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
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
166 JH/01 Avoid costly startup code when not strictly needed. This reduces time
167 for some exim process initialisations. It does mean that the logging
168 of TLS configuration problems is only done for the daemon startup.
170 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
172 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
173 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
176 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
177 DKIM signing is being done. Previously a transport_filter would always
178 disable CHUNKING, falling back to traditional DATA.
180 JH/05 Regard command-line recipients as tainted.
182 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
184 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
185 PAM library frees one of the arguments given to it, despite the
186 documentation. Therefore a plain malloc must be used.
188 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
189 on-stack buffers were used, resulting in a taint trap when DSN information
190 copied from a received message was written into the buffer.
192 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
193 the ordering of its ARC headers. This caused a crash.
195 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
196 a new record was being constructed with information from the peer, a trap
199 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
200 installation would get error messages from DMARC verify, when it hit the
201 nonexistent file indicated by the default. Distros wanting DMARC enabled
202 should both provide the file and set the option.
203 Also enforce no DMARC verification for command-line sourced messages.
205 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
206 could, depending on the platform, hang at the STARTTLS response.
208 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
209 message on a connection. Previously if one message had ARC headers and
210 the following one did not, a crash could result when adding an
211 Authentication-Results: header.
213 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
214 Exim main code and Exim-related utities. The introduction of taint
215 tracking also did many adjustments to string handling. Since then, eximon
216 frequently terminated with an assert failure.
218 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
219 check for 452 responses. This slightly helps the inefficieny of doing
220 a large alias-expansion into a recipient-limited target. The max_rcpt
221 transport option still applies (and at the current default, will override
222 the new feature). The check is done for either cause of synch, and forces
223 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
224 connection. The new facility is not tunable at this time.
226 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
227 library live data was being used, so the results became garbage. Make
228 copies while it is still usable.
230 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
231 on delivery deferred (==) and failed (**) lines (if a delivery was
232 attemtped). Previously it was only on completion (=>) lines.
234 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
235 for the expansion of the server_scram_iter and server_scram_salt options.
237 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
238 are now specifically given a NO_DATA response without hitting the system
239 resolver. The library goes on to do the now-standard TXT lookup.
240 Use of dnsdb lookups is not affected.
242 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
243 only retrieve the errormessage once. Previously two calls to dlerror()
244 were used, and the second one (for mainlog/paniclog) retrieved null
247 JH/20 Taint checking: disallow use of tainted data for
248 - the appendfile transport file and directory options
249 - the pipe transport command
250 - the autoreply transport file, log and once options
251 - file names used by the redirect router (including filter files)
253 - paths used by single-key lookups
254 Previously this was permitted.
256 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
257 adjusted the size of a major service buffer; this failed because the
258 buffer was in use at the time. Change to a compile-time increase in the
259 buffer size, when this authenticator is compiled into exim.
261 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
262 previous fast-mode was untenable in the face of glibs using mmap to
263 support larger malloc requests.
265 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
266 New values supported, if defined on system where compiled:
267 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
268 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
270 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
271 running a limited number of proceses in parallel, a benefit is gained. The
272 amount varies with the platform hardware and load. The use of the option
273 queue_run_in_order means we cannot do this, as ordering becomes
276 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
277 had introduced a string-copy (for ensuring NUL-termination) which was not
278 appropriate for that case, which can include embedded NUL bytes in the
279 block of data. Investigation showed the copy to actually be needless, the
280 data being length-specified.
282 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
283 done during a receiving connection, and both used TLS, global info was
284 used rather than per-connection info for tracking the state of data
285 queued for transmission. This could result in a connection hang.
287 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
288 Previously, when delivering serveral messages down a single connection
289 only the first would provide a SIZE. This was due to the size information
290 not being properly tracked.
292 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
293 TAI (at 37 seconds currently), pretend to be in UTC for time-related
294 expansion and logging. Previously, spurious values such as a future
295 minute could be seen.
297 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
298 it could crash from a null-deref. This could also affect the
299 ${addresses: } operator and ${readsock } item.
301 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
302 message following a mime one, the variable was not reset.
304 JH/30 When an pipelined-connect fails at the first response, assume incorrect
305 cached capability (perhaps the peer reneged?) and immediately retry in
308 JH/31 Fix spurious detection of timeout while writing to transport filter.
310 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
311 an attempt to copy the string was made before checking it.
313 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
314 taint of the lookup key was maintained; we now regard the presence in the
315 filesystem as sufficient validation.
317 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
318 argument is supplied.
320 JH/35 The dsearch lookup now requires that the directory is an absolute path.
321 Previously this was not checked, and nonempty relative paths made an
322 access under Exim's current working directory.
324 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
325 Previously no event was raised.
327 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
328 parameter supplied by the sender MAIL FROM command. Previously it was
329 ignored, and only the check_spool_space option value for the required
332 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
333 the size of the signing public-key. Previously it was instead giving
334 the size of the signature hash.
336 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
337 the default. See the (new) dkim_verify_min_keysizes option.
339 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
340 and an ACL use a lookup for checking either the local_part or domain,
341 stale data could be accessed. Ensure that variable references are
342 dropped between messages.
344 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
345 by the client was not checked as pointing within response data before
346 being used. A malicious client could thus cause an out-of-bounds read and
347 possibly gain authentication. Fix by adding the check.
349 JH/42 Internationalisation: change the default for downconversion in the smtp
350 transport to be "if needed". Previously it was "as previously set" for
351 the message, which usually meant "if needed" for message-submission but
352 "no" for everything else. However, MTAs have been seen using SMTPUTF8
353 even when the envelope addresses did not need it, resulting in forwarding
354 failures to non-supporting MTAs. A downconvert in such cases will be
355 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
356 transport. The change does mean that addresses needing conversion will
357 be converted when previously a delivery failure would occur.
359 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
360 response was received it would be used unchecked in a fail-DSN, violating
361 standards on line-length limits. Truncate if needed.
363 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
364 doesn't work. (Additionally add a new main config option to configure the
371 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
372 side, to decode a TLS 1.3 packet capture.
374 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
375 Previously the default library behaviour applied, sending two, each in
378 JH/03 Debug output for ACL now gives the config file name and line number for
381 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
383 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
385 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
386 buffer overrun for (non-chunking) other transports.
388 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
389 TLS1.3, means that a server rejecting a client certificate is not visible
390 to the client until the first read of encrypted data (typically the
391 response to EHLO). Add detection for that case and treat it as a failed
392 TLS connection attempt, so that the normal retry-in-clear can work (if
393 suitably configured).
395 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
396 and/or domain. Found and fixed by Jason Betts.
398 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
399 configuration). If a CNAME target was not a wellformed name pattern, a
402 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
403 the OS reports them interleaved with other addresses.
405 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
406 used both for input and for a verify callout, both encrypted, SMTP
407 responses being sent by the server could be lost. This resulted in
408 dropped connections and sometimes bounces generated by a peer sending
411 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
412 string. Previously it was possible to cause undefined behaviour in a
413 library routine (usually a crash). Found by "zerons".
415 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
418 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
419 API was removed, so update to use the newer ones.
421 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
422 any timeout set, is taking a long time. Previously we would hang on to a
423 rotated logfile "forever" if the input was arriving with long gaps
424 (a previous attempt to fix addressed lack, for a long time, of initial
427 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
428 shared (NFS) environment. The length of the tempfile name is now
429 4 + 16 ("hdr.$message_exim_id") which might break on file
430 systems which restrict the file name length to lower values.
433 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
434 shared (NFS) environment.
436 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
437 did for all versions <4.90). Notably -M, -m, --invert, -I may be
440 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
441 on some platforms for bit 31.
443 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
444 to changes apparently associated with TLS1.3 handling some of the APIs
445 previously used were either nonfunctional or inappropriate. Strings
446 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
447 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
448 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
449 This affects log line X= elements, the $tls_{in,out}_cipher variables,
450 and the use of specific cipher names in the encrypted= ACL condition.
452 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
454 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
455 verification result was not updated unless hosts_require_ocsp applied.
457 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
458 queue_list_requires_admin set to false, non-admin users were denied the
461 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
462 directory-of-certs mode. Previously they were advertised despite the
465 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
466 A single TCP connection by a client will now hold a TLS connection open
467 for multiple message deliveries, by default. Previously the default was to
470 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
471 default. If built with the facility, DANE will be used. The facility
472 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
474 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
475 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
476 must be defined and you must still, unless you define DISABLE_TLS, manage
477 the the include-dir and library-file requirements that go with that
478 choice. Non-TLS builds are still supported.
480 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
483 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
484 default. If the platform supports and has the facility enabled, it will
485 be requested on all coneections.
487 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
488 controlled by the build-time option SUPPORT_PIPE_CONNECT.
490 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
492 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
493 success-DSN messages. Previously the From: header was always the default
494 one for these; the option was ignored.
496 JH/28 Fix the timeout on smtp response to apply to the whole response.
497 Previously it was reset for every read, so a teergrubing peer sending
498 single bytes within the time limit could extend the connection for a
499 long time. Credit to Qualsys Security Advisory Team for the discovery.
501 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
502 delivery address, which leaked information of the results of local
503 forwarding. Change to the original envelope recipient address, per
506 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
507 requested. Previously not bounce was generated and a log entry of
508 error ignored was made.
510 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
512 JH/32 Introduce a general tainting mechanism for values read from the input
513 channel, and values derived from them. Refuse to expand any tainted
514 values, to catch one form of exploit.
516 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
517 was unused and the unexpanded text used for the test. Found and
518 fixed by Ruben Jenster.
520 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
521 an attempt to use a TLS library read routine dereffed a nul pointer,
524 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
525 them in our smtp response.
527 JH/36 Have the general router option retry_use_local_part default to true when
528 any of the restrictive preconditions are set (to anything). Previously it
529 was only for check_local user. The change removes one item of manual
530 configuration which is required for proper retries when a remote router
531 handles a subset of addresses for a domain.
533 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
534 link count into consideration.
536 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
537 caused the extension of big_buffer, the following lines were ignored.
539 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
540 accordance with RFC 2308. Previously there was no expiry, so a longlived
541 receive process (eg. due to ACL delays) versus a short SOA value could
544 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
546 JH/39 Promote DMARC support to mainline.
548 JH/40 Bug 2452: Add a References: header to DSNs.
550 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
551 parameters. The relevant library call is documented as "Deprecated: This
552 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
553 3.6.0, DH parameters are negotiated following RFC7919."
555 HS/06 Change the default of dnssec_request_domains to "*"
557 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
558 carried on and emitted a BDAT command, even when PIPELINING was not
561 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
562 buffer was used for the filename, resulting in a trap when tainted
563 arguments (eg. $domain) were used.
565 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
566 recommended to avoid a possible server-load attack. The feature can be
567 re-enabled via the openssl_options main cofiguration option.
569 JH/45 local_scan API: documented the current smtp_printf() call. This changed
570 for version 4.90 - adding a "more data" boolean to the arguments.
571 Bumped the ABI version number also, this having been missed previously;
572 release versions 4.90 to 4.92.3 inclusive were effectively broken in
573 respect of usage of smtp_printf() by either local_scan code or libraries
574 accessed via the ${dlfunc } expansion item. Both will need coding
575 adjustment for any calls to smtp_printf() to match the new function
576 signature; a FALSE value for the new argument is always safe.
578 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
579 the file-offset (which the Linux syscall does, and exim expects); this
580 resulted in an indefinite loop.
582 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
583 to do ARC verification. The Authentication-Results: header line added
584 by the configuration then had no ARC item.
590 JH/01 Remove code calling the customisable local_scan function, unless a new
591 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
593 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
594 non-signal-safe functions being used.
596 JH/03 Bug 2269: When presented with a received message having a stupidly large
597 number of DKIM-Signature headers, disable DKIM verification to avoid
598 a resource-consumption attack. The limit is set at twenty.
600 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
601 report of oldest_pass in ${authres } in consequence, and separate out
602 some descriptions of reasons for verification fail.
604 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
605 files in the spool were present and unlocked. A queue-runner could spot
606 them, resulting in a duplicate delivery. Fix that by doing the unlock
607 after the unlink. Investigation by Tim Stewart. Take the opportunity to
608 add more error-checking on spoolfile handling while that code is being
611 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
612 No known attacks, no CVE, this is defensive hardening.
614 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
615 a queue-runner could start a delivery while other operations were ongoing.
616 Cutthrough delivery was a common victim, resulting in duplicate delivery.
617 Found and investigated by Tim Stewart. Fix by using the open message data
618 file handle rather than opening another, and not locally closing it (which
619 releases a lock) for that case, while creating the temporary .eml format
620 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
622 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
623 $sender_verify_failure/$recipient_verify_failure to "random".
625 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
628 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
629 Previously this would segfault.
631 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
634 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
635 like zero, since the resolver should be doing this for us, But we need one
636 as a CNAME but no MX presence gets the CNAME returned; we need to check
637 that doesn't point to an MX to declare it "no MX returned" rather than
638 "error, loop". A new main option is added so the older capability of
639 following some limited number of chain links is maintained.
641 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
643 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
644 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
645 now-deprecated earlier definitions used only specified the range up to TLS
646 1.2 (in the older-version library docs).
648 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
650 JH/15 Rework TLS client-side context management. Stop using a global, and
651 explicitly pass a context around. This enables future use of TLS for
652 connections to service-daemons (eg. malware scanning) while a client smtp
653 connection is using TLS; with cutthrough connections this is quite likely.
655 JH/16 Fix ARC verification to do AS checks in reverse order.
657 JH/17 Support a "tls" option on the ${readsocket } expansion item.
659 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
660 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
661 Previously the "utf8" would be re-prepended for every additional message.
663 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
664 Previously thery were accepted, resulting in issues when attempting to
665 forward messages to a non-supporting MTA.
667 PP/02 Let -n work with printing macros too, not just options.
669 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
670 one parent address was copied, and bogus data was used at delivery-logging
671 time. Either a crash (after delivery) or bogus log data could result.
672 Discovery and analysis by Tim Stewart.
674 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
675 Previously if the string ended mid-character, we did not insert the
676 promised '?' replacement.
678 PP/04 Documentation: current string operators work on bytes, not codepoints.
680 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
681 should pack well giving a smaller memory footprint so better caching and
682 therefore performance. Group the declarations where this can't be done so
683 that the byte-sized flag variables are not interspersed among pointer
684 variables, giving a better chance of good packing by the compiler.
686 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
687 non-null, to avoid issues with sites running BATV. Previously reports were
688 sent with an empty envelope sender so looked like bounces.
690 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
691 The ignore_error flag wasn't being returned from the filter subprocess so
692 was not set for later routers. Investigation and fix by Matthias Kurz.
694 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
695 and a msg:complete for the whole, when a message is manually removed using
696 -Mrm. Developement by Matthias Kurz, hacked on by JH.
698 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
699 a "Gnu special" function, asprintf() in the DB utility binary builds; I
700 hope that is portable enough.
702 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
703 requiring a known-CA anchor certificate; make it now rely entirely on the
704 TLSA as an anchor. Checking the name on the leaf cert against the name
705 on the A-record for the host is still done for TA (but not for EE mode).
707 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
708 included in delivery lines for non-proxied connections, when compiled with
709 SUPPORT_SOCKS and running with proxy logging enabled.
711 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
712 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
713 move the existing event to fire before the normal logging of message
714 failure so that custom logging is bracketed by normal logging.
716 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
717 msg:fail:internal event. Developement by Matthias Kurz.
719 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
720 far too small for todays use of crypto signatures stored there. Go all
721 the way to the max DNS message size of 64kB, even though this might be
722 overmuch for IOT constrained device use.
724 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
725 copy a string over itself. The library routine is documented as not
726 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
728 JH/32 For main options check_spool_space and check_inode_space, where the
729 platform supports 64b integers, support more than the previous 2^31 kB
730 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
731 the previous G, M, k.
733 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
734 $authenticated_fail_id variable on authentication failure. Previously
737 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
738 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
739 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
740 GNUTLS_SEC_PARAM_MEDIUM.
742 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
743 side. Previously we would continue as if no SNI had been received.
745 JH/36 Harden the handling of string-lists. When a list consisted of a sole
746 "<" character, which should be a list-separator specification, we walked
747 off past the nul-terimation.
749 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
750 causes) even when the retry time is not yet met. Previously they were
751 not, meaning that when (say) an account was over-quota and temp-rejecting,
752 and multiple senders' messages were queued, only one sender would get
753 notified on each configured delay_warning cycle.
755 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
757 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
758 cipher-suites, an error can be left on the stack even for a succeeding
759 accept; this results in impossible error messages when a later operation
762 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
763 return error codes indicating retry. Under TLS1.3 this becomes required.
765 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
766 it only wrote the new authenticators, resulting in a lack of tracking of
767 peer changes of ESMTP extensions until the next cache flush.
769 JH/41 Fix the loop reading a message header line to check for integer overflow,
770 and more-often against header_maxsize. Previously a crafted message could
771 induce a crash of the recive process; now the message is cleanly rejected.
773 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
774 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
780 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
781 When redis_servers is set to a list of > 1 element, and the Redis servers
782 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
783 case of MOVED into a DEFER case instead, thus moving the query onto the
784 next server in the list. For a cluster of N elements, all N servers must
785 be defined in redis_servers.
787 GF/02 Catch and remove uninitialized value warning in exiqsumm
788 Check for existence of @ARGV before looking at $ARGV[0]
790 JH/01 Replace the store_release() internal interface with store_newblock(),
791 which internalises the check required to safely use the old one, plus
792 the allocate and data copy operations duplicated in both (!) of the
793 extant use locations.
795 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
796 modifier. This matches the restriction on the commandline.
798 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
799 Previously only the last row was returned.
801 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
802 we assumed that tags in the header were well-formed, and parsed the
803 element content after inspecting only the first char of the tag.
804 Assumptions at that stage could crash the receive process on malformed
807 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
808 While running the DKIM ACL we operate on the Permanent memory pool so that
809 variables created with "set" persist to the DATA ACL. Also (at any time)
810 DNS lookups that fail create cache records using the Permanent pool. But
811 expansions release any allocations made on the current pool - so a dnsdb
812 lookup expansion done in the DKIM ACL releases the memory used for the
813 DNS negative-cache, and bad things result. Solution is to switch to the
814 Main pool for expansions.
815 While we're in that code, add checks on the DNS cache during store_reset,
816 active in the testsuite.
817 Problem spotted, and debugging aided, by Wolfgang Breyha.
819 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
820 When none of the hosts presented to a transport match an already-open
821 connection, close it and proceed with the list. Previously we would
822 queue the message. Spotted by Lena with Yahoo, probably involving
825 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
826 Previously a spurious "250 OK id=" response was appended to the proper
829 JH/08 The "support for" informational output now, which built with Content
830 Scanning support, has a line for the malware scanner interfaces compiled
831 in. Interface can be individually included or not at build time.
833 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
834 by the template makefile "src/EDITME". The "STREAM" support for an older
835 ClamAV interface method is removed.
837 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
838 rows affected is given instead).
840 JH/11 The runtime Berkeley DB library version is now additionally output by
841 "exim -d -bV". Previously only the compile-time version was shown.
843 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
844 SMTP connection. Previously, when one had more recipients than the
845 first, an abortive onward connection was made. Move to full support for
846 multiple onward connections in sequence, handling cutthrough connection
847 for all multi-message initiating connections.
849 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
850 routers. Previously, a multi-recipient message would fail to match the
851 onward-connection opened for the first recipient, and cause its closure.
853 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
854 a timeout on read on a GnuTLS initiating connection, resulting in the
855 initiating connection being dropped. This mattered most when the callout
856 was marked defer_ok. Fix to keep the two timeout-detection methods
859 JH/15 Relax results from ACL control request to enable cutthrough, in
860 unsupported situations, from error to silently (except under debug)
861 ignoring. This covers use with PRDR, frozen messages, queue-only and
864 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
866 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
867 metadata, resulting in a crash in free().
869 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
870 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
871 Broken also in d185889f4, with init system revamp.
873 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
874 Previously we abruptly closed the connection after reading a malware-
875 found indication; now we go on to read the "scan ok" response line,
878 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
879 ACL. Previously, a crash would result.
881 JH/19 Speed up macro lookups during configuration file read, by skipping non-
882 macro text after a replacement (previously it was only once per line) and
883 by skipping builtin macros when searching for an uppercase lead character.
885 JH/20 DANE support moved from Experimental to mainline. The Makefile control
886 for the build is renamed.
888 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
889 was allocated for every new TLS startup, meaning one per message. Fix
890 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
892 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
893 reported the original. Fix to report (as far as possible) the ACL
894 result replacing the original.
896 JH/23 Fix memory leak during multi-message connections using STARTTLS under
897 OpenSSL. Certificate information is loaded for every new TLS startup,
898 and the resources needed to be freed.
900 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
902 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
905 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
906 DATA response info to the (existing) per-recipient response info for
907 the "C=" log element. It can have useful tracking info from the
908 destination system. Patch from Simon Arlott.
910 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
911 length value. Previously this would segfault.
913 HS/02 Support Avast multiline protoocol, this allows passing flags to
914 newer versions of the scanner.
916 JH/28 Ensure that variables possibly set during message acceptance are marked
917 dead before release of memory in the daemon loop. This stops complaints
918 about them when the debug_store option is enabled. Discovered specifically
919 for sender_rate_period, but applies to a whole set of variables.
920 Do the same for the queue-runner and queue-list loops, for variables set
921 from spool message files. Do the same for the SMTP per-message loop, for
922 certain variables indirectly set in ACL operations.
924 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
925 as a multi-recipient message from a mailinglist manager). The coding had
926 an arbitrary cutoff number of characters while checking for more input;
927 enforced by writing a NUL into the buffer. This corrupted long / fast
928 input. The problem was exposed more widely when more pipelineing of SMTP
929 responses was introduced, and one Exim system was feeding another.
930 The symptom is log complaints of SMTP syntax error (NUL chars) on the
931 receiving system, and refused recipients seen by the sending system
932 (propating to people being dropped from mailing lists).
933 Discovered and pinpointed by David Carter.
935 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
936 replaced by the ${authresults } expansion.
938 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
940 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
941 allows proper process termination in container environments.
943 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
944 Previously the "final dot" had a newline after it; ensure it is CR,LF.
946 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
947 and "err_perm", deprecated since 4.83 when the RFC-defined words
948 "temperror" and "permerror" were introduced.
950 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
951 transport-filters or DKIM-signing. The restriction was lost in the
952 consolidation of verify-callout and delivery SMTP handling.
953 Extend the restriction to also cover ARC-signing.
955 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
956 in defer=pass mode supply a 450 to the initiator. Previously the message
959 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
960 tls_require_ciphers is used as before.
962 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
963 "pass_unscanned". Only tmpfails from the scanner are written to
964 the paniclog, as they may require admin intervention (permission
965 denied, license issues). Other scanner errors (like decompression
966 bombs) do not cause a paniclog entry.
968 JH/36 Fix reinitialisation of DKIM logging variable between messages.
969 Previously it was possible to log spurious information in receive log
972 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
973 triggered odd behaviour from Outlook Express clients.
975 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
978 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
979 since the IETF WG has not yet settled on that versus the original
980 "bare" representation.
982 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
983 Previously the millisecond value corrupted the output.
984 Fix also for syslog_pid=no and log_selector +pid, for which the pid
985 corrupted the output.
991 JH/01 Rework error string handling in TLS interface so that the caller in
992 more cases is responsible for logging. This permits library-sourced
993 string to be attached to addresses during delivery, and collapses
994 pairs of long lines into single ones.
996 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
997 during configuration. Wildcards are allowed and expanded.
999 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1000 This permits better logging.
1002 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1003 do not close it down and have the child transport start it up again on
1004 the passed-on TCP connection. Instead, proxy the child (and any
1005 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1006 affected: the continued delivery log lines do not have any DNSSEC, TLS
1007 Certificate or OCSP information. TLS cipher information is still logged.
1009 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1010 identical IP addresses on different listening ports. Will also affect
1013 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1014 add noisy ifdef guards to special-case this sillyness.
1015 Patch from Bernd Kuhls.
1017 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1018 than 255 are no longer allowed.
1020 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1021 Disable the session-cache too, which might reduce our load. Since we
1022 currrectly use a new context for every connection, both as server and
1023 client, there is no benefit for these.
1024 GnuTLS appears to not support tickets server-side by default (we don't
1025 call gnutls_session_ticket_enable_server()) but client side is enabled
1026 by default on recent versions (3.1.3 +) unless the PFS priority string
1029 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1030 <https://reproducible-builds.org/specs/source-date-epoch/>.
1032 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1033 the check for any unsuccessful recipients did not notice the limit, and
1034 erroneously found still-pending ones.
1036 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1037 MSG_MORE. Only in-clear (not on TLS connections).
1039 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1040 transport-filter is involved we can buffer the headers in memory for
1041 creating the signature, and read the spool data file once for the
1042 signature and again for transmission.
1044 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1045 4.77 as the kernel support then wasn't solid, having issues in 64bit
1046 mode. Now, it's been long enough. Add support for FreeBSD also.
1048 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1049 case where the routing stage had gathered several addresses to send to
1050 a host before calling the transport for the first, we previously failed
1051 to close down TLS in the old transport process before passing the TCP
1052 connection to the new process. The new one sent a STARTTLS command
1053 which naturally failed, giving a failed delivery and bloating the retry
1054 database. Investigation and fix prototype from Wolfgang Breyha.
1056 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1057 false-negatives in the check that the sender had not preempted a response
1058 or prompt from Exim (running as a server), due to that code's lack of
1059 awareness of the SMTP input buffering.
1061 PP/04 Add commandline_checks_require_admin option.
1062 Exim drops privileges sanely, various checks such as -be aren't a
1063 security problem, as long as you trust local users with access to their
1064 own account. When invoked by services which pass untrusted data to
1065 Exim, this might be an issue. Set this option in main configuration
1066 AND make fixes to the calling application, such as using `--` to stop
1069 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1070 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1071 responses to those, into a single TLS record each way (this usually means
1072 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1075 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1076 affects you only if you're dancing at the edge of the param size limits.
1077 If you are, and this message makes sense to you, then: raise the
1078 configured limit or use OpenSSL 1.1. Nothing we can do for older
1081 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1082 cmdline element to get the documented default one. Previously it was
1085 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1086 are unsuable for non-message-specific reasons - eg. connection timeout,
1087 banner-time rejection.
1089 JH/16 Fix logging of delivery remote port, when specified by router, under
1092 PP/06 Repair manualroute's ability to take options in any order, even if one
1093 is the name of a transport.
1096 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1098 JH/17 Change the list-building routines interface to use the expanding-string
1099 triplet model, for better allocation and copying behaviour.
1101 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1102 Previously it was constructed the first time a possibly-matching string
1103 was met in the configuration file input during startup; now it is done
1106 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1107 compatible one, to avoid the (poorly documented) possibility of a config
1108 file in the working directory redirecting the DB files, possibly correpting
1109 some existing file. CVE-2017-10140 assigned for BDB.
1111 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1112 cache-hot. Previously, although the result was properly cached, the
1113 initial verify call returned a defer.
1115 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1116 the main verify for receipient in uncached-mode.
1118 JH/22 Retire historical build files to an "unsupported" subdir. These are
1119 defined as "ones for which we have no current evidence of testing".
1121 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1122 if present. Previously it was ignored.
1124 JH/24 Start using specified-initialisers in C structure init coding. This is
1125 a C99 feature (it's 2017, so now considered safe).
1127 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1128 if was a fixed-sized field and bitmask ops via macros; it is now more
1131 PP/07 GitHub PR 56: Apply MariaDB build fix.
1132 Patch provided by Jaroslav Škarvada.
1134 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1135 during Coverity cleanups [4.87 JH/47]
1136 Diagnosis and fix provided by Michael Fischer v. Mollard.
1138 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1139 the right size to place the terminating semicolon on its own folded
1140 line, the header hash was calculated to an incorrect value thanks to
1141 the (relaxed) space the fold became.
1143 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1144 and confused the parent.
1146 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1147 which could crash as a result. This could lead to undeliverable messages.
1149 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1152 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1153 matters on fast-turnover and PID-randomising systems, which were getting
1154 out-of-order delivery.
1156 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1157 a possibly-overlapping copy. The symptom was that "Remote host closed
1158 connection in response to HELO" was logged instead of the actual 4xx
1161 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1162 Previously only that bufferd was discarded, resulting in SYMTP command
1165 JH/32 DKIM: when a message has multiple signatures matching an identity given
1166 in dkim_verify_signers, run the dkim acl once for each. Previously only
1167 one run was done. Bug 2189.
1169 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1170 "panic the current process" to "deliberately defer". The panic log is
1171 still written with the problem list name; the mail and reject logs now
1172 get a temp-reject line for the message that was being handled, saying
1173 something like "domains check lookup or other defer". The SMTP 451
1174 message is still "Temporary local problem".
1176 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1177 A crafted sequence of BDAT commands could result in in-use memory beeing
1178 freed. CVE-2017-16943.
1180 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1181 from SMTP input. Previously it was always done; now only done for DATA
1182 and not BDAT commands. CVE-2017-16944.
1184 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1185 to the message (such as an overlong header line). Previously this was
1186 not done and we did not exit BDAT mode. Followon from the previous item
1187 though a different problem.
1193 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1194 than -2003 did; needs libidn2 in addition to libidn.
1196 JH/02 The path option on a pipe transport is now expanded before use.
1198 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1199 Patch provided by "Björn", documentation fix added too.
1201 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1202 missing a wire-to-host endian conversion.
1204 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1205 close after a BDAT command line could be taken as a following command,
1206 giving a synch failure. Fix by only checking for synch immediately
1207 before acknowledging the chunk.
1209 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1210 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1211 macro. Patches provided by Josh Soref.
1213 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1214 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1215 by IANA as a service extension. However, John Klensin suggests that we
1218 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1219 the dkim code may be unix-mode line endings rather than smtp wire-format
1220 CRLF, so prepend a CR to any bare LF.
1222 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1223 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1225 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1226 the first were themselves being wrongly included in the feed into dkim
1227 processing; with most chunk sizes in use this resulted in an incorrect
1228 body hash calculated value.
1230 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1231 DKIM signature block, for verification. Although advised against by
1232 standards it is specifically not ruled illegal.
1234 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1236 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1237 missing a body hash (the bh= tag).
1239 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1240 It seems that HAProxy sends the Proxy Protocol information in clear and
1241 only then does a TLS startup, so do the same.
1243 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1244 TCP connections (such as for Spamd) unless the daemon successfully set
1245 Fast Open mode on its listening sockets. This fixes breakage seen on
1246 too-old kernels or those not configured for Fast Open, at the cost of
1247 requiring both directions being enabled for TFO, and TFO never being used
1248 by non-daemon-related Exim processes.
1250 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1251 endings, at least on the first header line. Try to canonify any that get
1252 past that check, despite the cost.
1254 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1255 now limited to an arbitrary five deep, while parsing addresses with the
1256 strip_excess_angle_brackets option enabled.
1258 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1259 instead leave the unprompted TLS handshake in socket buffer for the
1260 TLS library to consume.
1262 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1264 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1266 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1267 region in one operation, for speed, and this leaves assigned pointers
1268 dangling. Add checks run only under the testsuite which checks all
1269 variables at a store-reset and panics on a dangling pointer; add code
1270 explicitly nulling out all the variables discovered. Fixes one known
1271 bug: a transport crash, where a dangling pointer for $sending_ip_address
1272 originally assigned in a verify callout, is re-used.
1274 PP/06 Drop '.' from @INC in various Perl scripts.
1276 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1278 PP/08 Reduce a number of compilation warnings under clang; building with
1279 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1280 should be warning-free.
1282 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1284 HS/01 Fix portability problems introduced by PP/08 for platforms where
1285 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1287 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1288 chunk. This allows us to accept broken chunked messages. We need a more
1289 general solution here.
1291 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1292 already-broken messages in the queue.
1294 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1296 JH/19 Fix reference counting bug in routing-generated-address tracking.
1302 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1303 supports it and a size is available (ie. the sending peer gave us one).
1305 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1306 years of being deprecated). The replacements are the ACLs
1307 acl_smtp_mime and acl_not_smtp_mime.
1309 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1310 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1311 or even an in-clear connection were permitted. Now, if the host lookup
1312 was dnssec and dane was requested then the host is only used if the
1313 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1314 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1315 if one fails this test.
1316 This means that a poorly-configured remote DNS will make it incommunicado;
1317 but it protects against a DNS-interception attack on it.
1319 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1320 non-noisy when a race steals the message being considered.
1322 JH/05 If main configuration option tls_certificate is unset, generate a
1323 self-signed certificate for inbound TLS connections.
1325 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1326 in rewrites and routers.
1328 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1329 and logged a warning sing 4.83; now they are a configuration file error.
1331 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1332 (lacking @domain). Apply the same qualification processing as RCPT.
1334 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1336 JH/10 Support ${sha256:} applied to a string (as well as the previous
1339 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1340 a cutthrough deliver is pending, as we always want to make a connection.
1341 This also avoids re-routing the message when later placing the cutthrough
1342 connection after a verify cache hit.
1343 Do not update it with the verify result either.
1345 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1346 when routing results in more than one destination address.
1348 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1349 signing (which inhibits the cutthrough capability). Previously only
1350 the presence of an option was tested; now an expansion evaluating as
1351 empty is permissible (obviously it should depend only on data available
1352 when the cutthrough connection is made).
1354 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1355 the relevant preceding SMTP command did not note the pipelining mode.
1357 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1358 Previously they were not counted.
1360 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1361 as one having no matching records. Previously we deferred the message
1362 that needed the lookup.
1364 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1365 distinguished as "(=".
1367 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1368 for missing MX records. Previously it only worked for missing A records.
1370 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1372 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1373 after the data-go-ahead and data-ack. Patch from Jason Betts.
1375 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1376 even for a "none" policy. Patch from Tony Meyer.
1378 JH/22 Fix continued use of a connection for further deliveries. If a port was
1379 specified by a router, it must also match for the delivery to be
1382 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1383 When one of the recipients of a message was unsuitable for the connection
1384 (has no matching addresses), we lost track of needing to mark it
1385 deferred. As a result mail would be lost.
1387 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1389 JH/25 Decoding ACL controls is now done using a binary search; the source code
1390 takes up less space and should be simpler to maintain. Merge the ACL
1391 condition decode tables also, with similar effect.
1393 JH/26 Fix problem with one_time used on a redirect router which returned the
1394 parent address unchanged. A retry would see the parent address marked as
1395 delivered, so not attempt the (identical) child. As a result mail would
1398 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1399 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1400 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1403 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1404 Main config options check_{log,spool}_{inodes,space} are now
1405 100 inodes, 10MB unless set otherwise in the configuration.
1407 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1408 Previously it only applied to the main-section connection policy
1411 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1413 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1414 by me. Added RFC7919 DH primes as an alternative.
1416 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1417 are not in the system include path.
1419 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1420 GnuTLS, when a session startup failed (eg because the client disconnected)
1421 Exim did stdio operations after fclose. This was exposed by a recent
1422 change which nulled out the file handle after the fclose.
1424 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1425 signed directly by the cert-signing cert, rather than an intermediate
1426 OCSP-signing cert. This is the model used by LetsEncrypt.
1428 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1430 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1431 an incoming connection.
1433 HS/02 Bug 1802: Do not half-close the connection after sending a request
1436 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1437 fallback to "prime256v1".
1439 JH/34 SECURITY: Use proper copy of DATA command in error message.
1440 Could leak key material. Remotely exploitable. CVE-2016-9963.
1446 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1447 and 3.4.4 - once the server is enabled to respond to an OCSP request
1448 it does even when not requested, resulting in a stapling non-aware
1449 client dropping the TLS connection.
1451 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1452 support variable-length bit vectors. No functional change.
1454 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1455 The I= interface field on outgoing lines is now after the H= remote
1456 host field, same as incoming lines. There is a separate
1457 outgoing_interface log selector which allows you to disable the
1460 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1461 If not running log_selector +smtp_connection the mainlog would be held
1462 open indefinitely after a "too many connections" event, including to a
1463 deleted file after a log rotate. Leave the per net connection logging
1464 leaving it open for efficiency as that will be quickly detected by the
1465 check on the next write.
1467 HS/01 Bug 1671: Fix post transport crash.
1468 Processing the wait-<transport> messages could crash the delivery
1469 process if the message IDs didn't exist for some reason. When
1470 using 'split_spool_directory=yes' the construction of the spool
1471 file name failed already, exposing the same netto behaviour.
1473 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1474 mime_regex ACL conditions.
1476 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1477 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1478 to HELO, local diagnostic string.
1480 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1481 log line to debug. Even when configured with a tls authenticator many
1482 client connections are expected to not authenticate in this way, so
1483 an authenticate fail is not an error.
1485 HS/02 Add the Exim version string to the process info. This way exiwhat
1486 gives some more detail about the running daemon.
1488 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1489 matter for fast-change records such as DNSBLs.
1491 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1492 retry record, even if constant. There may be multiple transports with
1493 different interface settings and the retry behaviour needs to be kept
1496 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1498 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1500 JH/10 Bug 840: fix log_defer_output option of pipe transport
1502 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1503 hosts_randomize. This matters a lot when combined with mua_wrapper.
1505 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1506 ${quote_pgsql:<string>} operator.
1508 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1510 JH/14 Change header file name for freeradius-client. Relevant if compiling
1511 with Radius support; from the Gentoo tree and checked under Fedora.
1513 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1515 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1516 option transport_filter, meaning no filtering.
1518 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1520 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1521 defaults to "*" (all hosts). The variable is now available when not built
1522 with TLS, default unset, mainly to enable keeping the testsuite sane.
1523 If a server certificate is not supplied (via tls_certificate) an error is
1524 logged, and clients will find TLS connections fail on startup. Presumably
1525 they will retry in-clear.
1526 Packagers of Exim are strongly encouraged to create a server certificate
1527 at installation time.
1529 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1530 with the $config_file variable.
1532 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1533 in transport context, after the attempt, and per-recipient. The latter type
1534 is per host attempted. The event data is the error message, and the errno
1535 information encodes the lookup type (A vs. MX) used for the (first) host,
1536 and the trailing two digits of the smtp 4xx response.
1538 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1539 to write to mainlog (or rejectlog, paniclog) in the window between file
1540 creation and permissions/ownership being changed. Particularly affects
1541 installations where exicyclog is run as root, rather than exim user;
1542 result is that the running daemon panics and dies.
1544 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1546 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1547 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1548 "pri" and "weight". Note that the previous implicit priority given by the
1549 list order is no longer honoured.
1551 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1552 for DKIM processing.
1554 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1555 by defining SUPPORT_SOCKS.
1557 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1558 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1559 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1560 variables are renamed to proxy_{local,external}_{address,port}.
1562 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1563 for a build by defining SUPPORT_I18N
1565 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1566 of the query string, and make ${quote_redis:} do that quoting.
1568 JH/29 Move Events support from Experimental to mainline, enabled by default
1569 and removable for a build by defining DISABLE_EVENT.
1571 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1573 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1574 cached by the daemon.
1576 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1577 by defining LOOKUP_REDIS. The libhiredis library is required.
1579 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1580 keys are given for lookup.
1582 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1583 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1584 only supported when built with TLS support. The PolarSSL SHA routines
1585 are still used when the TLS library is too old for convenient support.
1587 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1588 openssl_options), for security. OpenSSL forces this from version 1.1.0
1589 server-side so match that on older versions.
1591 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1592 allocation for $value could be released as the expansion processing
1593 concluded, but leaving the global pointer active for it.
1595 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1596 and to use the domains and local_parts ACL conditions.
1598 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1599 incorrectly not doubled on cutthrough transmission, hence seen as a
1600 body-termination at the receiving system - resulting in truncated mails.
1601 Commonly the sender saw a TCP-level error, and retransmitted the message
1602 via the normal store-and-forward channel. This could result in duplicates
1603 received - but deduplicating mailstores were liable to retain only the
1604 initial truncated version.
1606 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1608 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1610 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1611 we're in there, support oversigning also; bug 1309.
1613 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1615 HS/04 Add support for keep_environment and add_environment options.
1617 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1618 either intentional arithmetic overflow during PRNG, or testing config-
1621 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1622 delivery resulted in actual delivery. Cancel cutthrough before DATA
1625 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1626 rejects a recipient: pass the reject to the originator.
1628 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1629 Many were false-positives and ignorable, but it's worth fixing the
1632 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1633 for the new environment-manipulation done at startup. Move the routines
1634 from being local to tls.c to being global via the os.c file.
1636 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1637 an extract embedded as result-arg for a map, the first arg for extract
1638 is unavailable so we cannot tell if this is a numbered or keyed
1639 extraction. Accept either.
1645 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1648 JH/02 The smtp transport option "multi_domain" is now expanded.
1650 JH/03 The smtp transport now requests PRDR by default, if the server offers
1653 JH/04 Certificate name checking on server certificates, when exim is a client,
1654 is now done by default. The transport option tls_verify_cert_hostnames
1655 can be used to disable this per-host. The build option
1656 EXPERIMENTAL_CERTNAMES is withdrawn.
1658 JH/05 The value of the tls_verify_certificates smtp transport and main options
1659 default to the word "system" to access the system default CA bundle.
1660 For GnuTLS, only version 3.0.20 or later.
1662 JH/06 Verification of the server certificate for a TLS connection is now tried
1663 (but not required) by default. The verification status is now logged by
1664 default, for both outbound TLS and client-certificate supplying inbound
1667 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1670 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1671 Status Notification (bounce) messages are now MIME format per RFC 3464.
1672 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1673 under the control of the dsn_advertise_hosts option, and routers may
1674 have a dsn_lasthop option.
1676 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1677 default, modifiable by a malware= option. The list separator for
1678 the options can now be changed in the usual way. Bug 68.
1680 JH/10 The smtp_receive_timeout main option is now expanded before use.
1682 JH/11 The incoming_interface log option now also enables logging of the
1683 local interface on delivery outgoing connections.
1685 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1686 if the interface and destination host and port all match.
1688 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1691 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1692 Patch from Andrew Lewis.
1694 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1695 now supports optional time-restrictions, weighting, and priority
1696 modifiers per server. Patch originally by <rommer@active.by>.
1698 JH/16 The spamd_address main option now supports a mixed list of local
1699 and remote servers. Remote servers can be IPv6 addresses, and
1700 specify a port-range.
1702 JH/17 Bug 68: The spamd_address main option now supports an optional
1703 timeout value per server.
1705 JH/18 Bug 1581: Router and transport options headers_add/remove can
1706 now have the list separator specified.
1708 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1711 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1714 JH/21 Support for the A6 type of dns record is withdrawn.
1716 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1717 rather than the verbs used.
1719 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1720 from 255 to 1024 chars.
1722 JH/24 Verification callouts now attempt to use TLS by default.
1724 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1725 are generic router options now. The defaults didn't change.
1727 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1728 Original patch from Alexander Shikoff, worked over by JH.
1730 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1731 files if ZCAT_COMMAND is not executable.
1733 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1735 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1737 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1738 Normally benign, it bites when the pair was led to by a CNAME;
1739 modern usage is to not canonicalize the domain to a CNAME target
1740 (and we were inconsistent anyway for A-only vs AAAA+A).
1742 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1744 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1745 when evaluating $sender_host_dnssec.
1747 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1748 $sender_helo_dnssec variable.
1750 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1752 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1754 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1756 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1757 documented as working, but never had. Support all but $spam_report.
1759 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1760 added for tls authenticator.
1762 HS/03 Add perl_taintmode main config option
1768 TL/01 When running the test suite, the README says that variables such as
1769 no_msglog_check are global and can be placed anywhere in a specific
1770 test's script, however it was observed that placement needed to be near
1771 the beginning for it to behave that way. Changed the runtest perl
1772 script to read through the entire script once to detect and set these
1773 variables, reset to the beginning of the script, and then run through
1774 the script parsing/test process like normal.
1776 TL/02 The BSD's have an arc4random API. One of the functions to induce
1777 adding randomness was arc4random_stir(), but it has been removed in
1778 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1779 function when detected.
1781 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1782 cause callback expansion.
1784 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1785 syntax errors in an expansion can be treated as a string instead of
1786 logging or causing an error, due to the internal use of bool_lax
1787 instead of bool when processing it.
1789 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1790 server certificates when making smtp deliveries.
1792 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1794 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1796 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1798 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1799 Merged patch from Sebastian Wiedenroth.
1801 JH/05 Fix results-pipe from transport process. Several recipients, combined
1802 with certificate use, exposed issues where response data items split
1803 over buffer boundaries were not parsed properly. This eventually
1804 resulted in duplicates being sent. This issue only became common enough
1805 to notice due to the introduction of connection certificate information,
1806 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1808 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1809 size buffer was used, resulting in syntax errors when an expansion
1812 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1813 version 3.3.6 or later.
1815 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1816 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1817 both become "event_action", the variables become $event_name, $event_data
1818 and $event_defer_errno. There is a new variable $verify_mode, usable in
1819 routers, transports and related events. The tls:cert event is now also
1820 raised for inbound connections, if the main configuration event_action
1823 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1824 early OCSP support, but no stapling (appears to be less than 1.0.0).
1826 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1827 server certificate names available under the smtp transport option
1828 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1831 JH/10 Time-related extraction expansions from certificates now use the main
1832 option "timezone" setting for output formatting, and are consistent
1833 between OpenSSL and GnuTLS compilations. Bug 1541.
1835 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1836 encoded parameter in the incoming message. Bug 1558.
1838 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1839 include certificate info, eximon was claiming there were spoolfile
1842 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1844 JH/14 Log delivery-related information more consistently, using the sequence
1845 "H=<name> [<ip>]" wherever possible.
1847 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1848 are problematic for Debian distribution, omit them from the release
1851 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1853 JH/16 Fix string representation of time values on 64bit time_t architectures.
1856 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1857 output list separator was used.
1862 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1863 checkers that were complaining about end of non-void function with no
1866 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1867 This was a regression introduced in 4.83 by another bugfix.
1869 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1871 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1872 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1878 TF/01 Correctly close the server side of TLS when forking for delivery.
1880 When a message was received over SMTP with TLS, Exim failed to clear up
1881 the incoming connection properly after forking off the child process to
1882 deliver the message. In some situations the subsequent outgoing
1883 delivery connection happened to have the same fd number as the incoming
1884 connection previously had. Exim would try to use TLS and fail, logging
1885 a "Bad file descriptor" error.
1887 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1888 utilities have not been installed.
1890 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1891 temporary space as the ACL may create new global variables.
1893 TL/01 LDAP support uses per connection or global context settings, depending
1894 upon the detected version of the libraries at build time.
1896 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1897 to extract and use the src ip:port in logging and expansions as if it
1898 were a direct connection from the outside internet. PPv2 support was
1899 updated based on HAProxy spec change in May 2014.
1901 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1903 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1904 Properly escape header and check for NULL return.
1906 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1909 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1911 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1912 characters in header names, implemented as a verify condition.
1913 Contributed by Michael Fischer v. Mollard.
1915 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1916 results permerror and temperror. Previous values are deprecated but
1917 still accepted. In a future release, err_perm and err_temp will be
1918 completely removed, which will be a backward incompatibility if the
1919 ACL tests for either of these two old results. Patch contributed by
1920 user bes-internal on the mailing list.
1922 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1924 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1925 selectors, in both main and reject logs.
1927 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1930 JH/07 Add malware type "sock" for talking to simple daemon.
1932 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1934 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1935 routers/transports under cutthrough routing.
1937 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1938 numbers. Touch up "bool" conditional to keep the same definition.
1940 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1942 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1944 JH/12 Expand items in router/transport headers_add or headers_remove lists
1945 individually rather than the list as a whole. Bug 1452.
1947 Required for reasonable handling of multiple headers_ options when
1948 they may be empty; requires that headers_remove items with embedded
1949 colons must have them doubled (or the list-separator changed).
1951 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1952 view the policy declared in the DMARC record. Currently, $dmarc_status
1953 is a combined value of both the record presence and the result of the
1956 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1958 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1959 dnslookup router and the smtp transport (applying to the forward
1962 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1963 of ldap servers used for a specific lookup. Patch provided by Heiko
1966 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1967 New variable $lookup_dnssec_authenticated for observability.
1969 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1970 Patch submitted by Lars Timman.
1972 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1974 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1975 Requires trusted mode and valid format message id, aborts otherwise.
1976 Patch contributed by Heiko Schlichting.
1978 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1979 certextract with support for various fields. Bug 1358.
1981 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1982 is requested by default, modifiable by smtp transport option
1985 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1986 operate on certificate variables to give certificate fingerprints
1987 Also new ${sha256:cert_variable}.
1989 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1991 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1992 Christian Aistleitner.
1994 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1996 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1997 file. Patch from Wolfgang Breyha.
1999 JH/25 Expand the coverage of the delivery $host and $host_address to
2000 client authenticators run in verify callout. Bug 1476.
2002 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2003 align with daemon_smtp_ports. Bug 72.
2005 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2006 support and error reporting did not work properly.
2008 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2009 and is readable. Patch from Andrew Colin Kissa.
2011 TL/14 Enhance documentation of ${run expansion and how it parses the
2012 commandline after expansion, particularly in the case when an
2013 unquoted variable expansion results in an empty value.
2015 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2017 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2018 by renaming away from it. Addresses GH issue 15, reported by
2021 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2023 TL/15 SECURITY: prevent double expansion in math comparison functions
2024 (can expand unsanitized data). Not remotely exploitable.
2031 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2033 PP/02 Make -n do something, by making it not do something.
2034 When combined with -bP, the name of an option is not output.
2036 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2039 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2040 $sender_host_name and config options to manage this, and basic check
2043 PP/05 DSCP support for outbound connections and control modifier for inbound.
2045 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2046 (Only plugin which currently uses this is kerberos4, which nobody should
2047 be using, but we should make it available and other future plugins might
2048 conceivably use it, even though it would break NAT; stuff *should* be
2049 using channel bindings instead).
2051 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2052 name; added for Sendmail compatibility; requires admin caller.
2053 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2054 just ignore it); requires trusted caller.
2055 Also parse but ignore: -Ac -Am -X<logfile>
2058 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2060 TL/02 Add +smtp_confirmation as a default logging option.
2062 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2063 Patch by Magnus Holmgren from 2007-02-20.
2065 TL/04 Bugzilla 1281 - Spec typo.
2066 Bugzilla 1283 - Spec typo.
2067 Bugzilla 1290 - Spec grammar fixes.
2069 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2071 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2073 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2074 mailing list by by Dmitry Isaikin.
2076 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2078 JH/02 Support "G" suffix to numbers in ${if comparisons.
2080 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2082 NM/01 Bugzilla 1197 - Spec typo
2083 Bugzilla 1196 - Spec examples corrections
2085 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2087 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2088 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2091 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2092 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2094 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2095 "acl {{name}{arg}...}", and optional args on acl condition
2098 JH/05 Permit multiple router/transport headers_add/remove lines.
2100 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2102 JH/07 Avoid using a waiting database for a single-message-only transport.
2103 Performance patch from Paul Fisher. Bugzilla 1262.
2105 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2108 JH/09 Add $headers_added variable, with content from use of ACL modifier
2109 add_header (but not yet added to the message). Bugzilla 199.
2111 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2112 Pulled from Bugzilla 817 by Wolfgang Breyha.
2114 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2116 (nb: this is the same fix as in Exim 4.80.1)
2118 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2121 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2124 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2126 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2127 advertise SMTP AUTH mechanism to us, instead of a generic
2128 protocol violation error. Also, make Exim more robust to bad
2129 data from the Dovecot auth socket.
2131 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2133 When a queue runner is handling a message, Exim first routes the
2134 recipient addresses, during which it prunes them based on the retry
2135 hints database. After that it attempts to deliver the message to
2136 any remaining recipients. It then updates the hints database using
2139 So if a recipient address works intermittently, it can get repeatedly
2140 deferred at routing time. The retry hints record remains fresh so the
2141 address never reaches the final cutoff time.
2143 This is a fairly common occurrence when a user is bumping up against
2144 their storage quota. Exim had some logic in its local delivery code
2145 to deal with this. However it did not apply to per-recipient defers
2146 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2148 This change adds a proper retry rule check during routing so that the
2149 final cutoff time is checked against the message's age. We only do
2150 this check if there is an address retry record and there is not a
2151 domain retry record; this implies that previous attempts to handle
2152 the address had the retry_use_local_parts option turned on. We use
2153 this as an approximation for the destination being like a local
2154 delivery, as in LMTP.
2156 I suspect this new check makes the old local delivery cutoff check
2157 redundant, but I have not verified this so I left the code in place.
2159 TF/02 Correct gecos expansion when From: is a prefix of the username.
2161 Test 0254 submits a message to Exim with the header
2165 When I ran the test suite under the user fanf2, Exim expanded
2166 the header to contain my full name, whereas it should have added
2167 a Resent-Sender: header. It erroneously treats any prefix of the
2168 username as equal to the username.
2170 This change corrects that bug.
2172 GF/01 DCC debug and logging tidyup
2173 Error conditions log to paniclog rather than rejectlog.
2174 Debug lines prefixed by "DCC: " to remove any ambiguity.
2176 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2178 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2179 Bug spotted by Jeremy Harris; was flawed since initial commit.
2180 Would have resulted in OCSP responses post-SNI triggering an Exim
2181 NULL dereference and crash.
2183 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2185 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2186 Bug detection, analysis and fix by Samuel Thibault.
2187 Bugzilla 1331, Debian bug #698092.
2189 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2191 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2192 Server implementation by Todd Lyons, client by JH.
2193 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2194 config variable "prdr_enable" controls whether the server
2195 advertises the facility. If the client requests PRDR a new
2196 acl_data_smtp_prdr ACL is called once for each recipient, after
2197 the body content is received and before the acl_smtp_data ACL.
2198 The client is controlled by both of: a hosts_try_prdr option
2199 on the smtp transport, and the server advertisement.
2200 Default client logging of deliveries and rejections involving
2201 PRDR are flagged with the string "PRDR".
2203 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2204 fclose(). Diagnosis by Todd Lyons.
2206 PP/17 Update configure.default to handle IPv6 localhost better.
2207 Patch by Alain Williams (plus minor tweaks).
2210 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2211 This is now consistent with GnuTLS, and is now documented: the
2212 previous undocumented portable approach to treating the option as
2213 unset was to force an expansion failure. That still works, and
2214 an empty string is now equivalent.
2216 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2217 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2218 not performing validation itself.
2220 PP/20 Added force_command boolean option to pipe transport.
2221 Patch from Nick Koston, of cPanel Inc.
2223 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2226 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2228 PP/21 Fix eximon continuous updating with timestamped log-files.
2229 Broken in a format-string cleanup in 4.80, missed when I repaired the
2230 other false fix of the same issue.
2231 Report and fix from Heiko Schlichting.
2234 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2235 Report from Prashanth Katuri.
2237 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2238 It's SecureTransport, so affects any MacOS clients which use the
2239 system-integrated TLS libraries, including email clients.
2241 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2242 using a MIME ACL for non-SMTP local injection.
2243 Report and assistance in diagnosis by Warren Baker.
2245 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2247 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2249 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2250 last id that failed so it may be referenced in subsequent ACL's.
2252 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2255 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2256 ldap library initialization, allowing self-signed CA's to be
2257 used. Also properly sets require_cert option later in code by
2258 using NULL (global ldap config) instead of ldap handle (per
2259 session). Bug diagnosis and testing by alxgomz.
2261 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2262 the src/util/ subdirectory.
2264 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2265 renamed to Transport Post Delivery Action by Jeremy Harris, as
2268 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2269 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2270 redis_servers = needs to be configured which will be used by the redis
2271 lookup. Patch from Warren Baker, of The Packet Hub.
2273 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2275 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2276 hostname or reverse DNS when processing a host list. Used suggestions
2277 from multiple comments on this bug.
2279 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2281 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2282 Missed a few lines, added it to make the runtest require no keyboard
2285 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2286 contains upper case chars. Make router use caseful_local_part.
2288 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2289 support when GnuTLS has been built with p11-kit.
2295 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2297 This, or similar/improved, will also be change PP/11 of 4.82.
2303 PP/01 Handle short writes when writing local log-files.
2304 In practice, only affects FreeBSD (8 onwards).
2305 Bugzilla 1053, with thanks to Dmitry Isaikin.
2307 NM/01 Bugzilla 949 - Documentation tweak
2309 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2312 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2314 PP/02 Implemented gsasl authenticator.
2316 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2318 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2319 `pkg-config foo` for cflags/libs.
2321 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2322 with rest of GSASL and with heimdal_gssapi.
2324 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2325 `pkg-config foo` for cflags/libs for the TLS implementation.
2327 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2328 properties get this fed in as external SSF. A number of robustness
2329 and debugging improvements to the cyrus_sasl authenticator.
2331 PP/08 cyrus_sasl server now expands the server_realm option.
2333 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2334 Patch by Jeremy Harris.
2336 PP/10 Added dbmjz lookup type.
2338 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2340 PP/12 MAIL args handles TAB as well as SP, for better interop with
2341 non-compliant senders.
2342 Analysis and variant patch by Todd Lyons.
2344 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2345 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2346 Patch from Dirk Mueller <dmueller@suse.com>
2348 PP/13 tls_peerdn now print-escaped for spool files.
2349 Observed some $tls_peerdn in wild which contained \n, which resulted
2350 in spool file corruption.
2352 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2353 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2354 or write after TLS renegotiation, which otherwise led to messages
2357 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2358 as a tracking header (ie: a signed header comes before the signature).
2359 Patch from Wolfgang Breyha.
2361 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2362 comma-sep list; embedded commas doubled.
2364 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2366 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2368 Report and patch from Dmitry Banschikov.
2370 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2371 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2372 failures. We appear to now support TLS1.1+ with Exim.
2374 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2375 lets Exim select keys and certificates based upon TLS SNI from client.
2376 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2377 before an outbound SMTP session. New log_selector, +tls_sni.
2379 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2380 NULL dereference. Report and patch from Alun Jones.
2382 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2383 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2384 Not seeing resolver debug output on NetBSD, but suspect this is a
2385 resolver implementation change.
2387 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2388 Left warnings. Added "eximon gdb" invocation mode.
2390 PP/21 Defaulting "accept_8bitmime" to true, not false.
2392 PP/22 Added -bw for inetd wait mode support.
2394 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2395 locate the relevant includes and libraries. Made this the default.
2397 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2398 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2400 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2401 This may cause build issues on older platforms.
2403 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2404 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2405 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2406 Added SNI support via GnuTLS too.
2407 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2409 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2411 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2412 Applied second patch from Janne, implementing suggestion to default
2413 multiple-strings-in-record handling to match SPF spec.
2415 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2417 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2418 read-only, out of scope).
2419 Patch from Wolfgang Breyha, report from Stuart Northfield.
2421 PP/29 Fix three issues highlighted by clang analyser static analysis.
2422 Only crash-plausible issue would require the Cambridge-specific
2423 iplookup router and a misconfiguration.
2424 Report from Marcin Mirosław.
2426 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2428 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2429 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2430 As part of this, removing so much warning spew let me fix some minor
2431 real issues in debug logging.
2433 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2434 assignment on my part. Fixed.
2436 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2437 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2438 Janne Snabb (who went above and beyond: thank you).
2440 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2441 string otherwise requires a connection and a bunch more work and it's
2442 relatively easy to get wrong. Should also expose TLS library linkage
2445 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2446 64-bit ${eval} (JH/03).
2448 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2449 GNU libc to support some of the 64-bit stuff, should not lead to
2450 conflicts. Defined before os.h is pulled in, so if a given platform
2451 needs to override this, it can.
2453 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2454 protection layer was required, which is not implemented.
2455 Bugzilla 1254, patch from Wolfgang Breyha.
2457 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2458 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2459 tls_dhparam take prime identifiers. Also unbreak combination of
2460 OpenSSL+DH_params+TLSSNI.
2462 PP/39 Disable SSLv2 by default in OpenSSL support.
2468 PP/01 Solaris build fix for Oracle's LDAP libraries.
2469 Bugzilla 1109, patch from Stephen Usher.
2471 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2473 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2476 TF/02 Fix a couple more cases where we did not log the error message
2477 when unlink() failed. See also change 4.74-TF/03.
2479 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2480 lock up or crash if it happened to be inside a call to libc when it
2481 got a SIGUSR1 from exiwhat.
2483 The SIGUSR1 handler appends the current process status to the process
2484 log which is later printed by exiwhat. It used to use the general
2485 purpose logging code to do this, but several functions it calls are
2486 not safe for signals.
2488 The new output code in the SIGUSR1 handler is specific to the process
2489 log, and simple enough that it's easy to inspect for signal safety.
2490 Removing some special cases also simplifies the general logging code.
2491 Removing the spurious timestamps from the process log simplifies
2494 TF/04 Improved ratelimit ACL condition.
2496 The /noupdate option has been deprecated in favour of /readonly which
2497 has clearer semantics. The /leaky, /strict, and /readonly update modes
2498 are mutually exclusive. The update mode is no longer included in the
2499 database key; it just determines when the database is updated. (This
2500 means that when you upgrade Exim will forget old rate measurements.)
2502 Exim now checks that the per_* options are used with an update mode that
2503 makes sense for the current ACL. For example, when Exim is processing a
2504 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2505 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2506 must specify per_mail/readonly. If you omit the update mode it defaults to
2507 /leaky where that makes sense (as before) or /readonly where required.
2509 The /noupdate option is now undocumented but still supported for
2510 backwards compatibility. It is equivalent to /readonly except that in
2511 ACLs where /readonly is required you may specify /leaky/noupdate or
2512 /strict/noupdate which are treated the same as /readonly.
2514 A useful new feature is the /count= option. This is a generalization
2515 of the per_byte option, so that you can measure the throughput of other
2516 aggregate values. For example, the per_byte option is now equivalent
2517 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2519 The per_rcpt option has been generalized using the /count= mechanism
2520 (though it's more complicated than the per_byte equivalence). When it is
2521 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2522 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2523 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2524 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2525 Note that using per_rcpt with a non-readonly update mode in more than
2526 one ACL will cause the recipients to be double-counted. (The per_mail
2527 and per_byte options don't have this problem.)
2529 The handling of very low rates has changed slightly. If the computed rate
2530 is less than the event's count (usually one) then this event is the first
2531 after a long gap. In this case the rate is set to the same as this event's
2532 count, so that the first message of a spam run is counted properly.
2534 The major new feature is a mechanism for counting the rate of unique
2535 events. The new per_addr option counts the number of different
2536 recipients that someone has sent messages to in the last time period. It
2537 behaves like per_rcpt if all the recipient addresses are different, but
2538 duplicate recipient addresses do not increase the measured rate. Like
2539 the /count= option this is a general mechanism, so the per_addr option
2540 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2541 example, measure the rate that a client uses different sender addresses
2542 with the options per_mail/unique=$sender_address. There are further
2543 details in the main documentation.
2545 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2547 TF/06 Removed a few PCRE remnants.
2549 TF/07 Automatically extract Exim's version number from tags in the git
2550 repository when doing development or release builds.
2552 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2553 Bugzilla 879. Patch from Paul Fisher.
2555 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2556 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2559 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2561 PP/05 Variable $av_failed, true if the AV scanner deferred.
2562 Bugzilla 1078. Patch from John Horne.
2564 PP/06 Stop make process more reliably on build failure.
2565 Bugzilla 1087. Patch from Heiko Schlittermann.
2567 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2568 Bugzilla 1089. Patch from Heiko Schlittermann.
2570 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2571 Bugzilla 1131. Patch from Holger Weiß.
2573 PP/09 Handle IPv6 addresses with SPF.
2574 Bugzilla 860. Patch from Wolfgang Breyha.
2576 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2578 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2581 PP/11 match_* no longer expand right-hand-side by default.
2582 New compile-time build option, EXPAND_LISTMATCH_RHS.
2583 New expansion conditions, "inlist", "inlisti".
2585 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2587 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2589 PP/14 fix log_write() format string regression from TF/03.
2590 Bugzilla 1152. Patch from Dmitry Isaikin.
2596 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2598 PP/02 Harmonised TLS library version reporting; only show if debugging.
2599 Layout now matches that introduced for other libraries in 4.74 PP/03.
2601 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2603 PP/04 New "dns_use_edns0" global option.
2605 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2608 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2609 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2611 TK/01 Updated PolarSSL code to 0.14.2.
2612 Bugzilla 1097. Patch from Andreas Metzler.
2614 PP/07 Catch divide-by-zero in ${eval:...}.
2615 Fixes bugzilla 1102.
2617 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2620 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2621 format-string attack -- SECURITY: remote arbitrary code execution.
2623 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2624 time unintentionally subject to list matching rules, letting the header
2625 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2626 arbitrary string expansion). This allowed for information disclosure.
2628 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2629 INT_MIN/-1 -- value coerced to INT_MAX.
2635 NM/01 Workaround for PCRE version dependency in version reporting
2638 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2639 This fixes portability to compilers other than gcc, notably
2640 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2642 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2643 makefiles for portability to HP-UX and POSIX correctness.
2645 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2646 Also via indented variable definition in the Makefile.
2647 (Debugging by Oliver Heesakkers).
2649 PP/02 Restore caching of spamd results with expanded spamd_address.
2650 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2652 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2653 Improves build reliability. Fix from: Frank Elsner
2655 NM/02 Fix wide character breakage in the rfc2047 coding
2656 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2658 NM/03 Allow underscore in dnslist lookups
2659 Fixes bug 1026. Patch from Graeme Fowler
2661 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2662 Code patches from Adam Ciarcinski of NetBSD.
2664 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2667 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2668 is logged, to avoid truncation. Patch from John Horne.
2670 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2671 Patch from Jakob Hirsch.
2673 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2674 SQL string expansion failure details.
2675 Patch from Andrey Oktyabrski.
2677 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2678 Patch from Simon Arlott.
2680 PP/09 New lookups functionality failed to compile on old gcc which rejects
2681 extern declarations in function scope.
2682 Patch from Oliver Fleischmann
2684 PP/10 Use sig_atomic_t for flags set from signal handlers.
2685 Check getgroups() return and improve debugging.
2686 Fixed developed for diagnosis in bug 927 (which turned out to be
2689 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2690 Patch from Mark Zealey.
2692 PP/12 Bugzilla 1056: Improved spamd server selection.
2693 Patch from Mark Zealey.
2695 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2696 Based on patch from Heiko Schlittermann.
2698 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2699 Patch from Uwe Doering, sign-off by Michael Haardt.
2701 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2702 variable declaration deep within a block. Bug and patch from
2705 PP/15 lookups-Makefile IRIX compatibility coercion.
2707 PP/16 Make DISABLE_DKIM build knob functional.
2709 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2710 Patch by Simon Arlott
2712 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2713 variable argument macros. Our copy now differs from upstream.
2719 TF/01 Failure to get a lock on a hints database can have serious
2720 consequences so log it to the panic log.
2722 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2723 controlled using the smtp_confirmation log selector.
2725 TF/03 Include the error message when we fail to unlink a spool file.
2727 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2728 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2729 for maintaining out-of-tree patches for some time.
2731 PP/01 Bugzilla 139: Documentation and portability issues.
2732 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2733 Handle per-OS dynamic-module compilation flags.
2735 PP/02 Let /dev/null have normal permissions.
2736 The 4.73 fixes were a little too stringent and complained about the
2737 permissions on /dev/null. Exempt it from some checks.
2738 Reported by Andreas M. Kirchwitz.
2740 PP/03 Report version information for many libraries, including
2741 Exim version information for dynamically loaded libraries. Created
2742 version.h, now support a version extension string for distributors
2743 who patch heavily. Dynamic module ABI change.
2745 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2746 privilege escalation vulnerability whereby the Exim run-time user
2747 can cause root to append content of the attacker's choosing to
2750 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2753 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2754 If dropping privileges for untrusted macros, we disabled normal logging
2755 on the basis that it would fail; for the Exim run-time user, this is not
2756 the case, and it resulted in successful deliveries going unlogged.
2757 Fixed. Reported by Andreas Metzler.
2763 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2764 only prepend for the Resent-* case. Fixes regression introduced in
2765 Exim 4.70 by NM/22 for Bugzilla 607.
2767 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2768 increasing numbers of administrators be bitten by this.
2770 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2772 PP/03 Bugzilla 994: added openssl_options main configuration option.
2774 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2776 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2778 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2780 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2781 without a peer certificate, leading to a segfault because of an
2782 assumption that peers always have certificates. Be a little more
2783 paranoid. Problem reported by Martin Tscholak.
2785 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2786 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2787 NB: ClamAV planning to remove STREAM in "middle of 2010".
2788 CL also introduces -bmalware, various -d+acl logging additions and
2789 more caution in buffer sizes.
2791 PP/09 Implemented reverse_ip expansion operator.
2793 PP/10 Bugzilla 937: provide a "debug" ACL control.
2795 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2797 PP/12 Bugzilla 973: Implement --version.
2799 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2801 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2803 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2805 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2806 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2807 ignore trailing whitespace.
2809 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2811 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2814 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2815 Notification from Dr Andrew Aitchison.
2817 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2818 ExtendedDetectionInfo response format.
2819 Notification from John Horne.
2821 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2824 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2825 XSL and documented dependency on system catalogs, with examples of how
2828 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2831 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2832 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2833 configuration file which is writeable by the Exim user or group.
2835 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2836 of configuration files to cover files specified with the -C option if
2837 they are going to be used with root privileges, not just the default
2840 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2841 option (effectively making it always true).
2843 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2844 files to be used while preserving root privileges.
2846 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2847 that rogue child processes cannot use them.
2849 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2850 run-time user, instead of root.
2852 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2853 Exim run-time user without dropping privileges.
2855 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2856 result string, instead of calling string_vformat() twice with the same
2859 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2860 for other users. Others should always drop root privileges if they use
2861 -C on the command line, even for a whitelisted configure file.
2863 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2865 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2871 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2872 $data_path, and $header_path variables; fixed documentation bugs and
2875 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2876 exipick to access non-standard spools, including the "frozen" queue
2879 NM/01 Bugzilla 965: Support mysql stored procedures.
2880 Patch from Alain Williams
2882 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2884 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2885 Patch from Andreas Metzler
2887 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2888 Patch from Kirill Miazine
2890 NM/05 Bugzilla 671: Added umask to procmail example.
2892 JJ/03 installed exipick 20100323.0, fixing doc bug
2894 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2895 directory. Notification and patch from Dan Rosenberg.
2897 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2899 TK/02 Improve log output when DKIM signing operation fails.
2901 MH/01 Treat the transport option dkim_domain as a colon separated
2902 list, not as a single string, and sign the message with each element,
2903 omitting multiple occurences of the same signer.
2905 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2906 Bugzilla 985, 986. Patch by Simon Arlott
2908 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2909 Patch by Simon Arlott
2911 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2912 MBX locking. Notification from Dan Rosenberg.
2918 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2920 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2922 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2924 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2926 NM/04 Bugzilla 915: Flag broken perl installation during build.
2932 TK/01 Added patch by Johannes Berg that expands the main option
2933 "spamd_address" if it starts with a dollar sign.
2935 TK/02 Write list of recipients to X-Envelope-Sender header when building
2936 the mbox-format spool file for content scanning (suggested by Jakob
2939 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2940 (http://www.dcc-servers.net/) support via dccifd. Activated by
2941 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2943 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2944 by Mark Daniel Reidel <mr@df.eu>.
2946 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2947 When building exim an external PCRE library is now needed -
2948 PCRE is a system library on the majority of modern systems.
2949 See entry on PCRE_LIBS in EDITME file.
2951 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2952 conversation. Added nologin parameter to request.
2953 Patch contributed by Kirill Miazine.
2955 TF/01 Do not log submission mode rewrites if they do not change the address.
2957 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2959 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2960 log files in place. Contributed by Roberto Lima.
2962 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2964 TF/03 Bugzilla 615: When checking the local_parts router precondition
2965 after a local_part_suffix or local_part_prefix option, Exim now
2966 does not use the address's named list lookup cache, since this
2967 contains cached lookups for the whole local part.
2969 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2970 Robert Millan. Documentation is in experimental-spec.txt.
2972 TF/04 Bugzilla 668: Fix parallel build (make -j).
2974 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2976 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2977 Patch provided by Jan Srzednicki.
2979 TF/05 Leading white space used to be stripped from $spam_report which
2980 wrecked the formatting. Now it is preserved.
2982 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2983 that they are available at delivery time.
2985 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2987 TF/08 TLS error reporting now respects the incoming_interface and
2988 incoming_port log selectors.
2990 TF/09 Produce a more useful error message if an SMTP transport's hosts
2991 setting expands to an empty string.
2993 NM/06 Bugzilla 744: EXPN did not work under TLS.
2994 Patch provided by Phil Pennock.
2996 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2997 Patch provided by Richard Godbee.
2999 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3000 acl_smtp_notquit, added index entry.
3002 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3003 Patch provided by Eugene Bujak.
3005 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3006 accept(). Patch provided by Maxim Dounin.
3008 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3009 Patch provided by Phil Pennock.
3011 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3013 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3014 Patch provided by Brad "anomie" Jorsch.
3016 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3017 Patch provided by Dean Brooks.
3019 TK/05 Add native DKIM support (does not depend on external libraries).
3021 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3022 Patch provided by Graeme Fowler.
3024 NM/16 Bugzilla 851: Documentation example syntax fix.
3026 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3028 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3031 NM/19 Bugzilla 745: TLS version reporting.
3032 Patch provided by Phil Pennock.
3034 NM/20 Bugzilla 167: bool: condition support.
3035 Patch provided by Phil Pennock.
3037 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3038 clients. Patch provided by Phil Pennock.
3040 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3041 Patch provided by Brad "anomie" Jorsch.
3043 NM/23 Bugzilla 687: Fix misparses in eximstats.
3044 Patch provided by Heiko Schlittermann.
3046 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3047 Patch provided by Heiko Schlittermann.
3049 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3050 plus update to original patch.
3052 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3054 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3055 Patch provided by David Brownlee.
3057 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3059 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3061 NM/30 Bugzilla 888: TLS documentation bugfixes.
3063 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3065 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3066 Unlike the original bugzilla I have changed all shell scripts in src tree.
3068 NM/33 Bugzilla 898: Transport filter timeout fix.
3069 Patch by Todd Rinaldo.
3071 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3072 Patch by Serge Demonchaux.
3074 NM/35 Bugzilla 39: Base64 decode bug fixes.
3075 Patch by Jakob Hirsch.
3077 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3079 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3081 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3083 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3089 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3090 ALT-N's libdkim that I have put here:
3091 http://duncanthrax.net/exim-experimental/
3093 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3094 were called 'true' and it seems that C99 defines that as a reserved
3095 keyword to be used with 'bool' variable types. That means you could
3096 not include C99-style headers which use bools without triggering
3097 build errors in sieve.c.
3099 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3100 as mailq or other aliases. Changed the --help handling significantly
3101 to do whats expected. exim_usage() emits usage/help information.
3103 SC/01 Added the -bylocaldomain option to eximstats.
3105 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3107 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3109 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3115 PH/01 Another patch from the Sieve maintainer.
3117 PH/02 When an IPv6 address is converted to a string for single-key lookup
3118 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3119 dots are used instead of colons so that keys in lsearch files need not
3120 contain colons. This was done some time before quoting was made available
3121 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3122 (notated using the quote facility) so as to distinguish them from IPv4
3123 keys. This meant that lookups for IP addresses in host lists did not work
3124 for iplsearch lookups.
3126 This has been fixed by arranging for IPv6 addresses to be expressed with
3127 colons if the lookup type is iplsearch. This is not incompatible, because
3128 previously such lookups could never work.
3130 The situation is now rather anomalous, since one *can* have colons in
3131 ordinary lsearch keys. However, making the change in all cases is
3132 incompatible and would probably break a number of configurations.
3134 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3137 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3138 conversion specification without a maximum field width, thereby enabling
3139 a rogue spamd server to cause a buffer overflow. While nobody in their
3140 right mind would setup Exim to query an untrusted spamd server, an
3141 attacker that gains access to a server running spamd could potentially
3142 exploit this vulnerability to run arbitrary code as the Exim user.
3144 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3145 $primary_hostname instead of what libspf2 thinks the hosts name is.
3147 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3148 a directory entry by the name of the lookup key. Previously, if a
3149 symlink pointed to a non-existing file or a file in a directory that
3150 Exim lacked permissions to read, a lookup for a key matching that
3151 symlink would fail. Now it is enough that a matching directory entry
3152 exists, symlink or not. (Bugzilla 503.)
3154 PH/03 The body_linecount and body_zerocount variables are now exported in the
3157 PH/04 Added the $dnslist_matched variable.
3159 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3160 This means they are set thereafter only if the connection becomes
3163 PH/06 Added the client_condition to authenticators so that some can be skipped
3164 by clients under certain conditions.
3166 PH/07 The error message for a badly-placed control=no_multiline_responses left
3167 "_responses" off the end of the name.
3169 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3171 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3172 (without spaces) instead of just copying the configuration text.
3174 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3176 PH/11 Added $max_received_linelength.
3178 PH/12 Added +ignore_defer and +include_defer to host lists.
3180 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3181 way in which PCRE > 7.0 is built.
3183 PH/14 Implemented queue_only_load_latch.
3185 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3186 MAIL command. The effect was to mangle the value on 64-bit systems.
3188 PH/16 Another patch from the Sieve maintainer.
3190 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3192 PH/18 If a system quota error occurred while trying to create the file for
3193 a maildir delivery, the message "Mailbox is full" was not appended to the
3194 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3195 only to a quota excession during the actual writing of the file.
3197 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3198 characters?) which causes problems in log lines. The DN values are now
3199 passed through string_printing() before being added to log lines.
3201 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3202 and InterBase are left for another time.)
3204 PH/21 Added message_body_newlines option.
3206 PH/22 Guard against possible overflow in moan_check_errorcopy().
3208 PH/23 POSIX allows open() to be a macro; guard against that.
3210 PH/24 If the recipient of an error message contained an @ in the local part
3211 (suitably quoted, of course), incorrect values were put in $domain and
3212 $local_part during the evaluation of errors_copy.
3218 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3219 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3222 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3223 issue a MAIL command.
3225 PH/02 In an ACL statement such as
3227 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3229 if a client was not listed at all, or was listed with a value other than
3230 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3231 the condition was not true (as it should be), so access was not denied.
3232 The bug was that the ! inversion was incorrectly passed on to the second
3233 item. This has been fixed.
3235 PH/03 Added additional dnslists conditions == and =& which are different from
3236 = and & when the dns lookup returns more than one IP address.
3238 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3239 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3241 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3242 FSYNC, which compiles an option called disable_fsync that allows for
3243 bypassing fsync(). The documentation is heavily laced with warnings.
3245 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3247 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3248 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3249 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3250 including adding "make clean"; (3) Added -fPIC when compiling the test
3251 dynamically loaded module, to get rid of a warning.
3253 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3254 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3255 = 0s. The bug is otherwise harmless.
3257 PH/07 There was a bug in the dovecot authenticator such that the value of
3258 $auth1 could be overwritten, and so not correctly preserved, after a
3259 successful authentication. This usually meant that the value preserved by
3260 the server_setid option was incorrect.
3262 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3264 PH/09 Installed PCRE release 7.0.
3266 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3267 run for batched SMTP input. It is now run at the start of every message
3268 in the batch. While fixing this I discovered that the process information
3269 (output by running exiwhat) was not always getting set for -bs and -bS
3270 input. This is fixed, and it now also says "batched" for BSMTP.
3272 PH/11 Added control=no_pipelining.
3274 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3275 patch, slightly modified), and move the expansion of helo_data till after
3276 the connection is made in the smtp transport (so it can use these
3279 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3281 PH/14 Added log_selector = +pid.
3283 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3285 PH/16 Add ${if forany and ${if forall.
3287 PH/17 Added dsn_from option to vary the From: line in DSNs.
3289 PH/18 Flush SMTP output before performing a callout, unless control =
3290 no_callout_flush is set.
3292 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3293 was true (the default) a successful delivery failed to delete the retry
3294 item, thus causing premature timeout of the address. The bug is now
3297 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3299 PH/21 Long custom messages for fakedefer and fakereject are now split up
3300 into multiline responses in the same way that messages for "deny" and
3301 other ACL rejections are.
3303 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3304 with slight modification.
3306 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3307 draft, changing the syntax and factoring some duplicate code".
3309 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3310 for deliveries of the second and subsequent messages over the same SMTP
3313 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3314 ${reduce, with only minor "tidies".
3316 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3318 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3319 expansion side effects.
3321 PH/27 When a message times out after an over-quota error from an Exim-imposed
3322 quota, the bounce message says "mailbox is full". This message was not
3323 being given when it was a system quota that was exceeded. It now should
3326 MH/03 Made $recipients available in local_scan(). local_scan() already has
3327 better access to the recipient list through recipients_list[], but
3328 $recipients can be useful in postmaster-provided expansion strings.
3330 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3331 in the case of a MAIL command with additional options following the
3332 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3333 were accidentally chopped off.
3335 PH/29 SMTP synchronization checks are implemented when a command is read -
3336 there is a check that no more input is waiting when there shouldn't be
3337 any. However, for some commands, a delay in an ACL can mean that it is
3338 some time before the response is written. In this time, more input might
3339 arrive, invalidly. So now there are extra checks after an ACL has run for
3340 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3341 pipelining has not been advertised.
3343 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3345 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3346 returned in the SMTP response when smtp_return_error_details was false.
3347 This has been fixed.
3349 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3350 instead of the C I/O that was originally supplied, because problems were
3351 reported on Solaris.
3353 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3354 Exim which did not show up earlier: it was assuming that a call to
3355 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3356 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3357 was a macro that became an assignment, so it seemed to work. This has
3358 changed to a proper function call with a void return, hence the compile
3359 error. Exim's code has been fixed.
3361 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3364 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3366 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3368 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3369 because it turns out to be too big or there is a timeout) there is no
3370 "Completed" line in the log. When some messages of this type were
3371 selected by exigrep, they were listed as "not completed". Others were
3372 picked up by some special patterns. I have improved the selection
3373 criteria to be more general.
3375 PH/38 The host_find_failed option in the manualroute router can now be set
3376 to "ignore", to completely ignore a host whose IP address cannot be
3377 found. If all hosts are ignored, the behaviour is controlled by the new
3378 host_all_ignored option.
3380 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3381 homing or because of multiple MX records with /mx) generated more than
3382 one IP address, and the following item turned out to be the local host,
3383 all the secondary addresses of the first item were incorrectly removed
3384 from the list, along with the local host and any following hosts (which
3385 is what is supposed to happen).
3387 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3388 whoever called Exim into the -H file. In the case of the daemon it was
3389 behaving confusingly. When first started, it used values for whoever
3390 started the daemon, but after a SIGHUP it used the Exim user (because it
3391 calls itself on a restart). I have changed the code so that it now always
3394 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3395 message are rejected with the same error (e.g. no authentication or bad
3396 sender address), and a DATA command is nevertheless sent (as can happen
3397 with PIPELINING or a stupid MUA), the error message that was given to the
3398 RCPT commands is included in the rejection of the DATA command. This is
3399 intended to be helpful for MUAs that show only the final error to their
3402 PH/42 Another patch from the Sieve maintainer.
3404 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3405 Eximstats - Fixed some broken HTML links and added missing column headers
3407 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3408 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3410 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3412 PH/43 Yet another patch from the Sieve maintainer.
3414 PH/44 I found a way to check for a TCP/IP connection going away before sending
3415 the response to the final '.' that terminates a message, but only in the
3416 case where the client has not sent further data following the '.'
3417 (unfortunately, this is allowed). However, in many cases there won't be
3418 any further data because there won't be any more messages to send. A call
3419 to select() can be used: if it shows that the input is "ready", there is
3420 either input waiting, or the socket has been closed. An attempt to read
3421 the next input character can distinguish the two cases. Previously, Exim
3422 would have sent an OK response which the client would never have see.
3423 This could lead to message repetition. This fix should cure that, at
3424 least in a lot of common cases.
3426 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3427 advertised in response to EHLO.
3433 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3434 fixed by 4.65/MH/01 (is this a record?) are fixed:
3436 (i) An empty string was always treated as zero by the numeric comparison
3437 operators. This behaviour has been restored.
3439 (ii) It is documented that the numeric comparison operators always treat
3440 their arguments as decimal numbers. This was broken in that numbers
3441 starting with 0 were being interpreted as octal.
3443 While fixing these problems I realized that there was another issue that
3444 hadn't been noticed. Values of message_size_limit (both the global option
3445 and the transport option) were treated as octal if they started with 0.
3446 The documentation was vague. These values are now always treated as
3447 decimal, and I will make that clear in the documentation.
3453 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3454 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3457 MH/01 Don't check that the operands of numeric comparison operators are
3458 integers when their expansion is in "skipping" mode (fixes bug
3459 introduced by 4.64-PH/07).
3461 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3462 child addresses, Exim now panics and dies. Previously, because the count
3463 is held in a short int, deliveries were likely to be lost. As such a
3464 large number of recipients for a single message is ridiculous
3465 (performance will be very, very poor), I have chosen to impose a limit
3466 rather than extend the field.
3472 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3473 leftover -K file (the existence of which was triggered by #402).
3474 While we were at it, introduced process PID as part of the -K
3475 filename. This should rule out race conditions when creating
3478 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3479 processing considerably. Previous code took too long for large mails,
3480 triggering a timeout which in turn triggers #401.
3482 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3483 in the DK code in transports.c. sendfile() is not really portable,
3484 hence the _LINUX specificness.
3486 TF/01 In the add_headers option to the mail command in an Exim filter,
3487 there was a bug that Exim would claim a syntax error in any
3488 header after the first one which had an odd number of characters
3491 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3492 callout verification, Exim cached a "reject" for the entire domain. This
3493 is correct for most verifications, but it is not correct for a recipient
3494 verification with use_sender or use_postmaster set, because in that case
3495 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3496 case of MAIL FROM:<> rejection from other early rejections (e.g.
3497 rejection of HELO). When verifying a recipient using a non-null MAIL
3498 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3499 Whatever the result of the callout, the value of the domain cache is
3500 left unchanged (for any other kind of callout, getting as far as trying
3501 RCPT means that the domain itself is ok).
3503 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3506 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3507 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3508 session, the daemon ignores ECONNECT errors and logs others; it now
3509 ignores EPIPE as well.
3511 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3512 (quoted-printable decoding).
3514 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3515 later the small subsequent patch to fix an introduced bug.
3517 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3519 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3521 PH/08 An error is now given if message_size_limit is specified negative.
3523 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3524 to be given (somewhat) arbitrary names.
3526 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3529 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3530 miscellaneous code fixes
3532 PH/10 Added the log_reject_target ACL modifier to specify where to log
3535 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3536 hostname. This is wrong, because it relates to the incoming message (and
3537 probably the interface on which it is arriving) and not to the outgoing
3538 callout (which could be using a different interface). This has been
3539 changed to use the value of the helo_data option from the smtp transport
3540 instead - this is what is used when a message is actually being sent. If
3541 there is no remote transport (possible with a router that sets up host
3542 addresses), $smtp_active_hostname is used.
3544 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3545 tweaks were necessary in order to get it to work (see also 21 below):
3546 (a) The code assumed that strncpy() returns a negative number on buffer
3547 overflow, which isn't the case. Replaced with Exim's string_format()
3549 (b) There were several signed/unsigned issues. I just did the minimum
3550 hacking in of casts. There is scope for a larger refactoring.
3551 (c) The code used strcasecmp() which is not a standard C function.
3552 Replaced with Exim's strcmpic() function.
3553 (d) The code set only $1; it now sets $auth1 as well.
3554 (e) A simple test gave the error "authentication client didn't specify
3555 service in request". It would seem that Dovecot has changed its
3556 interface. Fortunately there's a specification; I followed it and
3557 changed what the client sends and it appears to be working now.
3559 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3562 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3563 address A is aliased to B and C, where B exists and C does not. Without
3564 -v the output is "A verified" because verification stops after a
3565 successful redirection if more than one address is generated. However,
3566 with -v the child addresses are also verified. Exim was outputting "A
3567 failed to verify" and then showing the successful verification for C,
3568 with its parentage. It now outputs "B failed to verify", showing B's
3569 parentage before showing the successful verification of C.
3571 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3572 look up a TXT record in a specific list after matching in a combined
3575 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3576 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3577 they consult the DNS. I had assumed they would set it the way they
3578 wanted; and indeed my experiments on Linux seem to show that in some
3579 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3580 To be on the safe side, however, I have now made the interface to
3581 host_find_byname() similar to host_find_bydns(), with an argument
3582 containing the DNS resolver options. The host_find_byname() function now
3583 sets these options at its start, just as host_find_bydns() does. The smtp
3584 transport options dns_qualify_single and dns_search_parents are passed to
3585 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3586 of host_find_byname() use the default settings of RES_DEFNAMES
3587 (qualify_single) but not RES_DNSRCH (search_parents).
3589 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3590 spool_read_header() do less string testing, by means of a preliminary
3591 switch on the second character of optional "-foo" lines. (This is
3592 overdue, caused by the large number of possibilities that now exist.
3593 Originally there were few.) While I was there, I also converted the
3594 str(n)cmp tests so they don't re-test the leading "-" and the first
3595 character, in the hope this might squeeze out yet more improvement.
3597 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3598 flag allowing group syntax was set by the header_syntax check but not
3599 turned off, possible causing trouble later; (2) The flag was not being
3600 set at all for the header_verify test, causing "group"-style headers to
3601 be rejected. I have now set it in this case, and also caused header_
3602 verify to ignore an empty address taken from a group. While doing this, I
3603 came across some other cases where the code for allowing group syntax
3604 while scanning a header line wasn't quite right (mostly, not resetting
3605 the flag correctly in the right place). These bugs could have caused
3606 trouble for malformed header lines. I hope it is now all correct.
3608 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3609 with the "reply" argument non-NULL. The code, however (which originally
3610 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3611 but it didn't always do it. This confused somebody who was copying the
3612 code for some other use. I have removed all the tests.
3614 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3615 feature that was used to support insecure browsers during the U.S. crypto
3616 embargo. It requires special client support, and Exim is probably the
3617 only MTA that supported it -- and would never use it because real RSA is
3618 always available. This code has been removed, because it had the bad
3619 effect of slowing Exim down by computing (never used) parameters for the
3620 RSA_EXPORT functionality.
3622 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3623 authenticator to fail if there's a tab character in the incoming data
3624 (there should never be unless someone is messing about, as it's supposed
3625 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3626 if the connection is using TLS or if the remote IP is the same as the
3627 local IP, and the "valid-client-cert option" if a client certificate has
3630 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3631 authenticators. This can be used for authorization after authentication
3632 succeeds. (In the case of plaintext, it servers for both authentication
3635 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3636 if any retry times were supplied.
3638 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3639 connection, where there is no client IP address to check. In this
3640 situation, the verify now always succeeds.
3642 PH/25 Applied John Jetmore's -Mset patch.
3644 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3646 PH/27 In a string expansion for a processed (not raw) header when multiple
3647 headers of the same name were present, leading whitespace was being
3648 removed from all of them, but trailing whitespace was being removed only
3649 from the last one. Now trailing whitespace is removed from each header
3650 before concatenation. Completely empty headers in a concatenation (as
3651 before) are ignored.
3653 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3654 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3656 PH/29 [Removed. This was a change that I later backed out, and forgot to
3657 correct the ChangeLog entry (that I had efficiently created) before
3658 committing the later change.]
3660 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3661 address errors (4xx response to RCPT) over the same connection as other
3662 messages routed to the same hosts. Such deliveries are always "forced",
3663 so retry times are not inspected. This resulted in far too many retries
3664 for the affected addresses. The effect occurred only when there were more
3665 hosts than the hosts_max_try setting in the smtp transport when it had
3666 the 4xx errors. Those hosts that it had tried were not added to the list
3667 of hosts for which the message was waiting, so if all were tried, there
3668 was no problem. Two fixes have been applied:
3670 (i) If there are any address or message errors in an SMTP delivery, none
3671 of the hosts (tried or untried) are now added to the list of hosts
3672 for which the message is waiting, so the message should not be a
3673 candidate for sending over the same connection that was used for a
3674 successful delivery of some other message. This seems entirely
3675 reasonable: after all the message is NOT "waiting for some host".
3676 This is so "obvious" that I'm not sure why it wasn't done
3677 previously. Hope I haven't missed anything, but it can't do any
3678 harm, as the worst effect is to miss an optimization.
3680 (ii) If, despite (i), such a delivery is accidentally attempted, the
3681 routing retry time is respected, so at least it doesn't keep
3682 hammering the server.
3684 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3685 in ${readsocket because some servers need this prod.
3687 PH/32 Added some extra debug output when updating a wait-xxx database.
3689 PH/33 The hint "could be header name not terminated by colon", which has been
3690 given for certain expansion errors for a long time, was not being given
3691 for the ${if def:h_colon_omitted{... case.
3693 PH/34 The spec says: "With one important exception, whenever a domain list is
3694 being scanned, $domain contains the subject domain." There was at least
3695 one case where this was not true.
3697 PH/35 The error "getsockname() failed: connection reset by peer" was being
3698 written to the panic log as well as the main log, but it isn't really
3699 panic-worthy as it just means the connection died rather early on. I have
3700 removed the panic log writing for the ECONNRESET error when getsockname()
3703 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3704 runs only) independently of the message's sender address. This meant
3705 that, if the 4xx error was in fact related to the sender, a different
3706 message to the same recipient with a different sender could confuse
3707 things. In particular, this can happen when sending to a greylisting
3708 server, but other circumstances could also provoke similar problems.
3709 I have changed the default so that the retry time for these errors is now
3710 based a combination of the sender and recipient addresses. This change
3711 can be overridden by setting address_retry_include_sender=false in the
3714 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3715 remote server are returned as part of bounce messages. This was not
3716 happening for LMTP over a pipe (the lmtp transport), but now it is the
3717 same for both kinds of LMTP.
3719 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3720 in header lines that were in fact CNAMEs. This is no longer the case.
3722 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3723 and queue runs started by the daemon processed all messages. This has
3724 been fixed so that -R and -S can now usefully be given with -q<time>.
3726 PH/40 Import PCRE release 6.7 (fixes some bugs).
3728 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3730 PH/42 Give an error if -q is specified more than once.
3732 PH/43 Renamed the variables $interface_address and $interface_port as
3733 $received_ip_address and $received_port, to make it clear that these
3734 values apply to message reception, and not to the outgoing interface when
3735 a message is delivered. (The old names remain recognized, of course.)
3737 PH/44 There was no timeout on the connect() call when using a Unix domain
3738 socket in the ${readsocket expansion. There now is.
3740 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3741 be meaningful with "accept".
3743 SC/01 Eximstats V1.43
3744 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3746 SC/02 Eximstats V1.44
3747 Use a glob alias rather than an array ref in the generated
3748 parser. This improves both readability and performance.
3750 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3751 Collect SpamAssassin and rejection statistics.
3752 Don't display local sender or destination tables unless
3753 there is data to show.
3754 Added average volumes into the top table text output.
3756 SC/04 Eximstats V1.46
3757 Collect data on the number of addresses (recipients)
3758 as well as the number of messages.
3760 SC/05 Eximstats V1.47
3761 Added 'Message too big' to the list of mail rejection
3762 reasons (thanks to Marco Gaiarin).
3764 SC/06 Eximstats V1.48
3765 Mainlog lines which have GMT offsets and are too short to
3766 have a flag are now skipped.
3768 SC/07 Eximstats V1.49 (Alain Williams)
3769 Added the -emptyok flag.
3771 SC/08 Eximstats V1.50
3772 Fixes for obtaining the IP address from reject messages.
3774 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3775 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3776 whitespace changes from 4.64-PH/27
3778 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3781 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3782 are found, allow negative numbers in numeric criteria)
3784 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3786 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3787 to match changes made in 4.64-PH/43
3789 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3791 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3792 "domain" is actually an IP address, and force a failure. This locks out
3793 those revolvers/nameservers that support "A-for-A" lookups, in
3794 contravention of the specifications.
3796 PH/48 When a host name was looked up from an IP address, and the subsequent
3797 forward lookup of the name timed out, the host name was left in
3798 $sender_host_name, contrary to the specification.
3800 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3801 restricted to single-key lookups, Exim was not diagnosing an error if
3802 * or *@ was used with a query-style lookup.
3804 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3806 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3807 long ago, but noone interested enough thought of it. Let's just say that
3808 the "1.1" means that there are some new functions that weren't there at
3809 some point in the past.
3811 PH/51 Error processing for expansion failure of helo_data from an smtp
3812 transport during callout processing was broken.
3814 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3815 tested/used via the -bh/-bhc/-bs options.
3817 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3818 bug, fixed in subsequent PCRE releases).
3820 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3821 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3823 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3829 SC/01 Use a glob alias rather than an array ref in eximstats generated
3830 parser. This improves both readability and performance.
3832 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3833 Don't display local sender or destination tables in eximstats unless
3834 there is data to show.
3835 Added average volumes into the eximstats top table text output.
3837 SC/03 Collect data on the number of addresses (recipients) as well
3838 as the number of messages in eximstats.
3840 TF/01 Correct an error in the documentation for the redirect router. Exim
3841 does (usually) call initgroups() when daemonizing.
3843 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3844 with consistent privilege compared to when running as a daemon.
3846 TF/03 Note in the spec that $authenticated_id is not set for local
3847 submissions from trusted users.
3849 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3850 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3852 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3853 by adding some example configuration directives to the default
3854 configuration file. A little bit of work is required to uncomment the
3855 directives and define how usernames and passwords are checked, but
3856 there is now a framework to start from.
3858 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3859 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3860 without this. I don't know how relevant this is to other LDAP libraries.
3862 PH/02 Add the verb name to the "unknown ACL verb" error.
3864 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3866 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3868 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3869 directory not expanded when it should be if an expanded home directory
3870 was set for the address (which is overridden by the transport).
3872 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3875 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3876 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3877 because it is too late at that time, and has no effect.
3879 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3880 security issue with \' (bugzilla #107). I could not use the
3881 PQescapeStringConn() function, because it needs a PGconn value as one of
3884 PH/08 When testing addresses using -bt, indicate those final addresses that
3885 are duplicates that would not cause an additional delivery. At least one
3886 person was confused, thinking that -bt output corresponded to deliveries.
3887 (Suppressing duplicates isn't a good idea as you lose the information
3888 about possibly different redirections that led to the duplicates.)
3890 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3891 systems where poll() doesn't work, in particular OS X.
3893 PH/10 Added more information to debugging output for retry time not reached.
3895 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3896 operations in malware.c.
3898 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3901 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3902 syslog_duplication set false, log lines that would normally be written
3903 both the the main log and to the reject log were not written to syslog at
3906 PH/14 In the default configuration, change the use of "message" in ACL warn
3907 statements to "add_header".
3909 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3910 not followed by a command (e.g. "seen endif").
3912 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3913 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3916 PH/17 Added extra conditions to the default value of delay_warning_condition
3920 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3921 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3922 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3925 The Auto-Submitted: and various List- headers are standardised, whereas I
3926 don't think Precedence: ever was.
3928 PH/18 Refactored debugging code in route_finduser() to show more information,
3929 in particular, the error code if getpwnam() issues one.
3931 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3932 This is apparently needed in addition to the PH/07 change above to avoid
3933 any possible encoding problems.
3935 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3936 but not after initializing Perl.
3938 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3939 output them only if debugging. By default they are written stderr,
3940 apparently, which is not desirable.
3942 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3945 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3948 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3950 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3951 authenticated or an ident call has been made. Suppress the default
3952 values for $authenticated_id and $authenticated_sender (but permit -oMai
3953 and -oMas) when testing with -bh.
3955 PH/24 Re-jigged the order of the tests in the default configuration so that the
3956 tests for valid domains and recipients precede the DNS black list and CSA
3957 tests, on the grounds that those ones are more expensive.
3959 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3960 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3961 This bug exists in every version of Exim that I still have, right back to
3964 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3965 However, an attempt to turn on case-sensitivity in a regex key by
3966 including (?-i) didn't work because the subject string was already
3967 lowercased, and the effects were non-intuitive. It turns out that a
3968 one-line patch can be used to allow (?-i) to work as expected.
3974 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3975 other effects) broke the use of negated acl sub-conditions.
3977 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3980 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3981 "Deny" causes Exim to reject the incoming connection with a 554 error.
3982 Unfortunately, if there is a major crisis, such as a disk failure,
3983 tcp-wrappers gives "deny", whereas what one would like would be some
3984 kind of temporary error. A kludge has been added to help with this.
3985 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3986 554 error is used if errno is still zero or contains ENOENT (which occurs
3987 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3990 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3992 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3993 errors. Otherwise a message that provokes a temporary error (when other
3994 messages do not) can cause a whole host to time out.
3996 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3997 addresses were routed directly to files or pipes from a redirect router.
3998 File deliveries just didn't batch; pipe deliveries might have suffered
4001 PH/06 A failure to get a lock for a hints database would erroneously always say
4002 "Failed to get write lock", even when it was really a read lock.
4004 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4005 of 0600. This has been changed to use the value of the lockfile_mode
4006 option (which defaults to 0600).
4008 PH/08 Applied small patch from the Sieve maintainer.
4010 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4011 folder from quota calculations, a direct delivery into this folder messed
4012 up the contents of the maildirsize file. This was because the regex was
4013 used only to exclude .Trash (or whatever) when the size of the mailbox
4014 was calculated. There was no check that a delivery was happening into an
4015 excluded directory. This bug has been fixed by ignoring all quota
4016 processing for deliveries into excluded directories.
4018 PH/10 Added the maildirfolder_create_regex option to appendfile.
4024 PH/01 The code for finding all the local interface addresses on a FreeBSD
4025 system running IPv6 was broken. This may well have applied to all BSD
4026 systems, as well as to others that have similar system calls. The broken
4027 code found IPv4 interfaces correctly, but gave incorrect values for the
4028 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4029 that it would not match correctly against @[] and not recognize the IPv6
4032 PH/02 The ipliteral router was not recognizing addresses of the form user@
4033 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4035 PH/03 Added disable_ipv6.
4037 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4038 decoded form, because it is most often used to construct To: headers
4039 lines in autoreplies, and the decoded form may well be syntactically
4040 invalid. However, $reply_address has leading white space removed, and all
4041 newlines turned into spaces so that the autoreply transport does not
4044 PH/05 If group was specified without a user on a router, and no group or user
4045 was specified on a transport, the group from the router was ignored.
4047 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4048 for visible compile-time settings that can be used to change these
4049 numbers, for those that want even more. Backwards compatibility with old
4050 spool files has been maintained. However, going back to a previous Exim
4051 release will lost any variables that are in spool files.
4053 PH/07 Two small changes when running in the test harness: increase delay when
4054 passing a TCP/IP connection to a new process, in case the original
4055 process has to generate a bounce, and remove special handling of
4056 127.0.0.2 (sic), which is no longer necessary.
4058 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4059 be the same on different OS.
4061 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4064 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4065 whether --show-vars was specified or not
4067 JJ/02 exipick: Added support for new ACL variable spool format introduced
4070 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4071 syntactically invalid From: or Reply-to: line, and a filter used this to
4072 generate an autoreply, and therefore failed to obtain an address for the
4073 autoreply, Exim could try to deliver to a non-existent relative file
4074 name, causing unrelated and misleading errors. What now happens is that
4075 it logs this as a hard delivery error, but does not attempt to create a
4078 PH/11 The exinext utility has a -C option for testing purposes, but although
4079 the given file was scanned by exinext itself; it wasn't being passed on
4080 when Exim was called.
4082 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4083 an end-of-file indication when reading a command response.
4085 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4086 compiled. In many other places in Exim, IPv6 addresses are always
4087 recognized, so I have changed this. It also means that IPv4 domain
4088 literals of the form [IPV4:n.n.n.n] are now always recognized.
4090 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4091 used if the router is not running as root, for example, when verifying at
4092 ACL time, or when using -bh. The debugging output from this situation was
4093 non-existent - all you got was a failure to exec. I have made two
4096 (a) Failures to set uid/gid, the current directory, or a process leader
4097 in a subprocess such as that created by queryprogram now generate
4098 suitable debugging output when -d is set.
4100 (b) The queryprogram router detects when it is not running as root,
4101 outputs suitable debugging information if -d is set, and then runs
4102 the subprocess without attempting to change uid/gid.
4104 PH/15 Minor change to Makefile for building test_host (undocumented testing
4107 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4108 additional section of a DNS packet that returns MX or SRV records.
4109 Instead, it always explicitly searches for A/AAAA records. This avoids
4110 major problems that occur when a DNS server includes only records of one
4111 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4112 fixed another bug: if SRV records were looked up and the corresponding
4113 address records were *not* found in the additional section, the port
4114 values from the SRV records were lost.
4116 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4117 using the correct key (the original address) when searching the retry
4118 rules in order to find which one to use for generating the retry hint.
4120 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4121 adding the default one. Similarly, if it contains a Reply-To: header, the
4122 errors_reply_to option, if set, is not used.
4124 PH/19 When calculating a retry time, Exim used to measure the "time since
4125 failure" by looking at the "first failed" field in the retry record. Now
4126 it does not use this if it is later than than the arrival time of the
4127 message. Instead it uses the arrival time. This makes for better
4128 behaviour in cases where some deliveries succeed, thus re-setting the
4129 "first failed" field. An example is a quota failure for a huge message
4130 when small messages continue to be delivered. Without this change, the
4131 "time since failure" will always be short, possible causing more frequent
4132 delivery attempts for the huge message than are intended.
4133 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4135 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4136 $1, $2, $3) because the numerical variables can be reset during some
4137 expansion items (e.g. "match"), thereby losing the authentication data.
4139 PH/21 Make -bV show the size of off_t variables so that the test suite can
4140 decide whether to run tests for quotas > 2G.
4142 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4143 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4144 filecount value is greater than 2G or if a quota value is greater than 2G
4145 on a system where the size of off_t is not greater than 4, a panic error
4148 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4149 never match. The debug and -bh output now contains an explicit error
4150 message indicating a malformed IPv4 address or mask.
4152 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4153 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4154 PH/23 above applies.
4156 PH/25 Do not write to syslog when running in the test harness. The only
4157 occasion when this arises is a failure to open the main or panic logs
4158 (for which there is an explicit test).
4160 PH/26 Added the /no_tell option to "control=freeze".
4162 PH/27 If a host name lookup failed very early in a connection, for example, if
4163 the IP address matched host_lookup and the reverse lookup yielded a name
4164 that did not have a forward lookup, an error message of the form "no IP
4165 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4166 could be logged. Now it outputs the IP address instead of "NULL".
4168 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4169 allows the sender and the authenticated sender to be set when
4170 submitting a message from within Exim. Since child_open_exim() is
4171 documented for local_scan(), the new function should be too.
4173 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4174 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4175 results in an empty string is now treated as unset.
4177 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4179 PH/31 Added sender_verify_fail logging option.
4181 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4182 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4183 tidied the source and removed it altogether.
4185 PH/33 When a queue run was abandoned because the load average was too high, a
4186 log line was always written; now it is written only if the queue_run log
4187 selector is set. In addition, the log line for abandonment now contains
4188 information about the queue run such as the pid. This is always present
4189 in "start" and "stop" lines but was omitted from the "abandon" line.
4191 PH/34 Omit spaces between a header name and the colon in the error message that
4192 is given when verify = headers_syntax fails (if there are lots of them,
4193 the message gets confusing).
4195 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4196 names, as there are now some PTR records that contain slashes. This check
4197 is only to protect against broken name servers that fall over on strange
4198 characters, so the fact that it applies to all lookups doesn't matter.
4200 PH/36 Now that the new test suite is complete, we can remove some of the
4201 special code in Exim that was needed for the old test suite. For example,
4202 sorting DNS records because real resolvers return them in an arbitrary
4203 order. The new test suite's fake resolver always returns records in the
4206 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4207 bounces) except when queue_only is set, to avoid logging races between
4208 the different processes.
4210 PH/38 Panic-die if .include specifies a non-absolute path.
4212 PH/39 A tweak to the "H" retry rule from its user.
4214 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4215 a label. They prevented compilation on older perls.
4217 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4218 a warning to be raised on newish perls.
4220 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4221 on queue. Changes to match documented behaviour of showing count of
4222 messages matching specified criteria.
4224 PH/40 Changed the default ident timeout from 30s to 5s.
4226 PH/41 Added support for the use of login_cap features, on those BSD systems
4227 that have them, for controlling the resources used by pipe deliveries.
4229 PH/42 The content-scanning code uses fopen() to create files in which to put
4230 message data. Previously it was not paying any attention to the mode of
4231 the files. Exim runs with umask(0) because the rest of the code creates
4232 files with open(), and sets the required mode explicitly. Thus, these
4233 files were ending up world-writeable. This was not a big issue, because,
4234 being within the spool directory, they were not world-accessible. I have
4235 created a function called modefopen, which takes an additional mode
4236 argument. It sets umask(777), creates the file, chmods it to the required
4237 mode, then resets the umask. All the relevant calls to fopen() in the
4238 content scanning code have been changed to use this function.
4240 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4241 to 24 hours. This avoids potential overflow problems when processing G
4242 and H retry rules. I suspect nobody ever tinkers with this value.
4244 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4246 PH/45 When the plaintext authenticator is running as a client, the server's
4247 challenges are checked to ensure they are valid base64 strings. By
4248 default, the authentication attempt is cancelled if an invalid string is
4249 received. Setting client_ignore_invalid_base64 true ignores these errors.
4250 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4251 they are received. Thus, the responses can be made to depend on the
4252 challenges. If an invalid string is ignored, an empty string is placed in
4255 PH/46 Messages that are created by the autoreply transport now contains a
4256 References: header, in accordance with RFCs 2822 and 3834.
4258 PH/47 Added authenticated_sender_force to the smtp transport.
4260 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4262 PH/49 Installed latest patch from the Sieve maintainer.
4264 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4265 was also set, the appendfile transport was unnecessarily scanning a
4266 directory of message files (e.g. for maildir delivery) to find the count
4267 of files (along with the size), even though it did not need this
4268 information. It now does the scan only if it needs to find either the
4269 size of the count of files.
4271 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4273 PH/52 Two bugs concerned with error handling when the smtp transport is
4276 (i) Exim was not creating retry information for temporary errors given
4277 for individual recipients after the DATA command when the smtp transport
4278 was used in LMTP mode. This meant that they could be retried too
4279 frequently, and not timed out correctly.
4281 (ii) Exim was setting the flag that allows error details to be returned
4282 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4283 recipients that were returned after the DATA command.
4285 PH/53 This is related to PH/52, but is more general: for any failing address,
4286 when detailed error information was permitted to be returned to the
4287 sender, but the error was temporary, then after the final timeout, only
4288 "retry timeout exceeded" was returned. Now it returns the full error as
4289 well as "retry timeout exceeded".
4291 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4292 do this, and (what is worse) MTAs that accept it.
4294 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4295 will now be deprecated.
4297 PH/56 New os.c-cygwin from the Cygwin maintainer.
4299 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4300 formats (previously only available in exim formats via -bpr, -bpru,
4301 and -bpra. Now also available in native and exiqgrep formats)
4303 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4304 with very large, slow to parse queues
4306 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4308 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4310 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4311 responses to authentication challenges, though it was showing the
4312 challenges; (ii) I've removed the CR characters from the debug output for
4315 PH/58 Allow for the insertion of a newline as well as a space when a string
4316 is turned into more than one encoded-word during RFC 2047 encoding. The
4317 Sieve code now uses this.
4319 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4320 data_4xx, lost_connection, tls_required.
4322 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4323 message was being sent as an SMTP response.
4325 PH/61 Add -l and -k options to exicyclog.
4327 PH/62 When verifying, if an address was redirected to one new address, so that
4328 verification continued, and the new address failed or deferred after
4329 having set something in $address_data, the value of $address_data was not
4330 passed back to the ACL. This was different to the case when no
4331 redirection occurred. The value is now passed back in both cases.
4333 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4334 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4335 use login_cap.h, so on its own it isn't the distinguishing feature. The
4336 new name refers directly to the setclassresources() function.
4338 PH/65 Added configuration files for NetBSD3.
4340 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4342 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4343 is preferred over IPv4.
4345 PH/68 The bounce_return_message and bounce_return_body options were not being
4346 honoured for bounces generated during the reception of non-SMTP messages.
4347 In particular, this applied to messages rejected by the ACL. This bug has
4348 been fixed. However, if bounce_return_message is true and bounce_return_
4349 body is false, the headers that are returned for a non-SMTP message
4350 include only those that have been read before the error was detected.
4351 (In the case of an ACL rejection, they have all been read.)
4353 PH/69 The HTML version of the specification is now built in a directory called
4354 spec_html instead of spec.html, because the latter looks like a path with
4355 a MIME-type, and this confuses some software.
4357 PH/70 Catch two compiler warnings in sieve.c.
4359 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4360 function verify_get_ident() calls ip_connect() to connect a socket, but
4361 if the "connect()" function timed out, ip_connect() used to close the
4362 socket. However, verify_get_ident() also closes the socket later, and in
4363 between Exim writes to the log, which may get opened at this point. When
4364 the socket was closed in ip_connect(), the log could get the same file
4365 descriptor number as the socket. This naturally causes chaos. The fix is
4366 not to close the socket in ip_connect(); the socket should be closed by
4367 the function that creates it. There was only one place in the code where
4368 this was missing, in the iplookup router, which I don't think anybody now
4369 uses, but I've fixed it anyway.
4371 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4372 well as to direct DNS lookups. Otherwise the handling of names in host
4373 lists is inconsistent and therefore confusing.
4379 PH/01 Two changes to the default runtime configuration:
4381 (1) Move the checks for relay_from_hosts and authenticated clients from
4382 after to before the (commented out) DNS black list checks.
4384 (2) Add control=submission to the relay_from_hosts and authenticated
4385 clients checks, on the grounds that messages accepted by these
4386 statements are most likely to be submissions.
4388 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4390 (1) Generate an error if the third argument for the ${prvs expansion is
4393 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4396 (3) Reset the variables that are obtained from the first argument of
4397 ${prvscheck and used in the second argument before leaving the code,
4398 because their memory is reclaimed, so using them afterwards may do
4401 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4402 one (it's much easier than Tom thought :-).
4404 (5) Because of (4), we can now allow for the use of $prvscheck_result
4405 inside the third argument.
4407 PH/03 For some reason, the default setting of PATH when running a command from
4408 a pipe transport was just "/usr/bin". I have changed it to
4411 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4412 anything to be listed in the output from -bV.
4414 PH/05 When a filter generated an autoreply, the entire To: header line was
4415 quoted in the delivery log line, like this:
4417 => >A.N.Other <ano@some.domain> <original@ddress> ...
4419 This has been changed so that it extracts the operative address. There
4420 may be more than one such address. If so, they are comma-separated, like
4423 => >ano@some.domain,ona@other.domain <original@ddress> ...
4425 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4426 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4427 not being looked up in the rDNS to get a host name, Exim was showing the
4428 IP address twice in Received: lines, even though the IP addresses were
4429 identical. For example:
4431 Received: from [1.2.3.4] (helo=[1.2.3.4])
4433 However, if the real host name was known, it was omitting the HELO data
4434 if it matched the actual IP address. This has been tidied up so that it
4435 doesn't show the same IP address twice.
4437 PH/07 When both +timestamp and +memory debugging was on, the value given by
4438 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4439 called by the debug printing, thereby overwriting the timestamp buffer.
4440 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4442 PH/08 When the original message was included in an autoreply transport, it
4443 always said "this is a copy of the message, including all the headers",
4444 even if body_only or headers_only was set. It now gives an appropriate
4447 PH/09 Applied a patch from the Sieve maintainer which:
4449 o fixes some comments
4450 o adds the (disabled) notify extension core
4451 o adds some debug output for the result of if/elsif tests
4452 o points to the current vacation draft in the documentation
4453 and documents the missing references header update
4457 o fixes a bug in processing the envelope test (when testing
4458 multiple envelope elements, the last element determined the
4461 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4462 Electronic Mail") by including:
4464 Auto-submitted: auto-generated
4466 in the messages that it generates (bounce messages and others, such as
4467 warnings). In the case of bounce messages for non-SMTP messages, there was
4468 also a typo: it was using "Auto_submitted" (underscore instead of
4469 hyphen). Since every message generated by Exim is necessarily in response
4470 to another message, thes have all been changed to:
4472 Auto-Submitted: auto-replied
4474 in accordance with these statements in the RFC:
4476 The auto-replied keyword:
4478 - SHOULD be used on messages sent in direct response to another
4479 message by an automatic process,
4481 - MUST NOT be used on manually-generated messages,
4483 - MAY be used on Delivery Status Notifications (DSNs) and Message
4484 Disposition Notifications (MDNs),
4486 - MUST NOT be used on messages generated by automatic or periodic
4487 processes, except for messages which are automatic responses to
4490 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4491 to the default Received: header definition.
4493 PH/12 Added log selector acl_warn_skipped (default on).
4495 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4496 variables because (a) they are in the wrong storage pool and (b) even if
4497 they were copied, it wouldn't work properly because of the caching.
4499 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4500 checking when decoding. Apparently there are clients that generate
4501 overlong encoded strings. Why am I not surprised?
4503 PH/15 If the first argument of "${if match_address" was not empty, but did not
4504 contain an "@" character, Exim crashed. Now it writes a panic log message
4505 and treats the condition as false.
4507 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4509 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4510 extension "envelope-auth". The code is finished and in agreement with
4511 other implementations, but there is no documentation so far and in fact,
4512 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4513 not changing the active code.
4515 Print executed "if" and "elsif" statements when debugging is used. This
4516 helps a great deal to understand what a filter does.
4518 Document more things not specified clearly in RFC3028. I had all this
4519 sorted out, when out of a sudden new issues came to my mind. Oops."
4521 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4524 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4525 canonical form (as documented). However, after a host name lookup from
4526 the IP address, check_host() was doing a simple string comparison with
4527 addresses acquired from the DNS when checking that the found name did
4528 have the original IP as one of its addresses. Since any found IPv6
4529 addresses are likely to be in abbreviated form, the comparison could
4530 fail. Luckily, there already exists a function for doing the comparison
4531 by converting both addresses to binary, so now that is used instead of
4532 the text comparison.
4534 PH/20 There was another similar case to PH/19, when a complete host name was
4535 given in a host list; looking up its IP address could give an abbreviated
4536 form, whereas the current host's name might or might not be abbreviated.
4537 The same fix has been applied.
4543 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4544 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4547 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4548 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4550 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4552 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4553 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4554 Also, for "Auto-Submitted", treat anything other than "no" as
4555 non-personal, in accordance with RFC 3834. (Previously it treated
4556 anything starting "auto-" as non-personal.)
4558 TF/01 The control=submission/name=... option had a problem with syntax
4559 errors if the name included a slash character. The /name= option
4560 now slurps the rest of the string, so it can include any characters
4561 but it must come last in the list of options (after /sender_retain
4564 PH/05 Some modifications to the interface to the fake nameserver for the new
4572 TK/01 Added the "success_on_redirect" address verification option. See
4573 NewStuff for rationale and an example.
4575 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4577 PH/02 Patch to exigrep to allow it to work on syslog lines.
4579 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4580 fread() to skip over the body file's header line, because in Cygwin the
4581 header line is locked and is inaccessible.
4583 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4584 co-exist for some time) to make it clear that it is the Exim ID that is
4585 referenced, not the Message-ID: header line.
4587 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4588 string_format() function, because snprintf() does not exist on all
4591 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4592 ${stat: expansion item.
4594 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4595 protocol synchronization error", to keep the pedants happy.
4597 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4598 well as for IRIX systems, when gcc is being used. See the host.c source
4601 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4603 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4606 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4607 transport and to the smtp transport in LMTP mode.
4609 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4611 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4612 run a filter in a subprocess. This could lead to confusion in subsequent
4613 lookups in the parent process. There should also be a search_tidyup() at
4614 the end of the subprocess.
4616 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4617 only if the host matched helo_try_verify_hosts, which caused the
4618 verification to occur when the EHLO/HELO command was issued. The ACL just
4619 tested the remembered result. Now, if a previous verification attempt has
4620 not happened, "verify = helo" does it there and then.
4622 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4624 TK/03 Fix log output including CR from clamd.
4626 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4627 exist provoked a memory error which could cause a segfault.
4629 PH/15 Installed PCRE 6.2
4631 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4633 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4634 of the problem. Specifically, suggested +O2 rather than +O1 for the
4637 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4639 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4640 record, the port was not correctly logged when the outgoing_port log
4641 selector was set (it logged the transort's default port).
4643 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4644 fallback_hosts, and "hosts" in the smtp transport.
4646 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4647 host errors such as "Connection refused".
4649 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4650 authentication with radiusclient 0.4.9:
4652 - Error returned from rc_read_config was caught wrongly
4653 - Username/password not passed on to radius server due to wrong length.
4655 The presumption is that some radiusclient API changes for 4.51/PH/17
4656 were not taken care of correctly. The code is still untested by me (my
4657 Linux distribution still has 0.3.2 of radiusclient), but it was
4658 contributed by a Radius user.
4660 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4661 expanding the "port" option of the smtp transport.
4663 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4664 while reading a MIME header. Thanks to Tom Hughes for a patch.
4666 PH/24 Include config.h inside local_scan.h so that configuration settings are
4669 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4670 that in an ACL for RCPT (for example), you can examine exactly what was
4673 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4674 commands, but it was not correctly comparing the address with the actual
4675 client host address. Thus, it would show the EHLO address in Received:
4676 header lines when this was not necessary.
4678 PH/27 Added the % operator to ${eval:}.
4680 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4681 it should be ignoring failures (because with -C, for example, it has lost
4682 privilege). It wasn't ignoring creation failures other than "already
4685 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4688 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4689 because an input line was too long, either on its own, or by virtue of
4690 too many continuations, the temporary file was not being removed, and the
4691 return code was incorrect.
4693 PH/31 Missing "BOOL" in function definition in filtertest.c.
4695 PH/32 Applied Sieve patches from the maintainer.
4697 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4699 PH/33 Added "verify = not_blind".
4701 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4702 Local/Makefile (with some defaults set). These are used in built scripts
4703 such as exicyclog, but they have never been used in the exim_install
4704 script (though there are many overriding facilities there). I have
4705 arranged that the exim_install script now takes note of these two
4708 PH/35 Installed configuration files for Dragonfly.
4710 PH/36 When a locally submitted message by a trusted user did not contain a
4711 From: header, and the sender address was obtained from -f or from an SMTP
4712 MAIL command, and the trusted user did not use -F to supply a sender
4713 name, $originator_name was incorrectly used when constructing a From:
4714 header. Furthermore, $originator_name was used for submission mode
4715 messages from external hosts without From: headers in a similar way,
4716 which is clearly wrong.
4718 PH/37 Added control=suppress_local_fixups.
4720 PH/38 When log_selector = +received_sender was set, and the addition of the
4721 sender made the log line's construction buffer exactly full, or one byte
4722 less than full, an overflow happened when the terminating "\n" was
4725 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4726 when the result of a list match is failure because a DNS lookup failed.
4728 PH/40 RM_COMMAND is now used in the building process.
4730 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4731 the "build-* directories that it finds.
4733 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4734 address in a domain literal was a prefix of an interface address.
4736 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4737 when verifying a sender address, unless rewrite_headers is false.
4739 PH/44 Wrote a long comment about why errors_to addresses are verified as
4740 recipients, not senders.
4742 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4743 the ratelimit ACL was added.
4745 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4747 PH/46 Added extra information about PostgreSQL errors to the error string.
4749 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4750 avoiding the need to install special zones in a real server. This is
4751 backwards compatible; if it can't find the fake resolver, it drops back.
4752 Thus, both old and new test suites can be run.
4754 TF/02 Added util/ratelimit.pl
4756 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4757 clock is set back in time.
4759 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4760 Candler <B.Candler@pobox.com>.
4762 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4763 OK for addresses that are the result of redirections.
4765 PH/48 A number of further additions for the benefit of the new test suite,
4766 including a fake gethostbyname() that interfaces to the fake DNS resolver
4769 TF/06 The fix for widen_domains has also been applied to qualify_single and
4770 search_parents which are the other dnslookup options that can cause
4773 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4776 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4778 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4779 service" if it encountered an MX record with an empty target hostname.
4780 The message is now "an MX or SRV record indicated no SMTP service".
4782 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4783 if the DNS of the sending site is misconfigured. This is quite a
4784 common situation. This change restores the behaviour of treating a
4785 helo verification defer as a failure.
4787 PH/51 If self=fail was set on a router, the bounce message did not include the
4788 actual error message.
4794 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4796 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4797 command itself ended in error, the underlying message about the transport
4798 filter timeout was being overwritten with the pipe command error. Now the
4799 underlying error message should be appended to the second error message.
4801 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4803 PH/02 Reduce the amount of output that "make" produces by default. Full output
4804 can still be requested.
4806 PH/03 The warning log line about a condition test deferring for a "warn" verb
4807 was being output only once per connection, rather than after each
4808 occurrence (because it was using the same function as for successful
4809 "warn" verbs). This seems wrong, so I have changed it.
4811 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4812 it should not have, which might have caused a crash in the right
4813 circumstances, but probably never did.
4815 PH/04 Installed a modified version of Tony Finch's patch to make submission
4816 mode fix the return path as well as the Sender: header line, and to
4817 add a /name= option so that you can make the user's friendly name appear
4820 TF/03 Added the control = fakedefer ACL modifier.
4822 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4823 Mark Lowes for thorough testing.
4825 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4827 TK/03 Merged latest SRS patch from Miles Wilton.
4829 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4830 with the definition in sysexits.h (which is #included earlier).
4831 Fortunately, Exim does not actually use EX_OK. The code used to try to
4832 preserve the sysexits.h value, by assuming that macro definitions were
4833 scanned for macro replacements. I have been disabused of this notion,
4834 so now the code just undefines EX_OK before #including unistd.h.
4836 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4837 in the smtp transport. When a block could not be written in a single
4838 write() function, the timeout was being re-applied to each part-write.
4839 This seems wrong - if the receiver was accepting one byte at a time it
4840 would take for ever. The timeout is now adjusted when this happens. It
4841 doesn't have to be particularly precise.
4843 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4844 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4846 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4847 without a domain if the check to <postmaster@domain> fails.
4849 SC/01 Eximstats: added -xls and the ability to specify output files
4850 (patch written by Frank Heydlauf).
4852 SC/02 Eximstats: use FileHandles for outputting results.
4854 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4856 SC/04 Eximstats: fixed display of large numbers with -nvr option
4858 SC/05 Eximstats: fixed merging of reports with empty tables.
4860 SC/06 Eximstats: added the -include_original_destination flag
4862 SC/07 Eximstats: removed tabs and trailing whitespace.
4864 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4866 TK/06 MBOX spool code: Add real "From " MBOX separator line
4867 so the .eml file is really in mbox format (even though
4868 most programs do not really care). Patch from Alex Miller.
4870 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4871 The latter is generated from $received_to and is only set if the
4872 message has one envelope recipient. SA can use these headers,
4873 obviously out-of-the-box. Patch from Alex Miller.
4875 PH/08 The ${def test on a variable was returning false if the variable's
4876 value was "0", contrary to what the specification has always said!
4877 The result should be true unless the variable is empty.
4879 PH/09 The syntax error of a character other than { following "${if
4880 def:variable_name" (after optional whitespace) was not being diagnosed.
4881 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4882 accidental colon was present, for example, could give incorrect results.
4884 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4885 result is used (not including appendfile, where other changes are about
4888 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4889 This involved changing a lot of size variables from int to off_t. It
4890 should work with maildirs and everything.
4892 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4893 spamd dying while we are connected to it.
4895 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4898 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4899 function for BDB 4.3.
4901 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4903 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4904 the sources for PCRE have been re-arranged and more files are now
4907 PH/15 The code I had for printing potentially long long variables in PH/11
4908 above was not the best (it lost precision). The length of off_t variables
4909 is now inspected at build time, and an appropriate printing format (%ld
4910 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4911 to be "long long int" or "long int". This is needed for the internal
4912 formatting function string_vformat().
4914 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4915 the configuration file to be ":syslog", then the script "guesses" where
4916 the logs files are, rather than using the compiled in default. In our
4917 case the guess is not the same as the compiled default, so the script
4918 suddenly stopped working when I started to use syslog. The patch checks
4919 to see if log_file_path is "". If so, it attempts to read it from exim
4920 with no configuration file to get the compiled in version, before it
4921 falls back to the previous guessing code."
4923 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4924 implementing BATV in an Exim configuration. See NewStuff for the gory
4927 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4928 Makefile that are specific to HP-UX.
4930 PH/18 If the "use_postmaster" option was set for a recipient callout together
4931 with the "random" option, the postmaster address was used as the MAIL
4932 FROM address for the random test, but not for the subsequent recipient
4933 test. It is now used for both.
4935 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4936 patch removes a few documentation additions to RFC 3028, because the
4937 latest draft now contains them. It adds the new en;ascii-case comparator
4938 and a new error check for 8bit text in MIME parts. Comparator and
4939 require names are now matched exactly. I enabled the subaddress
4940 extension, but it is not well tested yet (read: it works for me)."
4942 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4943 rework some of the code of TK/09 above to avoid the hardwired use of
4944 "%lld" and "long long". Replaced the call to snprintf() with a call to
4947 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4948 records point to non-existent hosts", "retry timeout exceeded", and
4949 "retry time not reached for any host after a long failure period".
4951 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4952 experimental DomainKeys support:
4954 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4955 (2) On an error such as an illegally used "control", the wrong name for
4956 the control was given.
4958 These problems did NOT occur unless DomainKeys support was compiled.
4960 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4962 PH/24 Added ${if match_ip condition.
4964 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4965 databases so that it will be absolutely obvious if a crash occurs in the
4966 DB library. This is a regular occurrence (often caused by mis-matched
4969 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4970 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4971 picked up on a user's system that detects such things. There doesn't seem
4972 to be a gcc warning option for this - only an attribute that has to be
4973 put on the function's prototype. It seems that in Fedora Core 4 they have
4974 set this on a number of new functions. No doubt there will be more in due
4977 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4978 specify a transport. However, if an address that was verified by such a
4979 router was the subject of a callout, Exim crashed because it tried to
4980 read the rcpt_include_affixes from the non-existent transport. Now it
4981 just assumes that the setting of that option is false. This bug was
4982 introduced by 4.51/PH/31.
4984 PH/28 Changed -d+all to exclude +memory, because that information is very
4985 rarely of interest, but it makes the output a lot bigger. People tend to
4986 do -d+all out of habit.
4988 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4989 code in os-type was giving problems when libc.so lives in lib64, like on
4992 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4993 aren't the modern standard, and it seems that some systems' include files
4994 don't always have them. Exim was already checking for some of the newer
4995 ones like T_AAAA, and defining it itself. I've added checks for all the
4996 record types that Exim uses.
4998 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4999 not automatically generating a new one, as it is supposed to. This
5000 prevented TLS from working. If the file did exist, but contained invalid
5001 data, a new version was generated, as expected. It was only the case of a
5002 non-existent file that was broken.
5004 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5005 with a change in libdomainkeys > 0.64.
5007 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5008 from DNS. If the selector record carries the flag, it now has
5009 precedence over the domain-wide flag.
5011 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5013 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5014 the use of a port name that isn't defined in /etc/services) occurred, the
5015 message was deferred as in a normal delivery, and thus remained on the
5016 spool, instead of being failed because of the mua_wrapper setting. This
5017 is now fixed, and I tidied up some of the mua_wrapper messages at the
5020 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5021 the messages in a hash of arrays rather than using individual hashes.
5022 This is a bit cleaner and results in dramatic memory savings, albeit
5023 at a slight CPU cost.
5025 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5026 as requested by Marc Sherman.
5028 SC/10 Eximstats: added histograms for user specified patterns as requested
5031 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5033 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5034 fopen() in the content-scanning modules that did not already have it.
5040 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5041 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5043 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5045 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5047 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5048 to test Sieve filters that use "vacation".
5050 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5051 that changes the way the GnuTLS parameters are stored in the cache file.
5052 The new format can be generated externally. For backward compatibility,
5053 if the data in the cache doesn't make sense, Exim assumes it has read an
5054 old-format file, and it generates new data and writes a new file. This
5055 means that you can't go back to an older release without removing the
5058 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5059 work if there are any delivery delays because "one_time" forces the
5060 parent to be marked "delivered", so its unseen clone is never tried
5061 again. For this reason, Exim now forbids the simultaneous setting of
5064 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5065 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5066 entry if you want to know the details. The fix, however, neglected to
5067 consider the case when local delivery batching is involved. The test for
5068 "previously delivered" was not happening when checking to see if an
5069 address could be batched with a previous (undelivered) one; under
5070 certain circumstances this could lead to multiple deliveries to the same
5073 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5074 in its include files, and this causes problems building Exim.
5076 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5077 header_syntax) but Exim was just ignoring anything given after a slash.
5078 In particular, this caused confusion with an attempt to use "verify =
5079 reverse_host_lookup/defer_ok". An error is now given when options are
5080 supplied for verify items that do not have them. (Maybe reverse_host_
5081 lookup should have a defer_ok option, but that's a different point.)
5083 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5084 defined by RFC 821) to 2048, because there were problems with some AUTH
5085 commands, and RFC 1869 says the size should be increased for extended
5086 SMTP commands that take arguments.
5088 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5091 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5092 "unknown" error; now it says that the functionality isn't in the binary.
5094 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5095 an address' error message when a string expansion fails (syntax or
5096 whatever). Otherwise the password may appear in the log. Following change
5097 PH/42 below, there is no longer a chance of it appearing in a bounce
5100 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5102 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5103 of its addresses was ever tried. (Bugzilla bug #2.)
5105 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5106 the result incorrectly in the debug output. (It correctly added a newline
5107 to what was transported.)
5109 TF/01 Added $received_time.
5111 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5112 commented out examples of how to interface to a virus scanner and to
5113 SpamAssassin. Also added commented examples of av_scanner and
5114 spamd_address settings.
5116 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5117 and controls are allowed in which ACLs. There were a couple of minor
5118 errors. Some of the entries in the conditions table (which is a table of
5119 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5120 negation of where the condition IS allowed.
5122 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5124 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5125 header file does not have a version number, so I've had to invent a new
5126 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5127 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5128 radiusclient), but it was contributed by a Radius user.
5130 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5131 files or directories, for OpenSSL.
5133 PH/19 When an Exim process that is running as root has to create an Exim log
5134 file, it does so in a subprocess that runs as exim:exim so as to get the
5135 ownership right at creation (otherwise, other Exim processes might see
5136 the file with the wrong ownership). There was no test for failure of this
5137 fork() call, which would lead to the process getting stuck as it waited
5138 for a non-existent subprocess. Forks do occasionally fail when resources
5139 run out. I reviewed all the other calls to fork(); they all seem to check
5142 PH/20 When checking for unexpected SMTP input at connect time (before writing
5143 the banner), Exim was not dealing correctly with a non-positive return
5144 from the read() function. If the client had disconnected by this time,
5145 the result was a log entry for a synchronization error with an empty
5146 string after "input=" when read() returned zero. If read() returned -1
5147 (an event I could not check), uninitialized data bytes were printed.
5148 There were reports of junk text (parts of files, etc) appearing after
5151 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5153 PH/22 Added support for macro redefinition, and (re)definition in between
5154 driver and ACL definitions.
5156 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5157 forgetting to use the resulting value; it was using the unexpanded value.
5159 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5160 hadn't been configured. The fix is from Juergen Kreileder, who
5161 understands it better than I do:
5163 "Here's what I see happening with three configured cyrus_sasl
5164 authenticators configured (plain, login, cram-md5):
5166 On startup auth_cyrus_sasl_init() gets called for each of these.
5167 This means three calls to sasl_listmech() without a specified mech_list.
5168 => SASL tests which mechs of all available mechs actually work
5169 => three warnings about OTP not working
5170 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5172 With the patch, sasl_listmech() also gets called three times. But now
5173 SASL's mech_list option is set to the server_mech specified in the the
5174 authenticator. Or in other words, the answer from sasl_listmech()
5175 gets limited to just the mech you're testing for (which is different
5177 => the return list contains just 'plain' or 'login', 'cram-md5' or
5178 nothing depending on the value of ob->server_mech.
5180 I've just tested the patch: Authentication still works fine,
5181 unavailable mechs specified in the exim configuration are still
5182 caught, and the auth.log warnings about OTP are gone."
5184 PH/25 When debugging is enabled, the contents of the command line are added
5185 to the debugging output, even when log_selector=+arguments is not
5188 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5189 answer is "GNU", and only if the return is "GNU/something" is the answer
5192 PH/27 $acl_verify_message is now set immediately after the failure of a
5193 verification in an ACL, and so is available in subsequent modifiers. In
5194 particular, the message can be preserved by coding like this:
5196 warn !verify = sender
5197 set acl_m0 = $acl_verify_message
5199 Previously, $acl_verify_message was set only while expanding "message"
5200 and "log_message" when a very denied access.
5202 PH/28 Modified OS/os.c-Linux with
5204 -#ifndef OS_LOAD_AVERAGE
5205 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5207 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5208 nomenclature these days.)
5210 PH/29 Installed patch from the Sieve maintainer that adds the options
5211 sieve_useraddress and sieve_subaddress to the redirect router.
5213 PH/30 In these circumstances:
5214 . Two addresses routed to the same list of hosts;
5215 . First host does not offer TLS;
5216 . First host accepts first address;
5217 . First host gives temporary error to second address;
5218 . Second host offers TLS and a TLS session is established;
5219 . Second host accepts second address.
5220 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5221 and peerdn, if requested) that were in fact used only for the second
5224 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5225 attention to any local part prefix or suffix that was matched by the
5226 router that accepted the address. It now behaves in the same way as it
5227 does for delivery: the affixes are removed from the local part unless
5228 rcpt_include_affixes is set on the transport.
5230 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5231 timeout during the DATA phase of an incoming message.
5233 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5234 applied a patch sanctioned by the Sieve maintainer.
5236 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5237 the uid or gid is negative. A case of a negative gid caused this to be
5238 noticed. The fix allows for either to be negative.
5240 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5241 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5244 PH/36 The change PH/12 above was broken. Fixed it.
5246 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5247 the grounds that routing the same address twice would always produce the
5248 same answer. This might have been true once, but it is certainly no
5249 longer true now. Routing a child address may depend on the previous
5250 routing that produced that child. Some complicated redirection strategies
5251 went wrong when messages had multiple recipients, and made Exim's
5252 behaviour dependent on the order in which the addresses were given.
5254 I have moved the duplicate checking until after the routing is complete.
5255 Exim scans the addresses that are assigned to local and remote
5256 transports, and removes any duplicates. This means that more work will be
5257 done, as duplicates will always all be routed, but duplicates are
5258 presumably rare, so I don't expect this is of any significance.
5260 For deliveries to pipes, files, and autoreplies, the duplicate checking
5261 still happens during the routing process, since they are not going to be
5264 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5265 It corrects a timeout issue with spamd. This is Ian's comment: "The
5266 background is that sometimes spamd either never reads data from a
5267 connection it has accepted, or it never writes response data. The exiscan
5268 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5269 blindly assumes that writes won't block so it may never time out."
5271 PH/39 Allow G after quota size as well as K and M.
5273 PH/40 The value set for $authenticated_id in an authenticator may not contain
5274 binary zeroes or newlines because the value is written to log lines and
5275 to spool files. There was no check on this. Now the value is run through
5276 the string_printing() function so that such characters are converted to
5277 printable escape sequences.
5279 PH/41 $message_linecount is a new variable that contains the total number of
5280 lines in the message. Compare $body_linecount, which is the count for the
5283 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5284 bounce and delay warning messages, except in certain special cases, which
5287 (a) An SMTP error message from a remote host;
5288 (b) A message specified in a :fail: redirection;
5289 (c) A message specified in a "fail" command in a system filter;
5290 (d) A message specified in a FAIL return from the queryprogram router;
5291 (e) A message specified by the cannot_route_message router option.
5293 In these cases only, Exim does include the error details in bounce and
5294 warning messages. There are also a few cases where bland messages such
5295 as "unrouteable address" or "local delivery error" are given.
5297 PH/43 $value is now also set for the "else" part of a ${run expansion.
5299 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5300 being worked on, but at least Exim now implements the latest version to
5303 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5304 process to complete was treated as a delivery failure, a timeout while
5305 writing the message to the pipe was logged, but erroneously treated as a
5306 successful delivery. Such timeouts include transport filter timeouts. For
5307 consistency with the overall process timeout, these timeouts are now
5308 treated as errors, giving rise to delivery failures by default. However,
5309 there is now a new Boolean option for the pipe transport called
5310 timeout_defer, which, if set TRUE, converts the failures into defers for
5311 both kinds of timeout. A transport filter timeout is now identified in
5314 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5315 systems where "make" and "gmake" are different, calling "gmake" at top
5316 level broke things. I've arranged for the value of $(MAKE) to be passed
5317 from the Makefile to this script so that it can call the same version of
5321 A note about Exim versions 4.44 and 4.50
5322 ----------------------------------------
5324 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5325 changes of various kinds. As a consequence, a big documentation update was
5326 needed. This delayed the release for rather longer than seemed good, especially
5327 in the light of a couple of (minor) security issues. Therefore, the changes
5328 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5329 release. So 4.44 and 4.50 are in effect two different branches that both start
5332 I have left the 4.50 change log unchanged; it contains all the changes since
5333 4.43. The change log for 4.44 is below; many of its items are identical to
5334 those for 4.50. This seems to be the most sensible way to preserve the
5335 historical information.
5341 1. Minor wording change to the doc/README.SIEVE file.
5343 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5344 computation of the current number of files was incorrect.
5346 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5347 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5350 4. Give more explanation in the error message when the command for a transport
5351 filter fails to execute.
5353 5. There are several places where Exim runs a non-Exim command in a
5354 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5355 was being done only for the command run by the queryprogram router. It is
5356 now done for all such subprocesses. The other cases are: ${run, transport
5357 filters, and the commands run by the lmtp and pipe transports.
5359 6. Added CONFIGURE_GROUP build-time option.
5361 7. Some older OS have a limit of 256 on the maximum number of file
5362 descriptors. Exim was using setrlimit() to set 1000 as a large value
5363 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5364 systems. I've change it so that if it can't get 1000, it tries for 256.
5366 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5367 was an oversight, and furthermore, ever since the addition of extra
5368 controls (e.g. 4.43/32), the checks on when to allow different forms of
5369 "control" were broken. There should now be diagnostics for all cases when a
5370 control that does not make sense is encountered.
5372 9. Added the /retain_sender option to "control=submission".
5374 10. $recipients is now available in the predata ACL (oversight).
5376 11. Tidy the search cache before the fork to do a delivery from a message
5377 received from the command line. Otherwise the child will trigger a lookup
5378 failure and thereby defer the delivery if it tries to use (for example) a
5379 cached ldap connection that the parent has called unbind on.
5381 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5382 of $address_data from the recipient verification was clobbered by the
5383 sender verification.
5385 13. The value of address_data from a sender verification is now available in
5386 $sender_address_data in subsequent conditions in the ACL statement.
5388 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5390 15. Added a new option "connect=<time>" to callout options, to set a different
5393 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5394 was its contents. (It was OK if the option was not defined at all.)
5396 17. A "Completed" log line is now written for messages that are removed from
5397 the spool by the -Mrm option.
5399 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5400 information about exactly what failed.
5402 19. Added -dd to debug only the daemon process.
5404 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5405 handles timeouts, both on the server side and network timeouts. Renamed the
5406 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5408 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5409 It is now set to "smtps".
5411 22. $host_address is now set to the target address during the checking of
5412 ignore_target_hosts.
5414 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5415 being passed; this would have caused $sender_host_name to have been used if
5416 matching the list had actually called for a host name (not very likely,
5417 since this list is usually IP addresses). A host name is now passed as
5420 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5421 code that specifies a non-restarting handler (typically sigaction() in
5422 modern systems) in an attempt to fix a rare and obscure crash bug.
5424 25. Narrowed the window for a race in the daemon that could cause it to ignore
5425 SIGCHLD signals. This is not a major problem, because they are used only to
5426 wake it up if nothing else does.
5428 26. A malformed maildirsize file could cause Exim to calculate negative values
5429 for the mailbox size or file count. Odd effects could occur as a result.
5430 The maildirsize information is now recalculated if the size or filecount
5433 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5434 support for a long time. Removed HAVE_SYS_VFS_H.
5436 28. Installed the latest version of exipick from John Jetmore.
5438 29. In an address list, if the pattern was not a regular expression, an empty
5439 subject address (from a bounce message) matched only if the pattern was an
5440 empty string. Non-empty patterns were not even tested. This was the wrong
5441 because it is perfectly reasonable to use an empty address as part of a
5442 database query. An empty address is now tested by patterns that are
5443 lookups. However, all the other forms of pattern expect the subject to
5444 contain a local part and a domain, and therefore, for them, an empty
5445 address still always fails if the pattern is not itself empty.
5447 30. Exim went into a mad DNS loop when attempting to do a callout where the
5448 host was specified on an smtp transport, and looking it up yielded more
5449 than one IP address.
5451 31. Re-factored the code for checking spool and log partition space into a
5452 function that finds that data and another that does the check. The former
5453 is then used to implement four new variables: $spool_space, $log_space,
5454 $spool_inodes, and $log_inodes.
5456 32. The RFC2047 encoding function was originally intended for short strings
5457 such as real names; it was not keeping to the 75-character limit for
5458 encoded words that the RFC imposes. It now respects the limit, and
5459 generates multiple encoded words if necessary. To be on the safe side, I
5460 have increased the buffer size for the ${rfc2047: expansion operator from
5463 33. It is now permitted to omit both strings after an "if" condition; if the
5464 condition is true, the result is "true". As before, when the second string
5465 is omitted, a false condition yields an empty string. This makes it less
5466 cumbersome to write custom ACL and router conditions.
5468 34. Failure to deliver a bounce message always caused it to be frozen, even if
5469 there was an errors_to setting on the router. The errors_to setting is now
5472 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5473 canonical form (fully expanded) before being placed in
5474 $sender_host_address.
5476 36. The table in the code that translates DNS record types into text (T_A to
5477 "A" for instance) was missing entries for NS and CNAME. It is just possible
5478 that this could have caused confusion if both these types were looked up
5479 for the same domain, because the text type is used as part of Exim's
5480 per-process caching. But the chance of anyone hitting this buglet seems
5483 37. The dnsdb lookup has been extended in a number of ways.
5485 (1) There is a new type, "zns", which walks up the domain tree until it
5486 finds some nameserver records. It should be used with care.
5488 (2) There is a new type, "mxh", which is like "mx" except that it returns
5489 just the host names, not the priorities.
5491 (3) It is now possible to give a list of domains (or IP addresses) to be
5492 looked up. The behaviour when one of the lookups defers can be
5493 controlled by a keyword.
5495 (4) It is now possible to specify the separator character for use when
5496 multiple records are returned.
5498 38. The dnslists ACL condition has been extended: it is now possible to supply
5499 a list of IP addresses and/or domains to be looked up in a particular DNS
5502 39. Added log_selector=+queue_time_overall.
5504 40. When running the queue in the test harness, wait just a tad after forking a
5505 delivery process, to get repeatability of debugging output.
5507 41. Include certificate and key file names in error message when GnuTLS fails
5508 to set them up, because the GnuTLS error message doesn't include the name
5509 of the failing file when there is a problem reading it.
5511 42. Allow both -bf and -bF in the same test run.
5513 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5515 44. The "Exiscan patch" is now merged into the mainline Exim source.
5517 45. Sometimes the final signoff response after QUIT could fail to get
5518 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5519 < 0 before doing a fflush(). This bug looks as though it goes back to the
5520 introduction of TLS in release 3.20, but "sometimes" must have been rare
5521 because the tests only now provoked it.
5523 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5524 (this can affect the format of dates).
5526 47. exim_tidydb, when checking for the continued existence of a message for
5527 which it has found a message-specific retry record, was not finding
5528 messages that were in split spool directories. Consequently, it was
5529 deleting retry records that should have stayed in existence.
5531 48. Steve fixed some bugs in eximstats.
5533 49. The SPA authentication driver was not abandoning authentication and moving
5534 on to the next authenticator when an expansion was forced to fail,
5535 contradicting the general specification for all authenticators. Instead it
5536 was generating a temporary error. It now behaves as specified.
5538 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5539 (the order specifies the preference for clients). The order is now AES256,
5540 AES128, 3DES, ARCFOUR128.
5542 51. Small patch to Sieve code - explicitly set From: when generating an
5545 52. Exim crashed if a remote delivery caused a very long error message to be
5546 recorded - for instance if somebody sent an entire SpamAssassin report back
5547 as a large number of 550 error lines. This bug was coincidentally fixed by
5548 increasing the size of one of Exim's internal buffers (big_buffer) that
5549 happened as part of the Exiscan merge. However, to be on the safe side, I
5550 have made the code more robust (and fixed the comments that describe what
5553 53. Now that there can be additional text after "Completed" in log lines (if
5554 the queue_time_overall log selector is set), a one-byte patch to exigrep
5555 was needed to allow it to recognize "Completed" as not the last thing in
5558 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5559 patch that reportedly fixes this has been added. I am not expert enough to
5560 create a test for it. This is what the patch creator wrote:
5562 "I found a little strange behaviour of ldap code when working with
5563 Windows 2003 AD Domain, where users was placed in more than one
5564 Organization Units. When I tried to give exim partial DN, the exit code
5565 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5566 But simultaneously result of request was absolutely normal ldap result,
5567 so I produce this patch..."
5569 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5570 REFERENCE, so I have modified the code to exclude the patch when that macro
5573 55. Some experimental protocols are using DNS PTR records for new purposes. The
5574 keys for these records are domain names, not reversed IP addresses. The
5575 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5576 leaves it alone. Component reversal etc. now happens only for IP addresses.
5579 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5581 57. Double the size of the debug message buffer (to 2048) so that more of very
5582 long debug lines gets shown.
5584 58. The exicyclog utility now does better if the number of log files to keep
5585 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5587 59. Two changes related to the smtp_active_hostname option:
5589 (1) $smtp_active_hostname is now available as a variable.
5590 (2) The default for smtp_banner uses $smtp_active_hostname instead
5591 of $primary_hostname.
5593 60. The host_aton() function is supposed to be passed a string that is known
5594 to be a valid IP address. However, in the case of IPv6 addresses, it was
5595 not checking this. This is a hostage to fortune. Exim now panics and dies
5596 if the condition is not met. A case was found where this could be provoked
5597 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5598 components; fortuitously, this particular loophole had already been fixed
5599 by change 4.50/55 above.
5601 If there are any other similar loopholes, the new check in host_aton()
5602 itself should stop them being exploited. The report I received stated that
5603 data on the command line could provoke the exploit when Exim was running as
5604 exim, but did not say which command line option was involved. All I could
5605 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5606 running as the user.
5609 61. There was a buffer overflow vulnerability in the SPA authentication code
5610 (which came originally from the Samba project). I have added a test to the
5611 spa_base64_to_bits() function which I hope fixes it.
5614 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5615 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5617 63. The daemon start-up calls getloadavg() while still root for those OS that
5618 need the first call to be done as root, but it missed one case: when
5619 deliver_queue_load_max is set with deliver_drop_privilege. This is
5620 necessary for the benefit of the queue runner, because there is no re-exec
5621 when deliver_drop_privilege is set.
5623 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5624 This has been fixed.
5626 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5627 was in use, was not putting the data itself into the right store pool;
5628 consequently, it could be overwritten for a subsequent message in the same
5629 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5632 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5634 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5635 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5636 the function were treating the return as a boolean value, which happened to
5637 work because 0=false and not-0=true, but is not correct code.
5639 68. The host_aton() function was not handling scoped IPv6 addresses (those
5640 with, for example, "%eth0" on the end) correctly.
5642 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5643 negated items (that is, ~something) in unsigned ints. Some compilers
5644 apparently mutter when there is no cast.
5646 70. If an address verification called from an ACL failed, and did not produce a
5647 user-specific message (i.e. there was only a "system" message), nothing was
5648 put in $acl_verify_message. In this situation, it now puts the system
5651 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5652 session; change 4.31/43 added the unwanted input to the log line - except
5653 that it did not do this in the start of session case. It now does.
5655 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5656 This is wrong and can cause the other end to generate a synchronization
5657 error if it is another Exim or anything else that does the synchronization
5658 check. A QUIT command is no longer sent after a timeout.
5660 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5661 during host lookups.
5663 74. The defer_ok option of callout verification was not working if it was used
5664 when verifying addresses in header lines, that is, for this case:
5666 verify = header_sender/callout=defer_ok
5668 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5669 those file descriptors could be used for SMTP connections. If anything
5670 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5671 could be sent to the SMTP client, causing chaos. The daemon now opens
5672 stdin, stdout, and stderr to /dev/null when it puts itself into the
5675 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5676 log by default. The user can override this with suitable Perl magic.
5678 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5679 the log message when discard triggers, was not working for the DATA ACL or
5680 for the non-SMTP ACL.
5682 78. Error message wording change in sieve.c.
5684 79. If smtp_accept_max_per_host was set, the number of connections could be
5685 restricted to fewer than expected, because the daemon was trying to set up
5686 a new connection before checking whether the processes handling previous
5687 connections had finished. The check for completed processes is now done
5688 earlier. On busy systems, this bug wouldn't be noticed because something
5689 else would have woken the daemon, and it would have reaped the completed
5692 80. If a message was submitted locally by a user whose login name contained one
5693 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5694 It caused a spool format error. I have fixed the spool reading code. A
5695 related problem was that the "from" clause in the Received: line became
5696 illegal because of the space(s). It is now covered by ${quote_local_part.
5698 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5701 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5703 83. Patch from Sieve maintainer to fix unterminated string problem in
5704 "vacation" handling.
5706 84. Some minor changes to the Linux configuration files to help with other
5707 OS variants using glibc.
5709 85. One more patch for Sieve to update vacation handling to latest spec.
5712 ----------------------------------------------------
5713 See the note above about the 4.44 and 4.50 releases.
5714 ----------------------------------------------------
5720 1. Change 4.43/35 introduced a bug that caused file counts to be
5721 incorrectly computed when quota_filecount was set in an appendfile
5724 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5725 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5728 3. Give more explanation in the error message when the command for a transport
5729 filter fails to execute.
5731 4. There are several places where Exim runs a non-Exim command in a
5732 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5733 was being done only for the command run by the queryprogram router. It is
5734 now done for all such subprocesses. The other cases are: ${run, transport
5735 filters, and the commands run by the lmtp and pipe transports.
5737 5. Some older OS have a limit of 256 on the maximum number of file
5738 descriptors. Exim was using setrlimit() to set 1000 as a large value
5739 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5740 systems. I've change it so that if it can't get 1000, it tries for 256.
5742 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5743 was an oversight, and furthermore, ever since the addition of extra
5744 controls (e.g. 4.43/32), the checks on when to allow different forms of
5745 "control" were broken. There should now be diagnostics for all cases when a
5746 control that does not make sense is encountered.
5748 7. $recipients is now available in the predata ACL (oversight).
5750 8. Tidy the search cache before the fork to do a delivery from a message
5751 received from the command line. Otherwise the child will trigger a lookup
5752 failure and thereby defer the delivery if it tries to use (for example) a
5753 cached ldap connection that the parent has called unbind on.
5755 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5756 of $address_data from the recipient verification was clobbered by the
5757 sender verification.
5759 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5760 was its contents. (It was OK if the option was not defined at all.)
5762 11. A "Completed" log line is now written for messages that are removed from
5763 the spool by the -Mrm option.
5765 12. $host_address is now set to the target address during the checking of
5766 ignore_target_hosts.
5768 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5769 being passed; this would have caused $sender_host_name to have been used if
5770 matching the list had actually called for a host name (not very likely,
5771 since this list is usually IP addresses). A host name is now passed as
5774 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5775 code that specifies a non-restarting handler (typically sigaction() in
5776 modern systems) in an attempt to fix a rare and obscure crash bug.
5778 15. Narrowed the window for a race in the daemon that could cause it to ignore
5779 SIGCHLD signals. This is not a major problem, because they are used only to
5780 wake it up if nothing else does.
5782 16. A malformed maildirsize file could cause Exim to calculate negative values
5783 for the mailbox size or file count. Odd effects could occur as a result.
5784 The maildirsize information is now recalculated if the size or filecount
5787 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5788 support for a long time. Removed HAVE_SYS_VFS_H.
5790 18. Updated exipick to current release from John Jetmore.
5792 19. Allow an empty sender to be matched against a lookup in an address list.
5793 Previously the only cases considered were a regular expression, or an
5796 20. Exim went into a mad DNS lookup loop when doing a callout where the
5797 host was specified on the transport, if the DNS lookup yielded more than
5800 21. The RFC2047 encoding function was originally intended for short strings
5801 such as real names; it was not keeping to the 75-character limit for
5802 encoded words that the RFC imposes. It now respects the limit, and
5803 generates multiple encoded words if necessary. To be on the safe side, I
5804 have increased the buffer size for the ${rfc2047: expansion operator from
5807 22. Failure to deliver a bounce message always caused it to be frozen, even if
5808 there was an errors_to setting on the router. The errors_to setting is now
5811 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5812 canonical form (fully expanded) before being placed in
5813 $sender_host_address.
5815 24. Updated eximstats to version 1.33
5817 25. Include certificate and key file names in error message when GnuTLS fails
5818 to set them up, because the GnuTLS error message doesn't include the name
5819 of the failing file when there is a problem reading it.
5821 26. Expand error message when OpenSSL has problems setting up cert/key files.
5824 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5825 (this can affect the format of dates).
5827 28. exim_tidydb, when checking for the continued existence of a message for
5828 which it has found a message-specific retry record, was not finding
5829 messages that were in split spool directories. Consequently, it was
5830 deleting retry records that should have stayed in existence.
5832 29. eximstats updated to version 1.35
5833 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5834 1.35 - bugfix such that pie charts by volume are generated correctly
5836 30. The SPA authentication driver was not abandoning authentication and moving
5837 on to the next authenticator when an expansion was forced to fail,
5838 contradicting the general specification for all authenticators. Instead it
5839 was generating a temporary error. It now behaves as specified.
5841 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5842 (the order specifies the preference for clients). The order is now AES256,
5843 AES128, 3DES, ARCFOUR128.
5845 31. Small patch to Sieve code - explicitly set From: when generating an
5848 32. Exim crashed if a remote delivery caused a very long error message to be
5849 recorded - for instance if somebody sent an entire SpamAssassin report back
5850 as a large number of 550 error lines. This bug was coincidentally fixed by
5851 increasing the size of one of Exim's internal buffers (big_buffer) that
5852 happened as part of the Exiscan merge. However, to be on the safe side, I
5853 have made the code more robust (and fixed the comments that describe what
5856 33. Some experimental protocols are using DNS PTR records for new purposes. The
5857 keys for these records are domain names, not reversed IP addresses. The
5858 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5859 leaves it alone. Component reversal etc. now happens only for IP addresses.
5862 34. The host_aton() function is supposed to be passed a string that is known
5863 to be a valid IP address. However, in the case of IPv6 addresses, it was
5864 not checking this. This is a hostage to fortune. Exim now panics and dies
5865 if the condition is not met. A case was found where this could be provoked
5866 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5867 components; fortuitously, this particular loophole had already been fixed
5868 by change 4.50/55 or 4.44/33 above.
5870 If there are any other similar loopholes, the new check in host_aton()
5871 itself should stop them being exploited. The report I received stated that
5872 data on the command line could provoke the exploit when Exim was running as
5873 exim, but did not say which command line option was involved. All I could
5874 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5875 running as the user.
5878 35. There was a buffer overflow vulnerability in the SPA authentication code
5879 (which came originally from the Samba project). I have added a test to the
5880 spa_base64_to_bits() function which I hope fixes it.
5883 36. The daemon start-up calls getloadavg() while still root for those OS that
5884 need the first call to be done as root, but it missed one case: when
5885 deliver_queue_load_max is set with deliver_drop_privilege. This is
5886 necessary for the benefit of the queue runner, because there is no re-exec
5887 when deliver_drop_privilege is set.
5889 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5890 was in use, was not putting the data itself into the right store pool;
5891 consequently, it could be overwritten for a subsequent message in the same
5892 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5895 38. Sometimes the final signoff response after QUIT could fail to get
5896 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5897 < 0 before doing a fflush(). This bug looks as though it goes back to the
5898 introduction of TLS in release 3.20, but "sometimes" must have been rare
5899 because the tests only now provoked it.
5905 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5906 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5907 Now it returns an integer. A number of calls were still expecting a T/F
5908 return. Fortuitously, in all cases, the tests worked in OK situations,
5909 which is the norm. However, things would have gone wrong on any write
5910 failures on the smtp file descriptor. This function is used when sending
5911 messages over SMTP and also when doing verify callouts.
5913 2. When Exim is called to do synchronous delivery of a locally submitted
5914 message (the -odf or -odi options), it no longer closes stderr before doing
5917 3. Implemented the mua_wrapper option.
5919 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5921 5. Implemented the functions header_remove(), header_testname(),
5922 header_add_at_position(), and receive_remove_recipient(), and exported them
5925 6. If an ACL "warn" statement specified the addition of headers, Exim already
5926 inserted X-ACL-Warn: at the start if there was no header name. However, it
5927 was not making this test for the second and subsequent header lines if
5928 there were newlines in the string. This meant that an invalid header could
5929 be inserted if Exim was badly configured.
5931 7. Allow an ACL "warn" statement to add header lines at the start or after all
5932 the Received: headers, as well as at the end.
5934 8. Added the rcpt_4xx retry error code.
5936 9. Added postmaster_mailfrom=xxx to callout verification option.
5938 10. Added mailfrom=xxxx to the callout verification option, for verify=
5941 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5942 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5944 12. Inserted some casts to stop certain compilers warning when using pointer
5945 differences as field lengths or precisions in printf-type calls (mostly
5946 affecting debugging statements).
5948 13. Added optional readline() support for -be (dynamically loaded).
5950 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5951 same clock tick as a message's arrival, so that its received time was the
5952 same as the "first fail" time on the retry record, and that message
5953 remained on the queue past the ultimate address timeout, every queue runner
5954 would try a delivery (because it was past the ultimate address timeout) but
5955 after another failure, the ultimate address timeout, which should have then
5956 bounced the address, did not kick in. This was a "< instead of <=" error;
5957 in most cases the first failure would have been in the next clock tick
5958 after the received time, and all would be well.
5960 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5961 being recognized when the domain list was tested by the match_domain
5962 condition in an expansion string.
5964 16. Added the ${str2b64: operator.
5966 17. Exim was always calling setrlimit() to set a large limit for the number of
5967 processes, without checking whether the existing limit was already
5968 adequate. (It did check for the limit on file descriptors.) Furthermore,
5969 errors from getrlimit() and setrlimit() were being ignored. Now they are
5970 logged to the main and panic logs, but Exim does carry on, to try to do its
5971 job under whatever limits there are.
5973 18. Imported PCRE 5.0.
5975 19. Trivial typo in log message " temporarily refused connection" (the leading
5978 20. If the log selector return_path_on_delivery was set and an address was
5979 redirected to /dev/null, the delivery process crashed because it assumed
5980 that a return path would always be set for a "successful" delivery. In this
5981 case, the whole delivery is bypassed as an optimization, and therefore no
5984 21. Internal re-arrangement: the function for sending a challenge and reading
5985 a response while authentication was assuming a zero-terminated challenge
5986 string. It's now changed to take a pointer and a length, to allow for
5987 binary data in such strings.
5989 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5991 23. Exim was not respecting finduser_retries when seeking the login of the
5992 uid under which it was called; it was always trying 10 times. (The default
5993 setting of finduser_retries is zero.) Also, it was sleeping after the final
5994 failure, which is pointless.
5996 24. Implemented tls_on_connect_ports.
5998 25. Implemented acl_smtp_predata.
6000 26. If the domain in control=submission is set empty, Exim assumes that the
6001 authenticated id is a complete email address when it generates From: or
6002 Sender: header lines.
6004 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6005 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6006 chown and chgrp in /bin and hostname in /usr/bin.
6008 28. Exim was keeping the "process log" file open after each use, just as it
6009 does for the main log. This opens the possibility of it remaining open for
6010 long periods when the USR1 signal hits a daemon. Occasional processlog
6011 errors were reported, that could have been caused by this. Anyway, it seems
6012 much more sensible not to leave this file open at all, so that is what now
6015 29. The long-running daemon process does not normally write to the log once it
6016 has entered its main loop, and it closes the log before doing so. This is
6017 so that log files can straightforwardly be renamed and moved. However,
6018 there are a couple of unusual error situations where the daemon does write
6019 log entries, and I had neglected to close the log afterwards.
6021 30. The text of an SMTP error response that was received during a remote
6022 delivery was being truncated at 512 bytes. This is too short for some of
6023 the long messages that one sometimes sees. I've increased the limit to
6026 31. It is now possible to make retry rules that apply only when a message has a
6027 specific sender, in particular, an empty sender.
6029 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6030 it possible to be selective about when SMTP synchronization is enforced.
6032 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6034 32. Implemented hosts_connection_nolog.
6036 33. Added an ACL for QUIT.
6038 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6041 35. Added mailbox_size and mailbox_filecount to appendfile.
6043 36. Added control = no_multiline_responses to ACLs.
6045 37. There was a bug in the logic of the code that waits for the clock to tick
6046 in the case where the clock went backwards by a substantial amount such
6047 that the microsecond fraction of "now" was more than the microsecond
6048 fraction of "then" (but the whole seconds number was less).
6050 38. Added support for the libradius Radius client library this is found on
6051 FreeBSD (previously only the radiusclient library was supported).
6057 1. When certain lookups returned multiple values in the form name=value, the
6058 quoting of the values was not always being done properly. Specifically:
6059 (a) If the value started with a double quote, but contained no whitespace,
6061 (b) If the value contained whitespace other than a space character (i.e.
6062 tabs or newlines or carriage returns) it was not quoted.
6063 This fix has been applied to the mysql and pgsql lookups by writing a
6064 separate quoting function and calling it from the lookup code. The fix
6065 should probably also be applied to nisplus, ibase and oracle lookups, but
6066 since I cannot test any of those, I have not disturbed their existing code.
6068 2. A hit in the callout cache for a specific address caused a log line with no
6069 reason for rejecting RCPT. Now it says "Previous (cached) callout
6070 verification failure".
6072 3. There was an off-by-one bug in the queryprogram router. An over-long
6073 return line was truncated at 256 instead of 255 characters, thereby
6074 overflowing its buffer with the terminating zero. As well as fixing this, I
6075 have increased the buffer size to 1024 (and made a note to document this).
6077 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6078 when Exim is waiting for an SMTP response from a remote server, Exim
6079 restarts its select() call on the socket, thereby resetting its timeout.
6080 This is not a problem when such interrupts are rare. Somebody set up a cron
6081 job to run exiwhat every 2 minutes, which is less than the normal select()
6082 timeout (5 or 10 minutes). This meant that the select() timeout never
6083 kicked in because it was always reset. I have fixed this by comparing the
6084 time when an interrupt arrives with the time at the start of the first call
6085 to select(). If more time than the timeout has elapsed, the interrupt is
6086 treated as a timeout.
6088 5. Some internal re-factoring in preparation for the addition of Sieve
6089 extensions (by MH). In particular, the "personal" test is moved to a
6090 separate function, and given an option for scanning Cc: and Bcc: (which is
6091 not set for Exim filters).
6093 6. When Exim created an email address using the login of the caller as the
6094 local part (e.g. when creating a From: or Sender: header line), it was not
6095 quoting the local part when it contained special characters such as @.
6097 7. Installed new OpenBSD configuration files.
6099 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6100 try to make them clearer.
6102 9. Callout options, other than the timeout value, were being ignored when
6103 verifying sender addresses in header lines. For example, when using
6105 verify = header_sender/callout=no_cache
6107 the cache was (incorrectly) being used.
6109 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6110 only the Cygwin environment.
6112 11. When return_path_on_delivery was set as a log selector, if different remote
6113 addresses in the same message used different return paths and parallel
6114 remote delivery occurred, the wrong values would sometimes be logged.
6115 (Whenever a remote delivery process finished, the return path value from
6116 the most recently started remote delivery process was logged.)
6118 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6119 lines when AUTH and/or TLS are in use. This is the "received protocol"
6120 field. Exim used to use "asmtp" for authenticated SMTP, without any
6121 indication (in the protocol name) for TLS use. Now it follows the RFC and
6122 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6123 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6124 names appear in log lines as well as in Received: header lines.
6126 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6127 extensions, and comparison tests, and to fix some bugs.
6129 14. Changes to the "personal" filter test:
6131 (1) The test was buggy in that it was just doing the equivalent of
6132 "contains" tests on header lines. For example, if a user's address was
6133 anne@some.where, the "personal" test would incorrectly be true for
6135 To: susanne@some.where
6137 This test is now done by extracting each address from the header in turn,
6138 and checking the entire address. Other tests that are part of "personal"
6139 are now done using regular expressions (for example, to check local parts
6140 of addresses in From: header lines).
6142 (2) The list of non-personal local parts in From: addresses has been
6143 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6144 taken from the Sieve specification recommendations.
6146 (3) If the message contains any header line starting with "List-" it is
6147 treated as non-personal.
6149 (4) The test for "circular" in the Subject: header line has been removed
6150 because it now seems ill-conceived.
6152 15. Minor typos in src/EDITME comments corrected.
6154 16. Installed latest exipick from John Jetmore.
6156 17. If headers_add on a router specified a text string that was too long for
6157 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6158 of string_sprintf() is now avoided.
6160 18. $message_body_size was not set (it was always zero) when running the DATA
6161 ACL and the local_scan() function.
6163 19. For the "mail" command in an Exim filter, no default was being set for
6164 the once_repeat time, causing a random time value to be used if "once" was
6165 specified. (If the value happened to be <= 0, no repeat happened.) The
6166 default is now 0s, meaning "never repeat". The "vacation" command was OK
6167 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6168 (I found it when inspecting the code).
6170 20. There is now an overall timeout for performing a callout verification. It
6171 defaults to 4 times the callout timeout, which applies to individual SMTP
6172 commands during the callout. The overall timeout applies when there is more
6173 than one host that can be tried. The timeout is checked before trying the
6174 next host. This prevents very long delays if there are a large number of
6175 hosts and all are timing out (e.g. when the network connections are timing
6176 out). The value of the overall timeout can be changed by specifying an
6177 additional sub-option for "callout", called "maxwait". For example:
6179 verify = sender/callout=5s,maxwait=20s
6181 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6182 to the end before writing, but this should make it even safer).
6184 22. Exim was forgetting that it had advertised PIPELINING for the second and
6185 subsequent messages on an SMTP connection. It was also not resetting its
6186 memory on STARTTLS and an internal HELO.
6188 23. When Exim logs an SMTP synchronization error within a session, it now
6189 records whether PIPELINING has been advertised or not.
6191 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6192 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6193 rather than long int.
6195 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6197 26. Added the never_mail option to autoreply.
6203 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6204 crash if the getsockname() call failed; this can happen if a connection is
6205 closed very soon after it is established. The problem was simply in the
6206 order in which certain operations were done, causing Exim to try to write
6207 to the SMTP stream before it had set up the file descriptor. The bug has
6208 been fixed by making things happen in the correct order.
6214 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6215 before the connection was closed, thus losing the rejection response.
6217 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6218 some early Solaris releases, but causes trouble in current releases where
6219 socklen_t is defined.
6221 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6224 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6227 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6228 that releasing the top of it at the end releases what was used for sub-
6229 expansions (unless the block got too big). However, discard this block if
6230 the first thing is a variable or header, so that we can use its block when
6231 it is dynamic (useful for very large $message_headers, for example).
6233 6. Lookups now cache *every* query, not just the most recent. A new, separate
6234 store pool is used for this. It can be recovered when all lookup caches are
6235 flushed. Lookups now release memory at the end of their result strings.
6236 This has involved some general refactoring of the lookup sources.
6238 7. Some code has been added to the store_xxx() functions to reduce the amount
6239 of flapping under certain conditions.
6241 8. log_incoming_interface used to affect only the <= reception log lines. Now
6242 it causes the local interface and port to be added to several more SMTP log
6243 lines, for example "SMTP connection from", and rejection lines.
6245 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6247 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6249 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6250 could be overwritten at the end of the current message (or the start of a
6251 new message if it was set in a HELO ACL). The value is now preserved for
6252 the duration of the SMTP connection.
6254 12. If a transport had a headers_rewrite setting, and a matching header line
6255 contained an unqualified address, that address was qualified, even if it
6256 did not match any rewriting rules. The underlying bug was that the values
6257 of the flags that permit the existence of unqualified sender and recipient
6258 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6259 non-local messages, and by -bnq for local messages) were not being
6260 preserved with the message after it was received.
6262 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6263 "next input=" as part of the text comprising the host identity instead of
6264 the correct text. The code was using the same buffer for two different
6265 strings. However, depending on which order the printing function evaluated
6266 its arguments, the bug did not always show up. Under Linux, for example, my
6267 test suite worked just fine.
6269 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6270 doesn't work with some older versions of Perl. It has been changed to "my",
6271 which in any case is probably the better facility to use.
6273 15. A really picky compiler found some instances of statements for creating
6274 error messages that either had too many or two few arguments for the format
6277 16. The size of the buffer for calls to the DNS resolver has been increased
6278 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6279 for addresses that have a lot of PTR records. This alleviates a problem; it
6280 does not fully solve it.
6282 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6283 buffer now truncates the list and logs the incident, which is the same
6284 action as happens when Exim is looking up a host name and its aliases.
6285 Previously in this situation something unpredictable would happen;
6286 sometimes it was "internal error: store_reset failed".
6288 18. If a server dropped the connection unexpectedly when an Exim client was
6289 using GnuTLS and trying to read a response, the client delivery process
6290 crashed while trying to generate an error log message.
6292 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6293 string, for example:
6295 warn message = H1: something\nH2: something
6297 the text was added as a single header line from Exim's point of view
6298 though it ended up OK in the delivered message. However, searching for the
6299 second and subsequent header lines using $h_h2: did not work. This has been
6300 fixed. Similarly, if a system filter added multiple headers in this way,
6301 the routers could not see them.
6303 20. Expanded the error message when iplsearch is called with an invalid key to
6304 suggest using net-iplsearch in a host list.
6306 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6307 ACLs are no longer actually imposed (and a message to that effect is
6310 22. If a "gecos" field in a passwd entry contained escaped characters, in
6311 particular, if it contained a \" sequence, Exim got it wrong when building
6312 a From: or a Sender: header from that name. A second bug also caused
6313 incorrect handling when an unquoted " was present following a character
6314 that needed quoting.
6316 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6317 was not being matched caselessly.
6319 24. Arranged for all hyphens in the exim.8 source to be escaped with
6322 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6323 properly. Recipient callout cache records were still being keyed to include
6324 the sender, even when use_sender was set false. This led to far more
6325 callouts that were necessary. The sender is no longer included in the key
6326 when use_sender is false.
6328 26. Added "control = submission" modifier to ACLs.
6330 27. Added the ${base62d: operator to decode base 62 numbers.
6332 28. dnsdb lookups can now access SRV records.
6334 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6335 the configuration file.
6337 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6338 (-v) mode. This makes the output for a verbose queue run more intelligible.
6340 31. Added a use_postmaster feature to recipient callouts.
6342 32. Added the $body_zerocount variable, containing the number of binary zero
6343 bytes in the message body.
6345 33. The time of last modification of the "new" subdirectory is now used as the
6346 "mailbox time last read" when there is a quota error for a maildir
6349 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6351 35. Added +ignore_unknown as a special item in host lists.
6353 36. Code for decoding IPv6 addresses in host lists is now included, even if
6354 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6355 address was recognized as an IP address, but was then not correctly decoded
6356 into binary, causing unexpected and incorrect effects when compared with
6363 1. Very minor rewording of debugging text in manualroute to say "list of
6364 hosts" instead of "hostlist".
6366 2. If verify=header_syntax was set, and a header line with an unqualified
6367 address (no domain) and a large number of spaces between the end of the
6368 name and the colon was received, the reception process suffered a buffer
6369 overflow, and (when I tested it) crashed. This was caused by some obsolete
6370 code that should have been removed. The fix is to remove it!
6372 3. When running in the test harness, delay a bit after writing a bounce
6373 message to get a bit more predictability in the log output.
6375 4. Added a call to search_tidyup() just before forking a reception process. In
6376 theory, someone could use a lookup in the expansion of smtp_accept_max_
6377 per_host which, without the tidyup, could leave open a database connection.
6379 5. Added the variables $recipient_data and $sender_data which get set from a
6380 lookup success in an ACL "recipients" or "senders" condition, or a router
6381 "senders" option, similar to $domain_data and $local_part_data.
6383 6. Moved the writing of debug_print from before to after the "senders" test
6386 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6387 problems for message scanning, either using a data ACL, or using
6388 local_scan() because the Received: header was not generated till after they
6389 were called (in order to set the time as the time of reception completion).
6390 I have revised the way this works. The header is now generated after the
6391 body is received, but before the ACL or local_scan() are called. After they
6392 are run, the timestamp in the header is updated.
6398 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6399 before starting a queue runner without re-exec. This happened only when
6400 deliver_drop_privilege was set or when the Exim user was set to root. The
6401 effect of the bug was that timeouts during subsequent deliveries caused
6402 crashes instead of being properly handled. The handler is now left at its
6403 default (and expected) setting.
6405 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6406 message, again when deliver_drop_privilege is set or Exim is run as root.
6407 The bug described in (1) was not present in this case, but the tidying up
6408 of the other signals was missing. I have made the two cases consistent.
6410 3. The ignore_target_hosts setting on a manualroute router was being ignored
6411 for hosts that were looked up using the /MX notation.
6413 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6416 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6417 operated on the sender address. After changing the $sender_address to <>
6418 for the sender address verify, Exim was re-instated it as the original
6419 (before rewriting) address, but remembering that it had rewritten it, so it
6420 wasn't rewriting it again. This bug also had the effect of breaking the
6421 sender address verification caching when the sender address was rewritten.
6423 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6424 This has been changed so that if the ip literal address matches
6425 ignore_target_hosts, the router declines.
6427 7. Added expansion conditions match_domain, match_address, and match_local_
6428 part (NOT match_host).
6430 8. The placeholder for the Received: header didn't have a length field set.
6432 9. Added code to Exim itself and to exim_lock to test for a specific race
6433 condition that could lead to file corruption when using MBX delivery. The
6434 issue is with the lockfile that is created in /tmp. If this file is removed
6435 after a process has opened it but before that process has acquired a lock,
6436 there is the potential for a second process to recreate the file and also
6437 acquire a lock. This could lead to two Exim processes writing to the file
6438 at the same time. The added code performs the same test as UW imapd; it
6439 checks after acquiring the lock that its file descriptor still refers to
6440 the same named file.
6442 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6443 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6446 11. Added the smtp_active_hostname option. If used, this will typically be made
6447 to depend on the incoming interface address. Because $interface_address is
6448 not set up until the daemon has forked a reception process, error responses
6449 that can happen earlier (such as "too many connections") no longer contain
6452 12. If an expansion in a condition on a "warn" statement fails because a lookup
6453 defers, the "warn" statement is abandoned, and the next ACL statement is
6454 processed. Previously this caused the whole ACL to be aborted.
6456 13. Added the iplsearch lookup type.
6458 14. Added ident_timeout as a log selector.
6460 15. Added tls_certificate_verified as a log selector.
6462 16. Added a global option tls_require_ciphers (compare the smtp transport
6463 option of the same name). This controls incoming TLS connections.
6465 17. I finally figured out how to make tls_require_ciphers do a similar thing
6466 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6467 before starting the TLS session.
6469 18. Tabs are now shown as \t in -bP output.
6471 19. If the log selector return_path_on_delivery was set, Exim crashed when
6472 bouncing a message because it had too many Received: header lines.
6474 20. If two routers both had headers_remove settings, and the first one included
6475 a superfluous trailing colon, the final name in the first list and the
6476 first name in the second list were incorrectly joined into one item (with a
6477 colon in the middle).
6483 1. Added -C and -D options to the exinext utility, mainly to make it easier
6484 to include in the automated testing, but these could be helpful when
6485 multiple configurations are in use.
6487 2. The exinext utility was not formatting the output nicely when there was
6488 an alternate port involved in the retry record key, nor when there was a
6489 message id as well (for retries that were specific to a specific message
6490 and a specific host). It was also confused by IPv6 addresses, because of
6491 the additional colons they contain. I have fixed the IPv4 problem, and
6492 patched it up to do a reasonable job for IPv6.
6494 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6495 delivery, the log line now contains "pipelined" if PIPELINING was used.
6497 4. An SMTP transport process used to panic and die if the bind() call to set
6498 an explicit outgoing interface failed. This has been changed; it is now
6499 treated in the same way as a connect() failure.
6501 5. A reference to $sender_host_name in the part of a conditional expansion
6502 that was being skipped was still causing a DNS lookup. This no longer
6505 6. The def: expansion condition was not recognizing references to header lines
6506 that used bh_ and bheader_.
6508 7. Added the _cache feature to named lists.
6510 8. The code for checking quota_filecount in the appendfile transport was
6511 allowing one more file than it should have been.
6513 9. For compatibility with Sendmail, the command line option
6521 and sets the incoming protocol and host name (for trusted callers). The
6522 host name and its colon can be omitted when only the protocol is to be set.
6523 Note the Exim already has two private options, -pd and -ps, that refer to
6524 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6525 "s", but I don't think that's a major issue.
6527 10. A number of refactoring changes to the code, none of which should affect
6530 (a) The number of logging options was getting close to filling up the
6531 32-bit word that was used as a bit map. I have split them into two classes:
6532 those that are passed in the argument to log_write(), and those that are
6533 only ever tested independently outside of that function. These are now in
6534 separate 32-bit words, so there is plenty of room for expansion again.
6535 There is no change in the user interface or the logging behaviour.
6537 (b) When building, for example, log lines, the code previously used a
6538 macro that called string_cat() twice, in order to add two strings. This is
6539 not really sufficiently general. Furthermore, there was one instance where
6540 it was actually wrong because one of the argument was used twice, and in
6541 one call a function was used. (As it happened, calling the function twice
6542 did not affect the overall behaviour.) The macro has been replaced by a
6543 function that can join an arbitrary number of extra strings onto a growing
6546 (c) The code for expansion conditions now uses a table and a binary chop
6547 instead of a serial search (which was left over from when there were very
6548 few conditions). Also, it now recognizes conditions like "pam" even when
6549 the relevant support is not compiled in: a suitably worded error message is
6550 given if an attempt is made to use such a condition.
6552 11. Added ${time_interval:xxxxx}.
6554 12. A bug was causing one of the ddress fields not to be passed back correctly
6555 from remote delivery subprocesses. The field in question was not being
6556 subsequently used, so this caused to problems in practice.
6558 13. Added new log selectors queue_time and deliver_time.
6560 14. Might have fixed a bug in maildirsizefile handling that threw up
6561 "unexpected character" debug warnings, and recalculated the data
6562 unnecessarily. In any case, I expanded the warning message to give more
6565 15. Added the message "Restricted characters in address" to the statements in
6566 the default ACL that block characters like @ and % in local parts.
6568 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6569 Three changes have been made:
6571 (a) There was a serious bug; a negative response to MAIL caused the whole
6572 recipient domain to be cached as invalid, thereby blocking all messages
6573 to all local parts at the same domain, from all senders. This bug has
6574 been fixed. The domain is no longer cached after a negative response to
6575 MAIL if the sender used is not empty.
6577 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6580 (c) A new callout option, "use_sender" has been added for people who want
6581 the modified behaviour.
6587 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6590 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6591 indeed breaks things for older releases.
6593 3. Added additional logging to the case where there is a problem reading data
6594 from a filter that is running in a subprocess using a pipe, in order to
6595 try to track down a specific problem.
6597 4. Testing facility fudge: when running in the test harness and attempting
6598 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6599 getting "No route to host". Convert this to a timeout.
6601 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6604 6. Some OS don't have socklen_t but use size_t instead. This affects the
6605 fifth argument of getsockopt() amongst other things. This is now
6606 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6607 can be set for individual OS. I have set it for SunOS5, OSF1, and
6608 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6609 some earlier ones do not.
6611 7. Change 4.30/15 was not doing the test caselessly.
6613 8. The standard form for an IPv6 address literal was being rejected by address
6614 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6615 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6616 this, as well as the form without the "IPv6" on the front (but only when
6617 address literals are enabled, of course).
6619 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6621 10. Exim crashed if a message with an empty sender address specified by -f
6622 encountered a router with an errors_to setting. This could be provoked only
6623 by a command such as
6627 where an empty string was supplied; "<>" did not hit this bug.
6629 11. Installed PCRE release 4.5.
6631 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6632 remained set. It is now erased.
6634 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6635 times from message ids (which are base 36 rather than the normal 62).
6637 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6638 were being counted as actual protocol errors, and logged if the log
6639 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6640 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6641 and DATA following a set of rejected RCPTs do not count as protocol errors.
6642 In other words, Exim assumes they were pipelined, though this may not
6643 actually be the case. Of course, in all cases the client gets an
6644 appropriate error code.
6646 15. If a lookup fails in an ACL condition, a message about the failure may
6647 be available; it is used if testing the ACL cannot continue, because most
6648 such messages specify what the cause of the deferral is. However, some
6649 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6650 that caused an old message to be retained and used if a later statement
6651 caused a defer, replacing the real cause of the deferral.
6653 16. If an IP address had so many PTR records that the DNS lookup buffer
6654 was not large enough to hold them, Exim could crash while trying to process
6655 the truncated data. It now detects and logs this case.
6657 17. Further to 4.21/58, another change has been made: if (and only if) the
6658 first line of a message (the first header line) ends with CRLF, a bare LF
6659 in a subsequent header line has a space inserted after it, so as not to
6660 terminate the header.
6662 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6663 unnecessarily, used atoi() instead of strtol(), and didn't check the
6664 termination when getting file sizes from file names by regex.
6666 19. Completely re-implemented the support for maildirsize files, in the light
6667 of a number of problems with the previous contributed implementation
6668 (4.30/29). In particular:
6670 . If the quota is zero, the maildirsize file is maintained, but no quota is
6673 . If the maildir directory does not exist, it is created before any attempt
6674 to write a maildirsize file.
6676 . The quota value in the file is just a cache; if the quota is changed in
6677 the transport, the new value overrides.
6679 . A regular expression is available for excluding directories from the
6682 20. The autoreply transport checks the characters in options that define the
6683 message's headers; it allows continued headers, but it was checking with
6684 isspace() after an embedded newline instead of explicitly looking for a
6687 21. If all the "regular" hosts to which an address was routed had passed their
6688 expiry times, and had not reached their retry times, the address was
6689 bounced, even if fallback hosts were defined. Now Exim should go on to try
6692 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6693 equivalent code in the SMTP transport. Some hosts send humungous responses
6694 to HELO/EHLO, more than 1024 it seems.
6696 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6697 gives compiler warnings in some environments. I've now done it "properly",
6700 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6701 (because of problems with the built-in one) was declared to return uschar *
6702 instead of char *, causing compiler failure.
6704 25. Fixed a file descriptor leak when processing alias/forward files.
6706 26. Fixed a minor format string issue in dbfn.c.
6708 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6710 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6711 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6714 29. When a sender address is verified, it is cached, to save repeating the test
6715 when there is more than one recipient in a message. However, when the
6716 verification involves a callout, it is possible for different callout
6717 options to be set for different recipients. It is too complicated to keep
6718 track of this in the cache, so now Exim always runs a verification when a
6719 callout is required, relying on the callout cache for the optimization.
6720 The overhead is duplication of the address routing, but this should not be
6723 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6724 to be verified with callout=postmaster, and the main callout worked but the
6725 postmaster check failed, the verification correctly failed. However, if a
6726 subsequent RCPT command asked for sender verification *without* the
6727 postmaster check, incorrect caching caused this verification also to fail,
6730 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6731 it was not caching the DNS options (qualify_single, search_parents) that
6732 were used when the lookup failed. A subsequent lookup with different
6733 options therefore always gave the same answer, though there were cases
6734 where it should not have. (Example: a "domains = !$mx_any" option on a
6735 dnslookup router: the "domains" option is always processed without any
6736 widening, but the router might have qualify_single set.) Now Exim uses the
6737 cached value only when the same options are set.
6739 32. Added John Jetmore's "exipick" utility to the distribution.
6741 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6742 than a timeout (e.g. a certificate is required, and is not provided), an
6743 Exim server now closes the connection immediately. Previously it waited for
6744 the client to close - but if the client is SSL, it seems that they each
6745 wait for each other, leading to a delay before one of them times out.
6747 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6748 maintained 0.8.x compatibility because I don't think many are using it, and
6749 it is clearly obsolete.
6751 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6754 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6755 client certificate was expired. A simple patch fixes this, though I don't
6756 understand the full logic of why the verify callback is called multiple
6759 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6760 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6761 which causes problems with some clients (such as the Certicom SSL Plus
6762 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6763 disables the coutermeasure allowing Eudora to connect."
6765 38. Exim was not checking that a write() to a log file succeeded. This could
6766 lead to Bad Things if a log got too big, in particular if it hit a file
6767 size limit. Exim now panics and dies if it cannot write to a log file, just
6768 as it does if it cannot open a log file.
6770 39. Modified OS/Makefile-Linux so that it now contains
6772 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6774 The two -D definitions ensure that Exim is compiled with large file
6775 support, which makes it possible to handle log files that are bigger than
6778 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6779 instance) a domain was checked against a named list that involved a lookup,
6780 causing $domain_data to be set, then another domain was checked against the
6781 same list, then the first domain was re-checked, the value of $domain_data
6782 after the final check could be wrong. In particular, if the second check
6783 failed, it could be set empty. This bug probably also applied to
6786 41. The strip_trailing_dot option was not being applied to the address given
6787 with the -f command-line option.
6789 42. The code for reading a message's header from the spool was incrementing
6790 $received_count, but never initializing it. This meant that the value was
6791 incorrect (doubled) while delivering a message in the same process in which
6792 it was received. In the most common configuration of Exim, this never
6793 happens - a fresh exec is done - but it can happen when
6794 deliver_drop_privilege is set.
6796 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6797 it now includes up to 150 characters of the unexpected data in the log
6800 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6801 and building data strings. The size of both of these buffers was 10 000
6802 bytes - far larger than anybody would *ever* want, thought I. Needless to
6803 say, somebody hit the limit. I have increased the maximum line length to
6804 20 000 and the maximum data length of concatenated lines to 100 000. I have
6805 also fixed two bugs, because there was no checking on these buffers. Tsk,
6806 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6807 buffer is too small.
6809 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6810 lsearch lookups. Now it does.
6812 46. When parsing a route_list item in a manualroute router, a fixed-length
6813 buffer was used for the list of hosts. I made this 1024 bytes long,
6814 thinking that nobody would ever have a list of hosts that long. Wrong.
6815 Somebody had a whole pile of complicated expansion conditions, and the
6816 string was silently truncated, leading to an expansion error. It turns out
6817 that it is easier to change to an unlimited length (owing to other changes
6818 that have happened since this code was originally written) than to build
6819 structure for giving a limitation error. The length of the item that
6820 expands into the list of hosts is now unlimited.
6822 47. The lsearch lookup could not handle data where the length of text line was
6823 more than 4095 characters. Such lines were truncated, leading to shortened
6824 data being returned. It should now handle lines of any length.
6826 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6827 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6830 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6831 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6832 no longer happens in comment lines. A list of replacements is now placed
6833 at the head of all of the source files, except those whose only change is
6834 to replace PERL_COMMAND in the very first #! line.
6836 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6837 messages on the queue, with a bottom-up merge sort, using code contributed
6838 by Michael Haardt. This should make operations like -bp somewhat faster on
6839 large queues. It won't affect queue runners, except when queue_run_in_order
6842 51. Installed eximstats 1.31 in the distribution.
6844 52. Added support for SRV lookups to the dnslookup router.
6846 53. If an ACL referred to $message_body or $message_body_end, the value was not
6847 reset for any messages that followed in the same SMTP session.
6849 54. The store-handling optimization for building very long strings was not
6850 differentiating between the different store pools. I don't think this
6851 actually made any difference in practice, but I've tidied it.
6853 55. While running the routers to verify a sender address, $sender_address
6854 was still set to the sender address. This is wrong, because when routing to
6855 send a bounce to the sender, it would be empty. Therefore, I have changed
6856 it so that, while verifying a sender address, $sender_address is set to <>.
6857 (There is no change to what happens when verifying a recipient address.)
6859 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6860 target A or AAAA records (if not already returned) without resetting the
6861 qualify_single or search_parents options of the DNS resolver. These are
6862 inappropriate in this case because the targets of MX and SRV records must
6863 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6864 target that, when qualified, matched something in the local domain. These
6865 two options are now turned off when doing these lookups.
6867 57. It seems that at least some releases of Reiserfs (which does not have the
6868 concept of a fixed number of inodes) returns zero and not -1 for the
6869 number of available inodes. This interacted badly with check_spool_inodes,
6870 which assumed that -1 was the "no such thing" setting. What I have done is
6871 to check that the total number of inodes is greater than zero before doing
6872 the test of how many are available.
6874 58. When a "warn" ACL statement has a log_message modifier, the message is
6875 remembered, and not repeated. This is to avoid a lot of repetition when a
6876 message has many recipients that cause the same warning to be written.
6877 However, Exim was preserving the list of already written lines for an
6878 entire SMTP session, which doesn't seem right. The memory is now reset if a
6879 new message is started.
6881 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6882 debugging output unless log_rewrite was also set.
6884 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6885 of (int)(handle) when we know that handle contains (void *)(-1).
6887 61. The Exim daemon panic-logs an error return when it closes the incoming
6888 connection. However "connection reset by peer" seems to be common, and
6889 isn't really an error worthy of noting specially, so that particular error
6892 62. When Exim is trying to find all the local interfaces, it used to panic and
6893 die if the ioctl to get the interface flags failed. However, it seems that
6894 on at least one OS (Solaris 9) it is possible to have an interface that is
6895 included in the list of interfaces, but for which you get a failure error
6896 for this call. This happens when the interface is not "plumbed" into a
6897 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6898 failure of the "get flags" call assumes that the interface is down.
6900 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6901 makes life easier for people who are doing arithmetic on fields extracted
6902 from dates, where you often get leading zeros that should not be
6903 interpreted as octal.
6905 64. Added qualify_domain to the redirect router, to override the global
6908 65. If a pathologically long header line contained very many addresses (the
6909 report of this problem mentioned 10 000) and each of them was rewritten,
6910 Exim could use up a very large amount of memory. (It kept on making new
6911 copies of the header line as it rewrote, and never released the old ones.)
6912 At the expense of a bit more processing, the header rewriting function has
6913 been changed so that it no longer eats memory in this way.
6915 66. The generation of the Received: header has been moved from the time that a
6916 message starts to be received, to the time that it finishes. The timestamp
6917 in the Received: header should now be very close to that of the <= log
6918 line. There are two side-effects of this change:
6920 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6921 logged header lines no longer include the local Received: line, because
6922 it has not yet been created. The same applies to a copy of the message
6923 that is returned to a non-SMTP sender when a message is rejected.
6925 (b) When a filter file is tested using -bf, no additional Received: header
6926 is added to the test message. After some thought, I decided that this
6929 This change does not affect the value of $received_for. It is still set
6930 after address rewriting, but before local_scan() is called.
6932 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6934 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6935 gave an unhelpful panic error message, and a defer error. I have managed to
6936 change this behaviour so that it now rejects any supplied certificate,
6937 which seems right, as the list of acceptable certificates is empty.
6939 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6940 gave an unhelpful defer error. I have not managed to make this reject any
6941 supplied certificates, but the error message it gives is "no certificate
6942 supplied", which is not helpful.
6944 70. exigrep's output now also includes lines that are not associated with any
6945 message, but which match the given pattern. Implemented by a patch from
6946 Martin Sluka, which also tidied up the Perl a bit.
6948 71. Recipient callout verification, like sender verification, was using <> in
6949 the MAIL FROM command. This isn't really the right thing, since the actual
6950 sender may affect whether the remote host accepts the recipient or not. I
6951 have changed it to use the actual sender in the callout; this means that
6952 the cache record is now keyed on a recipient/sender pair, not just the
6953 recipient address. There doesn't seem to be a real danger of callout loops,
6954 since a callout by the remote host to check the sender would use <>.
6955 [SEE ABOVE: changed after hitting problems.]
6957 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6958 temporary errors. However, in the case of such a code being given after
6959 the end of a data transmission (i.e. after ".") Exim was failing to write
6960 a retry record for the message. (Yes, there was some broken host that was
6961 actually sending 8xx at this point.)
6963 73. An unknown lookup type in a host list could cause Exim to panic-die when
6964 the list was checked. (An example that provoked this was putting <; in the
6965 middle of a list instead of at the start.) If this happened during a DATA
6966 ACL check, a -D file could be left lying around. This kind of configuration
6967 error no longer causes Exim to die; instead it causes a defer error. The
6968 incident is still logged to the main and panic logs.
6970 74. Buglet left over from Exim 3 conversion. The message "too many messages
6971 in one connection" was written to the rejectlog but not the mainlog, except
6972 when address rewriting (yes!) was being logged.
6974 75. Added write_rejectlog option.
6976 76. When a system filter was run not as root (that is, when system_filter_user
6977 was set), the values of the $n variables were not being returned to the
6978 main process; thus, they were not subsequently available in the $sn
6981 77. Added +return_path_on_delivery log selector.
6983 78. A connection timeout was being treated differently from recipients deferred
6984 when testing hosts_max_try with a message that was older than the host's
6985 retry timeout. (The host should not be counted, thus allowing all hosts to
6986 be tried at least once before bouncing.) This may have been the cause of an
6987 occasionally reported bug whereby a message would remain on the queue
6988 longer than the retry timeout, but would be bounced if a delivery was
6989 forced. I say "may" because I never totally pinned down the problem;
6990 setting up timeout/retry tests is difficult. See also the next item.
6992 79. The ultimate address timeout was not being applied to errors that involved
6993 a combination of host plus message (for example, a timeout on a MAIL
6994 command). When an address resolved to a number of possible hosts, and they
6995 were not all tried for each delivery (e.g. because of hosts_max_try), a
6996 message could remain on the queue longer than the retry timeout.
6998 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7001 81. Fixed an obscure SMTP outgoing bug which required at least the following
7002 conditions: (a) there was another message waiting for the same server;
7003 (b) the server returned 5xx to all RCPT commands in the first message so
7004 that the message was not completed; (c) the server dropped the connection
7005 or gave a negative response to the RSET that Exim sends to abort the
7006 transaction. The observed case was a dropped connection after DATA that had
7007 been sent in pipelining mode. That is, the server had advertised PIPELINING
7008 but was not implementing it correctly. The effect of the bug was incorrect
7009 behaviour, such as trying another host, and this could lead to a crash.
7015 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7016 and daemon.c were passed as pointers to ints; they should have been
7017 pointers to socklen_t variables (which are typically unsigned ints).
7019 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7022 3. Fixed a really odd bug that affected only the testing scheme; patching a
7023 certain fixed string in the binary changed the value of another string that
7024 happened to be identical to the end of the original first string.
7026 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7027 name", it returns that address as the IP address. On some operating
7028 systems (e.g. Solaris), it also passes back the IP address string as the
7029 "host name". However, on others (e.g. Linux), it passes back an empty
7030 string. Exim wasn't checking for this, and was changing the host name to an
7031 empty string, assuming it had been canonicalized.
7033 5. Although rare, it is permitted to have more than one PTR record for a given
7034 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7035 all the names associated with an address, because they do in Solaris.
7036 However, it seems that they do not in Linux for data that comes from the
7037 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7038 I found this out when I moved to a new Linux workstation and tried to run
7039 the Exim test suite.
7041 To get round this problem I have changed the code so that it now does its
7042 own call to the DNS to look up PTR records when searching for a host name.
7043 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7044 addresses that are only in /etc/hosts are still found.
7046 This behaviour is, however, controlled by an option called host_lookup_
7047 order, which defaults to "bydns:byaddr". If people want to use the other
7048 order, or indeed, just use one or the other means of lookup, they can
7049 specify it in this variable.
7051 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7052 some operating systems, this comes back from gethostbyaddr() as an empty
7053 string, and this is what Exim used to test for. However, it seems that in
7054 other systems, "." is yielded. Exim now tests for this case too.
7056 7. The values of check_spool_space and check_log_space are now held internally
7057 as a number of kilobytes instead of an absolute number of bytes. If a
7058 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7059 kilobyte. This means that much larger values can be stored.
7061 8. Exim monitor: an attempt to get the action menu when not actually pointing
7062 at a message produces an empty menu entitled "No message selected". This
7063 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7064 no entries in it ("Shell widget menu has zero width and/or height"). So I
7065 have added a single, blank menu entry in this case.
7067 9. Added ${quote_local_part.
7069 10. MIME decoding is now applied to the contents of Subject: header lines when
7072 11. Now that a reference to $sender_host_address automatically causes a reverse
7073 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7074 host lookup before query-style lookups in lists that might use this
7075 variable. This has therefore been abolished, and the "net-" prefix is no
7076 longer necessary for query-style lookups.
7078 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7079 have been a typo for LFLAGS, so it has been changed.
7081 13. The install script calls Exim with "-C /dev/null" in order to find the
7082 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7083 to be output. However, since Exim outputs its version number before the
7084 error, it didn't break the script. It just looked ugly. I fixed this by
7085 always allowing "-C /dev/null" if the caller is root.
7087 14. Ignore overlarge ACL variable number when reading spool file - insurance
7088 against a later release with more variables having written the file.
7090 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7091 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7092 this, as well as the form without the "IPv6" on the front.
7094 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7095 OS/Makefile-Darwin file.
7097 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7098 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7100 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7101 variable. However, some OS use ints for inodes. I've added cast to long int
7102 to get rid of the compiler warning.
7104 19. I had forgotten to lock out "/../" in configuration file names when
7105 ALT_CONFIG_PREFIX was set.
7107 20. Routers used for verification do not need to specify transports. However,
7108 if such a router generated a host list, and callout was configured, Exim
7109 crashed, because it could not find a port number from the (non-existent)
7110 transport. It now assumes port 25 in this circumstance.
7112 21. Added the -t option to exigrep.
7114 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7115 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7116 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7117 them is set, LOOKUP_LSEARCH is forced.
7119 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7121 24. Added sender and host information to the "rejected by local_scan()" log
7122 line; previously there was no indication of these.
7124 25. Added .include_if_exists.
7126 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7127 Linux. It turns out that not all file systems support this. Apparently some
7128 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7129 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7130 not supported on the file descriptor, is now ignored when Exim is trying to
7131 sync a directory. This applies only to Linux.
7133 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7135 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7136 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7139 29. Added support for maildirsize files from supplied patch (modified a bit).
7141 30. The use of :fail: followed by an empty string could lead Exim to respond to
7142 sender verification failures with (e.g.):
7144 550 Verification failed for <xxx>
7145 550 Sender verify failed
7147 where the first response line was missing the '-' that indicates it is not
7148 the final line of the response.
7150 31. The loop for finding the name of the user that called Exim had a hardwired
7151 limit of 10; it now uses the value of finduser_retries, which is used for
7152 all other user lookups.
7154 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7157 33. Exim was neglecting to zero errno before one call of strtol() when
7158 expanding a string and expecting an integer value. On some systems this
7159 resulted in spurious "integer overflow" errors. Also, it was casting the
7160 result into an int without checking.
7162 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7163 did not work. The code looks as if it has *never* worked, though it appears
7164 to have been documented since at least release 1.62. I have made it work.
7166 35. The "timeout_DNS" error in retry rules, also documented since at least
7167 1.62, also never worked. As it isn't clear exactly what this means, and
7168 clearly it isn't a major issue, I have abolished the feature by treating it
7169 as "timeout", and writing a warning to the main and panic logs.
7171 36. The display of retry rules for -brt wasn't always showing the error code
7174 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7175 timeout_connect_A, timeout_connect_MX.
7177 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7178 to the empty sender.
7180 39. The daemon was not analysing the content of -oX till after it had closed
7181 stderr and disconnected from the controlling terminal. This meant that any
7182 syntax errors were only noted on the panic log, and the return code from
7183 the command was 0. By re-arranging the code a little, I've made the
7184 decoding happen first, so such errors now appear on stderr, and the return
7185 code is 1. However, the actual setting up of the sockets still happens in
7186 the disconnected process, so errors there are still only recorded on the
7189 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7190 connections (as happens on some IP stacks) was logged at start up time as
7191 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7192 it from "IPv6 and IPv4", which means that two separate sockets are being
7195 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7196 says whether AF_INET or AF_INET6 was passed as an argument.
7198 42. Exiwhat output was messed up when time zones were included in log
7201 43. Exiwhat now gives more information about the daemon's listening ports,
7202 and whether -tls-on-connect was used.
7204 44. The "port" option of the smtp transport is now expanded.
7206 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7207 silently ignored. Now an error message is written to the main and panic
7210 46. There's a new ACL modifier called "logwrite" which writes to a log file
7211 as soon as it is encountered.
7213 47. Added $local_user_uid and $local_user_gid at routing time.
7215 48. Exim crashed when trying to verify a sender address that was being
7218 49. Exim was recognizing only a space character after ".include". It now also
7219 recognizes a tab character.
7221 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7222 extracting the relevant information from the specification. The man page no
7223 longer contains scrambled data for the -d option, and I've added a section
7224 at the front about calling Exim under different names.
7226 51. Added "extra_headers" argument to the "mail" command in filter files.
7228 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7231 53. Installed eximstats 1.29.
7233 54. Added transport_filter_timeout as a generic transport option.
7235 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7236 Cc: header lines. This was required by RFC 822, but it not required by RFC
7239 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7240 incoming messages that did not have them. Now it does so only if the
7241 message originates locally, that is, if there is no associated remote host
7242 address. When Resent- header lines are present, this applies to the Resent-
7243 lines rather than the non-Resent- lines.
7245 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7246 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7248 58. Messages for configuration errors now include the name of the main
7249 configuration file - useful now that there may be more than one file in a
7250 list (.included file names were always shown).
7252 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7253 for those rare installations that do not start the daemon as root or run it
7254 setuid root. I've cut out the call to initgroups() if the daemon is not
7257 60. The Exim user and group can now be bound into the binary as text strings
7258 that are looked up at the start of Exim's processing.
7260 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7262 62. Added $mailstore_basename variable.
7264 63. Installed patch to sieve.c from Michael Haardt.
7266 64. When Exim failed to open the panic log after failing to open the main log,
7267 the original message it was trying to log was written to stderr and debug
7268 output, but if they were not available (the usual case in production), it
7269 was lost. Now it is written to syslog before the two lines that record the
7270 failures to open the logs.
7272 65. Users' Exim filters run in subprocesses under the user's uid. It is
7273 possible for a "deliver" command or an alias in a "personal" command to
7274 provoke an address rewrite. If logging of address rewriting is configured,
7275 this fails because the process is not running as root or exim. There may be
7276 a better way of dealing with this, but for the moment (because 4.30 needs
7277 to be released), I have disabled address rewrite logging when running a
7278 filter in a non-root, non-exim process.
7284 1. The buildconfig auxiliary program wasn't quoting the value set for
7285 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7286 not defined. This bug was masked in 4.22 by the effect that was fixed in
7289 2. Some messages that were rejected after a message id was allocated were
7290 shown as "incomplete" by exigrep. It no longer does this for messages that
7291 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7293 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7294 have allow_domain_literals set, the ID did not get logged in the <= line.
7295 Domain literals are now always recognized in Message-ID: header lines.
7297 4. The first argument for a ${extract expansion item is the key name or field
7298 number. Leading and trailing spaces in this item were not being ignored,
7299 causing some misleading effects.
7301 5. When deliver_drop_privilege was set, single queue runner processes started
7302 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7303 same command in the process it spins off) were not dropping privilege.
7305 6. When the daemon running as "exim" started a queue runner, it always
7306 re-executed Exim in the spun-off process. This is a waste of effort when
7307 deliver_drop_privilege is set. The new process now just calls the
7308 queue-runner function directly.
7314 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7317 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7318 ignored. Though the use of -oP was forcing the writing of a pid file, it
7319 was always written to the default place.
7321 3. If the message "no IP address found for host xxxx" is generated during
7322 incoming verification, it is now followed by identification of the incoming
7323 connection (so you can more easily find what provoked it).
7325 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7327 5. Added some features to "harden" Exim a bit more against certain attacks:
7329 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7330 be put in Local/Makefile. This is like the never_users runtime option,
7331 but it cannot be overridden. The default setting is "root".
7333 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7334 prefix string with which any file named in a -C command line option
7337 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7338 is retained for -C and -D only if the caller of Exim is root. Without
7339 it, the exim user may also use -C and -D and retain privilege.
7341 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7342 command line option is disabled.
7344 6. Macro names set by the -D option must start with an upper case letter, just
7345 like macro names defined in the configuration file.
7347 7. Added "dereference=" facility to LDAP.
7349 8. Two instances of the typo "uknown" in the source files are fixed.
7351 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7352 the Configure-Makefile script screwed up while processing it.
7354 10. Incorporated PCRE 4.4.
7356 11. The SMTP synchronization check was not operating right at the start of an
7357 SMTP session. For example, it could not catch a HELO sent before the client
7358 waited for the greeting. There is now a check for outstanding input at the
7359 point when the greeting is written. Because of the duplex, asynchronous
7360 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7361 way, but not yet received, when the check is performed.
7363 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7364 on TCP/IP sockets, because this apparently causes some broken clients to
7367 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7368 unchanged) from the Cygwin maintainer.
7370 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7371 was supported instead of testing for mbx. Effectively a typo.
7373 15. The spa authenticator server code was not checking that the input it
7374 received was valid base64.
7376 16. The debug output line for the "set" modifier in ACLs was not showing the
7377 name of the variable that was being set.
7379 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7381 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7382 lookup on its own. Something else was needed to trigger the lookup. For
7383 example, a match in host_lookup or the need for a host name in a host list.
7384 Now, if $sender_host_name is referenced and the host name has not yet been
7385 looked up, a lookup is performed. If the lookup fails, the variable remains
7386 empty, and $host_lookup_failed is set to "1".
7388 19. Added "eqi" as a case-independent comparison operator.
7390 20. The saslauthd authentication condition could segfault if neither service
7391 nor realm was specified.
7393 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7394 error message that was logged was misleading, and incoming SMTP
7395 connections were dropped. The message is now more accurate, and temporary
7396 errors are given to SMTP connections.
7398 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7399 (except RSET) with a 421 error, until QUIT is received. However, it was
7400 failing to send a response to QUIT.
7402 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7403 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7404 ACL is now run after the helo_try_verify_hosts code.
7406 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7407 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7408 case-independent, so other case variants are also recognized). Apparently
7409 some systems use these upper case variants.
7411 25. If more than two messages were waiting for the same host, and a transport
7412 filter was specified for the transport, Exim sent two messages over the
7413 same TCP/IP connection, and then failed with "socket operation on non-
7414 socket" when it tried to send the third.
7416 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7418 27. The extern definition of crypt16() in expand.c was not being excluded when
7419 the OS had its own crypt16() function.
7421 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7422 as a preferred synonym for return_size_limit, both as an option and as an
7425 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7427 30. Changed the default configuration ACL to relax the local part checking rule
7428 for addresses that are not in any local domains. For these addresses,
7429 slashes and pipe symbols are allowed within local parts, but the sequence
7430 /../ is explicitly forbidden.
7432 31. SPA server authentication was not clearing the challenge buffer before
7435 32. log_message in a "warn" ACL statement was writing to the reject log as
7436 well as to the main log, which contradicts the documentation and doesn't
7437 seem right (because no rejection is happening). So I have stopped it.
7439 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7440 However, I am unable to do any testing of this.
7442 34. Fixed an infelicity in the appendfile transport. When checking directories
7443 for a mailbox, to see if any needed to be created, it was accidentally
7444 using path names with one or more superfluous leading slashes; tracing
7445 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7447 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7448 contents are added to the log line that is written for every discarded
7449 recipient. (Previously a log_message setting was ignored.)
7451 36. The ${quote: operator now quotes the string if it is empty.
7453 37. The install script runs exim in order to find its version number. If for
7454 some reason other than non-existence or emptiness, which it checks, it
7455 could not run './exim', it was installing it with an empty version number,
7456 i.e. as "exim-". This error state is now caught, and the installation is
7459 38. An argument was missing from the function that creates an error message
7460 when Exim fails to connect to the socket for saslauthd authentication.
7461 This could cause Exim to crash, or give a corrupted message.
7463 39. Added isip, isip4, and isip6 to ${if conditions.
7465 40. The ACL variables $acl_xx are now saved with the message, and can be
7466 accessed later in routers, transports, and filters.
7468 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7469 strings in the file are not string-expanded.
7471 42. If a MAIL command specified a SIZE value that was too large to fit into an
7472 int variable, the check against message_size_limit failed. Such values are
7473 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7474 day this will have to be increased, but I don't think I want to be around
7475 when emails are that large.
7482 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7483 iconv() is not standard in FreeBSD.
7485 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7486 IPv6 enabled. The observed symptom was a segmentation fault on return from
7487 the function os_common_find_running_interfaces() in src/os.c.
7489 3. In the check_special_case() function in daemon.c I had used "errno" as an
7490 argument name, which causes warnings on some systems. This was basically a
7491 typo, since it was named "eno" in the comments!
7493 4. The code that waits for the clock to tick (at a resolution of some fraction
7494 of a second) so as to ensure message-id uniqueness was always waiting for
7495 at least one whole tick, when it could have waited for less. [This is
7496 almost certainly not relevant at current processor speeds, where it is
7497 unlikely to ever wait at all. But we try to future-proof.]
7499 5. The function that sleeps for a time interval that includes fractions of a
7500 second contained a race. It did not block SIGALRM between setting the
7501 timer, and suspending (a couple of lines later). If the interval was short
7502 and the sigsuspend() was delayed until after it had expired, the suspension
7503 never ended. On busy systems this could lead to processes getting stuck for
7506 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7507 process, before it forks any delivery processes. The open lookup caching
7508 mechanism meant that the open file or database connection was passed into
7509 the delivery process. The problem was that delivery processes always tidy
7510 up cached lookup data. This could cause a problem for the next delivery
7511 process started by the queue runner, because the external queue runner
7512 process does not know about the closure. So the next delivery process
7513 still has data in the lookup cache. In the case of a file lookup, there was
7514 no problem because closing a file descriptor in a subprocess doesn't affect
7515 the parent. However, if the lookup was caching a connection to a database,
7516 the connection was closed, and the second delivery process was likely to
7517 see errors such as "PGSQL: query failed: server closed the connection
7518 unexpectedly". The problem has been fixed by closing all cached lookups
7519 in a queue runner before running a delivery process.
7521 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7522 seem to have the "const" qualifier which it has on other OS. I've
7525 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7526 *of the same type* (client or server) with the same public name that an
7527 error should be diagnosed.
7529 9. When Exim looked up a host name for an IP address, but failed to find the
7530 original IP address when looking up the host name (a safety check), it
7531 output the message "<ip address> does not match any IP for NULL", which was
7532 confusing, to say the least. The bug was that the host name should have
7533 appeared instead of "NULL".
7535 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7536 user that is built into the binary, and the -C or -D options is used, root
7537 privilege is dropped before the configuration file is read. In addition,
7538 logging is switched to stderr instead of the normal log files. If the
7539 configuration then re-defines the Exim user, the unprivileged environment
7540 is probably not what is expected, so Exim logs a panic warning message (but
7543 However, if deliver_drop_privilege is set, the unprivileged state may well
7544 be exactly what is intended, so the warning has been cut out in that case,
7545 and Exim is allowed to try to write to its normal log files.
7551 1. smtp_return_error_details was not giving details for temporary sender
7552 or receiver verification errors.
7554 2. Diagnose a configuration error if two authenticators have the same public
7557 3. Exim used not to create the message log file for a message until the first
7558 delivery attempt. This could be confusing when incoming messages were held
7559 for policy or load reasons. The message log file is now created at the time
7560 the message is received, and an initial "Received" line is written to it.
7562 4. The automatically generated man page for command line options had a minor
7563 bug that caused no ill effects; however, a more serious problem was that
7564 the procedure for building the man page automatically didn't always
7565 operate. Consequently, release 4.20 contains an out-of-date version. This
7566 shouldn't happen again.
7568 5. When building Exim with embedded Perl support, the script that builds the
7569 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7570 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7572 6. The freeze_tell option was not being used for messages that were frozen on
7573 arrival, either by an ACL or by local_scan().
7575 7. Added the smtp_incomplete_transaction log selector.
7577 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7578 was accepting AUTH without a new EHLO.
7580 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7581 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7584 10. Logging of TCP/IP connections (when configured) now happens in the main
7585 daemon process instead of the child process, so that the TCP/IP connection
7586 count is more accurate (but it can never be perfect).
7588 11. The use of "drop" in a nested ACL was not being handled correctly in the
7589 outer ACL. Now, if condition failure induced by the nested "drop" causes
7590 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7591 or "require"), the connection is dropped.
7593 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7594 that yield "discard" can now be used with an "accept" or a "discard" verb,
7595 but an error is generated for any others (because I can't see a useful way
7596 to define what should happen).
7598 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7599 are now processed in the same way as lines in the Exim configuration file.
7600 In particular, continuation lines are supported.
7602 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7604 15. Added -ti meaning -t -i.
7606 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7607 domains, and warn by logging if others are found.
7609 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7610 returned from GIFCONF when Exim is trying to find the list of interfaces on
7611 a host. The code in os.c has been modified to copy each ifreq to an aligned
7612 structure in all cases.
7614 Also, in some cases, the returned ifreq's were being copied to a 'struct
7615 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7616 means the last couple of bytes of an IPv6 address could be chopped if the
7617 ifreq contained only a normal sockaddr (14 bytes storage).
7619 18. Named domain lists were not supported in the hosts_treat_as_local option.
7620 An entry such as +xxxx was not recognized, and was treated as a literal
7623 19. Ensure that header lines added by a DATA ACL are included in the reject log
7624 if the ACL subsequently rejects the message.
7626 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7627 MD5 (which is deprecated).
7629 21. When testing a filter file using -bf, Exim was writing a message when it
7630 took the sender from a "From " line in the message, but it was not doing so
7631 when it took $return_path from a Return-Path: header line. It now does.
7633 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7634 with a valid header line field name (a series of printing characters
7635 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7637 23. Changed "disc" in the source to "disk" to conform to the documentation and
7638 the book and for uniformity.
7640 24. Ignore Sendmail's -Ooption=value command line item.
7642 25. When execve() failed while trying to run a command in a pipe transport,
7643 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7644 could be confused with a return value of 69 from the command itself. This
7645 has been changed to 127, the value the shell returns if it is asked to run
7646 a non-existent command. The wording for the related log line suggests a
7647 non-existent command as the problem.
7649 26. If received_header_text expands to an empty string, do not add a Received:
7650 header line to the message. (Well, it adds a token one on the spool, but
7651 marks it "old" so that it doesn't get used or transmitted.)
7653 27. Installed eximstats 1.28 (addition of -nt option).
7655 28. There was no check for failure on the call to getsockname() in the daemon
7656 code. This can fail if there is a shortage of resources on the system, with
7657 ENOMEM, for example. A temporary error is now given on failure.
7659 29. Contrary to the C standard, it seems that in some environments, the
7660 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7661 program. Exim now does this explicitly; it affects the formatting of
7662 timestamps using strftime().
7664 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7665 complaints. I've now initialized all the variables, to avoid this.
7667 32. Header lines added by a system filter were not being "seen" during
7668 transport-time rewrites.
7670 33. The info_callback() function passed to OpenSSL is set up with type void
7671 (*)(SSL *, int, int), as described somewhere. However, when calling the
7672 function (actually a macro) that sets it up, the type void(*)() is
7673 expected. I've put in a cast to prevent warnings from picky compilers.
7675 34. If a DNS black list lookup found a CNAME record, but there were no A
7676 records associated with the domain it pointed at, Exim crashed.
7678 35. If a DNS black list lookup returned more than one A record, Exim ignored
7679 all but the first. It now scans all returned addresses if a particular IP
7680 value is being sought. In this situation, the contents of the
7681 $dnslist_value variable are a list of all the addresses, separated by a
7684 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7685 to accept a host name and all its aliases if the forward lookup for any of
7686 them yielded the IP address of the incoming connection. Now it accepts only
7687 those names whose forward lookup yields the correct IP address. Any other
7688 names are discarded. This closes a loophole whereby a rogue DNS
7689 administrator could create reverse DNS records to break through a
7690 wildcarded host restriction in an ACL.
7692 37. If a user filter or a system filter that ran in a subprocess used any of
7693 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7694 the wrong values were passed to the pipe command ($thisaddress had the
7695 value of $0, $0 had the value of $1, etc). This bug was introduced by
7696 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7698 38. Improved the line breaking for long SMTP error messages from ACLs.
7699 Previously, if there was no break point between 40 and 75 characters, Exim
7700 left the rest of the message alone. Two changes have been made: (a) I've
7701 reduced the minimum length to 35 characters; (b) if it can't find a break
7702 point between 35 and 75 characters, it looks ahead and uses the first one
7703 that it finds. This may give the occasional overlong line, but at least the
7704 remaining text gets split now.
7706 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7707 file descriptors might be low, and that setting 1000 would always raise it.
7708 It turns out that in some environments, the limit is already over 1000 and
7709 that lowering it causes trouble. So now Exim takes care not to decrease it.
7711 40. When delivering a message, the value of $return_path is set to $sender_
7712 address at the start of routing (routers may change the value). By an
7713 oversight, this default was not being set up when an address was tested by
7714 -bt or -bv, which affected the outcome if any router or filter referred to
7717 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7718 log only when "message" or "log_message" are set. However, if one of the
7719 conditions was an address verification, or a call to a nested ACL, the
7720 messages generated by the underlying test were being passed through. This
7721 no longer happens. The underlying message is available in $acl_verify_
7722 message for both "message" and "log_message" expansions, so it can be
7723 passed through if needed.
7725 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7726 new expansion $bh_ to give the encoded byte string without charset
7727 translation. Translation happens only if iconv() is available; HAVE_ICONV
7728 indicates this at build time. HEADERS_CHARSET gives the charset to
7729 translate to; headers_charset can change it in the configuration, and
7730 "headers charset" can change it in an individual filter file.
7732 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7733 that creates RFC 2047 encoded "words" labels them as that charset instead
7734 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7735 expansion operator; (ii) when Exim creates a From: line for a local
7736 message; (iii) when a header line is rewritten to include a "phrase" part.
7738 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7739 buggy, causing it to skip the first lines of messages whose message ID
7740 ended in 'D'. This would not have bitten before Exim release 4.14, because
7741 message IDs were unlikely to end in 'D' before then. The effect was to have
7742 incorrect size information for certain domains.
7744 45. #include "config.h" was missing at the start of the crypt16.c module. This
7745 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7748 46. If there was a timeout during a "random" callout check, Exim treated it as
7749 a failure of the random address, and carried on sending RSET and the real
7750 address. If the delay was just some slowness somewhere, the response to the
7751 original RCPT would be taken as a response to RSET and so on, causing
7752 mayhem of various kinds.
7754 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7755 when I implemented it. It didn't allow for the fact that some option values
7756 may legitimately be negative (e.g. size_addition), and it didn't even do
7757 the right test for positive values.
7759 48. Domain names in DNS records are case-independent. Exim always looks them up
7760 in lower case. Some resolvers return domain names in exactly the case they
7761 appear in the zone file, that is, they may contain uppercase letters. Not
7762 all resolvers do this - some return always lower case. Exim was treating a
7763 change of case by a resolver as a change of domain, similar to a widening
7764 of a domain abbreviation. This triggered its re-routing code and so it was
7765 trying to route what was effectively the same domain again. This normally
7766 caused routing to fail (because the router wouldn't handle the domain
7767 twice). Now Exim checks for this case specially, and just changes the
7768 casing of the domain that it ultimately uses when it transmits the message
7771 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7774 50. If a filter generated a file delivery with a non-absolute name (possible if
7775 no home directory exists for the router), the forbid_file option was not
7778 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7779 the existing equality matching.
7781 52. Exim was using ints instead of ino_t variables in some places where it was
7782 dealing with inode numbers.
7784 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7785 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7786 called TMPDIR, and if it finds it is different, it changes its value.
7788 54. The smtp_printf() function is now made available to local_scan() so
7789 additional output lines can be written before returning. There is also an
7790 smtp_fflush() function to enable the detection of a dropped connection.
7791 The variables smtp_input and smtp_batched_input are exported to
7794 55. Changed the default runtime configuration: the message "Unknown user"
7795 has been removed from the ACL, and instead placed on the localuser router,
7796 using the cannot_route_message feature. This means that any verification
7797 failures that generate their own messages won't get overridden. Similarly,
7798 the "Unrouteable address" message that was in the ACL for unverifiable
7799 relay addresses has also been removed.
7801 56. Added hosts_avoid_esmtp to the smtp transport.
7803 57. The exicyclog script was not checking for the esoteric option
7804 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7805 will work only if exicyclog is run under the appropriate euid.
7807 58. Following a discussion on the list, the rules by which Exim recognises line
7808 endings on incoming messages have been changed. The -dropcr and drop_cr
7809 options are now no-ops, retained only for backwards compatibility. The
7810 following line terminators are recognized: LF CRLF CR. However, special
7811 processing applies to CR:
7813 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7814 nor a local message in the state where . is a terminator.
7816 (ii) If a bare CR is encountered in a header line, an extra space is added
7817 after the line terminator so as not to end the header. The reasoning
7818 behind this is that bare CRs in header lines are most likely either
7819 to be mistakes, or people trying to play silly games.
7821 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7822 was being incorrectly given as 18 bytes larger than it should have been.
7823 This is a VOB (very old bug).
7825 60. This may never have affected anything current, but just in case it has:
7826 When the local host is found other than at the start of a list of hosts,
7827 the local host, those with the same MX, and any that follow, are discarded.
7828 When the list in question was part of a longer list of hosts, the following
7829 hosts (not currently being processed) were also being discarded. This no
7830 longer happens. I'm not sure if this situation could ever has previously
7833 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7836 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7837 header. This is something that is recommended in a new Internet Draft, and
7838 is something that is documented as being done by Sendmail. There are two
7839 possible values. For messages generated by the autoreply transport, Exim
7842 Auto-Submitted: auto-replied
7844 whereas for all other generated messages (e.g. bounces) it adds
7846 Auto-Submitted: auto-generated
7848 63. The "personal" condition in filters now includes a test for the
7849 Auto-Submitted: header. If it contains the string "auto-" the message it
7850 not considered personal.
7852 64. Added rcpt_include_affixes as a generic transport option.
7854 65. Added queue_only_override (default true).
7856 66. Added the syslog_duplication option.
7858 67. If what should have been the first header line of a message consisted of
7859 a space followed by a colon, Exim was mis-interpreting it as a header line.
7860 It isn't of course - it is syntactically invalid and should therefore be
7861 treated as the start of the message body. The misbehaviour could have
7862 caused a number of strange effects, including loss of data in subsequent
7863 header lines, and spool format errors.
7865 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7866 client host had authenticated. This control can now be exercised by an ACL
7867 for more flexibility.
7869 69. By default, callouts do not happen when testing with -bh. There is now a
7870 variant, -bhc, which does actually run the callout code, including
7871 consulting and updating the callout cache.
7873 70. Added support for saslauthd authentication, courtesy of Alexander
7876 71. If statvfs() failed on the spool or log directories while checking their
7877 size for availability, Exim confusingly gave the error "space shortage".
7878 Furthermore, in debugging mode it crashed with a floating point exception.
7879 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7880 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7881 serious problem, Exim now writes to the main and panic logs when this
7882 happens, with details of the failure. It then refuses to accept the
7883 incoming message, giving the message "spool directory problem" or "log
7884 directory problem" with a 421 code for SMTP messages.
7886 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7887 0, 1, and 2 exist, because some OS complain for execs without them (see
7888 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7889 descriptors. However, the code omitted to check that the open succeeded,
7890 causing mysterious errors if for some reason the permissions on /dev/null
7891 got screwed. Now Exim writes a message to the main and panic logs, and
7892 bombs out if it can't open /dev/null.
7894 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7895 interact so that it is all more flexible. It is supposed to remain
7896 backwards compatible. Also added extra_local_interfaces.
7898 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7899 to bomb out with an assertion failure - to the client this appears as a
7900 connection drop. This problem occurs in the part of the code that was taken
7901 from the Samba project. Fortunately, the assertion is in a very simple
7902 function, so I have fixed this by reproducing the function inline in the
7903 one place where it is called, and arranging for authentication to fail
7904 instead of killing the process with assert().
7906 75. The SPA client code was not working when the server requested OEM rather
7907 than Unicode encoding.
7909 76. Added code to make require_files with a specific uid setting more usable in
7910 the case where statting the file as root fails - usually a non-root-mounted
7911 NFS file system. When this happens and the failure is EACCES, Exim now
7912 forks a subprocess and does the per-uid checking as the relevant uid.
7914 77. Added process_log_path.
7916 78. If log_file_path was not explicitly set, a setting of check_log_space or
7917 check_log_inodes was ignored.
7919 79. If a space check for the spool or log partitions fails, the incident is now
7920 logged. Of course, in the latter case the data may get lost...
7922 80. Added the %p formatting code to string_format() so that it can be used to
7923 print addresses in debug_print(). Adjusted all the address printing in the
7924 debugging in store.c to use %p rather than %d.
7926 81. There was a concern that a line of code in smtp_in.c could overflow a
7927 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7928 initially expressed, the concern was not well-founded, because trailing
7929 spaces are removed early. However, if the trailing spaces were followed by
7930 a NULL, they did not get removed, so the overflow was possible. Two fixes
7933 (a) I re-wrote the offending code in a cleaner fashion.
7934 (b) If an incoming SMTP command contains a NULL character, it is rejected
7937 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7938 runs initgroups(), so that if the Exim user is in any additional groups,
7939 they will be used during message reception.
7945 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob