlott
2908 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2909 MBX locking. Notification from Dan Rosenberg.
2915 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2917 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2919 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2921 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2923 NM/04 Bugzilla 915: Flag broken perl installation during build.
2929 TK/01 Added patch by Johannes Berg that expands the main option
2930 "spamd_address" if it starts with a dollar sign.
2932 TK/02 Write list of recipients to X-Envelope-Sender header when building
2933 the mbox-format spool file for content scanning (suggested by Jakob
2936 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2937 (http://www.dcc-servers.net/) support via dccifd. Activated by
2938 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2940 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2941 by Mark Daniel Reidel <mr@df.eu>.
2943 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2944 When building exim an external PCRE library is now needed -
2945 PCRE is a system library on the majority of modern systems.
2946 See entry on PCRE_LIBS in EDITME file.
2948 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2949 conversation. Added nologin parameter to request.
2950 Patch contributed by Kirill Miazine.
2952 TF/01 Do not log submission mode rewrites if they do not change the address.
2954 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2956 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2957 log files in place. Contributed by Roberto Lima.
2959 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2961 TF/03 Bugzilla 615: When checking the local_parts router precondition
2962 after a local_part_suffix or local_part_prefix option, Exim now
2963 does not use the address's named list lookup cache, since this
2964 contains cached lookups for the whole local part.
2966 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2967 Robert Millan. Documentation is in experimental-spec.txt.
2969 TF/04 Bugzilla 668: Fix parallel build (make -j).
2971 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2973 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2974 Patch provided by Jan Srzednicki.
2976 TF/05 Leading white space used to be stripped from $spam_report which
2977 wrecked the formatting. Now it is preserved.
2979 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2980 that they are available at delivery time.
2982 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2984 TF/08 TLS error reporting now respects the incoming_interface and
2985 incoming_port log selectors.
2987 TF/09 Produce a more useful error message if an SMTP transport's hosts
2988 setting expands to an empty string.
2990 NM/06 Bugzilla 744: EXPN did not work under TLS.
2991 Patch provided by Phil Pennock.
2993 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2994 Patch provided by Richard Godbee.
2996 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2997 acl_smtp_notquit, added index entry.
2999 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3000 Patch provided by Eugene Bujak.
3002 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3003 accept(). Patch provided by Maxim Dounin.
3005 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3006 Patch provided by Phil Pennock.
3008 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3010 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3011 Patch provided by Brad "anomie" Jorsch.
3013 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3014 Patch provided by Dean Brooks.
3016 TK/05 Add native DKIM support (does not depend on external libraries).
3018 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3019 Patch provided by Graeme Fowler.
3021 NM/16 Bugzilla 851: Documentation example syntax fix.
3023 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3025 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3028 NM/19 Bugzilla 745: TLS version reporting.
3029 Patch provided by Phil Pennock.
3031 NM/20 Bugzilla 167: bool: condition support.
3032 Patch provided by Phil Pennock.
3034 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3035 clients. Patch provided by Phil Pennock.
3037 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3038 Patch provided by Brad "anomie" Jorsch.
3040 NM/23 Bugzilla 687: Fix misparses in eximstats.
3041 Patch provided by Heiko Schlittermann.
3043 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3044 Patch provided by Heiko Schlittermann.
3046 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3047 plus update to original patch.
3049 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3051 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3052 Patch provided by David Brownlee.
3054 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3056 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3058 NM/30 Bugzilla 888: TLS documentation bugfixes.
3060 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3062 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3063 Unlike the original bugzilla I have changed all shell scripts in src tree.
3065 NM/33 Bugzilla 898: Transport filter timeout fix.
3066 Patch by Todd Rinaldo.
3068 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3069 Patch by Serge Demonchaux.
3071 NM/35 Bugzilla 39: Base64 decode bug fixes.
3072 Patch by Jakob Hirsch.
3074 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3076 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3078 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3080 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3086 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3087 ALT-N's libdkim that I have put here:
3088 http://duncanthrax.net/exim-experimental/
3090 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3091 were called 'true' and it seems that C99 defines that as a reserved
3092 keyword to be used with 'bool' variable types. That means you could
3093 not include C99-style headers which use bools without triggering
3094 build errors in sieve.c.
3096 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3097 as mailq or other aliases. Changed the --help handling significantly
3098 to do whats expected. exim_usage() emits usage/help information.
3100 SC/01 Added the -bylocaldomain option to eximstats.
3102 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3104 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3106 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3112 PH/01 Another patch from the Sieve maintainer.
3114 PH/02 When an IPv6 address is converted to a string for single-key lookup
3115 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3116 dots are used instead of colons so that keys in lsearch files need not
3117 contain colons. This was done some time before quoting was made available
3118 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3119 (notated using the quote facility) so as to distinguish them from IPv4
3120 keys. This meant that lookups for IP addresses in host lists did not work
3121 for iplsearch lookups.
3123 This has been fixed by arranging for IPv6 addresses to be expressed with
3124 colons if the lookup type is iplsearch. This is not incompatible, because
3125 previously such lookups could never work.
3127 The situation is now rather anomalous, since one *can* have colons in
3128 ordinary lsearch keys. However, making the change in all cases is
3129 incompatible and would probably break a number of configurations.
3131 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3134 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3135 conversion specification without a maximum field width, thereby enabling
3136 a rogue spamd server to cause a buffer overflow. While nobody in their
3137 right mind would setup Exim to query an untrusted spamd server, an
3138 attacker that gains access to a server running spamd could potentially
3139 exploit this vulnerability to run arbitrary code as the Exim user.
3141 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3142 $primary_hostname instead of what libspf2 thinks the hosts name is.
3144 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3145 a directory entry by the name of the lookup key. Previously, if a
3146 symlink pointed to a non-existing file or a file in a directory that
3147 Exim lacked permissions to read, a lookup for a key matching that
3148 symlink would fail. Now it is enough that a matching directory entry
3149 exists, symlink or not. (Bugzilla 503.)
3151 PH/03 The body_linecount and body_zerocount variables are now exported in the
3154 PH/04 Added the $dnslist_matched variable.
3156 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3157 This means they are set thereafter only if the connection becomes
3160 PH/06 Added the client_condition to authenticators so that some can be skipped
3161 by clients under certain conditions.
3163 PH/07 The error message for a badly-placed control=no_multiline_responses left
3164 "_responses" off the end of the name.
3166 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3168 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3169 (without spaces) instead of just copying the configuration text.
3171 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3173 PH/11 Added $max_received_linelength.
3175 PH/12 Added +ignore_defer and +include_defer to host lists.
3177 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3178 way in which PCRE > 7.0 is built.
3180 PH/14 Implemented queue_only_load_latch.
3182 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3183 MAIL command. The effect was to mangle the value on 64-bit systems.
3185 PH/16 Another patch from the Sieve maintainer.
3187 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3189 PH/18 If a system quota error occurred while trying to create the file for
3190 a maildir delivery, the message "Mailbox is full" was not appended to the
3191 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3192 only to a quota excession during the actual writing of the file.
3194 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3195 characters?) which causes problems in log lines. The DN values are now
3196 passed through string_printing() before being added to log lines.
3198 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3199 and InterBase are left for another time.)
3201 PH/21 Added message_body_newlines option.
3203 PH/22 Guard against possible overflow in moan_check_errorcopy().
3205 PH/23 POSIX allows open() to be a macro; guard against that.
3207 PH/24 If the recipient of an error message contained an @ in the local part
3208 (suitably quoted, of course), incorrect values were put in $domain and
3209 $local_part during the evaluation of errors_copy.
3215 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3216 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3219 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3220 issue a MAIL command.
3222 PH/02 In an ACL statement such as
3224 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3226 if a client was not listed at all, or was listed with a value other than
3227 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3228 the condition was not true (as it should be), so access was not denied.
3229 The bug was that the ! inversion was incorrectly passed on to the second
3230 item. This has been fixed.
3232 PH/03 Added additional dnslists conditions == and =& which are different from
3233 = and & when the dns lookup returns more than one IP address.
3235 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3236 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3238 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3239 FSYNC, which compiles an option called disable_fsync that allows for
3240 bypassing fsync(). The documentation is heavily laced with warnings.
3242 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3244 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3245 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3246 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3247 including adding "make clean"; (3) Added -fPIC when compiling the test
3248 dynamically loaded module, to get rid of a warning.
3250 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3251 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3252 = 0s. The bug is otherwise harmless.
3254 PH/07 There was a bug in the dovecot authenticator such that the value of
3255 $auth1 could be overwritten, and so not correctly preserved, after a
3256 successful authentication. This usually meant that the value preserved by
3257 the server_setid option was incorrect.
3259 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3261 PH/09 Installed PCRE release 7.0.
3263 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3264 run for batched SMTP input. It is now run at the start of every message
3265 in the batch. While fixing this I discovered that the process information
3266 (output by running exiwhat) was not always getting set for -bs and -bS
3267 input. This is fixed, and it now also says "batched" for BSMTP.
3269 PH/11 Added control=no_pipelining.
3271 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3272 patch, slightly modified), and move the expansion of helo_data till after
3273 the connection is made in the smtp transport (so it can use these
3276 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3278 PH/14 Added log_selector = +pid.
3280 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3282 PH/16 Add ${if forany and ${if forall.
3284 PH/17 Added dsn_from option to vary the From: line in DSNs.
3286 PH/18 Flush SMTP output before performing a callout, unless control =
3287 no_callout_flush is set.
3289 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3290 was true (the default) a successful delivery failed to delete the retry
3291 item, thus causing premature timeout of the address. The bug is now
3294 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3296 PH/21 Long custom messages for fakedefer and fakereject are now split up
3297 into multiline responses in the same way that messages for "deny" and
3298 other ACL rejections are.
3300 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3301 with slight modification.
3303 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3304 draft, changing the syntax and factoring some duplicate code".
3306 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3307 for deliveries of the second and subsequent messages over the same SMTP
3310 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3311 ${reduce, with only minor "tidies".
3313 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3315 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3316 expansion side effects.
3318 PH/27 When a message times out after an over-quota error from an Exim-imposed
3319 quota, the bounce message says "mailbox is full". This message was not
3320 being given when it was a system quota that was exceeded. It now should
3323 MH/03 Made $recipients available in local_scan(). local_scan() already has
3324 better access to the recipient list through recipients_list[], but
3325 $recipients can be useful in postmaster-provided expansion strings.
3327 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3328 in the case of a MAIL command with additional options following the
3329 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3330 were accidentally chopped off.
3332 PH/29 SMTP synchronization checks are implemented when a command is read -
3333 there is a check that no more input is waiting when there shouldn't be
3334 any. However, for some commands, a delay in an ACL can mean that it is
3335 some time before the response is written. In this time, more input might
3336 arrive, invalidly. So now there are extra checks after an ACL has run for
3337 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3338 pipelining has not been advertised.
3340 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3342 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3343 returned in the SMTP response when smtp_return_error_details was false.
3344 This has been fixed.
3346 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3347 instead of the C I/O that was originally supplied, because problems were
3348 reported on Solaris.
3350 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3351 Exim which did not show up earlier: it was assuming that a call to
3352 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3353 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3354 was a macro that became an assignment, so it seemed to work. This has
3355 changed to a proper function call with a void return, hence the compile
3356 error. Exim's code has been fixed.
3358 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3361 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3363 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3365 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3366 because it turns out to be too big or there is a timeout) there is no
3367 "Completed" line in the log. When some messages of this type were
3368 selected by exigrep, they were listed as "not completed". Others were
3369 picked up by some special patterns. I have improved the selection
3370 criteria to be more general.
3372 PH/38 The host_find_failed option in the manualroute router can now be set
3373 to "ignore", to completely ignore a host whose IP address cannot be
3374 found. If all hosts are ignored, the behaviour is controlled by the new
3375 host_all_ignored option.
3377 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3378 homing or because of multiple MX records with /mx) generated more than
3379 one IP address, and the following item turned out to be the local host,
3380 all the secondary addresses of the first item were incorrectly removed
3381 from the list, along with the local host and any following hosts (which
3382 is what is supposed to happen).
3384 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3385 whoever called Exim into the -H file. In the case of the daemon it was
3386 behaving confusingly. When first started, it used values for whoever
3387 started the daemon, but after a SIGHUP it used the Exim user (because it
3388 calls itself on a restart). I have changed the code so that it now always
3391 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3392 message are rejected with the same error (e.g. no authentication or bad
3393 sender address), and a DATA command is nevertheless sent (as can happen
3394 with PIPELINING or a stupid MUA), the error message that was given to the
3395 RCPT commands is included in the rejection of the DATA command. This is
3396 intended to be helpful for MUAs that show only the final error to their
3399 PH/42 Another patch from the Sieve maintainer.
3401 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3402 Eximstats - Fixed some broken HTML links and added missing column headers
3404 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3405 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3407 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3409 PH/43 Yet another patch from the Sieve maintainer.
3411 PH/44 I found a way to check for a TCP/IP connection going away before sending
3412 the response to the final '.' that terminates a message, but only in the
3413 case where the client has not sent further data following the '.'
3414 (unfortunately, this is allowed). However, in many cases there won't be
3415 any further data because there won't be any more messages to send. A call
3416 to select() can be used: if it shows that the input is "ready", there is
3417 either input waiting, or the socket has been closed. An attempt to read
3418 the next input character can distinguish the two cases. Previously, Exim
3419 would have sent an OK response which the client would never have see.
3420 This could lead to message repetition. This fix should cure that, at
3421 least in a lot of common cases.
3423 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3424 advertised in response to EHLO.
3430 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3431 fixed by 4.65/MH/01 (is this a record?) are fixed:
3433 (i) An empty string was always treated as zero by the numeric comparison
3434 operators. This behaviour has been restored.
3436 (ii) It is documented that the numeric comparison operators always treat
3437 their arguments as decimal numbers. This was broken in that numbers
3438 starting with 0 were being interpreted as octal.
3440 While fixing these problems I realized that there was another issue that
3441 hadn't been noticed. Values of message_size_limit (both the global option
3442 and the transport option) were treated as octal if they started with 0.
3443 The documentation was vague. These values are now always treated as
3444 decimal, and I will make that clear in the documentation.
3450 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3451 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3454 MH/01 Don't check that the operands of numeric comparison operators are
3455 integers when their expansion is in "skipping" mode (fixes bug
3456 introduced by 4.64-PH/07).
3458 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3459 child addresses, Exim now panics and dies. Previously, because the count
3460 is held in a short int, deliveries were likely to be lost. As such a
3461 large number of recipients for a single message is ridiculous
3462 (performance will be very, very poor), I have chosen to impose a limit
3463 rather than extend the field.
3469 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3470 leftover -K file (the existence of which was triggered by #402).
3471 While we were at it, introduced process PID as part of the -K
3472 filename. This should rule out race conditions when creating
3475 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3476 processing considerably. Previous code took too long for large mails,
3477 triggering a timeout which in turn triggers #401.
3479 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3480 in the DK code in transports.c. sendfile() is not really portable,
3481 hence the _LINUX specificness.
3483 TF/01 In the add_headers option to the mail command in an Exim filter,
3484 there was a bug that Exim would claim a syntax error in any
3485 header after the first one which had an odd number of characters
3488 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3489 callout verification, Exim cached a "reject" for the entire domain. This
3490 is correct for most verifications, but it is not correct for a recipient
3491 verification with use_sender or use_postmaster set, because in that case
3492 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3493 case of MAIL FROM:<> rejection from other early rejections (e.g.
3494 rejection of HELO). When verifying a recipient using a non-null MAIL
3495 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3496 Whatever the result of the callout, the value of the domain cache is
3497 left unchanged (for any other kind of callout, getting as far as trying
3498 RCPT means that the domain itself is ok).
3500 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3503 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3504 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3505 session, the daemon ignores ECONNECT errors and logs others; it now
3506 ignores EPIPE as well.
3508 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3509 (quoted-printable decoding).
3511 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3512 later the small subsequent patch to fix an introduced bug.
3514 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3516 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3518 PH/08 An error is now given if message_size_limit is specified negative.
3520 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3521 to be given (somewhat) arbitrary names.
3523 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3526 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3527 miscellaneous code fixes
3529 PH/10 Added the log_reject_target ACL modifier to specify where to log
3532 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3533 hostname. This is wrong, because it relates to the incoming message (and
3534 probably the interface on which it is arriving) and not to the outgoing
3535 callout (which could be using a different interface). This has been
3536 changed to use the value of the helo_data option from the smtp transport
3537 instead - this is what is used when a message is actually being sent. If
3538 there is no remote transport (possible with a router that sets up host
3539 addresses), $smtp_active_hostname is used.
3541 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3542 tweaks were necessary in order to get it to work (see also 21 below):
3543 (a) The code assumed that strncpy() returns a negative number on buffer
3544 overflow, which isn't the case. Replaced with Exim's string_format()
3546 (b) There were several signed/unsigned issues. I just did the minimum
3547 hacking in of casts. There is scope for a larger refactoring.
3548 (c) The code used strcasecmp() which is not a standard C function.
3549 Replaced with Exim's strcmpic() function.
3550 (d) The code set only $1; it now sets $auth1 as well.
3551 (e) A simple test gave the error "authentication client didn't specify
3552 service in request". It would seem that Dovecot has changed its
3553 interface. Fortunately there's a specification; I followed it and
3554 changed what the client sends and it appears to be working now.
3556 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3559 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3560 address A is aliased to B and C, where B exists and C does not. Without
3561 -v the output is "A verified" because verification stops after a
3562 successful redirection if more than one address is generated. However,
3563 with -v the child addresses are also verified. Exim was outputting "A
3564 failed to verify" and then showing the successful verification for C,
3565 with its parentage. It now outputs "B failed to verify", showing B's
3566 parentage before showing the successful verification of C.
3568 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3569 look up a TXT record in a specific list after matching in a combined
3572 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3573 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3574 they consult the DNS. I had assumed they would set it the way they
3575 wanted; and indeed my experiments on Linux seem to show that in some
3576 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3577 To be on the safe side, however, I have now made the interface to
3578 host_find_byname() similar to host_find_bydns(), with an argument
3579 containing the DNS resolver options. The host_find_byname() function now
3580 sets these options at its start, just as host_find_bydns() does. The smtp
3581 transport options dns_qualify_single and dns_search_parents are passed to
3582 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3583 of host_find_byname() use the default settings of RES_DEFNAMES
3584 (qualify_single) but not RES_DNSRCH (search_parents).
3586 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3587 spool_read_header() do less string testing, by means of a preliminary
3588 switch on the second character of optional "-foo" lines. (This is
3589 overdue, caused by the large number of possibilities that now exist.
3590 Originally there were few.) While I was there, I also converted the
3591 str(n)cmp tests so they don't re-test the leading "-" and the first
3592 character, in the hope this might squeeze out yet more improvement.
3594 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3595 flag allowing group syntax was set by the header_syntax check but not
3596 turned off, possible causing trouble later; (2) The flag was not being
3597 set at all for the header_verify test, causing "group"-style headers to
3598 be rejected. I have now set it in this case, and also caused header_
3599 verify to ignore an empty address taken from a group. While doing this, I
3600 came across some other cases where the code for allowing group syntax
3601 while scanning a header line wasn't quite right (mostly, not resetting
3602 the flag correctly in the right place). These bugs could have caused
3603 trouble for malformed header lines. I hope it is now all correct.
3605 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3606 with the "reply" argument non-NULL. The code, however (which originally
3607 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3608 but it didn't always do it. This confused somebody who was copying the
3609 code for some other use. I have removed all the tests.
3611 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3612 feature that was used to support insecure browsers during the U.S. crypto
3613 embargo. It requires special client support, and Exim is probably the
3614 only MTA that supported it -- and would never use it because real RSA is
3615 always available. This code has been removed, because it had the bad
3616 effect of slowing Exim down by computing (never used) parameters for the
3617 RSA_EXPORT functionality.
3619 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3620 authenticator to fail if there's a tab character in the incoming data
3621 (there should never be unless someone is messing about, as it's supposed
3622 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3623 if the connection is using TLS or if the remote IP is the same as the
3624 local IP, and the "valid-client-cert option" if a client certificate has
3627 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3628 authenticators. This can be used for authorization after authentication
3629 succeeds. (In the case of plaintext, it servers for both authentication
3632 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3633 if any retry times were supplied.
3635 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3636 connection, where there is no client IP address to check. In this
3637 situation, the verify now always succeeds.
3639 PH/25 Applied John Jetmore's -Mset patch.
3641 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3643 PH/27 In a string expansion for a processed (not raw) header when multiple
3644 headers of the same name were present, leading whitespace was being
3645 removed from all of them, but trailing whitespace was being removed only
3646 from the last one. Now trailing whitespace is removed from each header
3647 before concatenation. Completely empty headers in a concatenation (as
3648 before) are ignored.
3650 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3651 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3653 PH/29 [Removed. This was a change that I later backed out, and forgot to
3654 correct the ChangeLog entry (that I had efficiently created) before
3655 committing the later change.]
3657 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3658 address errors (4xx response to RCPT) over the same connection as other
3659 messages routed to the same hosts. Such deliveries are always "forced",
3660 so retry times are not inspected. This resulted in far too many retries
3661 for the affected addresses. The effect occurred only when there were more
3662 hosts than the hosts_max_try setting in the smtp transport when it had
3663 the 4xx errors. Those hosts that it had tried were not added to the list
3664 of hosts for which the message was waiting, so if all were tried, there
3665 was no problem. Two fixes have been applied:
3667 (i) If there are any address or message errors in an SMTP delivery, none
3668 of the hosts (tried or untried) are now added to the list of hosts
3669 for which the message is waiting, so the message should not be a
3670 candidate for sending over the same connection that was used for a
3671 successful delivery of some other message. This seems entirely
3672 reasonable: after all the message is NOT "waiting for some host".
3673 This is so "obvious" that I'm not sure why it wasn't done
3674 previously. Hope I haven't missed anything, but it can't do any
3675 harm, as the worst effect is to miss an optimization.
3677 (ii) If, despite (i), such a delivery is accidentally attempted, the
3678 routing retry time is respected, so at least it doesn't keep
3679 hammering the server.
3681 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3682 in ${readsocket because some servers need this prod.
3684 PH/32 Added some extra debug output when updating a wait-xxx database.
3686 PH/33 The hint "could be header name not terminated by colon", which has been
3687 given for certain expansion errors for a long time, was not being given
3688 for the ${if def:h_colon_omitted{... case.
3690 PH/34 The spec says: "With one important exception, whenever a domain list is
3691 being scanned, $domain contains the subject domain." There was at least
3692 one case where this was not true.
3694 PH/35 The error "getsockname() failed: connection reset by peer" was being
3695 written to the panic log as well as the main log, but it isn't really
3696 panic-worthy as it just means the connection died rather early on. I have
3697 removed the panic log writing for the ECONNRESET error when getsockname()
3700 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3701 runs only) independently of the message's sender address. This meant
3702 that, if the 4xx error was in fact related to the sender, a different
3703 message to the same recipient with a different sender could confuse
3704 things. In particular, this can happen when sending to a greylisting
3705 server, but other circumstances could also provoke similar problems.
3706 I have changed the default so that the retry time for these errors is now
3707 based a combination of the sender and recipient addresses. This change
3708 can be overridden by setting address_retry_include_sender=false in the
3711 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3712 remote server are returned as part of bounce messages. This was not
3713 happening for LMTP over a pipe (the lmtp transport), but now it is the
3714 same for both kinds of LMTP.
3716 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3717 in header lines that were in fact CNAMEs. This is no longer the case.
3719 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3720 and queue runs started by the daemon processed all messages. This has
3721 been fixed so that -R and -S can now usefully be given with -q<time>.
3723 PH/40 Import PCRE release 6.7 (fixes some bugs).
3725 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3727 PH/42 Give an error if -q is specified more than once.
3729 PH/43 Renamed the variables $interface_address and $interface_port as
3730 $received_ip_address and $received_port, to make it clear that these
3731 values apply to message reception, and not to the outgoing interface when
3732 a message is delivered. (The old names remain recognized, of course.)
3734 PH/44 There was no timeout on the connect() call when using a Unix domain
3735 socket in the ${readsocket expansion. There now is.
3737 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3738 be meaningful with "accept".
3740 SC/01 Eximstats V1.43
3741 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3743 SC/02 Eximstats V1.44
3744 Use a glob alias rather than an array ref in the generated
3745 parser. This improves both readability and performance.
3747 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3748 Collect SpamAssassin and rejection statistics.
3749 Don't display local sender or destination tables unless
3750 there is data to show.
3751 Added average volumes into the top table text output.
3753 SC/04 Eximstats V1.46
3754 Collect data on the number of addresses (recipients)
3755 as well as the number of messages.
3757 SC/05 Eximstats V1.47
3758 Added 'Message too big' to the list of mail rejection
3759 reasons (thanks to Marco Gaiarin).
3761 SC/06 Eximstats V1.48
3762 Mainlog lines which have GMT offsets and are too short to
3763 have a flag are now skipped.
3765 SC/07 Eximstats V1.49 (Alain Williams)
3766 Added the -emptyok flag.
3768 SC/08 Eximstats V1.50
3769 Fixes for obtaining the IP address from reject messages.
3771 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3772 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3773 whitespace changes from 4.64-PH/27
3775 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3778 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3779 are found, allow negative numbers in numeric criteria)
3781 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3783 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3784 to match changes made in 4.64-PH/43
3786 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3788 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3789 "domain" is actually an IP address, and force a failure. This locks out
3790 those revolvers/nameservers that support "A-for-A" lookups, in
3791 contravention of the specifications.
3793 PH/48 When a host name was looked up from an IP address, and the subsequent
3794 forward lookup of the name timed out, the host name was left in
3795 $sender_host_name, contrary to the specification.
3797 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3798 restricted to single-key lookups, Exim was not diagnosing an error if
3799 * or *@ was used with a query-style lookup.
3801 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3803 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3804 long ago, but noone interested enough thought of it. Let's just say that
3805 the "1.1" means that there are some new functions that weren't there at
3806 some point in the past.
3808 PH/51 Error processing for expansion failure of helo_data from an smtp
3809 transport during callout processing was broken.
3811 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3812 tested/used via the -bh/-bhc/-bs options.
3814 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3815 bug, fixed in subsequent PCRE releases).
3817 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3818 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3820 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3826 SC/01 Use a glob alias rather than an array ref in eximstats generated
3827 parser. This improves both readability and performance.
3829 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3830 Don't display local sender or destination tables in eximstats unless
3831 there is data to show.
3832 Added average volumes into the eximstats top table text output.
3834 SC/03 Collect data on the number of addresses (recipients) as well
3835 as the number of messages in eximstats.
3837 TF/01 Correct an error in the documentation for the redirect router. Exim
3838 does (usually) call initgroups() when daemonizing.
3840 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3841 with consistent privilege compared to when running as a daemon.
3843 TF/03 Note in the spec that $authenticated_id is not set for local
3844 submissions from trusted users.
3846 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3847 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3849 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3850 by adding some example configuration directives to the default
3851 configuration file. A little bit of work is required to uncomment the
3852 directives and define how usernames and passwords are checked, but
3853 there is now a framework to start from.
3855 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3856 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3857 without this. I don't know how relevant this is to other LDAP libraries.
3859 PH/02 Add the verb name to the "unknown ACL verb" error.
3861 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3863 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3865 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3866 directory not expanded when it should be if an expanded home directory
3867 was set for the address (which is overridden by the transport).
3869 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3872 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3873 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3874 because it is too late at that time, and has no effect.
3876 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3877 security issue with \' (bugzilla #107). I could not use the
3878 PQescapeStringConn() function, because it needs a PGconn value as one of
3881 PH/08 When testing addresses using -bt, indicate those final addresses that
3882 are duplicates that would not cause an additional delivery. At least one
3883 person was confused, thinking that -bt output corresponded to deliveries.
3884 (Suppressing duplicates isn't a good idea as you lose the information
3885 about possibly different redirections that led to the duplicates.)
3887 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3888 systems where poll() doesn't work, in particular OS X.
3890 PH/10 Added more information to debugging output for retry time not reached.
3892 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3893 operations in malware.c.
3895 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3898 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3899 syslog_duplication set false, log lines that would normally be written
3900 both the the main log and to the reject log were not written to syslog at
3903 PH/14 In the default configuration, change the use of "message" in ACL warn
3904 statements to "add_header".
3906 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3907 not followed by a command (e.g. "seen endif").
3909 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3910 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3913 PH/17 Added extra conditions to the default value of delay_warning_condition
3917 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3918 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3919 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3922 The Auto-Submitted: and various List- headers are standardised, whereas I
3923 don't think Precedence: ever was.
3925 PH/18 Refactored debugging code in route_finduser() to show more information,
3926 in particular, the error code if getpwnam() issues one.
3928 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3929 This is apparently needed in addition to the PH/07 change above to avoid
3930 any possible encoding problems.
3932 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3933 but not after initializing Perl.
3935 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3936 output them only if debugging. By default they are written stderr,
3937 apparently, which is not desirable.
3939 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3942 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3945 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3947 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3948 authenticated or an ident call has been made. Suppress the default
3949 values for $authenticated_id and $authenticated_sender (but permit -oMai
3950 and -oMas) when testing with -bh.
3952 PH/24 Re-jigged the order of the tests in the default configuration so that the
3953 tests for valid domains and recipients precede the DNS black list and CSA
3954 tests, on the grounds that those ones are more expensive.
3956 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3957 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3958 This bug exists in every version of Exim that I still have, right back to
3961 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3962 However, an attempt to turn on case-sensitivity in a regex key by
3963 including (?-i) didn't work because the subject string was already
3964 lowercased, and the effects were non-intuitive. It turns out that a
3965 one-line patch can be used to allow (?-i) to work as expected.
3971 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3972 other effects) broke the use of negated acl sub-conditions.
3974 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3977 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3978 "Deny" causes Exim to reject the incoming connection with a 554 error.
3979 Unfortunately, if there is a major crisis, such as a disk failure,
3980 tcp-wrappers gives "deny", whereas what one would like would be some
3981 kind of temporary error. A kludge has been added to help with this.
3982 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3983 554 error is used if errno is still zero or contains ENOENT (which occurs
3984 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3987 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3989 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3990 errors. Otherwise a message that provokes a temporary error (when other
3991 messages do not) can cause a whole host to time out.
3993 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3994 addresses were routed directly to files or pipes from a redirect router.
3995 File deliveries just didn't batch; pipe deliveries might have suffered
3998 PH/06 A failure to get a lock for a hints database would erroneously always say
3999 "Failed to get write lock", even when it was really a read lock.
4001 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4002 of 0600. This has been changed to use the value of the lockfile_mode
4003 option (which defaults to 0600).
4005 PH/08 Applied small patch from the Sieve maintainer.
4007 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4008 folder from quota calculations, a direct delivery into this folder messed
4009 up the contents of the maildirsize file. This was because the regex was
4010 used only to exclude .Trash (or whatever) when the size of the mailbox
4011 was calculated. There was no check that a delivery was happening into an
4012 excluded directory. This bug has been fixed by ignoring all quota
4013 processing for deliveries into excluded directories.
4015 PH/10 Added the maildirfolder_create_regex option to appendfile.
4021 PH/01 The code for finding all the local interface addresses on a FreeBSD
4022 system running IPv6 was broken. This may well have applied to all BSD
4023 systems, as well as to others that have similar system calls. The broken
4024 code found IPv4 interfaces correctly, but gave incorrect values for the
4025 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4026 that it would not match correctly against @[] and not recognize the IPv6
4029 PH/02 The ipliteral router was not recognizing addresses of the form user@
4030 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4032 PH/03 Added disable_ipv6.
4034 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4035 decoded form, because it is most often used to construct To: headers
4036 lines in autoreplies, and the decoded form may well be syntactically
4037 invalid. However, $reply_address has leading white space removed, and all
4038 newlines turned into spaces so that the autoreply transport does not
4041 PH/05 If group was specified without a user on a router, and no group or user
4042 was specified on a transport, the group from the router was ignored.
4044 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4045 for visible compile-time settings that can be used to change these
4046 numbers, for those that want even more. Backwards compatibility with old
4047 spool files has been maintained. However, going back to a previous Exim
4048 release will lost any variables that are in spool files.
4050 PH/07 Two small changes when running in the test harness: increase delay when
4051 passing a TCP/IP connection to a new process, in case the original
4052 process has to generate a bounce, and remove special handling of
4053 127.0.0.2 (sic), which is no longer necessary.
4055 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4056 be the same on different OS.
4058 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4061 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4062 whether --show-vars was specified or not
4064 JJ/02 exipick: Added support for new ACL variable spool format introduced
4067 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4068 syntactically invalid From: or Reply-to: line, and a filter used this to
4069 generate an autoreply, and therefore failed to obtain an address for the
4070 autoreply, Exim could try to deliver to a non-existent relative file
4071 name, causing unrelated and misleading errors. What now happens is that
4072 it logs this as a hard delivery error, but does not attempt to create a
4075 PH/11 The exinext utility has a -C option for testing purposes, but although
4076 the given file was scanned by exinext itself; it wasn't being passed on
4077 when Exim was called.
4079 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4080 an end-of-file indication when reading a command response.
4082 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4083 compiled. In many other places in Exim, IPv6 addresses are always
4084 recognized, so I have changed this. It also means that IPv4 domain
4085 literals of the form [IPV4:n.n.n.n] are now always recognized.
4087 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4088 used if the router is not running as root, for example, when verifying at
4089 ACL time, or when using -bh. The debugging output from this situation was
4090 non-existent - all you got was a failure to exec. I have made two
4093 (a) Failures to set uid/gid, the current directory, or a process leader
4094 in a subprocess such as that created by queryprogram now generate
4095 suitable debugging output when -d is set.
4097 (b) The queryprogram router detects when it is not running as root,
4098 outputs suitable debugging information if -d is set, and then runs
4099 the subprocess without attempting to change uid/gid.
4101 PH/15 Minor change to Makefile for building test_host (undocumented testing
4104 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4105 additional section of a DNS packet that returns MX or SRV records.
4106 Instead, it always explicitly searches for A/AAAA records. This avoids
4107 major problems that occur when a DNS server includes only records of one
4108 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4109 fixed another bug: if SRV records were looked up and the corresponding
4110 address records were *not* found in the additional section, the port
4111 values from the SRV records were lost.
4113 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4114 using the correct key (the original address) when searching the retry
4115 rules in order to find which one to use for generating the retry hint.
4117 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4118 adding the default one. Similarly, if it contains a Reply-To: header, the
4119 errors_reply_to option, if set, is not used.
4121 PH/19 When calculating a retry time, Exim used to measure the "time since
4122 failure" by looking at the "first failed" field in the retry record. Now
4123 it does not use this if it is later than than the arrival time of the
4124 message. Instead it uses the arrival time. This makes for better
4125 behaviour in cases where some deliveries succeed, thus re-setting the
4126 "first failed" field. An example is a quota failure for a huge message
4127 when small messages continue to be delivered. Without this change, the
4128 "time since failure" will always be short, possible causing more frequent
4129 delivery attempts for the huge message than are intended.
4130 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4132 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4133 $1, $2, $3) because the numerical variables can be reset during some
4134 expansion items (e.g. "match"), thereby losing the authentication data.
4136 PH/21 Make -bV show the size of off_t variables so that the test suite can
4137 decide whether to run tests for quotas > 2G.
4139 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4140 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4141 filecount value is greater than 2G or if a quota value is greater than 2G
4142 on a system where the size of off_t is not greater than 4, a panic error
4145 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4146 never match. The debug and -bh output now contains an explicit error
4147 message indicating a malformed IPv4 address or mask.
4149 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4150 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4151 PH/23 above applies.
4153 PH/25 Do not write to syslog when running in the test harness. The only
4154 occasion when this arises is a failure to open the main or panic logs
4155 (for which there is an explicit test).
4157 PH/26 Added the /no_tell option to "control=freeze".
4159 PH/27 If a host name lookup failed very early in a connection, for example, if
4160 the IP address matched host_lookup and the reverse lookup yielded a name
4161 that did not have a forward lookup, an error message of the form "no IP
4162 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4163 could be logged. Now it outputs the IP address instead of "NULL".
4165 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4166 allows the sender and the authenticated sender to be set when
4167 submitting a message from within Exim. Since child_open_exim() is
4168 documented for local_scan(), the new function should be too.
4170 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4171 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4172 results in an empty string is now treated as unset.
4174 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4176 PH/31 Added sender_verify_fail logging option.
4178 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4179 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4180 tidied the source and removed it altogether.
4182 PH/33 When a queue run was abandoned because the load average was too high, a
4183 log line was always written; now it is written only if the queue_run log
4184 selector is set. In addition, the log line for abandonment now contains
4185 information about the queue run such as the pid. This is always present
4186 in "start" and "stop" lines but was omitted from the "abandon" line.
4188 PH/34 Omit spaces between a header name and the colon in the error message that
4189 is given when verify = headers_syntax fails (if there are lots of them,
4190 the message gets confusing).
4192 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4193 names, as there are now some PTR records that contain slashes. This check
4194 is only to protect against broken name servers that fall over on strange
4195 characters, so the fact that it applies to all lookups doesn't matter.
4197 PH/36 Now that the new test suite is complete, we can remove some of the
4198 special code in Exim that was needed for the old test suite. For example,
4199 sorting DNS records because real resolvers return them in an arbitrary
4200 order. The new test suite's fake resolver always returns records in the
4203 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4204 bounces) except when queue_only is set, to avoid logging races between
4205 the different processes.
4207 PH/38 Panic-die if .include specifies a non-absolute path.
4209 PH/39 A tweak to the "H" retry rule from its user.
4211 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4212 a label. They prevented compilation on older perls.
4214 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4215 a warning to be raised on newish perls.
4217 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4218 on queue. Changes to match documented behaviour of showing count of
4219 messages matching specified criteria.
4221 PH/40 Changed the default ident timeout from 30s to 5s.
4223 PH/41 Added support for the use of login_cap features, on those BSD systems
4224 that have them, for controlling the resources used by pipe deliveries.
4226 PH/42 The content-scanning code uses fopen() to create files in which to put
4227 message data. Previously it was not paying any attention to the mode of
4228 the files. Exim runs with umask(0) because the rest of the code creates
4229 files with open(), and sets the required mode explicitly. Thus, these
4230 files were ending up world-writeable. This was not a big issue, because,
4231 being within the spool directory, they were not world-accessible. I have
4232 created a function called modefopen, which takes an additional mode
4233 argument. It sets umask(777), creates the file, chmods it to the required
4234 mode, then resets the umask. All the relevant calls to fopen() in the
4235 content scanning code have been changed to use this function.
4237 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4238 to 24 hours. This avoids potential overflow problems when processing G
4239 and H retry rules. I suspect nobody ever tinkers with this value.
4241 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4243 PH/45 When the plaintext authenticator is running as a client, the server's
4244 challenges are checked to ensure they are valid base64 strings. By
4245 default, the authentication attempt is cancelled if an invalid string is
4246 received. Setting client_ignore_invalid_base64 true ignores these errors.
4247 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4248 they are received. Thus, the responses can be made to depend on the
4249 challenges. If an invalid string is ignored, an empty string is placed in
4252 PH/46 Messages that are created by the autoreply transport now contains a
4253 References: header, in accordance with RFCs 2822 and 3834.
4255 PH/47 Added authenticated_sender_force to the smtp transport.
4257 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4259 PH/49 Installed latest patch from the Sieve maintainer.
4261 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4262 was also set, the appendfile transport was unnecessarily scanning a
4263 directory of message files (e.g. for maildir delivery) to find the count
4264 of files (along with the size), even though it did not need this
4265 information. It now does the scan only if it needs to find either the
4266 size of the count of files.
4268 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4270 PH/52 Two bugs concerned with error handling when the smtp transport is
4273 (i) Exim was not creating retry information for temporary errors given
4274 for individual recipients after the DATA command when the smtp transport
4275 was used in LMTP mode. This meant that they could be retried too
4276 frequently, and not timed out correctly.
4278 (ii) Exim was setting the flag that allows error details to be returned
4279 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4280 recipients that were returned after the DATA command.
4282 PH/53 This is related to PH/52, but is more general: for any failing address,
4283 when detailed error information was permitted to be returned to the
4284 sender, but the error was temporary, then after the final timeout, only
4285 "retry timeout exceeded" was returned. Now it returns the full error as
4286 well as "retry timeout exceeded".
4288 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4289 do this, and (what is worse) MTAs that accept it.
4291 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4292 will now be deprecated.
4294 PH/56 New os.c-cygwin from the Cygwin maintainer.
4296 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4297 formats (previously only available in exim formats via -bpr, -bpru,
4298 and -bpra. Now also available in native and exiqgrep formats)
4300 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4301 with very large, slow to parse queues
4303 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4305 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4307 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4308 responses to authentication challenges, though it was showing the
4309 challenges; (ii) I've removed the CR characters from the debug output for
4312 PH/58 Allow for the insertion of a newline as well as a space when a string
4313 is turned into more than one encoded-word during RFC 2047 encoding. The
4314 Sieve code now uses this.
4316 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4317 data_4xx, lost_connection, tls_required.
4319 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4320 message was being sent as an SMTP response.
4322 PH/61 Add -l and -k options to exicyclog.
4324 PH/62 When verifying, if an address was redirected to one new address, so that
4325 verification continued, and the new address failed or deferred after
4326 having set something in $address_data, the value of $address_data was not
4327 passed back to the ACL. This was different to the case when no
4328 redirection occurred. The value is now passed back in both cases.
4330 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4331 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4332 use login_cap.h, so on its own it isn't the distinguishing feature. The
4333 new name refers directly to the setclassresources() function.
4335 PH/65 Added configuration files for NetBSD3.
4337 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4339 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4340 is preferred over IPv4.
4342 PH/68 The bounce_return_message and bounce_return_body options were not being
4343 honoured for bounces generated during the reception of non-SMTP messages.
4344 In particular, this applied to messages rejected by the ACL. This bug has
4345 been fixed. However, if bounce_return_message is true and bounce_return_
4346 body is false, the headers that are returned for a non-SMTP message
4347 include only those that have been read before the error was detected.
4348 (In the case of an ACL rejection, they have all been read.)
4350 PH/69 The HTML version of the specification is now built in a directory called
4351 spec_html instead of spec.html, because the latter looks like a path with
4352 a MIME-type, and this confuses some software.
4354 PH/70 Catch two compiler warnings in sieve.c.
4356 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4357 function verify_get_ident() calls ip_connect() to connect a socket, but
4358 if the "connect()" function timed out, ip_connect() used to close the
4359 socket. However, verify_get_ident() also closes the socket later, and in
4360 between Exim writes to the log, which may get opened at this point. When
4361 the socket was closed in ip_connect(), the log could get the same file
4362 descriptor number as the socket. This naturally causes chaos. The fix is
4363 not to close the socket in ip_connect(); the socket should be closed by
4364 the function that creates it. There was only one place in the code where
4365 this was missing, in the iplookup router, which I don't think anybody now
4366 uses, but I've fixed it anyway.
4368 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4369 well as to direct DNS lookups. Otherwise the handling of names in host
4370 lists is inconsistent and therefore confusing.
4376 PH/01 Two changes to the default runtime configuration:
4378 (1) Move the checks for relay_from_hosts and authenticated clients from
4379 after to before the (commented out) DNS black list checks.
4381 (2) Add control=submission to the relay_from_hosts and authenticated
4382 clients checks, on the grounds that messages accepted by these
4383 statements are most likely to be submissions.
4385 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4387 (1) Generate an error if the third argument for the ${prvs expansion is
4390 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4393 (3) Reset the variables that are obtained from the first argument of
4394 ${prvscheck and used in the second argument before leaving the code,
4395 because their memory is reclaimed, so using them afterwards may do
4398 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4399 one (it's much easier than Tom thought :-).
4401 (5) Because of (4), we can now allow for the use of $prvscheck_result
4402 inside the third argument.
4404 PH/03 For some reason, the default setting of PATH when running a command from
4405 a pipe transport was just "/usr/bin". I have changed it to
4408 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4409 anything to be listed in the output from -bV.
4411 PH/05 When a filter generated an autoreply, the entire To: header line was
4412 quoted in the delivery log line, like this:
4414 => >A.N.Other <ano@some.domain> <original@ddress> ...
4416 This has been changed so that it extracts the operative address. There
4417 may be more than one such address. If so, they are comma-separated, like
4420 => >ano@some.domain,ona@other.domain <original@ddress> ...
4422 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4423 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4424 not being looked up in the rDNS to get a host name, Exim was showing the
4425 IP address twice in Received: lines, even though the IP addresses were
4426 identical. For example:
4428 Received: from [1.2.3.4] (helo=[1.2.3.4])
4430 However, if the real host name was known, it was omitting the HELO data
4431 if it matched the actual IP address. This has been tidied up so that it
4432 doesn't show the same IP address twice.
4434 PH/07 When both +timestamp and +memory debugging was on, the value given by
4435 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4436 called by the debug printing, thereby overwriting the timestamp buffer.
4437 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4439 PH/08 When the original message was included in an autoreply transport, it
4440 always said "this is a copy of the message, including all the headers",
4441 even if body_only or headers_only was set. It now gives an appropriate
4444 PH/09 Applied a patch from the Sieve maintainer which:
4446 o fixes some comments
4447 o adds the (disabled) notify extension core
4448 o adds some debug output for the result of if/elsif tests
4449 o points to the current vacation draft in the documentation
4450 and documents the missing references header update
4454 o fixes a bug in processing the envelope test (when testing
4455 multiple envelope elements, the last element determined the
4458 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4459 Electronic Mail") by including:
4461 Auto-submitted: auto-generated
4463 in the messages that it generates (bounce messages and others, such as
4464 warnings). In the case of bounce messages for non-SMTP messages, there was
4465 also a typo: it was using "Auto_submitted" (underscore instead of
4466 hyphen). Since every message generated by Exim is necessarily in response
4467 to another message, thes have all been changed to:
4469 Auto-Submitted: auto-replied
4471 in accordance with these statements in the RFC:
4473 The auto-replied keyword:
4475 - SHOULD be used on messages sent in direct response to another
4476 message by an automatic process,
4478 - MUST NOT be used on manually-generated messages,
4480 - MAY be used on Delivery Status Notifications (DSNs) and Message
4481 Disposition Notifications (MDNs),
4483 - MUST NOT be used on messages generated by automatic or periodic
4484 processes, except for messages which are automatic responses to
4487 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4488 to the default Received: header definition.
4490 PH/12 Added log selector acl_warn_skipped (default on).
4492 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4493 variables because (a) they are in the wrong storage pool and (b) even if
4494 they were copied, it wouldn't work properly because of the caching.
4496 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4497 checking when decoding. Apparently there are clients that generate
4498 overlong encoded strings. Why am I not surprised?
4500 PH/15 If the first argument of "${if match_address" was not empty, but did not
4501 contain an "@" character, Exim crashed. Now it writes a panic log message
4502 and treats the condition as false.
4504 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4506 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4507 extension "envelope-auth". The code is finished and in agreement with
4508 other implementations, but there is no documentation so far and in fact,
4509 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4510 not changing the active code.
4512 Print executed "if" and "elsif" statements when debugging is used. This
4513 helps a great deal to understand what a filter does.
4515 Document more things not specified clearly in RFC3028. I had all this
4516 sorted out, when out of a sudden new issues came to my mind. Oops."
4518 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4521 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4522 canonical form (as documented). However, after a host name lookup from
4523 the IP address, check_host() was doing a simple string comparison with
4524 addresses acquired from the DNS when checking that the found name did
4525 have the original IP as one of its addresses. Since any found IPv6
4526 addresses are likely to be in abbreviated form, the comparison could
4527 fail. Luckily, there already exists a function for doing the comparison
4528 by converting both addresses to binary, so now that is used instead of
4529 the text comparison.
4531 PH/20 There was another similar case to PH/19, when a complete host name was
4532 given in a host list; looking up its IP address could give an abbreviated
4533 form, whereas the current host's name might or might not be abbreviated.
4534 The same fix has been applied.
4540 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4541 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4544 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4545 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4547 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4549 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4550 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4551 Also, for "Auto-Submitted", treat anything other than "no" as
4552 non-personal, in accordance with RFC 3834. (Previously it treated
4553 anything starting "auto-" as non-personal.)
4555 TF/01 The control=submission/name=... option had a problem with syntax
4556 errors if the name included a slash character. The /name= option
4557 now slurps the rest of the string, so it can include any characters
4558 but it must come last in the list of options (after /sender_retain
4561 PH/05 Some modifications to the interface to the fake nameserver for the new
4569 TK/01 Added the "success_on_redirect" address verification option. See
4570 NewStuff for rationale and an example.
4572 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4574 PH/02 Patch to exigrep to allow it to work on syslog lines.
4576 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4577 fread() to skip over the body file's header line, because in Cygwin the
4578 header line is locked and is inaccessible.
4580 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4581 co-exist for some time) to make it clear that it is the Exim ID that is
4582 referenced, not the Message-ID: header line.
4584 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4585 string_format() function, because snprintf() does not exist on all
4588 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4589 ${stat: expansion item.
4591 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4592 protocol synchronization error", to keep the pedants happy.
4594 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4595 well as for IRIX systems, when gcc is being used. See the host.c source
4598 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4600 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4603 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4604 transport and to the smtp transport in LMTP mode.
4606 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4608 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4609 run a filter in a subprocess. This could lead to confusion in subsequent
4610 lookups in the parent process. There should also be a search_tidyup() at
4611 the end of the subprocess.
4613 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4614 only if the host matched helo_try_verify_hosts, which caused the
4615 verification to occur when the EHLO/HELO command was issued. The ACL just
4616 tested the remembered result. Now, if a previous verification attempt has
4617 not happened, "verify = helo" does it there and then.
4619 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4621 TK/03 Fix log output including CR from clamd.
4623 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4624 exist provoked a memory error which could cause a segfault.
4626 PH/15 Installed PCRE 6.2
4628 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4630 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4631 of the problem. Specifically, suggested +O2 rather than +O1 for the
4634 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4636 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4637 record, the port was not correctly logged when the outgoing_port log
4638 selector was set (it logged the transort's default port).
4640 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4641 fallback_hosts, and "hosts" in the smtp transport.
4643 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4644 host errors such as "Connection refused".
4646 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4647 authentication with radiusclient 0.4.9:
4649 - Error returned from rc_read_config was caught wrongly
4650 - Username/password not passed on to radius server due to wrong length.
4652 The presumption is that some radiusclient API changes for 4.51/PH/17
4653 were not taken care of correctly. The code is still untested by me (my
4654 Linux distribution still has 0.3.2 of radiusclient), but it was
4655 contributed by a Radius user.
4657 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4658 expanding the "port" option of the smtp transport.
4660 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4661 while reading a MIME header. Thanks to Tom Hughes for a patch.
4663 PH/24 Include config.h inside local_scan.h so that configuration settings are
4666 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4667 that in an ACL for RCPT (for example), you can examine exactly what was
4670 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4671 commands, but it was not correctly comparing the address with the actual
4672 client host address. Thus, it would show the EHLO address in Received:
4673 header lines when this was not necessary.
4675 PH/27 Added the % operator to ${eval:}.
4677 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4678 it should be ignoring failures (because with -C, for example, it has lost
4679 privilege). It wasn't ignoring creation failures other than "already
4682 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4685 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4686 because an input line was too long, either on its own, or by virtue of
4687 too many continuations, the temporary file was not being removed, and the
4688 return code was incorrect.
4690 PH/31 Missing "BOOL" in function definition in filtertest.c.
4692 PH/32 Applied Sieve patches from the maintainer.
4694 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4696 PH/33 Added "verify = not_blind".
4698 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4699 Local/Makefile (with some defaults set). These are used in built scripts
4700 such as exicyclog, but they have never been used in the exim_install
4701 script (though there are many overriding facilities there). I have
4702 arranged that the exim_install script now takes note of these two
4705 PH/35 Installed configuration files for Dragonfly.
4707 PH/36 When a locally submitted message by a trusted user did not contain a
4708 From: header, and the sender address was obtained from -f or from an SMTP
4709 MAIL command, and the trusted user did not use -F to supply a sender
4710 name, $originator_name was incorrectly used when constructing a From:
4711 header. Furthermore, $originator_name was used for submission mode
4712 messages from external hosts without From: headers in a similar way,
4713 which is clearly wrong.
4715 PH/37 Added control=suppress_local_fixups.
4717 PH/38 When log_selector = +received_sender was set, and the addition of the
4718 sender made the log line's construction buffer exactly full, or one byte
4719 less than full, an overflow happened when the terminating "\n" was
4722 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4723 when the result of a list match is failure because a DNS lookup failed.
4725 PH/40 RM_COMMAND is now used in the building process.
4727 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4728 the "build-* directories that it finds.
4730 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4731 address in a domain literal was a prefix of an interface address.
4733 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4734 when verifying a sender address, unless rewrite_headers is false.
4736 PH/44 Wrote a long comment about why errors_to addresses are verified as
4737 recipients, not senders.
4739 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4740 the ratelimit ACL was added.
4742 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4744 PH/46 Added extra information about PostgreSQL errors to the error string.
4746 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4747 avoiding the need to install special zones in a real server. This is
4748 backwards compatible; if it can't find the fake resolver, it drops back.
4749 Thus, both old and new test suites can be run.
4751 TF/02 Added util/ratelimit.pl
4753 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4754 clock is set back in time.
4756 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4757 Candler <B.Candler@pobox.com>.
4759 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4760 OK for addresses that are the result of redirections.
4762 PH/48 A number of further additions for the benefit of the new test suite,
4763 including a fake gethostbyname() that interfaces to the fake DNS resolver
4766 TF/06 The fix for widen_domains has also been applied to qualify_single and
4767 search_parents which are the other dnslookup options that can cause
4770 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4773 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4775 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4776 service" if it encountered an MX record with an empty target hostname.
4777 The message is now "an MX or SRV record indicated no SMTP service".
4779 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4780 if the DNS of the sending site is misconfigured. This is quite a
4781 common situation. This change restores the behaviour of treating a
4782 helo verification defer as a failure.
4784 PH/51 If self=fail was set on a router, the bounce message did not include the
4785 actual error message.
4791 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4793 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4794 command itself ended in error, the underlying message about the transport
4795 filter timeout was being overwritten with the pipe command error. Now the
4796 underlying error message should be appended to the second error message.
4798 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4800 PH/02 Reduce the amount of output that "make" produces by default. Full output
4801 can still be requested.
4803 PH/03 The warning log line about a condition test deferring for a "warn" verb
4804 was being output only once per connection, rather than after each
4805 occurrence (because it was using the same function as for successful
4806 "warn" verbs). This seems wrong, so I have changed it.
4808 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4809 it should not have, which might have caused a crash in the right
4810 circumstances, but probably never did.
4812 PH/04 Installed a modified version of Tony Finch's patch to make submission
4813 mode fix the return path as well as the Sender: header line, and to
4814 add a /name= option so that you can make the user's friendly name appear
4817 TF/03 Added the control = fakedefer ACL modifier.
4819 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4820 Mark Lowes for thorough testing.
4822 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4824 TK/03 Merged latest SRS patch from Miles Wilton.
4826 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4827 with the definition in sysexits.h (which is #included earlier).
4828 Fortunately, Exim does not actually use EX_OK. The code used to try to
4829 preserve the sysexits.h value, by assuming that macro definitions were
4830 scanned for macro replacements. I have been disabused of this notion,
4831 so now the code just undefines EX_OK before #including unistd.h.
4833 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4834 in the smtp transport. When a block could not be written in a single
4835 write() function, the timeout was being re-applied to each part-write.
4836 This seems wrong - if the receiver was accepting one byte at a time it
4837 would take for ever. The timeout is now adjusted when this happens. It
4838 doesn't have to be particularly precise.
4840 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4841 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4843 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4844 without a domain if the check to <postmaster@domain> fails.
4846 SC/01 Eximstats: added -xls and the ability to specify output files
4847 (patch written by Frank Heydlauf).
4849 SC/02 Eximstats: use FileHandles for outputting results.
4851 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4853 SC/04 Eximstats: fixed display of large numbers with -nvr option
4855 SC/05 Eximstats: fixed merging of reports with empty tables.
4857 SC/06 Eximstats: added the -include_original_destination flag
4859 SC/07 Eximstats: removed tabs and trailing whitespace.
4861 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4863 TK/06 MBOX spool code: Add real "From " MBOX separator line
4864 so the .eml file is really in mbox format (even though
4865 most programs do not really care). Patch from Alex Miller.
4867 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4868 The latter is generated from $received_to and is only set if the
4869 message has one envelope recipient. SA can use these headers,
4870 obviously out-of-the-box. Patch from Alex Miller.
4872 PH/08 The ${def test on a variable was returning false if the variable's
4873 value was "0", contrary to what the specification has always said!
4874 The result should be true unless the variable is empty.
4876 PH/09 The syntax error of a character other than { following "${if
4877 def:variable_name" (after optional whitespace) was not being diagnosed.
4878 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4879 accidental colon was present, for example, could give incorrect results.
4881 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4882 result is used (not including appendfile, where other changes are about
4885 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4886 This involved changing a lot of size variables from int to off_t. It
4887 should work with maildirs and everything.
4889 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4890 spamd dying while we are connected to it.
4892 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4895 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4896 function for BDB 4.3.
4898 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4900 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4901 the sources for PCRE have been re-arranged and more files are now
4904 PH/15 The code I had for printing potentially long long variables in PH/11
4905 above was not the best (it lost precision). The length of off_t variables
4906 is now inspected at build time, and an appropriate printing format (%ld
4907 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4908 to be "long long int" or "long int". This is needed for the internal
4909 formatting function string_vformat().
4911 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4912 the configuration file to be ":syslog", then the script "guesses" where
4913 the logs files are, rather than using the compiled in default. In our
4914 case the guess is not the same as the compiled default, so the script
4915 suddenly stopped working when I started to use syslog. The patch checks
4916 to see if log_file_path is "". If so, it attempts to read it from exim
4917 with no configuration file to get the compiled in version, before it
4918 falls back to the previous guessing code."
4920 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4921 implementing BATV in an Exim configuration. See NewStuff for the gory
4924 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4925 Makefile that are specific to HP-UX.
4927 PH/18 If the "use_postmaster" option was set for a recipient callout together
4928 with the "random" option, the postmaster address was used as the MAIL
4929 FROM address for the random test, but not for the subsequent recipient
4930 test. It is now used for both.
4932 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4933 patch removes a few documentation additions to RFC 3028, because the
4934 latest draft now contains them. It adds the new en;ascii-case comparator
4935 and a new error check for 8bit text in MIME parts. Comparator and
4936 require names are now matched exactly. I enabled the subaddress
4937 extension, but it is not well tested yet (read: it works for me)."
4939 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4940 rework some of the code of TK/09 above to avoid the hardwired use of
4941 "%lld" and "long long". Replaced the call to snprintf() with a call to
4944 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4945 records point to non-existent hosts", "retry timeout exceeded", and
4946 "retry time not reached for any host after a long failure period".
4948 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4949 experimental DomainKeys support:
4951 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4952 (2) On an error such as an illegally used "control", the wrong name for
4953 the control was given.
4955 These problems did NOT occur unless DomainKeys support was compiled.
4957 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4959 PH/24 Added ${if match_ip condition.
4961 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4962 databases so that it will be absolutely obvious if a crash occurs in the
4963 DB library. This is a regular occurrence (often caused by mis-matched
4966 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4967 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4968 picked up on a user's system that detects such things. There doesn't seem
4969 to be a gcc warning option for this - only an attribute that has to be
4970 put on the function's prototype. It seems that in Fedora Core 4 they have
4971 set this on a number of new functions. No doubt there will be more in due
4974 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4975 specify a transport. However, if an address that was verified by such a
4976 router was the subject of a callout, Exim crashed because it tried to
4977 read the rcpt_include_affixes from the non-existent transport. Now it
4978 just assumes that the setting of that option is false. This bug was
4979 introduced by 4.51/PH/31.
4981 PH/28 Changed -d+all to exclude +memory, because that information is very
4982 rarely of interest, but it makes the output a lot bigger. People tend to
4983 do -d+all out of habit.
4985 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4986 code in os-type was giving problems when libc.so lives in lib64, like on
4989 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4990 aren't the modern standard, and it seems that some systems' include files
4991 don't always have them. Exim was already checking for some of the newer
4992 ones like T_AAAA, and defining it itself. I've added checks for all the
4993 record types that Exim uses.
4995 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4996 not automatically generating a new one, as it is supposed to. This
4997 prevented TLS from working. If the file did exist, but contained invalid
4998 data, a new version was generated, as expected. It was only the case of a
4999 non-existent file that was broken.
5001 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5002 with a change in libdomainkeys > 0.64.
5004 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5005 from DNS. If the selector record carries the flag, it now has
5006 precedence over the domain-wide flag.
5008 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5010 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5011 the use of a port name that isn't defined in /etc/services) occurred, the
5012 message was deferred as in a normal delivery, and thus remained on the
5013 spool, instead of being failed because of the mua_wrapper setting. This
5014 is now fixed, and I tidied up some of the mua_wrapper messages at the
5017 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5018 the messages in a hash of arrays rather than using individual hashes.
5019 This is a bit cleaner and results in dramatic memory savings, albeit
5020 at a slight CPU cost.
5022 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5023 as requested by Marc Sherman.
5025 SC/10 Eximstats: added histograms for user specified patterns as requested
5028 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5030 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5031 fopen() in the content-scanning modules that did not already have it.
5037 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5038 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5040 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5042 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5044 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5045 to test Sieve filters that use "vacation".
5047 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5048 that changes the way the GnuTLS parameters are stored in the cache file.
5049 The new format can be generated externally. For backward compatibility,
5050 if the data in the cache doesn't make sense, Exim assumes it has read an
5051 old-format file, and it generates new data and writes a new file. This
5052 means that you can't go back to an older release without removing the
5055 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5056 work if there are any delivery delays because "one_time" forces the
5057 parent to be marked "delivered", so its unseen clone is never tried
5058 again. For this reason, Exim now forbids the simultaneous setting of
5061 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5062 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5063 entry if you want to know the details. The fix, however, neglected to
5064 consider the case when local delivery batching is involved. The test for
5065 "previously delivered" was not happening when checking to see if an
5066 address could be batched with a previous (undelivered) one; under
5067 certain circumstances this could lead to multiple deliveries to the same
5070 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5071 in its include files, and this causes problems building Exim.
5073 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5074 header_syntax) but Exim was just ignoring anything given after a slash.
5075 In particular, this caused confusion with an attempt to use "verify =
5076 reverse_host_lookup/defer_ok". An error is now given when options are
5077 supplied for verify items that do not have them. (Maybe reverse_host_
5078 lookup should have a defer_ok option, but that's a different point.)
5080 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5081 defined by RFC 821) to 2048, because there were problems with some AUTH
5082 commands, and RFC 1869 says the size should be increased for extended
5083 SMTP commands that take arguments.
5085 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5088 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5089 "unknown" error; now it says that the functionality isn't in the binary.
5091 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5092 an address' error message when a string expansion fails (syntax or
5093 whatever). Otherwise the password may appear in the log. Following change
5094 PH/42 below, there is no longer a chance of it appearing in a bounce
5097 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5099 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5100 of its addresses was ever tried. (Bugzilla bug #2.)
5102 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5103 the result incorrectly in the debug output. (It correctly added a newline
5104 to what was transported.)
5106 TF/01 Added $received_time.
5108 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5109 commented out examples of how to interface to a virus scanner and to
5110 SpamAssassin. Also added commented examples of av_scanner and
5111 spamd_address settings.
5113 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5114 and controls are allowed in which ACLs. There were a couple of minor
5115 errors. Some of the entries in the conditions table (which is a table of
5116 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5117 negation of where the condition IS allowed.
5119 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5121 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5122 header file does not have a version number, so I've had to invent a new
5123 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5124 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5125 radiusclient), but it was contributed by a Radius user.
5127 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5128 files or directories, for OpenSSL.
5130 PH/19 When an Exim process that is running as root has to create an Exim log
5131 file, it does so in a subprocess that runs as exim:exim so as to get the
5132 ownership right at creation (otherwise, other Exim processes might see
5133 the file with the wrong ownership). There was no test for failure of this
5134 fork() call, which would lead to the process getting stuck as it waited
5135 for a non-existent subprocess. Forks do occasionally fail when resources
5136 run out. I reviewed all the other calls to fork(); they all seem to check
5139 PH/20 When checking for unexpected SMTP input at connect time (before writing
5140 the banner), Exim was not dealing correctly with a non-positive return
5141 from the read() function. If the client had disconnected by this time,
5142 the result was a log entry for a synchronization error with an empty
5143 string after "input=" when read() returned zero. If read() returned -1
5144 (an event I could not check), uninitialized data bytes were printed.
5145 There were reports of junk text (parts of files, etc) appearing after
5148 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5150 PH/22 Added support for macro redefinition, and (re)definition in between
5151 driver and ACL definitions.
5153 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5154 forgetting to use the resulting value; it was using the unexpanded value.
5156 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5157 hadn't been configured. The fix is from Juergen Kreileder, who
5158 understands it better than I do:
5160 "Here's what I see happening with three configured cyrus_sasl
5161 authenticators configured (plain, login, cram-md5):
5163 On startup auth_cyrus_sasl_init() gets called for each of these.
5164 This means three calls to sasl_listmech() without a specified mech_list.
5165 => SASL tests which mechs of all available mechs actually work
5166 => three warnings about OTP not working
5167 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5169 With the patch, sasl_listmech() also gets called three times. But now
5170 SASL's mech_list option is set to the server_mech specified in the the
5171 authenticator. Or in other words, the answer from sasl_listmech()
5172 gets limited to just the mech you're testing for (which is different
5174 => the return list contains just 'plain' or 'login', 'cram-md5' or
5175 nothing depending on the value of ob->server_mech.
5177 I've just tested the patch: Authentication still works fine,
5178 unavailable mechs specified in the exim configuration are still
5179 caught, and the auth.log warnings about OTP are gone."
5181 PH/25 When debugging is enabled, the contents of the command line are added
5182 to the debugging output, even when log_selector=+arguments is not
5185 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5186 answer is "GNU", and only if the return is "GNU/something" is the answer
5189 PH/27 $acl_verify_message is now set immediately after the failure of a
5190 verification in an ACL, and so is available in subsequent modifiers. In
5191 particular, the message can be preserved by coding like this:
5193 warn !verify = sender
5194 set acl_m0 = $acl_verify_message
5196 Previously, $acl_verify_message was set only while expanding "message"
5197 and "log_message" when a very denied access.
5199 PH/28 Modified OS/os.c-Linux with
5201 -#ifndef OS_LOAD_AVERAGE
5202 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5204 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5205 nomenclature these days.)
5207 PH/29 Installed patch from the Sieve maintainer that adds the options
5208 sieve_useraddress and sieve_subaddress to the redirect router.
5210 PH/30 In these circumstances:
5211 . Two addresses routed to the same list of hosts;
5212 . First host does not offer TLS;
5213 . First host accepts first address;
5214 . First host gives temporary error to second address;
5215 . Second host offers TLS and a TLS session is established;
5216 . Second host accepts second address.
5217 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5218 and peerdn, if requested) that were in fact used only for the second
5221 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5222 attention to any local part prefix or suffix that was matched by the
5223 router that accepted the address. It now behaves in the same way as it
5224 does for delivery: the affixes are removed from the local part unless
5225 rcpt_include_affixes is set on the transport.
5227 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5228 timeout during the DATA phase of an incoming message.
5230 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5231 applied a patch sanctioned by the Sieve maintainer.
5233 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5234 the uid or gid is negative. A case of a negative gid caused this to be
5235 noticed. The fix allows for either to be negative.
5237 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5238 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5241 PH/36 The change PH/12 above was broken. Fixed it.
5243 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5244 the grounds that routing the same address twice would always produce the
5245 same answer. This might have been true once, but it is certainly no
5246 longer true now. Routing a child address may depend on the previous
5247 routing that produced that child. Some complicated redirection strategies
5248 went wrong when messages had multiple recipients, and made Exim's
5249 behaviour dependent on the order in which the addresses were given.
5251 I have moved the duplicate checking until after the routing is complete.
5252 Exim scans the addresses that are assigned to local and remote
5253 transports, and removes any duplicates. This means that more work will be
5254 done, as duplicates will always all be routed, but duplicates are
5255 presumably rare, so I don't expect this is of any significance.
5257 For deliveries to pipes, files, and autoreplies, the duplicate checking
5258 still happens during the routing process, since they are not going to be
5261 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5262 It corrects a timeout issue with spamd. This is Ian's comment: "The
5263 background is that sometimes spamd either never reads data from a
5264 connection it has accepted, or it never writes response data. The exiscan
5265 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5266 blindly assumes that writes won't block so it may never time out."
5268 PH/39 Allow G after quota size as well as K and M.
5270 PH/40 The value set for $authenticated_id in an authenticator may not contain
5271 binary zeroes or newlines because the value is written to log lines and
5272 to spool files. There was no check on this. Now the value is run through
5273 the string_printing() function so that such characters are converted to
5274 printable escape sequences.
5276 PH/41 $message_linecount is a new variable that contains the total number of
5277 lines in the message. Compare $body_linecount, which is the count for the
5280 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5281 bounce and delay warning messages, except in certain special cases, which
5284 (a) An SMTP error message from a remote host;
5285 (b) A message specified in a :fail: redirection;
5286 (c) A message specified in a "fail" command in a system filter;
5287 (d) A message specified in a FAIL return from the queryprogram router;
5288 (e) A message specified by the cannot_route_message router option.
5290 In these cases only, Exim does include the error details in bounce and
5291 warning messages. There are also a few cases where bland messages such
5292 as "unrouteable address" or "local delivery error" are given.
5294 PH/43 $value is now also set for the "else" part of a ${run expansion.
5296 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5297 being worked on, but at least Exim now implements the latest version to
5300 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5301 process to complete was treated as a delivery failure, a timeout while
5302 writing the message to the pipe was logged, but erroneously treated as a
5303 successful delivery. Such timeouts include transport filter timeouts. For
5304 consistency with the overall process timeout, these timeouts are now
5305 treated as errors, giving rise to delivery failures by default. However,
5306 there is now a new Boolean option for the pipe transport called
5307 timeout_defer, which, if set TRUE, converts the failures into defers for
5308 both kinds of timeout. A transport filter timeout is now identified in
5311 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5312 systems where "make" and "gmake" are different, calling "gmake" at top
5313 level broke things. I've arranged for the value of $(MAKE) to be passed
5314 from the Makefile to this script so that it can call the same version of
5318 A note about Exim versions 4.44 and 4.50
5319 ----------------------------------------
5321 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5322 changes of various kinds. As a consequence, a big documentation update was
5323 needed. This delayed the release for rather longer than seemed good, especially
5324 in the light of a couple of (minor) security issues. Therefore, the changes
5325 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5326 release. So 4.44 and 4.50 are in effect two different branches that both start
5329 I have left the 4.50 change log unchanged; it contains all the changes since
5330 4.43. The change log for 4.44 is below; many of its items are identical to
5331 those for 4.50. This seems to be the most sensible way to preserve the
5332 historical information.
5338 1. Minor wording change to the doc/README.SIEVE file.
5340 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5341 computation of the current number of files was incorrect.
5343 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5344 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5347 4. Give more explanation in the error message when the command for a transport
5348 filter fails to execute.
5350 5. There are several places where Exim runs a non-Exim command in a
5351 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5352 was being done only for the command run by the queryprogram router. It is
5353 now done for all such subprocesses. The other cases are: ${run, transport
5354 filters, and the commands run by the lmtp and pipe transports.
5356 6. Added CONFIGURE_GROUP build-time option.
5358 7. Some older OS have a limit of 256 on the maximum number of file
5359 descriptors. Exim was using setrlimit() to set 1000 as a large value
5360 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5361 systems. I've change it so that if it can't get 1000, it tries for 256.
5363 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5364 was an oversight, and furthermore, ever since the addition of extra
5365 controls (e.g. 4.43/32), the checks on when to allow different forms of
5366 "control" were broken. There should now be diagnostics for all cases when a
5367 control that does not make sense is encountered.
5369 9. Added the /retain_sender option to "control=submission".
5371 10. $recipients is now available in the predata ACL (oversight).
5373 11. Tidy the search cache before the fork to do a delivery from a message
5374 received from the command line. Otherwise the child will trigger a lookup
5375 failure and thereby defer the delivery if it tries to use (for example) a
5376 cached ldap connection that the parent has called unbind on.
5378 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5379 of $address_data from the recipient verification was clobbered by the
5380 sender verification.
5382 13. The value of address_data from a sender verification is now available in
5383 $sender_address_data in subsequent conditions in the ACL statement.
5385 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5387 15. Added a new option "connect=<time>" to callout options, to set a different
5390 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5391 was its contents. (It was OK if the option was not defined at all.)
5393 17. A "Completed" log line is now written for messages that are removed from
5394 the spool by the -Mrm option.
5396 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5397 information about exactly what failed.
5399 19. Added -dd to debug only the daemon process.
5401 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5402 handles timeouts, both on the server side and network timeouts. Renamed the
5403 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5405 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5406 It is now set to "smtps".
5408 22. $host_address is now set to the target address during the checking of
5409 ignore_target_hosts.
5411 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5412 being passed; this would have caused $sender_host_name to have been used if
5413 matching the list had actually called for a host name (not very likely,
5414 since this list is usually IP addresses). A host name is now passed as
5417 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5418 code that specifies a non-restarting handler (typically sigaction() in
5419 modern systems) in an attempt to fix a rare and obscure crash bug.
5421 25. Narrowed the window for a race in the daemon that could cause it to ignore
5422 SIGCHLD signals. This is not a major problem, because they are used only to
5423 wake it up if nothing else does.
5425 26. A malformed maildirsize file could cause Exim to calculate negative values
5426 for the mailbox size or file count. Odd effects could occur as a result.
5427 The maildirsize information is now recalculated if the size or filecount
5430 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5431 support for a long time. Removed HAVE_SYS_VFS_H.
5433 28. Installed the latest version of exipick from John Jetmore.
5435 29. In an address list, if the pattern was not a regular expression, an empty
5436 subject address (from a bounce message) matched only if the pattern was an
5437 empty string. Non-empty patterns were not even tested. This was the wrong
5438 because it is perfectly reasonable to use an empty address as part of a
5439 database query. An empty address is now tested by patterns that are
5440 lookups. However, all the other forms of pattern expect the subject to
5441 contain a local part and a domain, and therefore, for them, an empty
5442 address still always fails if the pattern is not itself empty.
5444 30. Exim went into a mad DNS loop when attempting to do a callout where the
5445 host was specified on an smtp transport, and looking it up yielded more
5446 than one IP address.
5448 31. Re-factored the code for checking spool and log partition space into a
5449 function that finds that data and another that does the check. The former
5450 is then used to implement four new variables: $spool_space, $log_space,
5451 $spool_inodes, and $log_inodes.
5453 32. The RFC2047 encoding function was originally intended for short strings
5454 such as real names; it was not keeping to the 75-character limit for
5455 encoded words that the RFC imposes. It now respects the limit, and
5456 generates multiple encoded words if necessary. To be on the safe side, I
5457 have increased the buffer size for the ${rfc2047: expansion operator from
5460 33. It is now permitted to omit both strings after an "if" condition; if the
5461 condition is true, the result is "true". As before, when the second string
5462 is omitted, a false condition yields an empty string. This makes it less
5463 cumbersome to write custom ACL and router conditions.
5465 34. Failure to deliver a bounce message always caused it to be frozen, even if
5466 there was an errors_to setting on the router. The errors_to setting is now
5469 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5470 canonical form (fully expanded) before being placed in
5471 $sender_host_address.
5473 36. The table in the code that translates DNS record types into text (T_A to
5474 "A" for instance) was missing entries for NS and CNAME. It is just possible
5475 that this could have caused confusion if both these types were looked up
5476 for the same domain, because the text type is used as part of Exim's
5477 per-process caching. But the chance of anyone hitting this buglet seems
5480 37. The dnsdb lookup has been extended in a number of ways.
5482 (1) There is a new type, "zns", which walks up the domain tree until it
5483 finds some nameserver records. It should be used with care.
5485 (2) There is a new type, "mxh", which is like "mx" except that it returns
5486 just the host names, not the priorities.
5488 (3) It is now possible to give a list of domains (or IP addresses) to be
5489 looked up. The behaviour when one of the lookups defers can be
5490 controlled by a keyword.
5492 (4) It is now possible to specify the separator character for use when
5493 multiple records are returned.
5495 38. The dnslists ACL condition has been extended: it is now possible to supply
5496 a list of IP addresses and/or domains to be looked up in a particular DNS
5499 39. Added log_selector=+queue_time_overall.
5501 40. When running the queue in the test harness, wait just a tad after forking a
5502 delivery process, to get repeatability of debugging output.
5504 41. Include certificate and key file names in error message when GnuTLS fails
5505 to set them up, because the GnuTLS error message doesn't include the name
5506 of the failing file when there is a problem reading it.
5508 42. Allow both -bf and -bF in the same test run.
5510 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5512 44. The "Exiscan patch" is now merged into the mainline Exim source.
5514 45. Sometimes the final signoff response after QUIT could fail to get
5515 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5516 < 0 before doing a fflush(). This bug looks as though it goes back to the
5517 introduction of TLS in release 3.20, but "sometimes" must have been rare
5518 because the tests only now provoked it.
5520 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5521 (this can affect the format of dates).
5523 47. exim_tidydb, when checking for the continued existence of a message for
5524 which it has found a message-specific retry record, was not finding
5525 messages that were in split spool directories. Consequently, it was
5526 deleting retry records that should have stayed in existence.
5528 48. Steve fixed some bugs in eximstats.
5530 49. The SPA authentication driver was not abandoning authentication and moving
5531 on to the next authenticator when an expansion was forced to fail,
5532 contradicting the general specification for all authenticators. Instead it
5533 was generating a temporary error. It now behaves as specified.
5535 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5536 (the order specifies the preference for clients). The order is now AES256,
5537 AES128, 3DES, ARCFOUR128.
5539 51. Small patch to Sieve code - explicitly set From: when generating an
5542 52. Exim crashed if a remote delivery caused a very long error message to be
5543 recorded - for instance if somebody sent an entire SpamAssassin report back
5544 as a large number of 550 error lines. This bug was coincidentally fixed by
5545 increasing the size of one of Exim's internal buffers (big_buffer) that
5546 happened as part of the Exiscan merge. However, to be on the safe side, I
5547 have made the code more robust (and fixed the comments that describe what
5550 53. Now that there can be additional text after "Completed" in log lines (if
5551 the queue_time_overall log selector is set), a one-byte patch to exigrep
5552 was needed to allow it to recognize "Completed" as not the last thing in
5555 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5556 patch that reportedly fixes this has been added. I am not expert enough to
5557 create a test for it. This is what the patch creator wrote:
5559 "I found a little strange behaviour of ldap code when working with
5560 Windows 2003 AD Domain, where users was placed in more than one
5561 Organization Units. When I tried to give exim partial DN, the exit code
5562 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5563 But simultaneously result of request was absolutely normal ldap result,
5564 so I produce this patch..."
5566 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5567 REFERENCE, so I have modified the code to exclude the patch when that macro
5570 55. Some experimental protocols are using DNS PTR records for new purposes. The
5571 keys for these records are domain names, not reversed IP addresses. The
5572 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5573 leaves it alone. Component reversal etc. now happens only for IP addresses.
5576 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5578 57. Double the size of the debug message buffer (to 2048) so that more of very
5579 long debug lines gets shown.
5581 58. The exicyclog utility now does better if the number of log files to keep
5582 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5584 59. Two changes related to the smtp_active_hostname option:
5586 (1) $smtp_active_hostname is now available as a variable.
5587 (2) The default for smtp_banner uses $smtp_active_hostname instead
5588 of $primary_hostname.
5590 60. The host_aton() function is supposed to be passed a string that is known
5591 to be a valid IP address. However, in the case of IPv6 addresses, it was
5592 not checking this. This is a hostage to fortune. Exim now panics and dies
5593 if the condition is not met. A case was found where this could be provoked
5594 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5595 components; fortuitously, this particular loophole had already been fixed
5596 by change 4.50/55 above.
5598 If there are any other similar loopholes, the new check in host_aton()
5599 itself should stop them being exploited. The report I received stated that
5600 data on the command line could provoke the exploit when Exim was running as
5601 exim, but did not say which command line option was involved. All I could
5602 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5603 running as the user.
5606 61. There was a buffer overflow vulnerability in the SPA authentication code
5607 (which came originally from the Samba project). I have added a test to the
5608 spa_base64_to_bits() function which I hope fixes it.
5611 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5612 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5614 63. The daemon start-up calls getloadavg() while still root for those OS that
5615 need the first call to be done as root, but it missed one case: when
5616 deliver_queue_load_max is set with deliver_drop_privilege. This is
5617 necessary for the benefit of the queue runner, because there is no re-exec
5618 when deliver_drop_privilege is set.
5620 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5621 This has been fixed.
5623 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5624 was in use, was not putting the data itself into the right store pool;
5625 consequently, it could be overwritten for a subsequent message in the same
5626 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5629 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5631 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5632 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5633 the function were treating the return as a boolean value, which happened to
5634 work because 0=false and not-0=true, but is not correct code.
5636 68. The host_aton() function was not handling scoped IPv6 addresses (those
5637 with, for example, "%eth0" on the end) correctly.
5639 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5640 negated items (that is, ~something) in unsigned ints. Some compilers
5641 apparently mutter when there is no cast.
5643 70. If an address verification called from an ACL failed, and did not produce a
5644 user-specific message (i.e. there was only a "system" message), nothing was
5645 put in $acl_verify_message. In this situation, it now puts the system
5648 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5649 session; change 4.31/43 added the unwanted input to the log line - except
5650 that it did not do this in the start of session case. It now does.
5652 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5653 This is wrong and can cause the other end to generate a synchronization
5654 error if it is another Exim or anything else that does the synchronization
5655 check. A QUIT command is no longer sent after a timeout.
5657 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5658 during host lookups.
5660 74. The defer_ok option of callout verification was not working if it was used
5661 when verifying addresses in header lines, that is, for this case:
5663 verify = header_sender/callout=defer_ok
5665 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5666 those file descriptors could be used for SMTP connections. If anything
5667 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5668 could be sent to the SMTP client, causing chaos. The daemon now opens
5669 stdin, stdout, and stderr to /dev/null when it puts itself into the
5672 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5673 log by default. The user can override this with suitable Perl magic.
5675 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5676 the log message when discard triggers, was not working for the DATA ACL or
5677 for the non-SMTP ACL.
5679 78. Error message wording change in sieve.c.
5681 79. If smtp_accept_max_per_host was set, the number of connections could be
5682 restricted to fewer than expected, because the daemon was trying to set up
5683 a new connection before checking whether the processes handling previous
5684 connections had finished. The check for completed processes is now done
5685 earlier. On busy systems, this bug wouldn't be noticed because something
5686 else would have woken the daemon, and it would have reaped the completed
5689 80. If a message was submitted locally by a user whose login name contained one
5690 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5691 It caused a spool format error. I have fixed the spool reading code. A
5692 related problem was that the "from" clause in the Received: line became
5693 illegal because of the space(s). It is now covered by ${quote_local_part.
5695 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5698 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5700 83. Patch from Sieve maintainer to fix unterminated string problem in
5701 "vacation" handling.
5703 84. Some minor changes to the Linux configuration files to help with other
5704 OS variants using glibc.
5706 85. One more patch for Sieve to update vacation handling to latest spec.
5709 ----------------------------------------------------
5710 See the note above about the 4.44 and 4.50 releases.
5711 ----------------------------------------------------
5717 1. Change 4.43/35 introduced a bug that caused file counts to be
5718 incorrectly computed when quota_filecount was set in an appendfile
5721 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5722 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5725 3. Give more explanation in the error message when the command for a transport
5726 filter fails to execute.
5728 4. There are several places where Exim runs a non-Exim command in a
5729 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5730 was being done only for the command run by the queryprogram router. It is
5731 now done for all such subprocesses. The other cases are: ${run, transport
5732 filters, and the commands run by the lmtp and pipe transports.
5734 5. Some older OS have a limit of 256 on the maximum number of file
5735 descriptors. Exim was using setrlimit() to set 1000 as a large value
5736 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5737 systems. I've change it so that if it can't get 1000, it tries for 256.
5739 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5740 was an oversight, and furthermore, ever since the addition of extra
5741 controls (e.g. 4.43/32), the checks on when to allow different forms of
5742 "control" were broken. There should now be diagnostics for all cases when a
5743 control that does not make sense is encountered.
5745 7. $recipients is now available in the predata ACL (oversight).
5747 8. Tidy the search cache before the fork to do a delivery from a message
5748 received from the command line. Otherwise the child will trigger a lookup
5749 failure and thereby defer the delivery if it tries to use (for example) a
5750 cached ldap connection that the parent has called unbind on.
5752 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5753 of $address_data from the recipient verification was clobbered by the
5754 sender verification.
5756 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5757 was its contents. (It was OK if the option was not defined at all.)
5759 11. A "Completed" log line is now written for messages that are removed from
5760 the spool by the -Mrm option.
5762 12. $host_address is now set to the target address during the checking of
5763 ignore_target_hosts.
5765 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5766 being passed; this would have caused $sender_host_name to have been used if
5767 matching the list had actually called for a host name (not very likely,
5768 since this list is usually IP addresses). A host name is now passed as
5771 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5772 code that specifies a non-restarting handler (typically sigaction() in
5773 modern systems) in an attempt to fix a rare and obscure crash bug.
5775 15. Narrowed the window for a race in the daemon that could cause it to ignore
5776 SIGCHLD signals. This is not a major problem, because they are used only to
5777 wake it up if nothing else does.
5779 16. A malformed maildirsize file could cause Exim to calculate negative values
5780 for the mailbox size or file count. Odd effects could occur as a result.
5781 The maildirsize information is now recalculated if the size or filecount
5784 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5785 support for a long time. Removed HAVE_SYS_VFS_H.
5787 18. Updated exipick to current release from John Jetmore.
5789 19. Allow an empty sender to be matched against a lookup in an address list.
5790 Previously the only cases considered were a regular expression, or an
5793 20. Exim went into a mad DNS lookup loop when doing a callout where the
5794 host was specified on the transport, if the DNS lookup yielded more than
5797 21. The RFC2047 encoding function was originally intended for short strings
5798 such as real names; it was not keeping to the 75-character limit for
5799 encoded words that the RFC imposes. It now respects the limit, and
5800 generates multiple encoded words if necessary. To be on the safe side, I
5801 have increased the buffer size for the ${rfc2047: expansion operator from
5804 22. Failure to deliver a bounce message always caused it to be frozen, even if
5805 there was an errors_to setting on the router. The errors_to setting is now
5808 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5809 canonical form (fully expanded) before being placed in
5810 $sender_host_address.
5812 24. Updated eximstats to version 1.33
5814 25. Include certificate and key file names in error message when GnuTLS fails
5815 to set them up, because the GnuTLS error message doesn't include the name
5816 of the failing file when there is a problem reading it.
5818 26. Expand error message when OpenSSL has problems setting up cert/key files.
5821 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5822 (this can affect the format of dates).
5824 28. exim_tidydb, when checking for the continued existence of a message for
5825 which it has found a message-specific retry record, was not finding
5826 messages that were in split spool directories. Consequently, it was
5827 deleting retry records that should have stayed in existence.
5829 29. eximstats updated to version 1.35
5830 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5831 1.35 - bugfix such that pie charts by volume are generated correctly
5833 30. The SPA authentication driver was not abandoning authentication and moving
5834 on to the next authenticator when an expansion was forced to fail,
5835 contradicting the general specification for all authenticators. Instead it
5836 was generating a temporary error. It now behaves as specified.
5838 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5839 (the order specifies the preference for clients). The order is now AES256,
5840 AES128, 3DES, ARCFOUR128.
5842 31. Small patch to Sieve code - explicitly set From: when generating an
5845 32. Exim crashed if a remote delivery caused a very long error message to be
5846 recorded - for instance if somebody sent an entire SpamAssassin report back
5847 as a large number of 550 error lines. This bug was coincidentally fixed by
5848 increasing the size of one of Exim's internal buffers (big_buffer) that
5849 happened as part of the Exiscan merge. However, to be on the safe side, I
5850 have made the code more robust (and fixed the comments that describe what
5853 33. Some experimental protocols are using DNS PTR records for new purposes. The
5854 keys for these records are domain names, not reversed IP addresses. The
5855 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5856 leaves it alone. Component reversal etc. now happens only for IP addresses.
5859 34. The host_aton() function is supposed to be passed a string that is known
5860 to be a valid IP address. However, in the case of IPv6 addresses, it was
5861 not checking this. This is a hostage to fortune. Exim now panics and dies
5862 if the condition is not met. A case was found where this could be provoked
5863 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5864 components; fortuitously, this particular loophole had already been fixed
5865 by change 4.50/55 or 4.44/33 above.
5867 If there are any other similar loopholes, the new check in host_aton()
5868 itself should stop them being exploited. The report I received stated that
5869 data on the command line could provoke the exploit when Exim was running as
5870 exim, but did not say which command line option was involved. All I could
5871 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5872 running as the user.
5875 35. There was a buffer overflow vulnerability in the SPA authentication code
5876 (which came originally from the Samba project). I have added a test to the
5877 spa_base64_to_bits() function which I hope fixes it.
5880 36. The daemon start-up calls getloadavg() while still root for those OS that
5881 need the first call to be done as root, but it missed one case: when
5882 deliver_queue_load_max is set with deliver_drop_privilege. This is
5883 necessary for the benefit of the queue runner, because there is no re-exec
5884 when deliver_drop_privilege is set.
5886 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5887 was in use, was not putting the data itself into the right store pool;
5888 consequently, it could be overwritten for a subsequent message in the same
5889 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5892 38. Sometimes the final signoff response after QUIT could fail to get
5893 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5894 < 0 before doing a fflush(). This bug looks as though it goes back to the
5895 introduction of TLS in release 3.20, but "sometimes" must have been rare
5896 because the tests only now provoked it.
5902 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5903 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5904 Now it returns an integer. A number of calls were still expecting a T/F
5905 return. Fortuitously, in all cases, the tests worked in OK situations,
5906 which is the norm. However, things would have gone wrong on any write
5907 failures on the smtp file descriptor. This function is used when sending
5908 messages over SMTP and also when doing verify callouts.
5910 2. When Exim is called to do synchronous delivery of a locally submitted
5911 message (the -odf or -odi options), it no longer closes stderr before doing
5914 3. Implemented the mua_wrapper option.
5916 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5918 5. Implemented the functions header_remove(), header_testname(),
5919 header_add_at_position(), and receive_remove_recipient(), and exported them
5922 6. If an ACL "warn" statement specified the addition of headers, Exim already
5923 inserted X-ACL-Warn: at the start if there was no header name. However, it
5924 was not making this test for the second and subsequent header lines if
5925 there were newlines in the string. This meant that an invalid header could
5926 be inserted if Exim was badly configured.
5928 7. Allow an ACL "warn" statement to add header lines at the start or after all
5929 the Received: headers, as well as at the end.
5931 8. Added the rcpt_4xx retry error code.
5933 9. Added postmaster_mailfrom=xxx to callout verification option.
5935 10. Added mailfrom=xxxx to the callout verification option, for verify=
5938 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5939 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5941 12. Inserted some casts to stop certain compilers warning when using pointer
5942 differences as field lengths or precisions in printf-type calls (mostly
5943 affecting debugging statements).
5945 13. Added optional readline() support for -be (dynamically loaded).
5947 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5948 same clock tick as a message's arrival, so that its received time was the
5949 same as the "first fail" time on the retry record, and that message
5950 remained on the queue past the ultimate address timeout, every queue runner
5951 would try a delivery (because it was past the ultimate address timeout) but
5952 after another failure, the ultimate address timeout, which should have then
5953 bounced the address, did not kick in. This was a "< instead of <=" error;
5954 in most cases the first failure would have been in the next clock tick
5955 after the received time, and all would be well.
5957 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5958 being recognized when the domain list was tested by the match_domain
5959 condition in an expansion string.
5961 16. Added the ${str2b64: operator.
5963 17. Exim was always calling setrlimit() to set a large limit for the number of
5964 processes, without checking whether the existing limit was already
5965 adequate. (It did check for the limit on file descriptors.) Furthermore,
5966 errors from getrlimit() and setrlimit() were being ignored. Now they are
5967 logged to the main and panic logs, but Exim does carry on, to try to do its
5968 job under whatever limits there are.
5970 18. Imported PCRE 5.0.
5972 19. Trivial typo in log message " temporarily refused connection" (the leading
5975 20. If the log selector return_path_on_delivery was set and an address was
5976 redirected to /dev/null, the delivery process crashed because it assumed
5977 that a return path would always be set for a "successful" delivery. In this
5978 case, the whole delivery is bypassed as an optimization, and therefore no
5981 21. Internal re-arrangement: the function for sending a challenge and reading
5982 a response while authentication was assuming a zero-terminated challenge
5983 string. It's now changed to take a pointer and a length, to allow for
5984 binary data in such strings.
5986 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5988 23. Exim was not respecting finduser_retries when seeking the login of the
5989 uid under which it was called; it was always trying 10 times. (The default
5990 setting of finduser_retries is zero.) Also, it was sleeping after the final
5991 failure, which is pointless.
5993 24. Implemented tls_on_connect_ports.
5995 25. Implemented acl_smtp_predata.
5997 26. If the domain in control=submission is set empty, Exim assumes that the
5998 authenticated id is a complete email address when it generates From: or
5999 Sender: header lines.
6001 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6002 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6003 chown and chgrp in /bin and hostname in /usr/bin.
6005 28. Exim was keeping the "process log" file open after each use, just as it
6006 does for the main log. This opens the possibility of it remaining open for
6007 long periods when the USR1 signal hits a daemon. Occasional processlog
6008 errors were reported, that could have been caused by this. Anyway, it seems
6009 much more sensible not to leave this file open at all, so that is what now
6012 29. The long-running daemon process does not normally write to the log once it
6013 has entered its main loop, and it closes the log before doing so. This is
6014 so that log files can straightforwardly be renamed and moved. However,
6015 there are a couple of unusual error situations where the daemon does write
6016 log entries, and I had neglected to close the log afterwards.
6018 30. The text of an SMTP error response that was received during a remote
6019 delivery was being truncated at 512 bytes. This is too short for some of
6020 the long messages that one sometimes sees. I've increased the limit to
6023 31. It is now possible to make retry rules that apply only when a message has a
6024 specific sender, in particular, an empty sender.
6026 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6027 it possible to be selective about when SMTP synchronization is enforced.
6029 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6031 32. Implemented hosts_connection_nolog.
6033 33. Added an ACL for QUIT.
6035 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6038 35. Added mailbox_size and mailbox_filecount to appendfile.
6040 36. Added control = no_multiline_responses to ACLs.
6042 37. There was a bug in the logic of the code that waits for the clock to tick
6043 in the case where the clock went backwards by a substantial amount such
6044 that the microsecond fraction of "now" was more than the microsecond
6045 fraction of "then" (but the whole seconds number was less).
6047 38. Added support for the libradius Radius client library this is found on
6048 FreeBSD (previously only the radiusclient library was supported).
6054 1. When certain lookups returned multiple values in the form name=value, the
6055 quoting of the values was not always being done properly. Specifically:
6056 (a) If the value started with a double quote, but contained no whitespace,
6058 (b) If the value contained whitespace other than a space character (i.e.
6059 tabs or newlines or carriage returns) it was not quoted.
6060 This fix has been applied to the mysql and pgsql lookups by writing a
6061 separate quoting function and calling it from the lookup code. The fix
6062 should probably also be applied to nisplus, ibase and oracle lookups, but
6063 since I cannot test any of those, I have not disturbed their existing code.
6065 2. A hit in the callout cache for a specific address caused a log line with no
6066 reason for rejecting RCPT. Now it says "Previous (cached) callout
6067 verification failure".
6069 3. There was an off-by-one bug in the queryprogram router. An over-long
6070 return line was truncated at 256 instead of 255 characters, thereby
6071 overflowing its buffer with the terminating zero. As well as fixing this, I
6072 have increased the buffer size to 1024 (and made a note to document this).
6074 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6075 when Exim is waiting for an SMTP response from a remote server, Exim
6076 restarts its select() call on the socket, thereby resetting its timeout.
6077 This is not a problem when such interrupts are rare. Somebody set up a cron
6078 job to run exiwhat every 2 minutes, which is less than the normal select()
6079 timeout (5 or 10 minutes). This meant that the select() timeout never
6080 kicked in because it was always reset. I have fixed this by comparing the
6081 time when an interrupt arrives with the time at the start of the first call
6082 to select(). If more time than the timeout has elapsed, the interrupt is
6083 treated as a timeout.
6085 5. Some internal re-factoring in preparation for the addition of Sieve
6086 extensions (by MH). In particular, the "personal" test is moved to a
6087 separate function, and given an option for scanning Cc: and Bcc: (which is
6088 not set for Exim filters).
6090 6. When Exim created an email address using the login of the caller as the
6091 local part (e.g. when creating a From: or Sender: header line), it was not
6092 quoting the local part when it contained special characters such as @.
6094 7. Installed new OpenBSD configuration files.
6096 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6097 try to make them clearer.
6099 9. Callout options, other than the timeout value, were being ignored when
6100 verifying sender addresses in header lines. For example, when using
6102 verify = header_sender/callout=no_cache
6104 the cache was (incorrectly) being used.
6106 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6107 only the Cygwin environment.
6109 11. When return_path_on_delivery was set as a log selector, if different remote
6110 addresses in the same message used different return paths and parallel
6111 remote delivery occurred, the wrong values would sometimes be logged.
6112 (Whenever a remote delivery process finished, the return path value from
6113 the most recently started remote delivery process was logged.)
6115 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6116 lines when AUTH and/or TLS are in use. This is the "received protocol"
6117 field. Exim used to use "asmtp" for authenticated SMTP, without any
6118 indication (in the protocol name) for TLS use. Now it follows the RFC and
6119 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6120 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6121 names appear in log lines as well as in Received: header lines.
6123 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6124 extensions, and comparison tests, and to fix some bugs.
6126 14. Changes to the "personal" filter test:
6128 (1) The test was buggy in that it was just doing the equivalent of
6129 "contains" tests on header lines. For example, if a user's address was
6130 anne@some.where, the "personal" test would incorrectly be true for
6132 To: susanne@some.where
6134 This test is now done by extracting each address from the header in turn,
6135 and checking the entire address. Other tests that are part of "personal"
6136 are now done using regular expressions (for example, to check local parts
6137 of addresses in From: header lines).
6139 (2) The list of non-personal local parts in From: addresses has been
6140 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6141 taken from the Sieve specification recommendations.
6143 (3) If the message contains any header line starting with "List-" it is
6144 treated as non-personal.
6146 (4) The test for "circular" in the Subject: header line has been removed
6147 because it now seems ill-conceived.
6149 15. Minor typos in src/EDITME comments corrected.
6151 16. Installed latest exipick from John Jetmore.
6153 17. If headers_add on a router specified a text string that was too long for
6154 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6155 of string_sprintf() is now avoided.
6157 18. $message_body_size was not set (it was always zero) when running the DATA
6158 ACL and the local_scan() function.
6160 19. For the "mail" command in an Exim filter, no default was being set for
6161 the once_repeat time, causing a random time value to be used if "once" was
6162 specified. (If the value happened to be <= 0, no repeat happened.) The
6163 default is now 0s, meaning "never repeat". The "vacation" command was OK
6164 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6165 (I found it when inspecting the code).
6167 20. There is now an overall timeout for performing a callout verification. It
6168 defaults to 4 times the callout timeout, which applies to individual SMTP
6169 commands during the callout. The overall timeout applies when there is more
6170 than one host that can be tried. The timeout is checked before trying the
6171 next host. This prevents very long delays if there are a large number of
6172 hosts and all are timing out (e.g. when the network connections are timing
6173 out). The value of the overall timeout can be changed by specifying an
6174 additional sub-option for "callout", called "maxwait". For example:
6176 verify = sender/callout=5s,maxwait=20s
6178 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6179 to the end before writing, but this should make it even safer).
6181 22. Exim was forgetting that it had advertised PIPELINING for the second and
6182 subsequent messages on an SMTP connection. It was also not resetting its
6183 memory on STARTTLS and an internal HELO.
6185 23. When Exim logs an SMTP synchronization error within a session, it now
6186 records whether PIPELINING has been advertised or not.
6188 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6189 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6190 rather than long int.
6192 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6194 26. Added the never_mail option to autoreply.
6200 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6201 crash if the getsockname() call failed; this can happen if a connection is
6202 closed very soon after it is established. The problem was simply in the
6203 order in which certain operations were done, causing Exim to try to write
6204 to the SMTP stream before it had set up the file descriptor. The bug has
6205 been fixed by making things happen in the correct order.
6211 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6212 before the connection was closed, thus losing the rejection response.
6214 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6215 some early Solaris releases, but causes trouble in current releases where
6216 socklen_t is defined.
6218 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6221 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6224 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6225 that releasing the top of it at the end releases what was used for sub-
6226 expansions (unless the block got too big). However, discard this block if
6227 the first thing is a variable or header, so that we can use its block when
6228 it is dynamic (useful for very large $message_headers, for example).
6230 6. Lookups now cache *every* query, not just the most recent. A new, separate
6231 store pool is used for this. It can be recovered when all lookup caches are
6232 flushed. Lookups now release memory at the end of their result strings.
6233 This has involved some general refactoring of the lookup sources.
6235 7. Some code has been added to the store_xxx() functions to reduce the amount
6236 of flapping under certain conditions.
6238 8. log_incoming_interface used to affect only the <= reception log lines. Now
6239 it causes the local interface and port to be added to several more SMTP log
6240 lines, for example "SMTP connection from", and rejection lines.
6242 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6244 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6246 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6247 could be overwritten at the end of the current message (or the start of a
6248 new message if it was set in a HELO ACL). The value is now preserved for
6249 the duration of the SMTP connection.
6251 12. If a transport had a headers_rewrite setting, and a matching header line
6252 contained an unqualified address, that address was qualified, even if it
6253 did not match any rewriting rules. The underlying bug was that the values
6254 of the flags that permit the existence of unqualified sender and recipient
6255 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6256 non-local messages, and by -bnq for local messages) were not being
6257 preserved with the message after it was received.
6259 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6260 "next input=" as part of the text comprising the host identity instead of
6261 the correct text. The code was using the same buffer for two different
6262 strings. However, depending on which order the printing function evaluated
6263 its arguments, the bug did not always show up. Under Linux, for example, my
6264 test suite worked just fine.
6266 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6267 doesn't work with some older versions of Perl. It has been changed to "my",
6268 which in any case is probably the better facility to use.
6270 15. A really picky compiler found some instances of statements for creating
6271 error messages that either had too many or two few arguments for the format
6274 16. The size of the buffer for calls to the DNS resolver has been increased
6275 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6276 for addresses that have a lot of PTR records. This alleviates a problem; it
6277 does not fully solve it.
6279 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6280 buffer now truncates the list and logs the incident, which is the same
6281 action as happens when Exim is looking up a host name and its aliases.
6282 Previously in this situation something unpredictable would happen;
6283 sometimes it was "internal error: store_reset failed".
6285 18. If a server dropped the connection unexpectedly when an Exim client was
6286 using GnuTLS and trying to read a response, the client delivery process
6287 crashed while trying to generate an error log message.
6289 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6290 string, for example:
6292 warn message = H1: something\nH2: something
6294 the text was added as a single header line from Exim's point of view
6295 though it ended up OK in the delivered message. However, searching for the
6296 second and subsequent header lines using $h_h2: did not work. This has been
6297 fixed. Similarly, if a system filter added multiple headers in this way,
6298 the routers could not see them.
6300 20. Expanded the error message when iplsearch is called with an invalid key to
6301 suggest using net-iplsearch in a host list.
6303 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6304 ACLs are no longer actually imposed (and a message to that effect is
6307 22. If a "gecos" field in a passwd entry contained escaped characters, in
6308 particular, if it contained a \" sequence, Exim got it wrong when building
6309 a From: or a Sender: header from that name. A second bug also caused
6310 incorrect handling when an unquoted " was present following a character
6311 that needed quoting.
6313 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6314 was not being matched caselessly.
6316 24. Arranged for all hyphens in the exim.8 source to be escaped with
6319 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6320 properly. Recipient callout cache records were still being keyed to include
6321 the sender, even when use_sender was set false. This led to far more
6322 callouts that were necessary. The sender is no longer included in the key
6323 when use_sender is false.
6325 26. Added "control = submission" modifier to ACLs.
6327 27. Added the ${base62d: operator to decode base 62 numbers.
6329 28. dnsdb lookups can now access SRV records.
6331 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6332 the configuration file.
6334 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6335 (-v) mode. This makes the output for a verbose queue run more intelligible.
6337 31. Added a use_postmaster feature to recipient callouts.
6339 32. Added the $body_zerocount variable, containing the number of binary zero
6340 bytes in the message body.
6342 33. The time of last modification of the "new" subdirectory is now used as the
6343 "mailbox time last read" when there is a quota error for a maildir
6346 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6348 35. Added +ignore_unknown as a special item in host lists.
6350 36. Code for decoding IPv6 addresses in host lists is now included, even if
6351 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6352 address was recognized as an IP address, but was then not correctly decoded
6353 into binary, causing unexpected and incorrect effects when compared with
6360 1. Very minor rewording of debugging text in manualroute to say "list of
6361 hosts" instead of "hostlist".
6363 2. If verify=header_syntax was set, and a header line with an unqualified
6364 address (no domain) and a large number of spaces between the end of the
6365 name and the colon was received, the reception process suffered a buffer
6366 overflow, and (when I tested it) crashed. This was caused by some obsolete
6367 code that should have been removed. The fix is to remove it!
6369 3. When running in the test harness, delay a bit after writing a bounce
6370 message to get a bit more predictability in the log output.
6372 4. Added a call to search_tidyup() just before forking a reception process. In
6373 theory, someone could use a lookup in the expansion of smtp_accept_max_
6374 per_host which, without the tidyup, could leave open a database connection.
6376 5. Added the variables $recipient_data and $sender_data which get set from a
6377 lookup success in an ACL "recipients" or "senders" condition, or a router
6378 "senders" option, similar to $domain_data and $local_part_data.
6380 6. Moved the writing of debug_print from before to after the "senders" test
6383 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6384 problems for message scanning, either using a data ACL, or using
6385 local_scan() because the Received: header was not generated till after they
6386 were called (in order to set the time as the time of reception completion).
6387 I have revised the way this works. The header is now generated after the
6388 body is received, but before the ACL or local_scan() are called. After they
6389 are run, the timestamp in the header is updated.
6395 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6396 before starting a queue runner without re-exec. This happened only when
6397 deliver_drop_privilege was set or when the Exim user was set to root. The
6398 effect of the bug was that timeouts during subsequent deliveries caused
6399 crashes instead of being properly handled. The handler is now left at its
6400 default (and expected) setting.
6402 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6403 message, again when deliver_drop_privilege is set or Exim is run as root.
6404 The bug described in (1) was not present in this case, but the tidying up
6405 of the other signals was missing. I have made the two cases consistent.
6407 3. The ignore_target_hosts setting on a manualroute router was being ignored
6408 for hosts that were looked up using the /MX notation.
6410 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6413 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6414 operated on the sender address. After changing the $sender_address to <>
6415 for the sender address verify, Exim was re-instated it as the original
6416 (before rewriting) address, but remembering that it had rewritten it, so it
6417 wasn't rewriting it again. This bug also had the effect of breaking the
6418 sender address verification caching when the sender address was rewritten.
6420 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6421 This has been changed so that if the ip literal address matches
6422 ignore_target_hosts, the router declines.
6424 7. Added expansion conditions match_domain, match_address, and match_local_
6425 part (NOT match_host).
6427 8. The placeholder for the Received: header didn't have a length field set.
6429 9. Added code to Exim itself and to exim_lock to test for a specific race
6430 condition that could lead to file corruption when using MBX delivery. The
6431 issue is with the lockfile that is created in /tmp. If this file is removed
6432 after a process has opened it but before that process has acquired a lock,
6433 there is the potential for a second process to recreate the file and also
6434 acquire a lock. This could lead to two Exim processes writing to the file
6435 at the same time. The added code performs the same test as UW imapd; it
6436 checks after acquiring the lock that its file descriptor still refers to
6437 the same named file.
6439 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6440 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6443 11. Added the smtp_active_hostname option. If used, this will typically be made
6444 to depend on the incoming interface address. Because $interface_address is
6445 not set up until the daemon has forked a reception process, error responses
6446 that can happen earlier (such as "too many connections") no longer contain
6449 12. If an expansion in a condition on a "warn" statement fails because a lookup
6450 defers, the "warn" statement is abandoned, and the next ACL statement is
6451 processed. Previously this caused the whole ACL to be aborted.
6453 13. Added the iplsearch lookup type.
6455 14. Added ident_timeout as a log selector.
6457 15. Added tls_certificate_verified as a log selector.
6459 16. Added a global option tls_require_ciphers (compare the smtp transport
6460 option of the same name). This controls incoming TLS connections.
6462 17. I finally figured out how to make tls_require_ciphers do a similar thing
6463 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6464 before starting the TLS session.
6466 18. Tabs are now shown as \t in -bP output.
6468 19. If the log selector return_path_on_delivery was set, Exim crashed when
6469 bouncing a message because it had too many Received: header lines.
6471 20. If two routers both had headers_remove settings, and the first one included
6472 a superfluous trailing colon, the final name in the first list and the
6473 first name in the second list were incorrectly joined into one item (with a
6474 colon in the middle).
6480 1. Added -C and -D options to the exinext utility, mainly to make it easier
6481 to include in the automated testing, but these could be helpful when
6482 multiple configurations are in use.
6484 2. The exinext utility was not formatting the output nicely when there was
6485 an alternate port involved in the retry record key, nor when there was a
6486 message id as well (for retries that were specific to a specific message
6487 and a specific host). It was also confused by IPv6 addresses, because of
6488 the additional colons they contain. I have fixed the IPv4 problem, and
6489 patched it up to do a reasonable job for IPv6.
6491 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6492 delivery, the log line now contains "pipelined" if PIPELINING was used.
6494 4. An SMTP transport process used to panic and die if the bind() call to set
6495 an explicit outgoing interface failed. This has been changed; it is now
6496 treated in the same way as a connect() failure.
6498 5. A reference to $sender_host_name in the part of a conditional expansion
6499 that was being skipped was still causing a DNS lookup. This no longer
6502 6. The def: expansion condition was not recognizing references to header lines
6503 that used bh_ and bheader_.
6505 7. Added the _cache feature to named lists.
6507 8. The code for checking quota_filecount in the appendfile transport was
6508 allowing one more file than it should have been.
6510 9. For compatibility with Sendmail, the command line option
6518 and sets the incoming protocol and host name (for trusted callers). The
6519 host name and its colon can be omitted when only the protocol is to be set.
6520 Note the Exim already has two private options, -pd and -ps, that refer to
6521 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6522 "s", but I don't think that's a major issue.
6524 10. A number of refactoring changes to the code, none of which should affect
6527 (a) The number of logging options was getting close to filling up the
6528 32-bit word that was used as a bit map. I have split them into two classes:
6529 those that are passed in the argument to log_write(), and those that are
6530 only ever tested independently outside of that function. These are now in
6531 separate 32-bit words, so there is plenty of room for expansion again.
6532 There is no change in the user interface or the logging behaviour.
6534 (b) When building, for example, log lines, the code previously used a
6535 macro that called string_cat() twice, in order to add two strings. This is
6536 not really sufficiently general. Furthermore, there was one instance where
6537 it was actually wrong because one of the argument was used twice, and in
6538 one call a function was used. (As it happened, calling the function twice
6539 did not affect the overall behaviour.) The macro has been replaced by a
6540 function that can join an arbitrary number of extra strings onto a growing
6543 (c) The code for expansion conditions now uses a table and a binary chop
6544 instead of a serial search (which was left over from when there were very
6545 few conditions). Also, it now recognizes conditions like "pam" even when
6546 the relevant support is not compiled in: a suitably worded error message is
6547 given if an attempt is made to use such a condition.
6549 11. Added ${time_interval:xxxxx}.
6551 12. A bug was causing one of the ddress fields not to be passed back correctly
6552 from remote delivery subprocesses. The field in question was not being
6553 subsequently used, so this caused to problems in practice.
6555 13. Added new log selectors queue_time and deliver_time.
6557 14. Might have fixed a bug in maildirsizefile handling that threw up
6558 "unexpected character" debug warnings, and recalculated the data
6559 unnecessarily. In any case, I expanded the warning message to give more
6562 15. Added the message "Restricted characters in address" to the statements in
6563 the default ACL that block characters like @ and % in local parts.
6565 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6566 Three changes have been made:
6568 (a) There was a serious bug; a negative response to MAIL caused the whole
6569 recipient domain to be cached as invalid, thereby blocking all messages
6570 to all local parts at the same domain, from all senders. This bug has
6571 been fixed. The domain is no longer cached after a negative response to
6572 MAIL if the sender used is not empty.
6574 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6577 (c) A new callout option, "use_sender" has been added for people who want
6578 the modified behaviour.
6584 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6587 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6588 indeed breaks things for older releases.
6590 3. Added additional logging to the case where there is a problem reading data
6591 from a filter that is running in a subprocess using a pipe, in order to
6592 try to track down a specific problem.
6594 4. Testing facility fudge: when running in the test harness and attempting
6595 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6596 getting "No route to host". Convert this to a timeout.
6598 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6601 6. Some OS don't have socklen_t but use size_t instead. This affects the
6602 fifth argument of getsockopt() amongst other things. This is now
6603 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6604 can be set for individual OS. I have set it for SunOS5, OSF1, and
6605 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6606 some earlier ones do not.
6608 7. Change 4.30/15 was not doing the test caselessly.
6610 8. The standard form for an IPv6 address literal was being rejected by address
6611 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6612 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6613 this, as well as the form without the "IPv6" on the front (but only when
6614 address literals are enabled, of course).
6616 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6618 10. Exim crashed if a message with an empty sender address specified by -f
6619 encountered a router with an errors_to setting. This could be provoked only
6620 by a command such as
6624 where an empty string was supplied; "<>" did not hit this bug.
6626 11. Installed PCRE release 4.5.
6628 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6629 remained set. It is now erased.
6631 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6632 times from message ids (which are base 36 rather than the normal 62).
6634 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6635 were being counted as actual protocol errors, and logged if the log
6636 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6637 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6638 and DATA following a set of rejected RCPTs do not count as protocol errors.
6639 In other words, Exim assumes they were pipelined, though this may not
6640 actually be the case. Of course, in all cases the client gets an
6641 appropriate error code.
6643 15. If a lookup fails in an ACL condition, a message about the failure may
6644 be available; it is used if testing the ACL cannot continue, because most
6645 such messages specify what the cause of the deferral is. However, some
6646 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6647 that caused an old message to be retained and used if a later statement
6648 caused a defer, replacing the real cause of the deferral.
6650 16. If an IP address had so many PTR records that the DNS lookup buffer
6651 was not large enough to hold them, Exim could crash while trying to process
6652 the truncated data. It now detects and logs this case.
6654 17. Further to 4.21/58, another change has been made: if (and only if) the
6655 first line of a message (the first header line) ends with CRLF, a bare LF
6656 in a subsequent header line has a space inserted after it, so as not to
6657 terminate the header.
6659 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6660 unnecessarily, used atoi() instead of strtol(), and didn't check the
6661 termination when getting file sizes from file names by regex.
6663 19. Completely re-implemented the support for maildirsize files, in the light
6664 of a number of problems with the previous contributed implementation
6665 (4.30/29). In particular:
6667 . If the quota is zero, the maildirsize file is maintained, but no quota is
6670 . If the maildir directory does not exist, it is created before any attempt
6671 to write a maildirsize file.
6673 . The quota value in the file is just a cache; if the quota is changed in
6674 the transport, the new value overrides.
6676 . A regular expression is available for excluding directories from the
6679 20. The autoreply transport checks the characters in options that define the
6680 message's headers; it allows continued headers, but it was checking with
6681 isspace() after an embedded newline instead of explicitly looking for a
6684 21. If all the "regular" hosts to which an address was routed had passed their
6685 expiry times, and had not reached their retry times, the address was
6686 bounced, even if fallback hosts were defined. Now Exim should go on to try
6689 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6690 equivalent code in the SMTP transport. Some hosts send humungous responses
6691 to HELO/EHLO, more than 1024 it seems.
6693 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6694 gives compiler warnings in some environments. I've now done it "properly",
6697 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6698 (because of problems with the built-in one) was declared to return uschar *
6699 instead of char *, causing compiler failure.
6701 25. Fixed a file descriptor leak when processing alias/forward files.
6703 26. Fixed a minor format string issue in dbfn.c.
6705 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6707 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6708 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6711 29. When a sender address is verified, it is cached, to save repeating the test
6712 when there is more than one recipient in a message. However, when the
6713 verification involves a callout, it is possible for different callout
6714 options to be set for different recipients. It is too complicated to keep
6715 track of this in the cache, so now Exim always runs a verification when a
6716 callout is required, relying on the callout cache for the optimization.
6717 The overhead is duplication of the address routing, but this should not be
6720 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6721 to be verified with callout=postmaster, and the main callout worked but the
6722 postmaster check failed, the verification correctly failed. However, if a
6723 subsequent RCPT command asked for sender verification *without* the
6724 postmaster check, incorrect caching caused this verification also to fail,
6727 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6728 it was not caching the DNS options (qualify_single, search_parents) that
6729 were used when the lookup failed. A subsequent lookup with different
6730 options therefore always gave the same answer, though there were cases
6731 where it should not have. (Example: a "domains = !$mx_any" option on a
6732 dnslookup router: the "domains" option is always processed without any
6733 widening, but the router might have qualify_single set.) Now Exim uses the
6734 cached value only when the same options are set.
6736 32. Added John Jetmore's "exipick" utility to the distribution.
6738 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6739 than a timeout (e.g. a certificate is required, and is not provided), an
6740 Exim server now closes the connection immediately. Previously it waited for
6741 the client to close - but if the client is SSL, it seems that they each
6742 wait for each other, leading to a delay before one of them times out.
6744 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6745 maintained 0.8.x compatibility because I don't think many are using it, and
6746 it is clearly obsolete.
6748 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6751 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6752 client certificate was expired. A simple patch fixes this, though I don't
6753 understand the full logic of why the verify callback is called multiple
6756 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6757 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6758 which causes problems with some clients (such as the Certicom SSL Plus
6759 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6760 disables the coutermeasure allowing Eudora to connect."
6762 38. Exim was not checking that a write() to a log file succeeded. This could
6763 lead to Bad Things if a log got too big, in particular if it hit a file
6764 size limit. Exim now panics and dies if it cannot write to a log file, just
6765 as it does if it cannot open a log file.
6767 39. Modified OS/Makefile-Linux so that it now contains
6769 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6771 The two -D definitions ensure that Exim is compiled with large file
6772 support, which makes it possible to handle log files that are bigger than
6775 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6776 instance) a domain was checked against a named list that involved a lookup,
6777 causing $domain_data to be set, then another domain was checked against the
6778 same list, then the first domain was re-checked, the value of $domain_data
6779 after the final check could be wrong. In particular, if the second check
6780 failed, it could be set empty. This bug probably also applied to
6783 41. The strip_trailing_dot option was not being applied to the address given
6784 with the -f command-line option.
6786 42. The code for reading a message's header from the spool was incrementing
6787 $received_count, but never initializing it. This meant that the value was
6788 incorrect (doubled) while delivering a message in the same process in which
6789 it was received. In the most common configuration of Exim, this never
6790 happens - a fresh exec is done - but it can happen when
6791 deliver_drop_privilege is set.
6793 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6794 it now includes up to 150 characters of the unexpected data in the log
6797 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6798 and building data strings. The size of both of these buffers was 10 000
6799 bytes - far larger than anybody would *ever* want, thought I. Needless to
6800 say, somebody hit the limit. I have increased the maximum line length to
6801 20 000 and the maximum data length of concatenated lines to 100 000. I have
6802 also fixed two bugs, because there was no checking on these buffers. Tsk,
6803 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6804 buffer is too small.
6806 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6807 lsearch lookups. Now it does.
6809 46. When parsing a route_list item in a manualroute router, a fixed-length
6810 buffer was used for the list of hosts. I made this 1024 bytes long,
6811 thinking that nobody would ever have a list of hosts that long. Wrong.
6812 Somebody had a whole pile of complicated expansion conditions, and the
6813 string was silently truncated, leading to an expansion error. It turns out
6814 that it is easier to change to an unlimited length (owing to other changes
6815 that have happened since this code was originally written) than to build
6816 structure for giving a limitation error. The length of the item that
6817 expands into the list of hosts is now unlimited.
6819 47. The lsearch lookup could not handle data where the length of text line was
6820 more than 4095 characters. Such lines were truncated, leading to shortened
6821 data being returned. It should now handle lines of any length.
6823 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6824 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6827 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6828 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6829 no longer happens in comment lines. A list of replacements is now placed
6830 at the head of all of the source files, except those whose only change is
6831 to replace PERL_COMMAND in the very first #! line.
6833 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6834 messages on the queue, with a bottom-up merge sort, using code contributed
6835 by Michael Haardt. This should make operations like -bp somewhat faster on
6836 large queues. It won't affect queue runners, except when queue_run_in_order
6839 51. Installed eximstats 1.31 in the distribution.
6841 52. Added support for SRV lookups to the dnslookup router.
6843 53. If an ACL referred to $message_body or $message_body_end, the value was not
6844 reset for any messages that followed in the same SMTP session.
6846 54. The store-handling optimization for building very long strings was not
6847 differentiating between the different store pools. I don't think this
6848 actually made any difference in practice, but I've tidied it.
6850 55. While running the routers to verify a sender address, $sender_address
6851 was still set to the sender address. This is wrong, because when routing to
6852 send a bounce to the sender, it would be empty. Therefore, I have changed
6853 it so that, while verifying a sender address, $sender_address is set to <>.
6854 (There is no change to what happens when verifying a recipient address.)
6856 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6857 target A or AAAA records (if not already returned) without resetting the
6858 qualify_single or search_parents options of the DNS resolver. These are
6859 inappropriate in this case because the targets of MX and SRV records must
6860 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6861 target that, when qualified, matched something in the local domain. These
6862 two options are now turned off when doing these lookups.
6864 57. It seems that at least some releases of Reiserfs (which does not have the
6865 concept of a fixed number of inodes) returns zero and not -1 for the
6866 number of available inodes. This interacted badly with check_spool_inodes,
6867 which assumed that -1 was the "no such thing" setting. What I have done is
6868 to check that the total number of inodes is greater than zero before doing
6869 the test of how many are available.
6871 58. When a "warn" ACL statement has a log_message modifier, the message is
6872 remembered, and not repeated. This is to avoid a lot of repetition when a
6873 message has many recipients that cause the same warning to be written.
6874 However, Exim was preserving the list of already written lines for an
6875 entire SMTP session, which doesn't seem right. The memory is now reset if a
6876 new message is started.
6878 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6879 debugging output unless log_rewrite was also set.
6881 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6882 of (int)(handle) when we know that handle contains (void *)(-1).
6884 61. The Exim daemon panic-logs an error return when it closes the incoming
6885 connection. However "connection reset by peer" seems to be common, and
6886 isn't really an error worthy of noting specially, so that particular error
6889 62. When Exim is trying to find all the local interfaces, it used to panic and
6890 die if the ioctl to get the interface flags failed. However, it seems that
6891 on at least one OS (Solaris 9) it is possible to have an interface that is
6892 included in the list of interfaces, but for which you get a failure error
6893 for this call. This happens when the interface is not "plumbed" into a
6894 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6895 failure of the "get flags" call assumes that the interface is down.
6897 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6898 makes life easier for people who are doing arithmetic on fields extracted
6899 from dates, where you often get leading zeros that should not be
6900 interpreted as octal.
6902 64. Added qualify_domain to the redirect router, to override the global
6905 65. If a pathologically long header line contained very many addresses (the
6906 report of this problem mentioned 10 000) and each of them was rewritten,
6907 Exim could use up a very large amount of memory. (It kept on making new
6908 copies of the header line as it rewrote, and never released the old ones.)
6909 At the expense of a bit more processing, the header rewriting function has
6910 been changed so that it no longer eats memory in this way.
6912 66. The generation of the Received: header has been moved from the time that a
6913 message starts to be received, to the time that it finishes. The timestamp
6914 in the Received: header should now be very close to that of the <= log
6915 line. There are two side-effects of this change:
6917 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6918 logged header lines no longer include the local Received: line, because
6919 it has not yet been created. The same applies to a copy of the message
6920 that is returned to a non-SMTP sender when a message is rejected.
6922 (b) When a filter file is tested using -bf, no additional Received: header
6923 is added to the test message. After some thought, I decided that this
6926 This change does not affect the value of $received_for. It is still set
6927 after address rewriting, but before local_scan() is called.
6929 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6931 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6932 gave an unhelpful panic error message, and a defer error. I have managed to
6933 change this behaviour so that it now rejects any supplied certificate,
6934 which seems right, as the list of acceptable certificates is empty.
6936 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6937 gave an unhelpful defer error. I have not managed to make this reject any
6938 supplied certificates, but the error message it gives is "no certificate
6939 supplied", which is not helpful.
6941 70. exigrep's output now also includes lines that are not associated with any
6942 message, but which match the given pattern. Implemented by a patch from
6943 Martin Sluka, which also tidied up the Perl a bit.
6945 71. Recipient callout verification, like sender verification, was using <> in
6946 the MAIL FROM command. This isn't really the right thing, since the actual
6947 sender may affect whether the remote host accepts the recipient or not. I
6948 have changed it to use the actual sender in the callout; this means that
6949 the cache record is now keyed on a recipient/sender pair, not just the
6950 recipient address. There doesn't seem to be a real danger of callout loops,
6951 since a callout by the remote host to check the sender would use <>.
6952 [SEE ABOVE: changed after hitting problems.]
6954 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6955 temporary errors. However, in the case of such a code being given after
6956 the end of a data transmission (i.e. after ".") Exim was failing to write
6957 a retry record for the message. (Yes, there was some broken host that was
6958 actually sending 8xx at this point.)
6960 73. An unknown lookup type in a host list could cause Exim to panic-die when
6961 the list was checked. (An example that provoked this was putting <; in the
6962 middle of a list instead of at the start.) If this happened during a DATA
6963 ACL check, a -D file could be left lying around. This kind of configuration
6964 error no longer causes Exim to die; instead it causes a defer error. The
6965 incident is still logged to the main and panic logs.
6967 74. Buglet left over from Exim 3 conversion. The message "too many messages
6968 in one connection" was written to the rejectlog but not the mainlog, except
6969 when address rewriting (yes!) was being logged.
6971 75. Added write_rejectlog option.
6973 76. When a system filter was run not as root (that is, when system_filter_user
6974 was set), the values of the $n variables were not being returned to the
6975 main process; thus, they were not subsequently available in the $sn
6978 77. Added +return_path_on_delivery log selector.
6980 78. A connection timeout was being treated differently from recipients deferred
6981 when testing hosts_max_try with a message that was older than the host's
6982 retry timeout. (The host should not be counted, thus allowing all hosts to
6983 be tried at least once before bouncing.) This may have been the cause of an
6984 occasionally reported bug whereby a message would remain on the queue
6985 longer than the retry timeout, but would be bounced if a delivery was
6986 forced. I say "may" because I never totally pinned down the problem;
6987 setting up timeout/retry tests is difficult. See also the next item.
6989 79. The ultimate address timeout was not being applied to errors that involved
6990 a combination of host plus message (for example, a timeout on a MAIL
6991 command). When an address resolved to a number of possible hosts, and they
6992 were not all tried for each delivery (e.g. because of hosts_max_try), a
6993 message could remain on the queue longer than the retry timeout.
6995 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6998 81. Fixed an obscure SMTP outgoing bug which required at least the following
6999 conditions: (a) there was another message waiting for the same server;
7000 (b) the server returned 5xx to all RCPT commands in the first message so
7001 that the message was not completed; (c) the server dropped the connection
7002 or gave a negative response to the RSET that Exim sends to abort the
7003 transaction. The observed case was a dropped connection after DATA that had
7004 been sent in pipelining mode. That is, the server had advertised PIPELINING
7005 but was not implementing it correctly. The effect of the bug was incorrect
7006 behaviour, such as trying another host, and this could lead to a crash.
7012 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7013 and daemon.c were passed as pointers to ints; they should have been
7014 pointers to socklen_t variables (which are typically unsigned ints).
7016 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7019 3. Fixed a really odd bug that affected only the testing scheme; patching a
7020 certain fixed string in the binary changed the value of another string that
7021 happened to be identical to the end of the original first string.
7023 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7024 name", it returns that address as the IP address. On some operating
7025 systems (e.g. Solaris), it also passes back the IP address string as the
7026 "host name". However, on others (e.g. Linux), it passes back an empty
7027 string. Exim wasn't checking for this, and was changing the host name to an
7028 empty string, assuming it had been canonicalized.
7030 5. Although rare, it is permitted to have more than one PTR record for a given
7031 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7032 all the names associated with an address, because they do in Solaris.
7033 However, it seems that they do not in Linux for data that comes from the
7034 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7035 I found this out when I moved to a new Linux workstation and tried to run
7036 the Exim test suite.
7038 To get round this problem I have changed the code so that it now does its
7039 own call to the DNS to look up PTR records when searching for a host name.
7040 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7041 addresses that are only in /etc/hosts are still found.
7043 This behaviour is, however, controlled by an option called host_lookup_
7044 order, which defaults to "bydns:byaddr". If people want to use the other
7045 order, or indeed, just use one or the other means of lookup, they can
7046 specify it in this variable.
7048 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7049 some operating systems, this comes back from gethostbyaddr() as an empty
7050 string, and this is what Exim used to test for. However, it seems that in
7051 other systems, "." is yielded. Exim now tests for this case too.
7053 7. The values of check_spool_space and check_log_space are now held internally
7054 as a number of kilobytes instead of an absolute number of bytes. If a
7055 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7056 kilobyte. This means that much larger values can be stored.
7058 8. Exim monitor: an attempt to get the action menu when not actually pointing
7059 at a message produces an empty menu entitled "No message selected". This
7060 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7061 no entries in it ("Shell widget menu has zero width and/or height"). So I
7062 have added a single, blank menu entry in this case.
7064 9. Added ${quote_local_part.
7066 10. MIME decoding is now applied to the contents of Subject: header lines when
7069 11. Now that a reference to $sender_host_address automatically causes a reverse
7070 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7071 host lookup before query-style lookups in lists that might use this
7072 variable. This has therefore been abolished, and the "net-" prefix is no
7073 longer necessary for query-style lookups.
7075 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7076 have been a typo for LFLAGS, so it has been changed.
7078 13. The install script calls Exim with "-C /dev/null" in order to find the
7079 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7080 to be output. However, since Exim outputs its version number before the
7081 error, it didn't break the script. It just looked ugly. I fixed this by
7082 always allowing "-C /dev/null" if the caller is root.
7084 14. Ignore overlarge ACL variable number when reading spool file - insurance
7085 against a later release with more variables having written the file.
7087 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7088 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7089 this, as well as the form without the "IPv6" on the front.
7091 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7092 OS/Makefile-Darwin file.
7094 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7095 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7097 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7098 variable. However, some OS use ints for inodes. I've added cast to long int
7099 to get rid of the compiler warning.
7101 19. I had forgotten to lock out "/../" in configuration file names when
7102 ALT_CONFIG_PREFIX was set.
7104 20. Routers used for verification do not need to specify transports. However,
7105 if such a router generated a host list, and callout was configured, Exim
7106 crashed, because it could not find a port number from the (non-existent)
7107 transport. It now assumes port 25 in this circumstance.
7109 21. Added the -t option to exigrep.
7111 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7112 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7113 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7114 them is set, LOOKUP_LSEARCH is forced.
7116 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7118 24. Added sender and host information to the "rejected by local_scan()" log
7119 line; previously there was no indication of these.
7121 25. Added .include_if_exists.
7123 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7124 Linux. It turns out that not all file systems support this. Apparently some
7125 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7126 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7127 not supported on the file descriptor, is now ignored when Exim is trying to
7128 sync a directory. This applies only to Linux.
7130 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7132 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7133 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7136 29. Added support for maildirsize files from supplied patch (modified a bit).
7138 30. The use of :fail: followed by an empty string could lead Exim to respond to
7139 sender verification failures with (e.g.):
7141 550 Verification failed for <xxx>
7142 550 Sender verify failed
7144 where the first response line was missing the '-' that indicates it is not
7145 the final line of the response.
7147 31. The loop for finding the name of the user that called Exim had a hardwired
7148 limit of 10; it now uses the value of finduser_retries, which is used for
7149 all other user lookups.
7151 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7154 33. Exim was neglecting to zero errno before one call of strtol() when
7155 expanding a string and expecting an integer value. On some systems this
7156 resulted in spurious "integer overflow" errors. Also, it was casting the
7157 result into an int without checking.
7159 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7160 did not work. The code looks as if it has *never* worked, though it appears
7161 to have been documented since at least release 1.62. I have made it work.
7163 35. The "timeout_DNS" error in retry rules, also documented since at least
7164 1.62, also never worked. As it isn't clear exactly what this means, and
7165 clearly it isn't a major issue, I have abolished the feature by treating it
7166 as "timeout", and writing a warning to the main and panic logs.
7168 36. The display of retry rules for -brt wasn't always showing the error code
7171 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7172 timeout_connect_A, timeout_connect_MX.
7174 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7175 to the empty sender.
7177 39. The daemon was not analysing the content of -oX till after it had closed
7178 stderr and disconnected from the controlling terminal. This meant that any
7179 syntax errors were only noted on the panic log, and the return code from
7180 the command was 0. By re-arranging the code a little, I've made the
7181 decoding happen first, so such errors now appear on stderr, and the return
7182 code is 1. However, the actual setting up of the sockets still happens in
7183 the disconnected process, so errors there are still only recorded on the
7186 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7187 connections (as happens on some IP stacks) was logged at start up time as
7188 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7189 it from "IPv6 and IPv4", which means that two separate sockets are being
7192 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7193 says whether AF_INET or AF_INET6 was passed as an argument.
7195 42. Exiwhat output was messed up when time zones were included in log
7198 43. Exiwhat now gives more information about the daemon's listening ports,
7199 and whether -tls-on-connect was used.
7201 44. The "port" option of the smtp transport is now expanded.
7203 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7204 silently ignored. Now an error message is written to the main and panic
7207 46. There's a new ACL modifier called "logwrite" which writes to a log file
7208 as soon as it is encountered.
7210 47. Added $local_user_uid and $local_user_gid at routing time.
7212 48. Exim crashed when trying to verify a sender address that was being
7215 49. Exim was recognizing only a space character after ".include". It now also
7216 recognizes a tab character.
7218 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7219 extracting the relevant information from the specification. The man page no
7220 longer contains scrambled data for the -d option, and I've added a section
7221 at the front about calling Exim under different names.
7223 51. Added "extra_headers" argument to the "mail" command in filter files.
7225 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7228 53. Installed eximstats 1.29.
7230 54. Added transport_filter_timeout as a generic transport option.
7232 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7233 Cc: header lines. This was required by RFC 822, but it not required by RFC
7236 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7237 incoming messages that did not have them. Now it does so only if the
7238 message originates locally, that is, if there is no associated remote host
7239 address. When Resent- header lines are present, this applies to the Resent-
7240 lines rather than the non-Resent- lines.
7242 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7243 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7245 58. Messages for configuration errors now include the name of the main
7246 configuration file - useful now that there may be more than one file in a
7247 list (.included file names were always shown).
7249 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7250 for those rare installations that do not start the daemon as root or run it
7251 setuid root. I've cut out the call to initgroups() if the daemon is not
7254 60. The Exim user and group can now be bound into the binary as text strings
7255 that are looked up at the start of Exim's processing.
7257 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7259 62. Added $mailstore_basename variable.
7261 63. Installed patch to sieve.c from Michael Haardt.
7263 64. When Exim failed to open the panic log after failing to open the main log,
7264 the original message it was trying to log was written to stderr and debug
7265 output, but if they were not available (the usual case in production), it
7266 was lost. Now it is written to syslog before the two lines that record the
7267 failures to open the logs.
7269 65. Users' Exim filters run in subprocesses under the user's uid. It is
7270 possible for a "deliver" command or an alias in a "personal" command to
7271 provoke an address rewrite. If logging of address rewriting is configured,
7272 this fails because the process is not running as root or exim. There may be
7273 a better way of dealing with this, but for the moment (because 4.30 needs
7274 to be released), I have disabled address rewrite logging when running a
7275 filter in a non-root, non-exim process.
7281 1. The buildconfig auxiliary program wasn't quoting the value set for
7282 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7283 not defined. This bug was masked in 4.22 by the effect that was fixed in
7286 2. Some messages that were rejected after a message id was allocated were
7287 shown as "incomplete" by exigrep. It no longer does this for messages that
7288 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7290 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7291 have allow_domain_literals set, the ID did not get logged in the <= line.
7292 Domain literals are now always recognized in Message-ID: header lines.
7294 4. The first argument for a ${extract expansion item is the key name or field
7295 number. Leading and trailing spaces in this item were not being ignored,
7296 causing some misleading effects.
7298 5. When deliver_drop_privilege was set, single queue runner processes started
7299 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7300 same command in the process it spins off) were not dropping privilege.
7302 6. When the daemon running as "exim" started a queue runner, it always
7303 re-executed Exim in the spun-off process. This is a waste of effort when
7304 deliver_drop_privilege is set. The new process now just calls the
7305 queue-runner function directly.
7311 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7314 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7315 ignored. Though the use of -oP was forcing the writing of a pid file, it
7316 was always written to the default place.
7318 3. If the message "no IP address found for host xxxx" is generated during
7319 incoming verification, it is now followed by identification of the incoming
7320 connection (so you can more easily find what provoked it).
7322 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7324 5. Added some features to "harden" Exim a bit more against certain attacks:
7326 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7327 be put in Local/Makefile. This is like the never_users runtime option,
7328 but it cannot be overridden. The default setting is "root".
7330 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7331 prefix string with which any file named in a -C command line option
7334 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7335 is retained for -C and -D only if the caller of Exim is root. Without
7336 it, the exim user may also use -C and -D and retain privilege.
7338 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7339 command line option is disabled.
7341 6. Macro names set by the -D option must start with an upper case letter, just
7342 like macro names defined in the configuration file.
7344 7. Added "dereference=" facility to LDAP.
7346 8. Two instances of the typo "uknown" in the source files are fixed.
7348 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7349 the Configure-Makefile script screwed up while processing it.
7351 10. Incorporated PCRE 4.4.
7353 11. The SMTP synchronization check was not operating right at the start of an
7354 SMTP session. For example, it could not catch a HELO sent before the client
7355 waited for the greeting. There is now a check for outstanding input at the
7356 point when the greeting is written. Because of the duplex, asynchronous
7357 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7358 way, but not yet received, when the check is performed.
7360 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7361 on TCP/IP sockets, because this apparently causes some broken clients to
7364 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7365 unchanged) from the Cygwin maintainer.
7367 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7368 was supported instead of testing for mbx. Effectively a typo.
7370 15. The spa authenticator server code was not checking that the input it
7371 received was valid base64.
7373 16. The debug output line for the "set" modifier in ACLs was not showing the
7374 name of the variable that was being set.
7376 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7378 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7379 lookup on its own. Something else was needed to trigger the lookup. For
7380 example, a match in host_lookup or the need for a host name in a host list.
7381 Now, if $sender_host_name is referenced and the host name has not yet been
7382 looked up, a lookup is performed. If the lookup fails, the variable remains
7383 empty, and $host_lookup_failed is set to "1".
7385 19. Added "eqi" as a case-independent comparison operator.
7387 20. The saslauthd authentication condition could segfault if neither service
7388 nor realm was specified.
7390 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7391 error message that was logged was misleading, and incoming SMTP
7392 connections were dropped. The message is now more accurate, and temporary
7393 errors are given to SMTP connections.
7395 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7396 (except RSET) with a 421 error, until QUIT is received. However, it was
7397 failing to send a response to QUIT.
7399 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7400 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7401 ACL is now run after the helo_try_verify_hosts code.
7403 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7404 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7405 case-independent, so other case variants are also recognized). Apparently
7406 some systems use these upper case variants.
7408 25. If more than two messages were waiting for the same host, and a transport
7409 filter was specified for the transport, Exim sent two messages over the
7410 same TCP/IP connection, and then failed with "socket operation on non-
7411 socket" when it tried to send the third.
7413 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7415 27. The extern definition of crypt16() in expand.c was not being excluded when
7416 the OS had its own crypt16() function.
7418 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7419 as a preferred synonym for return_size_limit, both as an option and as an
7422 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7424 30. Changed the default configuration ACL to relax the local part checking rule
7425 for addresses that are not in any local domains. For these addresses,
7426 slashes and pipe symbols are allowed within local parts, but the sequence
7427 /../ is explicitly forbidden.
7429 31. SPA server authentication was not clearing the challenge buffer before
7432 32. log_message in a "warn" ACL statement was writing to the reject log as
7433 well as to the main log, which contradicts the documentation and doesn't
7434 seem right (because no rejection is happening). So I have stopped it.
7436 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7437 However, I am unable to do any testing of this.
7439 34. Fixed an infelicity in the appendfile transport. When checking directories
7440 for a mailbox, to see if any needed to be created, it was accidentally
7441 using path names with one or more superfluous leading slashes; tracing
7442 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7444 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7445 contents are added to the log line that is written for every discarded
7446 recipient. (Previously a log_message setting was ignored.)
7448 36. The ${quote: operator now quotes the string if it is empty.
7450 37. The install script runs exim in order to find its version number. If for
7451 some reason other than non-existence or emptiness, which it checks, it
7452 could not run './exim', it was installing it with an empty version number,
7453 i.e. as "exim-". This error state is now caught, and the installation is
7456 38. An argument was missing from the function that creates an error message
7457 when Exim fails to connect to the socket for saslauthd authentication.
7458 This could cause Exim to crash, or give a corrupted message.
7460 39. Added isip, isip4, and isip6 to ${if conditions.
7462 40. The ACL variables $acl_xx are now saved with the message, and can be
7463 accessed later in routers, transports, and filters.
7465 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7466 strings in the file are not string-expanded.
7468 42. If a MAIL command specified a SIZE value that was too large to fit into an
7469 int variable, the check against message_size_limit failed. Such values are
7470 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7471 day this will have to be increased, but I don't think I want to be around
7472 when emails are that large.
7479 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7480 iconv() is not standard in FreeBSD.
7482 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7483 IPv6 enabled. The observed symptom was a segmentation fault on return from
7484 the function os_common_find_running_interfaces() in src/os.c.
7486 3. In the check_special_case() function in daemon.c I had used "errno" as an
7487 argument name, which causes warnings on some systems. This was basically a
7488 typo, since it was named "eno" in the comments!
7490 4. The code that waits for the clock to tick (at a resolution of some fraction
7491 of a second) so as to ensure message-id uniqueness was always waiting for
7492 at least one whole tick, when it could have waited for less. [This is
7493 almost certainly not relevant at current processor speeds, where it is
7494 unlikely to ever wait at all. But we try to future-proof.]
7496 5. The function that sleeps for a time interval that includes fractions of a
7497 second contained a race. It did not block SIGALRM between setting the
7498 timer, and suspending (a couple of lines later). If the interval was short
7499 and the sigsuspend() was delayed until after it had expired, the suspension
7500 never ended. On busy systems this could lead to processes getting stuck for
7503 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7504 process, before it forks any delivery processes. The open lookup caching
7505 mechanism meant that the open file or database connection was passed into
7506 the delivery process. The problem was that delivery processes always tidy
7507 up cached lookup data. This could cause a problem for the next delivery
7508 process started by the queue runner, because the external queue runner
7509 process does not know about the closure. So the next delivery process
7510 still has data in the lookup cache. In the case of a file lookup, there was
7511 no problem because closing a file descriptor in a subprocess doesn't affect
7512 the parent. However, if the lookup was caching a connection to a database,
7513 the connection was closed, and the second delivery process was likely to
7514 see errors such as "PGSQL: query failed: server closed the connection
7515 unexpectedly". The problem has been fixed by closing all cached lookups
7516 in a queue runner before running a delivery process.
7518 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7519 seem to have the "const" qualifier which it has on other OS. I've
7522 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7523 *of the same type* (client or server) with the same public name that an
7524 error should be diagnosed.
7526 9. When Exim looked up a host name for an IP address, but failed to find the
7527 original IP address when looking up the host name (a safety check), it
7528 output the message "<ip address> does not match any IP for NULL", which was
7529 confusing, to say the least. The bug was that the host name should have
7530 appeared instead of "NULL".
7532 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7533 user that is built into the binary, and the -C or -D options is used, root
7534 privilege is dropped before the configuration file is read. In addition,
7535 logging is switched to stderr instead of the normal log files. If the
7536 configuration then re-defines the Exim user, the unprivileged environment
7537 is probably not what is expected, so Exim logs a panic warning message (but
7540 However, if deliver_drop_privilege is set, the unprivileged state may well
7541 be exactly what is intended, so the warning has been cut out in that case,
7542 and Exim is allowed to try to write to its normal log files.
7548 1. smtp_return_error_details was not giving details for temporary sender
7549 or receiver verification errors.
7551 2. Diagnose a configuration error if two authenticators have the same public
7554 3. Exim used not to create the message log file for a message until the first
7555 delivery attempt. This could be confusing when incoming messages were held
7556 for policy or load reasons. The message log file is now created at the time
7557 the message is received, and an initial "Received" line is written to it.
7559 4. The automatically generated man page for command line options had a minor
7560 bug that caused no ill effects; however, a more serious problem was that
7561 the procedure for building the man page automatically didn't always
7562 operate. Consequently, release 4.20 contains an out-of-date version. This
7563 shouldn't happen again.
7565 5. When building Exim with embedded Perl support, the script that builds the
7566 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7567 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7569 6. The freeze_tell option was not being used for messages that were frozen on
7570 arrival, either by an ACL or by local_scan().
7572 7. Added the smtp_incomplete_transaction log selector.
7574 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7575 was accepting AUTH without a new EHLO.
7577 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7578 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7581 10. Logging of TCP/IP connections (when configured) now happens in the main
7582 daemon process instead of the child process, so that the TCP/IP connection
7583 count is more accurate (but it can never be perfect).
7585 11. The use of "drop" in a nested ACL was not being handled correctly in the
7586 outer ACL. Now, if condition failure induced by the nested "drop" causes
7587 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7588 or "require"), the connection is dropped.
7590 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7591 that yield "discard" can now be used with an "accept" or a "discard" verb,
7592 but an error is generated for any others (because I can't see a useful way
7593 to define what should happen).
7595 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7596 are now processed in the same way as lines in the Exim configuration file.
7597 In particular, continuation lines are supported.
7599 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7601 15. Added -ti meaning -t -i.
7603 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7604 domains, and warn by logging if others are found.
7606 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7607 returned from GIFCONF when Exim is trying to find the list of interfaces on
7608 a host. The code in os.c has been modified to copy each ifreq to an aligned
7609 structure in all cases.
7611 Also, in some cases, the returned ifreq's were being copied to a 'struct
7612 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7613 means the last couple of bytes of an IPv6 address could be chopped if the
7614 ifreq contained only a normal sockaddr (14 bytes storage).
7616 18. Named domain lists were not supported in the hosts_treat_as_local option.
7617 An entry such as +xxxx was not recognized, and was treated as a literal
7620 19. Ensure that header lines added by a DATA ACL are included in the reject log
7621 if the ACL subsequently rejects the message.
7623 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7624 MD5 (which is deprecated).
7626 21. When testing a filter file using -bf, Exim was writing a message when it
7627 took the sender from a "From " line in the message, but it was not doing so
7628 when it took $return_path from a Return-Path: header line. It now does.
7630 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7631 with a valid header line field name (a series of printing characters
7632 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7634 23. Changed "disc" in the source to "disk" to conform to the documentation and
7635 the book and for uniformity.
7637 24. Ignore Sendmail's -Ooption=value command line item.
7639 25. When execve() failed while trying to run a command in a pipe transport,
7640 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7641 could be confused with a return value of 69 from the command itself. This
7642 has been changed to 127, the value the shell returns if it is asked to run
7643 a non-existent command. The wording for the related log line suggests a
7644 non-existent command as the problem.
7646 26. If received_header_text expands to an empty string, do not add a Received:
7647 header line to the message. (Well, it adds a token one on the spool, but
7648 marks it "old" so that it doesn't get used or transmitted.)
7650 27. Installed eximstats 1.28 (addition of -nt option).
7652 28. There was no check for failure on the call to getsockname() in the daemon
7653 code. This can fail if there is a shortage of resources on the system, with
7654 ENOMEM, for example. A temporary error is now given on failure.
7656 29. Contrary to the C standard, it seems that in some environments, the
7657 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7658 program. Exim now does this explicitly; it affects the formatting of
7659 timestamps using strftime().
7661 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7662 complaints. I've now initialized all the variables, to avoid this.
7664 32. Header lines added by a system filter were not being "seen" during
7665 transport-time rewrites.
7667 33. The info_callback() function passed to OpenSSL is set up with type void
7668 (*)(SSL *, int, int), as described somewhere. However, when calling the
7669 function (actually a macro) that sets it up, the type void(*)() is
7670 expected. I've put in a cast to prevent warnings from picky compilers.
7672 34. If a DNS black list lookup found a CNAME record, but there were no A
7673 records associated with the domain it pointed at, Exim crashed.
7675 35. If a DNS black list lookup returned more than one A record, Exim ignored
7676 all but the first. It now scans all returned addresses if a particular IP
7677 value is being sought. In this situation, the contents of the
7678 $dnslist_value variable are a list of all the addresses, separated by a
7681 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7682 to accept a host name and all its aliases if the forward lookup for any of
7683 them yielded the IP address of the incoming connection. Now it accepts only
7684 those names whose forward lookup yields the correct IP address. Any other
7685 names are discarded. This closes a loophole whereby a rogue DNS
7686 administrator could create reverse DNS records to break through a
7687 wildcarded host restriction in an ACL.
7689 37. If a user filter or a system filter that ran in a subprocess used any of
7690 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7691 the wrong values were passed to the pipe command ($thisaddress had the
7692 value of $0, $0 had the value of $1, etc). This bug was introduced by
7693 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7695 38. Improved the line breaking for long SMTP error messages from ACLs.
7696 Previously, if there was no break point between 40 and 75 characters, Exim
7697 left the rest of the message alone. Two changes have been made: (a) I've
7698 reduced the minimum length to 35 characters; (b) if it can't find a break
7699 point between 35 and 75 characters, it looks ahead and uses the first one
7700 that it finds. This may give the occasional overlong line, but at least the
7701 remaining text gets split now.
7703 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7704 file descriptors might be low, and that setting 1000 would always raise it.
7705 It turns out that in some environments, the limit is already over 1000 and
7706 that lowering it causes trouble. So now Exim takes care not to decrease it.
7708 40. When delivering a message, the value of $return_path is set to $sender_
7709 address at the start of routing (routers may change the value). By an
7710 oversight, this default was not being set up when an address was tested by
7711 -bt or -bv, which affected the outcome if any router or filter referred to
7714 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7715 log only when "message" or "log_message" are set. However, if one of the
7716 conditions was an address verification, or a call to a nested ACL, the
7717 messages generated by the underlying test were being passed through. This
7718 no longer happens. The underlying message is available in $acl_verify_
7719 message for both "message" and "log_message" expansions, so it can be
7720 passed through if needed.
7722 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7723 new expansion $bh_ to give the encoded byte string without charset
7724 translation. Translation happens only if iconv() is available; HAVE_ICONV
7725 indicates this at build time. HEADERS_CHARSET gives the charset to
7726 translate to; headers_charset can change it in the configuration, and
7727 "headers charset" can change it in an individual filter file.
7729 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7730 that creates RFC 2047 encoded "words" labels them as that charset instead
7731 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7732 expansion operator; (ii) when Exim creates a From: line for a local
7733 message; (iii) when a header line is rewritten to include a "phrase" part.
7735 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7736 buggy, causing it to skip the first lines of messages whose message ID
7737 ended in 'D'. This would not have bitten before Exim release 4.14, because
7738 message IDs were unlikely to end in 'D' before then. The effect was to have
7739 incorrect size information for certain domains.
7741 45. #include "config.h" was missing at the start of the crypt16.c module. This
7742 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7745 46. If there was a timeout during a "random" callout check, Exim treated it as
7746 a failure of the random address, and carried on sending RSET and the real
7747 address. If the delay was just some slowness somewhere, the response to the
7748 original RCPT would be taken as a response to RSET and so on, causing
7749 mayhem of various kinds.
7751 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7752 when I implemented it. It didn't allow for the fact that some option values
7753 may legitimately be negative (e.g. size_addition), and it didn't even do
7754 the right test for positive values.
7756 48. Domain names in DNS records are case-independent. Exim always looks them up
7757 in lower case. Some resolvers return domain names in exactly the case they
7758 appear in the zone file, that is, they may contain uppercase letters. Not
7759 all resolvers do this - some return always lower case. Exim was treating a
7760 change of case by a resolver as a change of domain, similar to a widening
7761 of a domain abbreviation. This triggered its re-routing code and so it was
7762 trying to route what was effectively the same domain again. This normally
7763 caused routing to fail (because the router wouldn't handle the domain
7764 twice). Now Exim checks for this case specially, and just changes the
7765 casing of the domain that it ultimately uses when it transmits the message
7768 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7771 50. If a filter generated a file delivery with a non-absolute name (possible if
7772 no home directory exists for the router), the forbid_file option was not
7775 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7776 the existing equality matching.
7778 52. Exim was using ints instead of ino_t variables in some places where it was
7779 dealing with inode numbers.
7781 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7782 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7783 called TMPDIR, and if it finds it is different, it changes its value.
7785 54. The smtp_printf() function is now made available to local_scan() so
7786 additional output lines can be written before returning. There is also an
7787 smtp_fflush() function to enable the detection of a dropped connection.
7788 The variables smtp_input and smtp_batched_input are exported to
7791 55. Changed the default runtime configuration: the message "Unknown user"
7792 has been removed from the ACL, and instead placed on the localuser router,
7793 using the cannot_route_message feature. This means that any verification
7794 failures that generate their own messages won't get overridden. Similarly,
7795 the "Unrouteable address" message that was in the ACL for unverifiable
7796 relay addresses has also been removed.
7798 56. Added hosts_avoid_esmtp to the smtp transport.
7800 57. The exicyclog script was not checking for the esoteric option
7801 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7802 will work only if exicyclog is run under the appropriate euid.
7804 58. Following a discussion on the list, the rules by which Exim recognises line
7805 endings on incoming messages have been changed. The -dropcr and drop_cr
7806 options are now no-ops, retained only for backwards compatibility. The
7807 following line terminators are recognized: LF CRLF CR. However, special
7808 processing applies to CR:
7810 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7811 nor a local message in the state where . is a terminator.
7813 (ii) If a bare CR is encountered in a header line, an extra space is added
7814 after the line terminator so as not to end the header. The reasoning
7815 behind this is that bare CRs in header lines are most likely either
7816 to be mistakes, or people trying to play silly games.
7818 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7819 was being incorrectly given as 18 bytes larger than it should have been.
7820 This is a VOB (very old bug).
7822 60. This may never have affected anything current, but just in case it has:
7823 When the local host is found other than at the start of a list of hosts,
7824 the local host, those with the same MX, and any that follow, are discarded.
7825 When the list in question was part of a longer list of hosts, the following
7826 hosts (not currently being processed) were also being discarded. This no
7827 longer happens. I'm not sure if this situation could ever has previously
7830 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7833 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7834 header. This is something that is recommended in a new Internet Draft, and
7835 is something that is documented as being done by Sendmail. There are two
7836 possible values. For messages generated by the autoreply transport, Exim
7839 Auto-Submitted: auto-replied
7841 whereas for all other generated messages (e.g. bounces) it adds
7843 Auto-Submitted: auto-generated
7845 63. The "personal" condition in filters now includes a test for the
7846 Auto-Submitted: header. If it contains the string "auto-" the message it
7847 not considered personal.
7849 64. Added rcpt_include_affixes as a generic transport option.
7851 65. Added queue_only_override (default true).
7853 66. Added the syslog_duplication option.
7855 67. If what should have been the first header line of a message consisted of
7856 a space followed by a colon, Exim was mis-interpreting it as a header line.
7857 It isn't of course - it is syntactically invalid and should therefore be
7858 treated as the start of the message body. The misbehaviour could have
7859 caused a number of strange effects, including loss of data in subsequent
7860 header lines, and spool format errors.
7862 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7863 client host had authenticated. This control can now be exercised by an ACL
7864 for more flexibility.
7866 69. By default, callouts do not happen when testing with -bh. There is now a
7867 variant, -bhc, which does actually run the callout code, including
7868 consulting and updating the callout cache.
7870 70. Added support for saslauthd authentication, courtesy of Alexander
7873 71. If statvfs() failed on the spool or log directories while checking their
7874 size for availability, Exim confusingly gave the error "space shortage".
7875 Furthermore, in debugging mode it crashed with a floating point exception.
7876 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7877 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7878 serious problem, Exim now writes to the main and panic logs when this
7879 happens, with details of the failure. It then refuses to accept the
7880 incoming message, giving the message "spool directory problem" or "log
7881 directory problem" with a 421 code for SMTP messages.
7883 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7884 0, 1, and 2 exist, because some OS complain for execs without them (see
7885 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7886 descriptors. However, the code omitted to check that the open succeeded,
7887 causing mysterious errors if for some reason the permissions on /dev/null
7888 got screwed. Now Exim writes a message to the main and panic logs, and
7889 bombs out if it can't open /dev/null.
7891 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7892 interact so that it is all more flexible. It is supposed to remain
7893 backwards compatible. Also added extra_local_interfaces.
7895 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7896 to bomb out with an assertion failure - to the client this appears as a
7897 connection drop. This problem occurs in the part of the code that was taken
7898 from the Samba project. Fortunately, the assertion is in a very simple
7899 function, so I have fixed this by reproducing the function inline in the
7900 one place where it is called, and arranging for authentication to fail
7901 instead of killing the process with assert().
7903 75. The SPA client code was not working when the server requested OEM rather
7904 than Unicode encoding.
7906 76. Added code to make require_files with a specific uid setting more usable in
7907 the case where statting the file as root fails - usually a non-root-mounted
7908 NFS file system. When this happens and the failure is EACCES, Exim now
7909 forks a subprocess and does the per-uid checking as the relevant uid.
7911 77. Added process_log_path.
7913 78. If log_file_path was not explicitly set, a setting of check_log_space or
7914 check_log_inodes was ignored.
7916 79. If a space check for the spool or log partitions fails, the incident is now
7917 logged. Of course, in the latter case the data may get lost...
7919 80. Added the %p formatting code to string_format() so that it can be used to
7920 print addresses in debug_print(). Adjusted all the address printing in the
7921 debugging in store.c to use %p rather than %d.
7923 81. There was a concern that a line of code in smtp_in.c could overflow a
7924 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7925 initially expressed, the concern was not well-founded, because trailing
7926 spaces are removed early. However, if the trailing spaces were followed by
7927 a NULL, they did not get removed, so the overflow was possible. Two fixes
7930 (a) I re-wrote the offending code in a cleaner fashion.
7931 (b) If an incoming SMTP command contains a NULL character, it is rejected
7934 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7935 runs initgroups(), so that if the Exim user is in any additional groups,
7936 they will be used during message reception.
7942 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob