1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
133 JH/01 Avoid costly startup code when not strictly needed. This reduces time
134 for some exim process initialisations. It does mean that the logging
135 of TLS configuration problems is only done for the daemon startup.
137 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
139 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
140 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
143 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
144 DKIM signing is being done. Previously a transport_filter would always
145 disable CHUNKING, falling back to traditional DATA.
147 JH/05 Regard command-line receipients as tainted.
149 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
151 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
152 PAM library frees one of the arguments given to it, despite the
153 documentation. Therefore a plain malloc must be used.
155 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
156 on-stack buffers were used, resulting in a taint trap when DSN information
157 copied from a received message was written into the buffer.
159 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
160 the ordering of its ARC headers. This caused a crash.
162 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
163 a new record was being constructed with information from the peer, a trap
166 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
167 installation would get error messages from DMARC verify, when it hit the
168 nonexistent file indicated by the default. Distros wanting DMARC enabled
169 should both provide the file and set the option.
170 Also enforce no DMARC verification for command-line sourced messages.
172 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
173 could, depending on the platform, hang at the STARTTLS response.
175 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
176 message on a connection. Previously if one message had ARC headers and
177 the following one did not, a crash could result when adding an
178 Authentication-Results: header.
180 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
181 Exim main code and Exim-related utities. The introduction of taint
182 tracking also did many adjustments to string handling. Since then, eximon
183 frequently terminated with an assert failure.
185 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
186 check for 452 responses. This slightly helps the inefficieny of doing
187 a large alias-expansion into a recipient-limited target. The max_rcpt
188 transport option still applies (and at the current default, will override
189 the new feature). The check is done for either cause of synch, and forces
190 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
191 connection. The new facility is not tunable at this time.
193 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
194 library live data was being used, so the results became garbage. Make
195 copies while it is still usable.
197 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
198 on delivery deferred (==) and failed (**) lines (if a delivery was
199 attemtped). Previously it was only on completion (=>) lines.
201 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
202 for the expansion of the server_scram_iter and server_scram_salt options.
204 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
205 are now specifically given a NO_DATA response without hitting the system
206 resolver. The library goes on to do the now-standard TXT lookup.
207 Use of dnsdb lookups is not affected.
209 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
210 only retrieve the errormessage once. Previously two calls to dlerror()
211 were used, and the second one (for mainlog/paniclog) retrieved null
214 JH/20 Taint checking: disallow use of tainted data for
215 - the appendfile transport file and directory options
216 - the pipe transport command
217 - the autoreply transport file, log and once options
218 - file names used by the redirect router (including filter files)
220 - paths used by single-key lookups
221 Previously this was permitted.
223 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
224 adjusted the size of a major service buffer; this failed because the
225 buffer was in use at the time. Change to a compile-time increase in the
226 buffer size, when this authenticator is compiled into exim.
228 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
229 previous fast-mode was untenable in the face of glibs using mmap to
230 support larger malloc requests.
232 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
233 New values supported, if defined on system where compiled:
234 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
235 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
237 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
238 running a limited number of proceses in parallel, a benefit is gained. The
239 amount varies with the platform hardware and load. The use of the option
240 queue_run_in_order means we cannot do this, as ordering becomes
243 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
244 had introduced a string-copy (for ensuring NUL-termination) which was not
245 appropriate for that case, which can include embedded NUL bytes in the
246 block of data. Investigation showed the copy to actually be needless, the
247 data being length-specified.
249 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
250 done during a receiving connection, and both used TLS, global info was
251 used rather than per-connection info for tracking the state of data
252 queued for transmission. This could result in a connection hang.
254 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
255 Previously, when delivering serveral messages down a single connection
256 only the first would provide a SIZE. This was due to the size information
257 not being properly tracked.
259 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
260 TAI (at 37 seconds currently), pretend to be in UTC for time-related
261 expansion and logging. Previously, spurious values such as a future
262 minute could be seen.
264 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
265 it could crash from a null-deref. This could also affect the
266 ${addresses: } operator and ${readsock } item.
268 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
269 message following a mime one, the variable was not reset.
271 JH/30 When an pipelined-connect fails at the first response, assume incorrect
272 cached capability (perhaps the peer reneged?) and immediately retry in
275 JH/31 Fix spurious detection of timeout while writing to transport filter.
277 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
278 an attempt to copy the string was made before checking it.
280 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
281 taint of the lookup key was maintained; we now regard the presence in the
282 filesystem as sufficient validation.
284 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
285 argument is supplied.
287 JH/35 The dsearch lookup now requires that the directory is an absolute path.
288 Previously this was not checked, and nonempty relative paths made an
289 access under Exim's current working directory.
291 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
292 Previously no event was raised.
294 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
295 parameter supplied by the sender MAIL FROM command. Previously it was
296 ignored, and only the check_spool_space option value for the required
299 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
300 the size of the signing public-key. Previously it was instead giving
301 the size of the signature hash.
303 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
304 the default. See the (new) dkim_verify_min_keysizes option.
306 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
307 and an ACL use a lookup for checking either the local_part or domain,
308 stale data could be accessed. Ensure that variable references are
309 dropped between messages.
311 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
312 by the client was not checked as pointing within response data before
313 being used. A malicious client could thus cause an out-of-bounds read and
314 possibly gain authentication. Fix by adding the check.
316 JH/42 Internationalisation: change the default for downconversion in the smtp
317 transport to be "if needed". Previously it was "as previously set" for
318 the message, which usually meant "if needed" for message-submission but
319 "no" for everything else. However, MTAs have been seen using SMTPUTF8
320 even when the envelope addresses did not need it, resulting in forwarding
321 failures to non-supporting MTAs. A downconvert in such cases will be
322 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
323 transport. The change does mean that addresses needing conversion will
324 be converted when previously a delivery failure would occur.
326 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
327 response was received it would be used unchecked in a fail-DSN, violating
328 standards on line-length limits. Truncate if needed.
330 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
331 doesn't work. (Additionally add a new main config option to configure the
338 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
339 side, to decode a TLS 1.3 packet capture.
341 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
342 Previously the default library behaviour applied, sending two, each in
345 JH/03 Debug output for ACL now gives the config file name and line number for
348 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
350 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
352 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
353 buffer overrun for (non-chunking) other transports.
355 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
356 TLS1.3, means that a server rejecting a client certificate is not visible
357 to the client until the first read of encrypted data (typically the
358 response to EHLO). Add detection for that case and treat it as a failed
359 TLS connection attempt, so that the normal retry-in-clear can work (if
360 suitably configured).
362 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
363 and/or domain. Found and fixed by Jason Betts.
365 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
366 configuration). If a CNAME target was not a wellformed name pattern, a
369 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
370 the OS reports them interleaved with other addresses.
372 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
373 used both for input and for a verify callout, both encrypted, SMTP
374 responses being sent by the server could be lost. This resulted in
375 dropped connections and sometimes bounces generated by a peer sending
378 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
379 string. Previously it was possible to cause undefined behaviour in a
380 library routine (usually a crash). Found by "zerons".
382 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
385 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
386 API was removed, so update to use the newer ones.
388 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
389 any timeout set, is taking a long time. Previously we would hang on to a
390 rotated logfile "forever" if the input was arriving with long gaps
391 (a previous attempt to fix addressed lack, for a long time, of initial
394 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
395 shared (NFS) environment. The length of the tempfile name is now
396 4 + 16 ("hdr.$message_exim_id") which might break on file
397 systems which restrict the file name length to lower values.
400 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
401 shared (NFS) environment.
403 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
404 did for all versions <4.90). Notably -M, -m, --invert, -I may be
407 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
408 on some platforms for bit 31.
410 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
411 to changes apparently associated with TLS1.3 handling some of the APIs
412 previously used were either nonfunctional or inappropriate. Strings
413 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
414 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
415 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
416 This affects log line X= elements, the $tls_{in,out}_cipher variables,
417 and the use of specific cipher names in the encrypted= ACL condition.
419 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
421 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
422 verification result was not updated unless hosts_require_ocsp applied.
424 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
425 queue_list_requires_admin set to false, non-admin users were denied the
428 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
429 directory-of-certs mode. Previously they were advertised despite the
432 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
433 A single TCP connection by a client will now hold a TLS connection open
434 for multiple message deliveries, by default. Previously the default was to
437 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
438 default. If built with the facility, DANE will be used. The facility
439 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
441 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
442 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
443 must be defined and you must still, unless you define DISABLE_TLS, manage
444 the the include-dir and library-file requirements that go with that
445 choice. Non-TLS builds are still supported.
447 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
450 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
451 default. If the platform supports and has the facility enabled, it will
452 be requested on all coneections.
454 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
455 controlled by the build-time option SUPPORT_PIPE_CONNECT.
457 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
459 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
460 success-DSN messages. Previously the From: header was always the default
461 one for these; the option was ignored.
463 JH/28 Fix the timeout on smtp response to apply to the whole response.
464 Previously it was reset for every read, so a teergrubing peer sending
465 single bytes within the time limit could extend the connection for a
466 long time. Credit to Qualsys Security Advisory Team for the discovery.
468 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
469 delivery address, which leaked information of the results of local
470 forwarding. Change to the original envelope recipient address, per
473 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
474 requested. Previously not bounce was generated and a log entry of
475 error ignored was made.
477 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
479 JH/32 Introduce a general tainting mechanism for values read from the input
480 channel, and values derived from them. Refuse to expand any tainted
481 values, to catch one form of exploit.
483 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
484 was unused and the unexpanded text used for the test. Found and
485 fixed by Ruben Jenster.
487 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
488 an attempt to use a TLS library read routine dereffed a nul pointer,
491 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
492 them in our smtp response.
494 JH/36 Have the general router option retry_use_local_part default to true when
495 any of the restrictive preconditions are set (to anything). Previously it
496 was only for check_local user. The change removes one item of manual
497 configuration which is required for proper retries when a remote router
498 handles a subset of addresses for a domain.
500 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
501 link count into consideration.
503 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
504 caused the extension of big_buffer, the following lines were ignored.
506 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
507 accordance with RFC 2308. Previously there was no expiry, so a longlived
508 receive process (eg. due to ACL delays) versus a short SOA value could
511 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
513 JH/39 Promote DMARC support to mainline.
515 JH/40 Bug 2452: Add a References: header to DSNs.
517 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
518 parameters. The relevant library call is documented as "Deprecated: This
519 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
520 3.6.0, DH parameters are negotiated following RFC7919."
522 HS/06 Change the default of dnssec_request_domains to "*"
524 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
525 carried on and emitted a BDAT command, even when PIPELINING was not
528 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
529 buffer was used for the filename, resulting in a trap when tainted
530 arguments (eg. $domain) were used.
532 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
533 recommended to avoid a possible server-load attack. The feature can be
534 re-enabled via the openssl_options main cofiguration option.
536 JH/45 local_scan API: documented the current smtp_printf() call. This changed
537 for version 4.90 - adding a "more data" boolean to the arguments.
538 Bumped the ABI version number also, this having been missed previously;
539 release versions 4.90 to 4.92.3 inclusive were effectively broken in
540 respect of usage of smtp_printf() by either local_scan code or libraries
541 accessed via the ${dlfunc } expansion item. Both will need coding
542 adjustment for any calls to smtp_printf() to match the new function
543 signature; a FALSE value for the new argument is always safe.
545 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
546 the file-offset (which the Linux syscall does, and exim expects); this
547 resulted in an indefinite loop.
549 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
550 to do ARC verification. The Authentication-Results: header line added
551 by the configuration then had no ARC item.
557 JH/01 Remove code calling the customisable local_scan function, unless a new
558 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
560 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
561 non-signal-safe functions being used.
563 JH/03 Bug 2269: When presented with a received message having a stupidly large
564 number of DKIM-Signature headers, disable DKIM verification to avoid
565 a resource-consumption attack. The limit is set at twenty.
567 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
568 report of oldest_pass in ${authres } in consequence, and separate out
569 some descriptions of reasons for verification fail.
571 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
572 files in the spool were present and unlocked. A queue-runner could spot
573 them, resulting in a duplicate delivery. Fix that by doing the unlock
574 after the unlink. Investigation by Tim Stewart. Take the opportunity to
575 add more error-checking on spoolfile handling while that code is being
578 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
579 No known attacks, no CVE, this is defensive hardening.
581 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
582 a queue-runner could start a delivery while other operations were ongoing.
583 Cutthrough delivery was a common victim, resulting in duplicate delivery.
584 Found and investigated by Tim Stewart. Fix by using the open message data
585 file handle rather than opening another, and not locally closing it (which
586 releases a lock) for that case, while creating the temporary .eml format
587 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
589 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
590 $sender_verify_failure/$recipient_verify_failure to "random".
592 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
595 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
596 Previously this would segfault.
598 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
601 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
602 like zero, since the resolver should be doing this for us, But we need one
603 as a CNAME but no MX presence gets the CNAME returned; we need to check
604 that doesn't point to an MX to declare it "no MX returned" rather than
605 "error, loop". A new main option is added so the older capability of
606 following some limited number of chain links is maintained.
608 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
610 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
611 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
612 now-deprecated earlier definitions used only specified the range up to TLS
613 1.2 (in the older-version library docs).
615 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
617 JH/15 Rework TLS client-side context management. Stop using a global, and
618 explicitly pass a context around. This enables future use of TLS for
619 connections to service-daemons (eg. malware scanning) while a client smtp
620 connection is using TLS; with cutthrough connections this is quite likely.
622 JH/16 Fix ARC verification to do AS checks in reverse order.
624 JH/17 Support a "tls" option on the ${readsocket } expansion item.
626 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
627 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
628 Previously the "utf8" would be re-prepended for every additional message.
630 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
631 Previously thery were accepted, resulting in issues when attempting to
632 forward messages to a non-supporting MTA.
634 PP/02 Let -n work with printing macros too, not just options.
636 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
637 one parent address was copied, and bogus data was used at delivery-logging
638 time. Either a crash (after delivery) or bogus log data could result.
639 Discovery and analysis by Tim Stewart.
641 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
642 Previously if the string ended mid-character, we did not insert the
643 promised '?' replacement.
645 PP/04 Documentation: current string operators work on bytes, not codepoints.
647 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
648 should pack well giving a smaller memory footprint so better caching and
649 therefore performance. Group the declarations where this can't be done so
650 that the byte-sized flag variables are not interspersed among pointer
651 variables, giving a better chance of good packing by the compiler.
653 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
654 non-null, to avoid issues with sites running BATV. Previously reports were
655 sent with an empty envelope sender so looked like bounces.
657 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
658 The ignore_error flag wasn't being returned from the filter subprocess so
659 was not set for later routers. Investigation and fix by Matthias Kurz.
661 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
662 and a msg:complete for the whole, when a message is manually removed using
663 -Mrm. Developement by Matthias Kurz, hacked on by JH.
665 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
666 a "Gnu special" function, asprintf() in the DB utility binary builds; I
667 hope that is portable enough.
669 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
670 requiring a known-CA anchor certificate; make it now rely entirely on the
671 TLSA as an anchor. Checking the name on the leaf cert against the name
672 on the A-record for the host is still done for TA (but not for EE mode).
674 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
675 included in delivery lines for non-proxied connections, when compiled with
676 SUPPORT_SOCKS and running with proxy logging enabled.
678 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
679 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
680 move the existing event to fire before the normal logging of message
681 failure so that custom logging is bracketed by normal logging.
683 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
684 msg:fail:internal event. Developement by Matthias Kurz.
686 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
687 far too small for todays use of crypto signatures stored there. Go all
688 the way to the max DNS message size of 64kB, even though this might be
689 overmuch for IOT constrained device use.
691 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
692 copy a string over itself. The library routine is documented as not
693 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
695 JH/32 For main options check_spool_space and check_inode_space, where the
696 platform supports 64b integers, support more than the previous 2^31 kB
697 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
698 the previous G, M, k.
700 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
701 $authenticated_fail_id variable on authentication failure. Previously
704 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
705 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
706 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
707 GNUTLS_SEC_PARAM_MEDIUM.
709 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
710 side. Previously we would continue as if no SNI had been received.
712 JH/36 Harden the handling of string-lists. When a list consisted of a sole
713 "<" character, which should be a list-separator specification, we walked
714 off past the nul-terimation.
716 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
717 causes) even when the retry time is not yet met. Previously they were
718 not, meaning that when (say) an account was over-quota and temp-rejecting,
719 and multiple senders' messages were queued, only one sender would get
720 notified on each configured delay_warning cycle.
722 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
724 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
725 cipher-suites, an error can be left on the stack even for a succeeding
726 accept; this results in impossible error messages when a later operation
729 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
730 return error codes indicating retry. Under TLS1.3 this becomes required.
732 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
733 it only wrote the new authenticators, resulting in a lack of tracking of
734 peer changes of ESMTP extensions until the next cache flush.
736 JH/41 Fix the loop reading a message header line to check for integer overflow,
737 and more-often against header_maxsize. Previously a crafted message could
738 induce a crash of the recive process; now the message is cleanly rejected.
740 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
741 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
747 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
748 When redis_servers is set to a list of > 1 element, and the Redis servers
749 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
750 case of MOVED into a DEFER case instead, thus moving the query onto the
751 next server in the list. For a cluster of N elements, all N servers must
752 be defined in redis_servers.
754 GF/02 Catch and remove uninitialized value warning in exiqsumm
755 Check for existence of @ARGV before looking at $ARGV[0]
757 JH/01 Replace the store_release() internal interface with store_newblock(),
758 which internalises the check required to safely use the old one, plus
759 the allocate and data copy operations duplicated in both (!) of the
760 extant use locations.
762 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
763 modifier. This matches the restriction on the commandline.
765 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
766 Previously only the last row was returned.
768 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
769 we assumed that tags in the header were well-formed, and parsed the
770 element content after inspecting only the first char of the tag.
771 Assumptions at that stage could crash the receive process on malformed
774 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
775 While running the DKIM ACL we operate on the Permanent memory pool so that
776 variables created with "set" persist to the DATA ACL. Also (at any time)
777 DNS lookups that fail create cache records using the Permanent pool. But
778 expansions release any allocations made on the current pool - so a dnsdb
779 lookup expansion done in the DKIM ACL releases the memory used for the
780 DNS negative-cache, and bad things result. Solution is to switch to the
781 Main pool for expansions.
782 While we're in that code, add checks on the DNS cache during store_reset,
783 active in the testsuite.
784 Problem spotted, and debugging aided, by Wolfgang Breyha.
786 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
787 When none of the hosts presented to a transport match an already-open
788 connection, close it and proceed with the list. Previously we would
789 queue the message. Spotted by Lena with Yahoo, probably involving
792 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
793 Previously a spurious "250 OK id=" response was appended to the proper
796 JH/08 The "support for" informational output now, which built with Content
797 Scanning support, has a line for the malware scanner interfaces compiled
798 in. Interface can be individually included or not at build time.
800 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
801 by the template makefile "src/EDITME". The "STREAM" support for an older
802 ClamAV interface method is removed.
804 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
805 rows affected is given instead).
807 JH/11 The runtime Berkeley DB library version is now additionally output by
808 "exim -d -bV". Previously only the compile-time version was shown.
810 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
811 SMTP connection. Previously, when one had more receipients than the
812 first, an abortive onward connection was made. Move to full support for
813 multiple onward connections in sequence, handling cutthrough connection
814 for all multi-message initiating connections.
816 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
817 routers. Previously, a multi-recipient message would fail to match the
818 onward-connection opened for the first recipient, and cause its closure.
820 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
821 a timeout on read on a GnuTLS initiating connection, resulting in the
822 initiating connection being dropped. This mattered most when the callout
823 was marked defer_ok. Fix to keep the two timeout-detection methods
826 JH/15 Relax results from ACL control request to enable cutthrough, in
827 unsupported situations, from error to silently (except under debug)
828 ignoring. This covers use with PRDR, frozen messages, queue-only and
831 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
833 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
834 metadata, resulting in a crash in free().
836 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
837 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
838 Broken also in d185889f4, with init system revamp.
840 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
841 Previously we abruptly closed the connection after reading a malware-
842 found indication; now we go on to read the "scan ok" response line,
845 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
846 ACL. Previously, a crash would result.
848 JH/19 Speed up macro lookups during configuration file read, by skipping non-
849 macro text after a replacement (previously it was only once per line) and
850 by skipping builtin macros when searching for an uppercase lead character.
852 JH/20 DANE support moved from Experimental to mainline. The Makefile control
853 for the build is renamed.
855 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
856 was allocated for every new TLS startup, meaning one per message. Fix
857 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
859 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
860 reported the original. Fix to report (as far as possible) the ACL
861 result replacing the original.
863 JH/23 Fix memory leak during multi-message connections using STARTTLS under
864 OpenSSL. Certificate information is loaded for every new TLS startup,
865 and the resources needed to be freed.
867 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
869 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
872 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
873 DATA response info to the (existing) per-recipient response info for
874 the "C=" log element. It can have useful tracking info from the
875 destination system. Patch from Simon Arlott.
877 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
878 length value. Previously this would segfault.
880 HS/02 Support Avast multiline protoocol, this allows passing flags to
881 newer versions of the scanner.
883 JH/28 Ensure that variables possibly set during message acceptance are marked
884 dead before release of memory in the daemon loop. This stops complaints
885 about them when the debug_store option is enabled. Discovered specifically
886 for sender_rate_period, but applies to a whole set of variables.
887 Do the same for the queue-runner and queue-list loops, for variables set
888 from spool message files. Do the same for the SMTP per-message loop, for
889 certain variables indirectly set in ACL operations.
891 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
892 as a multi-recipient message from a mailinglist manager). The coding had
893 an arbitrary cutoff number of characters while checking for more input;
894 enforced by writing a NUL into the buffer. This corrupted long / fast
895 input. The problem was exposed more widely when more pipelineing of SMTP
896 responses was introduced, and one Exim system was feeding another.
897 The symptom is log complaints of SMTP syntax error (NUL chars) on the
898 receiving system, and refused recipients seen by the sending system
899 (propating to people being dropped from mailing lists).
900 Discovered and pinpointed by David Carter.
902 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
903 replaced by the ${authresults } expansion.
905 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
907 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
908 allows proper process termination in container environments.
910 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
911 Previously the "final dot" had a newline after it; ensure it is CR,LF.
913 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
914 and "err_perm", deprecated since 4.83 when the RFC-defined words
915 "temperror" and "permerror" were introduced.
917 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
918 transport-filters or DKIM-signing. The restriction was lost in the
919 consolidation of verify-callout and delivery SMTP handling.
920 Extend the restriction to also cover ARC-signing.
922 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
923 in defer=pass mode supply a 450 to the initiator. Previously the message
926 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
927 tls_require_ciphers is used as before.
929 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
930 "pass_unscanned". Only tmpfails from the scanner are written to
931 the paniclog, as they may require admin intervention (permission
932 denied, license issues). Other scanner errors (like decompression
933 bombs) do not cause a paniclog entry.
935 JH/36 Fix reinitialisation of DKIM logging variable between messages.
936 Previously it was possible to log spurious information in receive log
939 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
940 triggered odd behaviour from Outlook Express clients.
942 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
945 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
946 since the IETF WG has not yet settled on that versus the original
947 "bare" representation.
949 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
950 Previously the millisecond value corrupted the output.
951 Fix also for syslog_pid=no and log_selector +pid, for which the pid
952 corrupted the output.
958 JH/01 Rework error string handling in TLS interface so that the caller in
959 more cases is responsible for logging. This permits library-sourced
960 string to be attached to addresses during delivery, and collapses
961 pairs of long lines into single ones.
963 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
964 during configuration. Wildcards are allowed and expanded.
966 JH/02 Rework error string handling in DKIM to pass more info back to callers.
967 This permits better logging.
969 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
970 do not close it down and have the child transport start it up again on
971 the passed-on TCP connection. Instead, proxy the child (and any
972 subsequent ones) for TLS via a unix-domain socket channel. Logging is
973 affected: the continued delivery log lines do not have any DNSSEC, TLS
974 Certificate or OCSP information. TLS cipher information is still logged.
976 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
977 identical IP addresses on different listening ports. Will also affect
980 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
981 add noisy ifdef guards to special-case this sillyness.
982 Patch from Bernd Kuhls.
984 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
985 than 255 are no longer allowed.
987 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
988 Disable the session-cache too, which might reduce our load. Since we
989 currrectly use a new context for every connection, both as server and
990 client, there is no benefit for these.
991 GnuTLS appears to not support tickets server-side by default (we don't
992 call gnutls_session_ticket_enable_server()) but client side is enabled
993 by default on recent versions (3.1.3 +) unless the PFS priority string
996 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
997 <https://reproducible-builds.org/specs/source-date-epoch/>.
999 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1000 the check for any unsuccessful recipients did not notice the limit, and
1001 erroneously found still-pending ones.
1003 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1004 MSG_MORE. Only in-clear (not on TLS connections).
1006 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1007 transport-filter is involved we can buffer the headers in memory for
1008 creating the signature, and read the spool data file once for the
1009 signature and again for transmission.
1011 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1012 4.77 as the kernel support then wasn't solid, having issues in 64bit
1013 mode. Now, it's been long enough. Add support for FreeBSD also.
1015 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1016 case where the routing stage had gathered several addresses to send to
1017 a host before calling the transport for the first, we previously failed
1018 to close down TLS in the old transport process before passing the TCP
1019 connection to the new process. The new one sent a STARTTLS command
1020 which naturally failed, giving a failed delivery and bloating the retry
1021 database. Investigation and fix prototype from Wolfgang Breyha.
1023 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1024 false-negatives in the check that the sender had not preempted a response
1025 or prompt from Exim (running as a server), due to that code's lack of
1026 awareness of the SMTP input buffering.
1028 PP/04 Add commandline_checks_require_admin option.
1029 Exim drops privileges sanely, various checks such as -be aren't a
1030 security problem, as long as you trust local users with access to their
1031 own account. When invoked by services which pass untrusted data to
1032 Exim, this might be an issue. Set this option in main configuration
1033 AND make fixes to the calling application, such as using `--` to stop
1036 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1037 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1038 responses to those, into a single TLS record each way (this usually means
1039 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1042 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1043 affects you only if you're dancing at the edge of the param size limits.
1044 If you are, and this message makes sense to you, then: raise the
1045 configured limit or use OpenSSL 1.1. Nothing we can do for older
1048 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1049 cmdline element to get the documented default one. Previously it was
1052 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1053 are unsuable for non-message-specific reasons - eg. connection timeout,
1054 banner-time rejection.
1056 JH/16 Fix logging of delivery remote port, when specified by router, under
1059 PP/06 Repair manualroute's ability to take options in any order, even if one
1060 is the name of a transport.
1063 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1065 JH/17 Change the list-building routines interface to use the expanding-string
1066 triplet model, for better allocation and copying behaviour.
1068 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1069 Previously it was constructed the first time a possibly-matching string
1070 was met in the configuration file input during startup; now it is done
1073 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1074 compatible one, to avoid the (poorly documented) possibility of a config
1075 file in the working directory redirecting the DB files, possibly correpting
1076 some existing file. CVE-2017-10140 assigned for BDB.
1078 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1079 cache-hot. Previously, although the result was properly cached, the
1080 initial verify call returned a defer.
1082 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1083 the main verify for receipient in uncached-mode.
1085 JH/22 Retire historical build files to an "unsupported" subdir. These are
1086 defined as "ones for which we have no current evidence of testing".
1088 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1089 if present. Previously it was ignored.
1091 JH/24 Start using specified-initialisers in C structure init coding. This is
1092 a C99 feature (it's 2017, so now considered safe).
1094 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1095 if was a fixed-sized field and bitmask ops via macros; it is now more
1098 PP/07 GitHub PR 56: Apply MariaDB build fix.
1099 Patch provided by Jaroslav Škarvada.
1101 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1102 during Coverity cleanups [4.87 JH/47]
1103 Diagnosis and fix provided by Michael Fischer v. Mollard.
1105 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1106 the right size to place the terminating semicolon on its own folded
1107 line, the header hash was calculated to an incorrect value thanks to
1108 the (relaxed) space the fold became.
1110 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1111 and confused the parent.
1113 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1114 which could crash as a result. This could lead to undeliverable messages.
1116 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1119 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1120 matters on fast-turnover and PID-randomising systems, which were getting
1121 out-of-order delivery.
1123 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1124 a possibly-overlapping copy. The symptom was that "Remote host closed
1125 connection in response to HELO" was logged instead of the actual 4xx
1128 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1129 Previously only that bufferd was discarded, resulting in SYMTP command
1132 JH/32 DKIM: when a message has multiple signatures matching an identity given
1133 in dkim_verify_signers, run the dkim acl once for each. Previously only
1134 one run was done. Bug 2189.
1136 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1137 "panic the current process" to "deliberately defer". The panic log is
1138 still written with the problem list name; the mail and reject logs now
1139 get a temp-reject line for the message that was being handled, saying
1140 something like "domains check lookup or other defer". The SMTP 451
1141 message is still "Temporary local problem".
1143 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1144 A crafted sequence of BDAT commands could result in in-use memory beeing
1145 freed. CVE-2017-16943.
1147 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1148 from SMTP input. Previously it was always done; now only done for DATA
1149 and not BDAT commands. CVE-2017-16944.
1151 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1152 to the message (such as an overlong header line). Previously this was
1153 not done and we did not exit BDAT mode. Followon from the previous item
1154 though a different problem.
1160 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1161 than -2003 did; needs libidn2 in addition to libidn.
1163 JH/02 The path option on a pipe transport is now expanded before use.
1165 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1166 Patch provided by "Björn", documentation fix added too.
1168 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1169 missing a wire-to-host endian conversion.
1171 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1172 close after a BDAT command line could be taken as a following command,
1173 giving a synch failure. Fix by only checking for synch immediately
1174 before acknowledging the chunk.
1176 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1177 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1178 macro. Patches provided by Josh Soref.
1180 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1181 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1182 by IANA as a service extension. However, John Klensin suggests that we
1185 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1186 the dkim code may be unix-mode line endings rather than smtp wire-format
1187 CRLF, so prepend a CR to any bare LF.
1189 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1190 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1192 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1193 the first were themselves being wrongly included in the feed into dkim
1194 processing; with most chunk sizes in use this resulted in an incorrect
1195 body hash calculated value.
1197 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1198 DKIM signature block, for verification. Although advised against by
1199 standards it is specifically not ruled illegal.
1201 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1203 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1204 missing a body hash (the bh= tag).
1206 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1207 It seems that HAProxy sends the Proxy Protocol information in clear and
1208 only then does a TLS startup, so do the same.
1210 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1211 TCP connections (such as for Spamd) unless the daemon successfully set
1212 Fast Open mode on its listening sockets. This fixes breakage seen on
1213 too-old kernels or those not configured for Fast Open, at the cost of
1214 requiring both directions being enabled for TFO, and TFO never being used
1215 by non-daemon-related Exim processes.
1217 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1218 endings, at least on the first header line. Try to canonify any that get
1219 past that check, despite the cost.
1221 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1222 now limited to an arbitrary five deep, while parsing addresses with the
1223 strip_excess_angle_brackets option enabled.
1225 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1226 instead leave the unprompted TLS handshake in socket buffer for the
1227 TLS library to consume.
1229 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1231 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1233 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1234 region in one operation, for speed, and this leaves assigned pointers
1235 dangling. Add checks run only under the testsuite which checks all
1236 variables at a store-reset and panics on a dangling pointer; add code
1237 explicitly nulling out all the variables discovered. Fixes one known
1238 bug: a transport crash, where a dangling pointer for $sending_ip_address
1239 originally assigned in a verify callout, is re-used.
1241 PP/06 Drop '.' from @INC in various Perl scripts.
1243 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1245 PP/08 Reduce a number of compilation warnings under clang; building with
1246 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1247 should be warning-free.
1249 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1251 HS/01 Fix portability problems introduced by PP/08 for platforms where
1252 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1254 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1255 chunk. This allows us to accept broken chunked messages. We need a more
1256 general solution here.
1258 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1259 already-broken messages in the queue.
1261 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1263 JH/19 Fix reference counting bug in routing-generated-address tracking.
1269 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1270 supports it and a size is available (ie. the sending peer gave us one).
1272 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1273 years of being deprecated). The replacements are the ACLs
1274 acl_smtp_mime and acl_not_smtp_mime.
1276 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1277 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1278 or even an in-clear connection were permitted. Now, if the host lookup
1279 was dnssec and dane was requested then the host is only used if the
1280 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1281 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1282 if one fails this test.
1283 This means that a poorly-configured remote DNS will make it incommunicado;
1284 but it protects against a DNS-interception attack on it.
1286 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1287 non-noisy when a race steals the message being considered.
1289 JH/05 If main configuration option tls_certificate is unset, generate a
1290 self-signed certificate for inbound TLS connections.
1292 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1293 in rewrites and routers.
1295 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1296 and logged a warning sing 4.83; now they are a configuration file error.
1298 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1299 (lacking @domain). Apply the same qualification processing as RCPT.
1301 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1303 JH/10 Support ${sha256:} applied to a string (as well as the previous
1306 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1307 a cutthrough deliver is pending, as we always want to make a connection.
1308 This also avoids re-routing the message when later placing the cutthrough
1309 connection after a verify cache hit.
1310 Do not update it with the verify result either.
1312 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1313 when routing results in more than one destination address.
1315 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1316 signing (which inhibits the cutthrough capability). Previously only
1317 the presence of an option was tested; now an expansion evaluating as
1318 empty is permissible (obviously it should depend only on data available
1319 when the cutthrough connection is made).
1321 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1322 the relevant preceding SMTP command did not note the pipelining mode.
1324 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1325 Previously they were not counted.
1327 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1328 as one having no matching records. Previously we deferred the message
1329 that needed the lookup.
1331 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1332 distinguished as "(=".
1334 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1335 for missing MX records. Previously it only worked for missing A records.
1337 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1339 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1340 after the data-go-ahead and data-ack. Patch from Jason Betts.
1342 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1343 even for a "none" policy. Patch from Tony Meyer.
1345 JH/22 Fix continued use of a connection for further deliveries. If a port was
1346 specified by a router, it must also match for the delivery to be
1349 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1350 When one of the recipients of a message was unsuitable for the connection
1351 (has no matching addresses), we lost track of needing to mark it
1352 deferred. As a result mail would be lost.
1354 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1356 JH/25 Decoding ACL controls is now done using a binary search; the source code
1357 takes up less space and should be simpler to maintain. Merge the ACL
1358 condition decode tables also, with similar effect.
1360 JH/26 Fix problem with one_time used on a redirect router which returned the
1361 parent address unchanged. A retry would see the parent address marked as
1362 delivered, so not attempt the (identical) child. As a result mail would
1365 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1366 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1367 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1370 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1371 Main config options check_{log,spool}_{inodes,space} are now
1372 100 inodes, 10MB unless set otherwise in the configuration.
1374 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1375 Previously it only applied to the main-section connection policy
1378 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1380 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1381 by me. Added RFC7919 DH primes as an alternative.
1383 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1384 are not in the system include path.
1386 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1387 GnuTLS, when a session startup failed (eg because the client disconnected)
1388 Exim did stdio operations after fclose. This was exposed by a recent
1389 change which nulled out the file handle after the fclose.
1391 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1392 signed directly by the cert-signing cert, rather than an intermediate
1393 OCSP-signing cert. This is the model used by LetsEncrypt.
1395 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1397 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1398 an incoming connection.
1400 HS/02 Bug 1802: Do not half-close the connection after sending a request
1403 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1404 fallback to "prime256v1".
1406 JH/34 SECURITY: Use proper copy of DATA command in error message.
1407 Could leak key material. Remotely exploitable. CVE-2016-9963.
1413 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1414 and 3.4.4 - once the server is enabled to respond to an OCSP request
1415 it does even when not requested, resulting in a stapling non-aware
1416 client dropping the TLS connection.
1418 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1419 support variable-length bit vectors. No functional change.
1421 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1422 The I= interface field on outgoing lines is now after the H= remote
1423 host field, same as incoming lines. There is a separate
1424 outgoing_interface log selector which allows you to disable the
1427 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1428 If not running log_selector +smtp_connection the mainlog would be held
1429 open indefinitely after a "too many connections" event, including to a
1430 deleted file after a log rotate. Leave the per net connection logging
1431 leaving it open for efficiency as that will be quickly detected by the
1432 check on the next write.
1434 HS/01 Bug 1671: Fix post transport crash.
1435 Processing the wait-<transport> messages could crash the delivery
1436 process if the message IDs didn't exist for some reason. When
1437 using 'split_spool_directory=yes' the construction of the spool
1438 file name failed already, exposing the same netto behaviour.
1440 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1441 mime_regex ACL conditions.
1443 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1444 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1445 to HELO, local diagnostic string.
1447 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1448 log line to debug. Even when configured with a tls authenticator many
1449 client connections are expected to not authenticate in this way, so
1450 an authenticate fail is not an error.
1452 HS/02 Add the Exim version string to the process info. This way exiwhat
1453 gives some more detail about the running daemon.
1455 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1456 matter for fast-change records such as DNSBLs.
1458 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1459 retry record, even if constant. There may be multiple transports with
1460 different interface settings and the retry behaviour needs to be kept
1463 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1465 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1467 JH/10 Bug 840: fix log_defer_output option of pipe transport
1469 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1470 hosts_randomize. This matters a lot when combined with mua_wrapper.
1472 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1473 ${quote_pgsql:<string>} operator.
1475 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1477 JH/14 Change header file name for freeradius-client. Relevant if compiling
1478 with Radius support; from the Gentoo tree and checked under Fedora.
1480 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1482 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1483 option transport_filter, meaning no filtering.
1485 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1487 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1488 defaults to "*" (all hosts). The variable is now available when not built
1489 with TLS, default unset, mainly to enable keeping the testsuite sane.
1490 If a server certificate is not supplied (via tls_certificate) an error is
1491 logged, and clients will find TLS connections fail on startup. Presumably
1492 they will retry in-clear.
1493 Packagers of Exim are strongly encouraged to create a server certificate
1494 at installation time.
1496 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1497 with the $config_file variable.
1499 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1500 in transport context, after the attempt, and per-recipient. The latter type
1501 is per host attempted. The event data is the error message, and the errno
1502 information encodes the lookup type (A vs. MX) used for the (first) host,
1503 and the trailing two digits of the smtp 4xx response.
1505 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1506 to write to mainlog (or rejectlog, paniclog) in the window between file
1507 creation and permissions/ownership being changed. Particularly affects
1508 installations where exicyclog is run as root, rather than exim user;
1509 result is that the running daemon panics and dies.
1511 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1513 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1514 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1515 "pri" and "weight". Note that the previous implicit priority given by the
1516 list order is no longer honoured.
1518 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1519 for DKIM processing.
1521 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1522 by defining SUPPORT_SOCKS.
1524 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1525 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1526 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1527 variables are renamed to proxy_{local,external}_{address,port}.
1529 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1530 for a build by defining SUPPORT_I18N
1532 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1533 of the query string, and make ${quote_redis:} do that quoting.
1535 JH/29 Move Events support from Experimental to mainline, enabled by default
1536 and removable for a build by defining DISABLE_EVENT.
1538 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1540 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1541 cached by the daemon.
1543 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1544 by defining LOOKUP_REDIS. The libhiredis library is required.
1546 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1547 keys are given for lookup.
1549 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1550 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1551 only supported when built with TLS support. The PolarSSL SHA routines
1552 are still used when the TLS library is too old for convenient support.
1554 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1555 openssl_options), for security. OpenSSL forces this from version 1.1.0
1556 server-side so match that on older versions.
1558 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1559 allocation for $value could be released as the expansion processing
1560 concluded, but leaving the global pointer active for it.
1562 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1563 and to use the domains and local_parts ACL conditions.
1565 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1566 incorrectly not doubled on cutthrough transmission, hence seen as a
1567 body-termination at the receiving system - resulting in truncated mails.
1568 Commonly the sender saw a TCP-level error, and retransmitted the message
1569 via the normal store-and-forward channel. This could result in duplicates
1570 received - but deduplicating mailstores were liable to retain only the
1571 initial truncated version.
1573 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1575 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1577 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1578 we're in there, support oversigning also; bug 1309.
1580 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1582 HS/04 Add support for keep_environment and add_environment options.
1584 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1585 either intentional arithmetic overflow during PRNG, or testing config-
1588 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1589 delivery resulted in actual delivery. Cancel cutthrough before DATA
1592 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1593 rejects a recipient: pass the reject to the originator.
1595 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1596 Many were false-positives and ignorable, but it's worth fixing the
1599 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1600 for the new environment-manipulation done at startup. Move the routines
1601 from being local to tls.c to being global via the os.c file.
1603 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1604 an extract embedded as result-arg for a map, the first arg for extract
1605 is unavailable so we cannot tell if this is a numbered or keyed
1606 extraction. Accept either.
1612 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1615 JH/02 The smtp transport option "multi_domain" is now expanded.
1617 JH/03 The smtp transport now requests PRDR by default, if the server offers
1620 JH/04 Certificate name checking on server certificates, when exim is a client,
1621 is now done by default. The transport option tls_verify_cert_hostnames
1622 can be used to disable this per-host. The build option
1623 EXPERIMENTAL_CERTNAMES is withdrawn.
1625 JH/05 The value of the tls_verify_certificates smtp transport and main options
1626 default to the word "system" to access the system default CA bundle.
1627 For GnuTLS, only version 3.0.20 or later.
1629 JH/06 Verification of the server certificate for a TLS connection is now tried
1630 (but not required) by default. The verification status is now logged by
1631 default, for both outbound TLS and client-certificate supplying inbound
1634 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1637 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1638 Status Notification (bounce) messages are now MIME format per RFC 3464.
1639 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1640 under the control of the dsn_advertise_hosts option, and routers may
1641 have a dsn_lasthop option.
1643 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1644 default, modifiable by a malware= option. The list separator for
1645 the options can now be changed in the usual way. Bug 68.
1647 JH/10 The smtp_receive_timeout main option is now expanded before use.
1649 JH/11 The incoming_interface log option now also enables logging of the
1650 local interface on delivery outgoing connections.
1652 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1653 if the interface and destination host and port all match.
1655 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1658 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1659 Patch from Andrew Lewis.
1661 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1662 now supports optional time-restrictions, weighting, and priority
1663 modifiers per server. Patch originally by <rommer@active.by>.
1665 JH/16 The spamd_address main option now supports a mixed list of local
1666 and remote servers. Remote servers can be IPv6 addresses, and
1667 specify a port-range.
1669 JH/17 Bug 68: The spamd_address main option now supports an optional
1670 timeout value per server.
1672 JH/18 Bug 1581: Router and transport options headers_add/remove can
1673 now have the list separator specified.
1675 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1678 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1681 JH/21 Support for the A6 type of dns record is withdrawn.
1683 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1684 rather than the verbs used.
1686 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1687 from 255 to 1024 chars.
1689 JH/24 Verification callouts now attempt to use TLS by default.
1691 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1692 are generic router options now. The defaults didn't change.
1694 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1695 Original patch from Alexander Shikoff, worked over by JH.
1697 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1698 files if ZCAT_COMMAND is not executable.
1700 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1702 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1704 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1705 Normally benign, it bites when the pair was led to by a CNAME;
1706 modern usage is to not canonicalize the domain to a CNAME target
1707 (and we were inconsistent anyway for A-only vs AAAA+A).
1709 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1711 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1712 when evaluating $sender_host_dnssec.
1714 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1715 $sender_helo_dnssec variable.
1717 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1719 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1721 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1723 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1724 documented as working, but never had. Support all but $spam_report.
1726 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1727 added for tls authenticator.
1729 HS/03 Add perl_taintmode main config option
1735 TL/01 When running the test suite, the README says that variables such as
1736 no_msglog_check are global and can be placed anywhere in a specific
1737 test's script, however it was observed that placement needed to be near
1738 the beginning for it to behave that way. Changed the runtest perl
1739 script to read through the entire script once to detect and set these
1740 variables, reset to the beginning of the script, and then run through
1741 the script parsing/test process like normal.
1743 TL/02 The BSD's have an arc4random API. One of the functions to induce
1744 adding randomness was arc4random_stir(), but it has been removed in
1745 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1746 function when detected.
1748 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1749 cause callback expansion.
1751 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1752 syntax errors in an expansion can be treated as a string instead of
1753 logging or causing an error, due to the internal use of bool_lax
1754 instead of bool when processing it.
1756 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1757 server certificates when making smtp deliveries.
1759 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1761 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1763 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1765 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1766 Merged patch from Sebastian Wiedenroth.
1768 JH/05 Fix results-pipe from transport process. Several recipients, combined
1769 with certificate use, exposed issues where response data items split
1770 over buffer boundaries were not parsed properly. This eventually
1771 resulted in duplicates being sent. This issue only became common enough
1772 to notice due to the introduction of connection certificate information,
1773 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1775 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1776 size buffer was used, resulting in syntax errors when an expansion
1779 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1780 version 3.3.6 or later.
1782 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1783 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1784 both become "event_action", the variables become $event_name, $event_data
1785 and $event_defer_errno. There is a new variable $verify_mode, usable in
1786 routers, transports and related events. The tls:cert event is now also
1787 raised for inbound connections, if the main configuration event_action
1790 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1791 early OCSP support, but no stapling (appears to be less than 1.0.0).
1793 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1794 server certificate names available under the smtp transport option
1795 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1798 JH/10 Time-related extraction expansions from certificates now use the main
1799 option "timezone" setting for output formatting, and are consistent
1800 between OpenSSL and GnuTLS compilations. Bug 1541.
1802 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1803 encoded parameter in the incoming message. Bug 1558.
1805 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1806 include certificate info, eximon was claiming there were spoolfile
1809 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1811 JH/14 Log delivery-related information more consistently, using the sequence
1812 "H=<name> [<ip>]" wherever possible.
1814 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1815 are problematic for Debian distribution, omit them from the release
1818 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1820 JH/16 Fix string representation of time values on 64bit time_t architectures.
1823 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1824 output list separator was used.
1829 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1830 checkers that were complaining about end of non-void function with no
1833 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1834 This was a regression introduced in 4.83 by another bugfix.
1836 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1838 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1839 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1845 TF/01 Correctly close the server side of TLS when forking for delivery.
1847 When a message was received over SMTP with TLS, Exim failed to clear up
1848 the incoming connection properly after forking off the child process to
1849 deliver the message. In some situations the subsequent outgoing
1850 delivery connection happened to have the same fd number as the incoming
1851 connection previously had. Exim would try to use TLS and fail, logging
1852 a "Bad file descriptor" error.
1854 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1855 utilities have not been installed.
1857 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1858 temporary space as the ACL may create new global variables.
1860 TL/01 LDAP support uses per connection or global context settings, depending
1861 upon the detected version of the libraries at build time.
1863 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1864 to extract and use the src ip:port in logging and expansions as if it
1865 were a direct connection from the outside internet. PPv2 support was
1866 updated based on HAProxy spec change in May 2014.
1868 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1870 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1871 Properly escape header and check for NULL return.
1873 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1876 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1878 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1879 characters in header names, implemented as a verify condition.
1880 Contributed by Michael Fischer v. Mollard.
1882 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1883 results permerror and temperror. Previous values are deprecated but
1884 still accepted. In a future release, err_perm and err_temp will be
1885 completely removed, which will be a backward incompatibility if the
1886 ACL tests for either of these two old results. Patch contributed by
1887 user bes-internal on the mailing list.
1889 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1891 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1892 selectors, in both main and reject logs.
1894 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1897 JH/07 Add malware type "sock" for talking to simple daemon.
1899 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1901 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1902 routers/transports under cutthrough routing.
1904 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1905 numbers. Touch up "bool" conditional to keep the same definition.
1907 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1909 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1911 JH/12 Expand items in router/transport headers_add or headers_remove lists
1912 individually rather than the list as a whole. Bug 1452.
1914 Required for reasonable handling of multiple headers_ options when
1915 they may be empty; requires that headers_remove items with embedded
1916 colons must have them doubled (or the list-separator changed).
1918 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1919 view the policy declared in the DMARC record. Currently, $dmarc_status
1920 is a combined value of both the record presence and the result of the
1923 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1925 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1926 dnslookup router and the smtp transport (applying to the forward
1929 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1930 of ldap servers used for a specific lookup. Patch provided by Heiko
1933 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1934 New variable $lookup_dnssec_authenticated for observability.
1936 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1937 Patch submitted by Lars Timman.
1939 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1941 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1942 Requires trusted mode and valid format message id, aborts otherwise.
1943 Patch contributed by Heiko Schlichting.
1945 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1946 certextract with support for various fields. Bug 1358.
1948 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1949 is requested by default, modifiable by smtp transport option
1952 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1953 operate on certificate variables to give certificate fingerprints
1954 Also new ${sha256:cert_variable}.
1956 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1958 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1959 Christian Aistleitner.
1961 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1963 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1964 file. Patch from Wolfgang Breyha.
1966 JH/25 Expand the coverage of the delivery $host and $host_address to
1967 client authenticators run in verify callout. Bug 1476.
1969 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1970 align with daemon_smtp_ports. Bug 72.
1972 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1973 support and error reporting did not work properly.
1975 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1976 and is readable. Patch from Andrew Colin Kissa.
1978 TL/14 Enhance documentation of ${run expansion and how it parses the
1979 commandline after expansion, particularly in the case when an
1980 unquoted variable expansion results in an empty value.
1982 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1984 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1985 by renaming away from it. Addresses GH issue 15, reported by
1988 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1990 TL/15 SECURITY: prevent double expansion in math comparison functions
1991 (can expand unsanitized data). Not remotely exploitable.
1998 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2000 PP/02 Make -n do something, by making it not do something.
2001 When combined with -bP, the name of an option is not output.
2003 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2006 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2007 $sender_host_name and config options to manage this, and basic check
2010 PP/05 DSCP support for outbound connections and control modifier for inbound.
2012 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2013 (Only plugin which currently uses this is kerberos4, which nobody should
2014 be using, but we should make it available and other future plugins might
2015 conceivably use it, even though it would break NAT; stuff *should* be
2016 using channel bindings instead).
2018 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2019 name; added for Sendmail compatibility; requires admin caller.
2020 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2021 just ignore it); requires trusted caller.
2022 Also parse but ignore: -Ac -Am -X<logfile>
2025 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2027 TL/02 Add +smtp_confirmation as a default logging option.
2029 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2030 Patch by Magnus Holmgren from 2007-02-20.
2032 TL/04 Bugzilla 1281 - Spec typo.
2033 Bugzilla 1283 - Spec typo.
2034 Bugzilla 1290 - Spec grammar fixes.
2036 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2038 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2040 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2041 mailing list by by Dmitry Isaikin.
2043 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2045 JH/02 Support "G" suffix to numbers in ${if comparisons.
2047 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2049 NM/01 Bugzilla 1197 - Spec typo
2050 Bugzilla 1196 - Spec examples corrections
2052 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2054 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2055 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2058 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2059 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2061 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2062 "acl {{name}{arg}...}", and optional args on acl condition
2065 JH/05 Permit multiple router/transport headers_add/remove lines.
2067 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2069 JH/07 Avoid using a waiting database for a single-message-only transport.
2070 Performance patch from Paul Fisher. Bugzilla 1262.
2072 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2075 JH/09 Add $headers_added variable, with content from use of ACL modifier
2076 add_header (but not yet added to the message). Bugzilla 199.
2078 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2079 Pulled from Bugzilla 817 by Wolfgang Breyha.
2081 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2083 (nb: this is the same fix as in Exim 4.80.1)
2085 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2088 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2091 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2093 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2094 advertise SMTP AUTH mechanism to us, instead of a generic
2095 protocol violation error. Also, make Exim more robust to bad
2096 data from the Dovecot auth socket.
2098 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2100 When a queue runner is handling a message, Exim first routes the
2101 recipient addresses, during which it prunes them based on the retry
2102 hints database. After that it attempts to deliver the message to
2103 any remaining recipients. It then updates the hints database using
2106 So if a recipient address works intermittently, it can get repeatedly
2107 deferred at routing time. The retry hints record remains fresh so the
2108 address never reaches the final cutoff time.
2110 This is a fairly common occurrence when a user is bumping up against
2111 their storage quota. Exim had some logic in its local delivery code
2112 to deal with this. However it did not apply to per-recipient defers
2113 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2115 This change adds a proper retry rule check during routing so that the
2116 final cutoff time is checked against the message's age. We only do
2117 this check if there is an address retry record and there is not a
2118 domain retry record; this implies that previous attempts to handle
2119 the address had the retry_use_local_parts option turned on. We use
2120 this as an approximation for the destination being like a local
2121 delivery, as in LMTP.
2123 I suspect this new check makes the old local delivery cutoff check
2124 redundant, but I have not verified this so I left the code in place.
2126 TF/02 Correct gecos expansion when From: is a prefix of the username.
2128 Test 0254 submits a message to Exim with the header
2132 When I ran the test suite under the user fanf2, Exim expanded
2133 the header to contain my full name, whereas it should have added
2134 a Resent-Sender: header. It erroneously treats any prefix of the
2135 username as equal to the username.
2137 This change corrects that bug.
2139 GF/01 DCC debug and logging tidyup
2140 Error conditions log to paniclog rather than rejectlog.
2141 Debug lines prefixed by "DCC: " to remove any ambiguity.
2143 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2145 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2146 Bug spotted by Jeremy Harris; was flawed since initial commit.
2147 Would have resulted in OCSP responses post-SNI triggering an Exim
2148 NULL dereference and crash.
2150 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2152 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2153 Bug detection, analysis and fix by Samuel Thibault.
2154 Bugzilla 1331, Debian bug #698092.
2156 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2158 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2159 Server implementation by Todd Lyons, client by JH.
2160 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2161 config variable "prdr_enable" controls whether the server
2162 advertises the facility. If the client requests PRDR a new
2163 acl_data_smtp_prdr ACL is called once for each recipient, after
2164 the body content is received and before the acl_smtp_data ACL.
2165 The client is controlled by both of: a hosts_try_prdr option
2166 on the smtp transport, and the server advertisement.
2167 Default client logging of deliveries and rejections involving
2168 PRDR are flagged with the string "PRDR".
2170 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2171 fclose(). Diagnosis by Todd Lyons.
2173 PP/17 Update configure.default to handle IPv6 localhost better.
2174 Patch by Alain Williams (plus minor tweaks).
2177 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2178 This is now consistent with GnuTLS, and is now documented: the
2179 previous undocumented portable approach to treating the option as
2180 unset was to force an expansion failure. That still works, and
2181 an empty string is now equivalent.
2183 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2184 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2185 not performing validation itself.
2187 PP/20 Added force_command boolean option to pipe transport.
2188 Patch from Nick Koston, of cPanel Inc.
2190 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2193 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2195 PP/21 Fix eximon continuous updating with timestamped log-files.
2196 Broken in a format-string cleanup in 4.80, missed when I repaired the
2197 other false fix of the same issue.
2198 Report and fix from Heiko Schlichting.
2201 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2202 Report from Prashanth Katuri.
2204 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2205 It's SecureTransport, so affects any MacOS clients which use the
2206 system-integrated TLS libraries, including email clients.
2208 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2209 using a MIME ACL for non-SMTP local injection.
2210 Report and assistance in diagnosis by Warren Baker.
2212 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2214 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2216 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2217 last id that failed so it may be referenced in subsequent ACL's.
2219 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2222 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2223 ldap library initialization, allowing self-signed CA's to be
2224 used. Also properly sets require_cert option later in code by
2225 using NULL (global ldap config) instead of ldap handle (per
2226 session). Bug diagnosis and testing by alxgomz.
2228 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2229 the src/util/ subdirectory.
2231 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2232 renamed to Transport Post Delivery Action by Jeremy Harris, as
2235 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2236 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2237 redis_servers = needs to be configured which will be used by the redis
2238 lookup. Patch from Warren Baker, of The Packet Hub.
2240 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2242 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2243 hostname or reverse DNS when processing a host list. Used suggestions
2244 from multiple comments on this bug.
2246 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2248 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2249 Missed a few lines, added it to make the runtest require no keyboard
2252 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2253 contains upper case chars. Make router use caseful_local_part.
2255 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2256 support when GnuTLS has been built with p11-kit.
2262 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2264 This, or similar/improved, will also be change PP/11 of 4.82.
2270 PP/01 Handle short writes when writing local log-files.
2271 In practice, only affects FreeBSD (8 onwards).
2272 Bugzilla 1053, with thanks to Dmitry Isaikin.
2274 NM/01 Bugzilla 949 - Documentation tweak
2276 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2279 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2281 PP/02 Implemented gsasl authenticator.
2283 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2285 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2286 `pkg-config foo` for cflags/libs.
2288 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2289 with rest of GSASL and with heimdal_gssapi.
2291 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2292 `pkg-config foo` for cflags/libs for the TLS implementation.
2294 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2295 properties get this fed in as external SSF. A number of robustness
2296 and debugging improvements to the cyrus_sasl authenticator.
2298 PP/08 cyrus_sasl server now expands the server_realm option.
2300 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2301 Patch by Jeremy Harris.
2303 PP/10 Added dbmjz lookup type.
2305 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2307 PP/12 MAIL args handles TAB as well as SP, for better interop with
2308 non-compliant senders.
2309 Analysis and variant patch by Todd Lyons.
2311 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2312 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2313 Patch from Dirk Mueller <dmueller@suse.com>
2315 PP/13 tls_peerdn now print-escaped for spool files.
2316 Observed some $tls_peerdn in wild which contained \n, which resulted
2317 in spool file corruption.
2319 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2320 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2321 or write after TLS renegotiation, which otherwise led to messages
2324 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2325 as a tracking header (ie: a signed header comes before the signature).
2326 Patch from Wolfgang Breyha.
2328 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2329 comma-sep list; embedded commas doubled.
2331 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2333 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2335 Report and patch from Dmitry Banschikov.
2337 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2338 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2339 failures. We appear to now support TLS1.1+ with Exim.
2341 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2342 lets Exim select keys and certificates based upon TLS SNI from client.
2343 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2344 before an outbound SMTP session. New log_selector, +tls_sni.
2346 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2347 NULL dereference. Report and patch from Alun Jones.
2349 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2350 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2351 Not seeing resolver debug output on NetBSD, but suspect this is a
2352 resolver implementation change.
2354 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2355 Left warnings. Added "eximon gdb" invocation mode.
2357 PP/21 Defaulting "accept_8bitmime" to true, not false.
2359 PP/22 Added -bw for inetd wait mode support.
2361 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2362 locate the relevant includes and libraries. Made this the default.
2364 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2365 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2367 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2368 This may cause build issues on older platforms.
2370 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2371 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2372 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2373 Added SNI support via GnuTLS too.
2374 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2376 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2378 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2379 Applied second patch from Janne, implementing suggestion to default
2380 multiple-strings-in-record handling to match SPF spec.
2382 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2384 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2385 read-only, out of scope).
2386 Patch from Wolfgang Breyha, report from Stuart Northfield.
2388 PP/29 Fix three issues highlighted by clang analyser static analysis.
2389 Only crash-plausible issue would require the Cambridge-specific
2390 iplookup router and a misconfiguration.
2391 Report from Marcin Mirosław.
2393 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2395 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2396 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2397 As part of this, removing so much warning spew let me fix some minor
2398 real issues in debug logging.
2400 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2401 assignment on my part. Fixed.
2403 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2404 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2405 Janne Snabb (who went above and beyond: thank you).
2407 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2408 string otherwise requires a connection and a bunch more work and it's
2409 relatively easy to get wrong. Should also expose TLS library linkage
2412 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2413 64-bit ${eval} (JH/03).
2415 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2416 GNU libc to support some of the 64-bit stuff, should not lead to
2417 conflicts. Defined before os.h is pulled in, so if a given platform
2418 needs to override this, it can.
2420 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2421 protection layer was required, which is not implemented.
2422 Bugzilla 1254, patch from Wolfgang Breyha.
2424 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2425 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2426 tls_dhparam take prime identifiers. Also unbreak combination of
2427 OpenSSL+DH_params+TLSSNI.
2429 PP/39 Disable SSLv2 by default in OpenSSL support.
2435 PP/01 Solaris build fix for Oracle's LDAP libraries.
2436 Bugzilla 1109, patch from Stephen Usher.
2438 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2440 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2443 TF/02 Fix a couple more cases where we did not log the error message
2444 when unlink() failed. See also change 4.74-TF/03.
2446 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2447 lock up or crash if it happened to be inside a call to libc when it
2448 got a SIGUSR1 from exiwhat.
2450 The SIGUSR1 handler appends the current process status to the process
2451 log which is later printed by exiwhat. It used to use the general
2452 purpose logging code to do this, but several functions it calls are
2453 not safe for signals.
2455 The new output code in the SIGUSR1 handler is specific to the process
2456 log, and simple enough that it's easy to inspect for signal safety.
2457 Removing some special cases also simplifies the general logging code.
2458 Removing the spurious timestamps from the process log simplifies
2461 TF/04 Improved ratelimit ACL condition.
2463 The /noupdate option has been deprecated in favour of /readonly which
2464 has clearer semantics. The /leaky, /strict, and /readonly update modes
2465 are mutually exclusive. The update mode is no longer included in the
2466 database key; it just determines when the database is updated. (This
2467 means that when you upgrade Exim will forget old rate measurements.)
2469 Exim now checks that the per_* options are used with an update mode that
2470 makes sense for the current ACL. For example, when Exim is processing a
2471 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2472 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2473 must specify per_mail/readonly. If you omit the update mode it defaults to
2474 /leaky where that makes sense (as before) or /readonly where required.
2476 The /noupdate option is now undocumented but still supported for
2477 backwards compatibility. It is equivalent to /readonly except that in
2478 ACLs where /readonly is required you may specify /leaky/noupdate or
2479 /strict/noupdate which are treated the same as /readonly.
2481 A useful new feature is the /count= option. This is a generalization
2482 of the per_byte option, so that you can measure the throughput of other
2483 aggregate values. For example, the per_byte option is now equivalent
2484 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2486 The per_rcpt option has been generalized using the /count= mechanism
2487 (though it's more complicated than the per_byte equivalence). When it is
2488 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2489 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2490 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2491 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2492 Note that using per_rcpt with a non-readonly update mode in more than
2493 one ACL will cause the recipients to be double-counted. (The per_mail
2494 and per_byte options don't have this problem.)
2496 The handling of very low rates has changed slightly. If the computed rate
2497 is less than the event's count (usually one) then this event is the first
2498 after a long gap. In this case the rate is set to the same as this event's
2499 count, so that the first message of a spam run is counted properly.
2501 The major new feature is a mechanism for counting the rate of unique
2502 events. The new per_addr option counts the number of different
2503 recipients that someone has sent messages to in the last time period. It
2504 behaves like per_rcpt if all the recipient addresses are different, but
2505 duplicate recipient addresses do not increase the measured rate. Like
2506 the /count= option this is a general mechanism, so the per_addr option
2507 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2508 example, measure the rate that a client uses different sender addresses
2509 with the options per_mail/unique=$sender_address. There are further
2510 details in the main documentation.
2512 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2514 TF/06 Removed a few PCRE remnants.
2516 TF/07 Automatically extract Exim's version number from tags in the git
2517 repository when doing development or release builds.
2519 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2520 Bugzilla 879. Patch from Paul Fisher.
2522 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2523 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2526 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2528 PP/05 Variable $av_failed, true if the AV scanner deferred.
2529 Bugzilla 1078. Patch from John Horne.
2531 PP/06 Stop make process more reliably on build failure.
2532 Bugzilla 1087. Patch from Heiko Schlittermann.
2534 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2535 Bugzilla 1089. Patch from Heiko Schlittermann.
2537 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2538 Bugzilla 1131. Patch from Holger Weiß.
2540 PP/09 Handle IPv6 addresses with SPF.
2541 Bugzilla 860. Patch from Wolfgang Breyha.
2543 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2545 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2548 PP/11 match_* no longer expand right-hand-side by default.
2549 New compile-time build option, EXPAND_LISTMATCH_RHS.
2550 New expansion conditions, "inlist", "inlisti".
2552 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2554 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2556 PP/14 fix log_write() format string regression from TF/03.
2557 Bugzilla 1152. Patch from Dmitry Isaikin.
2563 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2565 PP/02 Harmonised TLS library version reporting; only show if debugging.
2566 Layout now matches that introduced for other libraries in 4.74 PP/03.
2568 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2570 PP/04 New "dns_use_edns0" global option.
2572 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2575 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2576 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2578 TK/01 Updated PolarSSL code to 0.14.2.
2579 Bugzilla 1097. Patch from Andreas Metzler.
2581 PP/07 Catch divide-by-zero in ${eval:...}.
2582 Fixes bugzilla 1102.
2584 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2587 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2588 format-string attack -- SECURITY: remote arbitrary code execution.
2590 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2591 time unintentionally subject to list matching rules, letting the header
2592 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2593 arbitrary string expansion). This allowed for information disclosure.
2595 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2596 INT_MIN/-1 -- value coerced to INT_MAX.
2602 NM/01 Workaround for PCRE version dependency in version reporting
2605 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2606 This fixes portability to compilers other than gcc, notably
2607 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2609 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2610 makefiles for portability to HP-UX and POSIX correctness.
2612 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2613 Also via indented variable definition in the Makefile.
2614 (Debugging by Oliver Heesakkers).
2616 PP/02 Restore caching of spamd results with expanded spamd_address.
2617 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2619 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2620 Improves build reliability. Fix from: Frank Elsner
2622 NM/02 Fix wide character breakage in the rfc2047 coding
2623 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2625 NM/03 Allow underscore in dnslist lookups
2626 Fixes bug 1026. Patch from Graeme Fowler
2628 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2629 Code patches from Adam Ciarcinski of NetBSD.
2631 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2634 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2635 is logged, to avoid truncation. Patch from John Horne.
2637 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2638 Patch from Jakob Hirsch.
2640 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2641 SQL string expansion failure details.
2642 Patch from Andrey Oktyabrski.
2644 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2645 Patch from Simon Arlott.
2647 PP/09 New lookups functionality failed to compile on old gcc which rejects
2648 extern declarations in function scope.
2649 Patch from Oliver Fleischmann
2651 PP/10 Use sig_atomic_t for flags set from signal handlers.
2652 Check getgroups() return and improve debugging.
2653 Fixed developed for diagnosis in bug 927 (which turned out to be
2656 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2657 Patch from Mark Zealey.
2659 PP/12 Bugzilla 1056: Improved spamd server selection.
2660 Patch from Mark Zealey.
2662 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2663 Based on patch from Heiko Schlittermann.
2665 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2666 Patch from Uwe Doering, sign-off by Michael Haardt.
2668 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2669 variable declaration deep within a block. Bug and patch from
2672 PP/15 lookups-Makefile IRIX compatibility coercion.
2674 PP/16 Make DISABLE_DKIM build knob functional.
2676 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2677 Patch by Simon Arlott
2679 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2680 variable argument macros. Our copy now differs from upstream.
2686 TF/01 Failure to get a lock on a hints database can have serious
2687 consequences so log it to the panic log.
2689 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2690 controlled using the smtp_confirmation log selector.
2692 TF/03 Include the error message when we fail to unlink a spool file.
2694 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2695 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2696 for maintaining out-of-tree patches for some time.
2698 PP/01 Bugzilla 139: Documentation and portability issues.
2699 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2700 Handle per-OS dynamic-module compilation flags.
2702 PP/02 Let /dev/null have normal permissions.
2703 The 4.73 fixes were a little too stringent and complained about the
2704 permissions on /dev/null. Exempt it from some checks.
2705 Reported by Andreas M. Kirchwitz.
2707 PP/03 Report version information for many libraries, including
2708 Exim version information for dynamically loaded libraries. Created
2709 version.h, now support a version extension string for distributors
2710 who patch heavily. Dynamic module ABI change.
2712 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2713 privilege escalation vulnerability whereby the Exim run-time user
2714 can cause root to append content of the attacker's choosing to
2717 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2720 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2721 If dropping privileges for untrusted macros, we disabled normal logging
2722 on the basis that it would fail; for the Exim run-time user, this is not
2723 the case, and it resulted in successful deliveries going unlogged.
2724 Fixed. Reported by Andreas Metzler.
2730 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2731 only prepend for the Resent-* case. Fixes regression introduced in
2732 Exim 4.70 by NM/22 for Bugzilla 607.
2734 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2735 increasing numbers of administrators be bitten by this.
2737 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2739 PP/03 Bugzilla 994: added openssl_options main configuration option.
2741 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2743 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2745 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2747 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2748 without a peer certificate, leading to a segfault because of an
2749 assumption that peers always have certificates. Be a little more
2750 paranoid. Problem reported by Martin Tscholak.
2752 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2753 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2754 NB: ClamAV planning to remove STREAM in "middle of 2010".
2755 CL also introduces -bmalware, various -d+acl logging additions and
2756 more caution in buffer sizes.
2758 PP/09 Implemented reverse_ip expansion operator.
2760 PP/10 Bugzilla 937: provide a "debug" ACL control.
2762 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2764 PP/12 Bugzilla 973: Implement --version.
2766 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2768 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2770 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2772 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2773 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2774 ignore trailing whitespace.
2776 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2778 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2781 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2782 Notification from Dr Andrew Aitchison.
2784 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2785 ExtendedDetectionInfo response format.
2786 Notification from John Horne.
2788 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2791 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2792 XSL and documented dependency on system catalogs, with examples of how
2795 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2798 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2799 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2800 configuration file which is writeable by the Exim user or group.
2802 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2803 of configuration files to cover files specified with the -C option if
2804 they are going to be used with root privileges, not just the default
2807 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2808 option (effectively making it always true).
2810 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2811 files to be used while preserving root privileges.
2813 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2814 that rogue child processes cannot use them.
2816 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2817 run-time user, instead of root.
2819 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2820 Exim run-time user without dropping privileges.
2822 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2823 result string, instead of calling string_vformat() twice with the same
2826 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2827 for other users. Others should always drop root privileges if they use
2828 -C on the command line, even for a whitelisted configure file.
2830 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2832 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2838 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2839 $data_path, and $header_path variables; fixed documentation bugs and
2842 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2843 exipick to access non-standard spools, including the "frozen" queue
2846 NM/01 Bugzilla 965: Support mysql stored procedures.
2847 Patch from Alain Williams
2849 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2851 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2852 Patch from Andreas Metzler
2854 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2855 Patch from Kirill Miazine
2857 NM/05 Bugzilla 671: Added umask to procmail example.
2859 JJ/03 installed exipick 20100323.0, fixing doc bug
2861 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2862 directory. Notification and patch from Dan Rosenberg.
2864 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2866 TK/02 Improve log output when DKIM signing operation fails.
2868 MH/01 Treat the transport option dkim_domain as a colon separated
2869 list, not as a single string, and sign the message with each element,
2870 omitting multiple occurences of the same signer.
2872 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2873 Bugzilla 985, 986. Patch by Simon Arlott
2875 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2876 Patch by Simon Arlott
2878 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2879 MBX locking. Notification from Dan Rosenberg.
2885 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2887 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2889 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2891 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2893 NM/04 Bugzilla 915: Flag broken perl installation during build.
2899 TK/01 Added patch by Johannes Berg that expands the main option
2900 "spamd_address" if it starts with a dollar sign.
2902 TK/02 Write list of recipients to X-Envelope-Sender header when building
2903 the mbox-format spool file for content scanning (suggested by Jakob
2906 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2907 (http://www.dcc-servers.net/) support via dccifd. Activated by
2908 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2910 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2911 by Mark Daniel Reidel <mr@df.eu>.
2913 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2914 When building exim an external PCRE library is now needed -
2915 PCRE is a system library on the majority of modern systems.
2916 See entry on PCRE_LIBS in EDITME file.
2918 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2919 conversation. Added nologin parameter to request.
2920 Patch contributed by Kirill Miazine.
2922 TF/01 Do not log submission mode rewrites if they do not change the address.
2924 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2926 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2927 log files in place. Contributed by Roberto Lima.
2929 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2931 TF/03 Bugzilla 615: When checking the local_parts router precondition
2932 after a local_part_suffix or local_part_prefix option, Exim now
2933 does not use the address's named list lookup cache, since this
2934 contains cached lookups for the whole local part.
2936 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2937 Robert Millan. Documentation is in experimental-spec.txt.
2939 TF/04 Bugzilla 668: Fix parallel build (make -j).
2941 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2943 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2944 Patch provided by Jan Srzednicki.
2946 TF/05 Leading white space used to be stripped from $spam_report which
2947 wrecked the formatting. Now it is preserved.
2949 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2950 that they are available at delivery time.
2952 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2954 TF/08 TLS error reporting now respects the incoming_interface and
2955 incoming_port log selectors.
2957 TF/09 Produce a more useful error message if an SMTP transport's hosts
2958 setting expands to an empty string.
2960 NM/06 Bugzilla 744: EXPN did not work under TLS.
2961 Patch provided by Phil Pennock.
2963 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2964 Patch provided by Richard Godbee.
2966 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2967 acl_smtp_notquit, added index entry.
2969 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2970 Patch provided by Eugene Bujak.
2972 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2973 accept(). Patch provided by Maxim Dounin.
2975 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2976 Patch provided by Phil Pennock.
2978 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2980 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2981 Patch provided by Brad "anomie" Jorsch.
2983 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2984 Patch provided by Dean Brooks.
2986 TK/05 Add native DKIM support (does not depend on external libraries).
2988 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2989 Patch provided by Graeme Fowler.
2991 NM/16 Bugzilla 851: Documentation example syntax fix.
2993 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2995 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2998 NM/19 Bugzilla 745: TLS version reporting.
2999 Patch provided by Phil Pennock.
3001 NM/20 Bugzilla 167: bool: condition support.
3002 Patch provided by Phil Pennock.
3004 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3005 clients. Patch provided by Phil Pennock.
3007 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3008 Patch provided by Brad "anomie" Jorsch.
3010 NM/23 Bugzilla 687: Fix misparses in eximstats.
3011 Patch provided by Heiko Schlittermann.
3013 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3014 Patch provided by Heiko Schlittermann.
3016 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3017 plus update to original patch.
3019 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3021 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3022 Patch provided by David Brownlee.
3024 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3026 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3028 NM/30 Bugzilla 888: TLS documentation bugfixes.
3030 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3032 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3033 Unlike the original bugzilla I have changed all shell scripts in src tree.
3035 NM/33 Bugzilla 898: Transport filter timeout fix.
3036 Patch by Todd Rinaldo.
3038 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3039 Patch by Serge Demonchaux.
3041 NM/35 Bugzilla 39: Base64 decode bug fixes.
3042 Patch by Jakob Hirsch.
3044 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3046 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3048 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3050 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3056 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3057 ALT-N's libdkim that I have put here:
3058 http://duncanthrax.net/exim-experimental/
3060 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3061 were called 'true' and it seems that C99 defines that as a reserved
3062 keyword to be used with 'bool' variable types. That means you could
3063 not include C99-style headers which use bools without triggering
3064 build errors in sieve.c.
3066 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3067 as mailq or other aliases. Changed the --help handling significantly
3068 to do whats expected. exim_usage() emits usage/help information.
3070 SC/01 Added the -bylocaldomain option to eximstats.
3072 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3074 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3076 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3082 PH/01 Another patch from the Sieve maintainer.
3084 PH/02 When an IPv6 address is converted to a string for single-key lookup
3085 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3086 dots are used instead of colons so that keys in lsearch files need not
3087 contain colons. This was done some time before quoting was made available
3088 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3089 (notated using the quote facility) so as to distinguish them from IPv4
3090 keys. This meant that lookups for IP addresses in host lists did not work
3091 for iplsearch lookups.
3093 This has been fixed by arranging for IPv6 addresses to be expressed with
3094 colons if the lookup type is iplsearch. This is not incompatible, because
3095 previously such lookups could never work.
3097 The situation is now rather anomalous, since one *can* have colons in
3098 ordinary lsearch keys. However, making the change in all cases is
3099 incompatible and would probably break a number of configurations.
3101 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3104 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3105 conversion specification without a maximum field width, thereby enabling
3106 a rogue spamd server to cause a buffer overflow. While nobody in their
3107 right mind would setup Exim to query an untrusted spamd server, an
3108 attacker that gains access to a server running spamd could potentially
3109 exploit this vulnerability to run arbitrary code as the Exim user.
3111 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3112 $primary_hostname instead of what libspf2 thinks the hosts name is.
3114 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3115 a directory entry by the name of the lookup key. Previously, if a
3116 symlink pointed to a non-existing file or a file in a directory that
3117 Exim lacked permissions to read, a lookup for a key matching that
3118 symlink would fail. Now it is enough that a matching directory entry
3119 exists, symlink or not. (Bugzilla 503.)
3121 PH/03 The body_linecount and body_zerocount variables are now exported in the
3124 PH/04 Added the $dnslist_matched variable.
3126 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3127 This means they are set thereafter only if the connection becomes
3130 PH/06 Added the client_condition to authenticators so that some can be skipped
3131 by clients under certain conditions.
3133 PH/07 The error message for a badly-placed control=no_multiline_responses left
3134 "_responses" off the end of the name.
3136 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3138 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3139 (without spaces) instead of just copying the configuration text.
3141 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3143 PH/11 Added $max_received_linelength.
3145 PH/12 Added +ignore_defer and +include_defer to host lists.
3147 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3148 way in which PCRE > 7.0 is built.
3150 PH/14 Implemented queue_only_load_latch.
3152 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3153 MAIL command. The effect was to mangle the value on 64-bit systems.
3155 PH/16 Another patch from the Sieve maintainer.
3157 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3159 PH/18 If a system quota error occurred while trying to create the file for
3160 a maildir delivery, the message "Mailbox is full" was not appended to the
3161 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3162 only to a quota excession during the actual writing of the file.
3164 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3165 characters?) which causes problems in log lines. The DN values are now
3166 passed through string_printing() before being added to log lines.
3168 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3169 and InterBase are left for another time.)
3171 PH/21 Added message_body_newlines option.
3173 PH/22 Guard against possible overflow in moan_check_errorcopy().
3175 PH/23 POSIX allows open() to be a macro; guard against that.
3177 PH/24 If the recipient of an error message contained an @ in the local part
3178 (suitably quoted, of course), incorrect values were put in $domain and
3179 $local_part during the evaluation of errors_copy.
3185 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3186 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3189 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3190 issue a MAIL command.
3192 PH/02 In an ACL statement such as
3194 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3196 if a client was not listed at all, or was listed with a value other than
3197 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3198 the condition was not true (as it should be), so access was not denied.
3199 The bug was that the ! inversion was incorrectly passed on to the second
3200 item. This has been fixed.
3202 PH/03 Added additional dnslists conditions == and =& which are different from
3203 = and & when the dns lookup returns more than one IP address.
3205 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3206 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3208 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3209 FSYNC, which compiles an option called disable_fsync that allows for
3210 bypassing fsync(). The documentation is heavily laced with warnings.
3212 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3214 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3215 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3216 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3217 including adding "make clean"; (3) Added -fPIC when compiling the test
3218 dynamically loaded module, to get rid of a warning.
3220 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3221 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3222 = 0s. The bug is otherwise harmless.
3224 PH/07 There was a bug in the dovecot authenticator such that the value of
3225 $auth1 could be overwritten, and so not correctly preserved, after a
3226 successful authentication. This usually meant that the value preserved by
3227 the server_setid option was incorrect.
3229 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3231 PH/09 Installed PCRE release 7.0.
3233 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3234 run for batched SMTP input. It is now run at the start of every message
3235 in the batch. While fixing this I discovered that the process information
3236 (output by running exiwhat) was not always getting set for -bs and -bS
3237 input. This is fixed, and it now also says "batched" for BSMTP.
3239 PH/11 Added control=no_pipelining.
3241 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3242 patch, slightly modified), and move the expansion of helo_data till after
3243 the connection is made in the smtp transport (so it can use these
3246 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3248 PH/14 Added log_selector = +pid.
3250 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3252 PH/16 Add ${if forany and ${if forall.
3254 PH/17 Added dsn_from option to vary the From: line in DSNs.
3256 PH/18 Flush SMTP output before performing a callout, unless control =
3257 no_callout_flush is set.
3259 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3260 was true (the default) a successful delivery failed to delete the retry
3261 item, thus causing premature timeout of the address. The bug is now
3264 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3266 PH/21 Long custom messages for fakedefer and fakereject are now split up
3267 into multiline responses in the same way that messages for "deny" and
3268 other ACL rejections are.
3270 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3271 with slight modification.
3273 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3274 draft, changing the syntax and factoring some duplicate code".
3276 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3277 for deliveries of the second and subsequent messages over the same SMTP
3280 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3281 ${reduce, with only minor "tidies".
3283 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3285 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3286 expansion side effects.
3288 PH/27 When a message times out after an over-quota error from an Exim-imposed
3289 quota, the bounce message says "mailbox is full". This message was not
3290 being given when it was a system quota that was exceeded. It now should
3293 MH/03 Made $recipients available in local_scan(). local_scan() already has
3294 better access to the recipient list through recipients_list[], but
3295 $recipients can be useful in postmaster-provided expansion strings.
3297 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3298 in the case of a MAIL command with additional options following the
3299 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3300 were accidentally chopped off.
3302 PH/29 SMTP synchronization checks are implemented when a command is read -
3303 there is a check that no more input is waiting when there shouldn't be
3304 any. However, for some commands, a delay in an ACL can mean that it is
3305 some time before the response is written. In this time, more input might
3306 arrive, invalidly. So now there are extra checks after an ACL has run for
3307 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3308 pipelining has not been advertised.
3310 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3312 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3313 returned in the SMTP response when smtp_return_error_details was false.
3314 This has been fixed.
3316 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3317 instead of the C I/O that was originally supplied, because problems were
3318 reported on Solaris.
3320 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3321 Exim which did not show up earlier: it was assuming that a call to
3322 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3323 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3324 was a macro that became an assignment, so it seemed to work. This has
3325 changed to a proper function call with a void return, hence the compile
3326 error. Exim's code has been fixed.
3328 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3331 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3333 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3335 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3336 because it turns out to be too big or there is a timeout) there is no
3337 "Completed" line in the log. When some messages of this type were
3338 selected by exigrep, they were listed as "not completed". Others were
3339 picked up by some special patterns. I have improved the selection
3340 criteria to be more general.
3342 PH/38 The host_find_failed option in the manualroute router can now be set
3343 to "ignore", to completely ignore a host whose IP address cannot be
3344 found. If all hosts are ignored, the behaviour is controlled by the new
3345 host_all_ignored option.
3347 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3348 homing or because of multiple MX records with /mx) generated more than
3349 one IP address, and the following item turned out to be the local host,
3350 all the secondary addresses of the first item were incorrectly removed
3351 from the list, along with the local host and any following hosts (which
3352 is what is supposed to happen).
3354 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3355 whoever called Exim into the -H file. In the case of the daemon it was
3356 behaving confusingly. When first started, it used values for whoever
3357 started the daemon, but after a SIGHUP it used the Exim user (because it
3358 calls itself on a restart). I have changed the code so that it now always
3361 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3362 message are rejected with the same error (e.g. no authentication or bad
3363 sender address), and a DATA command is nevertheless sent (as can happen
3364 with PIPELINING or a stupid MUA), the error message that was given to the
3365 RCPT commands is included in the rejection of the DATA command. This is
3366 intended to be helpful for MUAs that show only the final error to their
3369 PH/42 Another patch from the Sieve maintainer.
3371 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3372 Eximstats - Fixed some broken HTML links and added missing column headers
3374 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3375 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3377 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3379 PH/43 Yet another patch from the Sieve maintainer.
3381 PH/44 I found a way to check for a TCP/IP connection going away before sending
3382 the response to the final '.' that terminates a message, but only in the
3383 case where the client has not sent further data following the '.'
3384 (unfortunately, this is allowed). However, in many cases there won't be
3385 any further data because there won't be any more messages to send. A call
3386 to select() can be used: if it shows that the input is "ready", there is
3387 either input waiting, or the socket has been closed. An attempt to read
3388 the next input character can distinguish the two cases. Previously, Exim
3389 would have sent an OK response which the client would never have see.
3390 This could lead to message repetition. This fix should cure that, at
3391 least in a lot of common cases.
3393 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3394 advertised in response to EHLO.
3400 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3401 fixed by 4.65/MH/01 (is this a record?) are fixed:
3403 (i) An empty string was always treated as zero by the numeric comparison
3404 operators. This behaviour has been restored.
3406 (ii) It is documented that the numeric comparison operators always treat
3407 their arguments as decimal numbers. This was broken in that numbers
3408 starting with 0 were being interpreted as octal.
3410 While fixing these problems I realized that there was another issue that
3411 hadn't been noticed. Values of message_size_limit (both the global option
3412 and the transport option) were treated as octal if they started with 0.
3413 The documentation was vague. These values are now always treated as
3414 decimal, and I will make that clear in the documentation.
3420 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3421 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3424 MH/01 Don't check that the operands of numeric comparison operators are
3425 integers when their expansion is in "skipping" mode (fixes bug
3426 introduced by 4.64-PH/07).
3428 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3429 child addresses, Exim now panics and dies. Previously, because the count
3430 is held in a short int, deliveries were likely to be lost. As such a
3431 large number of recipients for a single message is ridiculous
3432 (performance will be very, very poor), I have chosen to impose a limit
3433 rather than extend the field.
3439 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3440 leftover -K file (the existence of which was triggered by #402).
3441 While we were at it, introduced process PID as part of the -K
3442 filename. This should rule out race conditions when creating
3445 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3446 processing considerably. Previous code took too long for large mails,
3447 triggering a timeout which in turn triggers #401.
3449 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3450 in the DK code in transports.c. sendfile() is not really portable,
3451 hence the _LINUX specificness.
3453 TF/01 In the add_headers option to the mail command in an Exim filter,
3454 there was a bug that Exim would claim a syntax error in any
3455 header after the first one which had an odd number of characters
3458 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3459 callout verification, Exim cached a "reject" for the entire domain. This
3460 is correct for most verifications, but it is not correct for a recipient
3461 verification with use_sender or use_postmaster set, because in that case
3462 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3463 case of MAIL FROM:<> rejection from other early rejections (e.g.
3464 rejection of HELO). When verifying a recipient using a non-null MAIL
3465 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3466 Whatever the result of the callout, the value of the domain cache is
3467 left unchanged (for any other kind of callout, getting as far as trying
3468 RCPT means that the domain itself is ok).
3470 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3473 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3474 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3475 session, the daemon ignores ECONNECT errors and logs others; it now
3476 ignores EPIPE as well.
3478 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3479 (quoted-printable decoding).
3481 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3482 later the small subsequent patch to fix an introduced bug.
3484 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3486 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3488 PH/08 An error is now given if message_size_limit is specified negative.
3490 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3491 to be given (somewhat) arbitrary names.
3493 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3496 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3497 miscellaneous code fixes
3499 PH/10 Added the log_reject_target ACL modifier to specify where to log
3502 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3503 hostname. This is wrong, because it relates to the incoming message (and
3504 probably the interface on which it is arriving) and not to the outgoing
3505 callout (which could be using a different interface). This has been
3506 changed to use the value of the helo_data option from the smtp transport
3507 instead - this is what is used when a message is actually being sent. If
3508 there is no remote transport (possible with a router that sets up host
3509 addresses), $smtp_active_hostname is used.
3511 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3512 tweaks were necessary in order to get it to work (see also 21 below):
3513 (a) The code assumed that strncpy() returns a negative number on buffer
3514 overflow, which isn't the case. Replaced with Exim's string_format()
3516 (b) There were several signed/unsigned issues. I just did the minimum
3517 hacking in of casts. There is scope for a larger refactoring.
3518 (c) The code used strcasecmp() which is not a standard C function.
3519 Replaced with Exim's strcmpic() function.
3520 (d) The code set only $1; it now sets $auth1 as well.
3521 (e) A simple test gave the error "authentication client didn't specify
3522 service in request". It would seem that Dovecot has changed its
3523 interface. Fortunately there's a specification; I followed it and
3524 changed what the client sends and it appears to be working now.
3526 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3529 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3530 address A is aliased to B and C, where B exists and C does not. Without
3531 -v the output is "A verified" because verification stops after a
3532 successful redirection if more than one address is generated. However,
3533 with -v the child addresses are also verified. Exim was outputting "A
3534 failed to verify" and then showing the successful verification for C,
3535 with its parentage. It now outputs "B failed to verify", showing B's
3536 parentage before showing the successful verification of C.
3538 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3539 look up a TXT record in a specific list after matching in a combined
3542 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3543 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3544 they consult the DNS. I had assumed they would set it the way they
3545 wanted; and indeed my experiments on Linux seem to show that in some
3546 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3547 To be on the safe side, however, I have now made the interface to
3548 host_find_byname() similar to host_find_bydns(), with an argument
3549 containing the DNS resolver options. The host_find_byname() function now
3550 sets these options at its start, just as host_find_bydns() does. The smtp
3551 transport options dns_qualify_single and dns_search_parents are passed to
3552 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3553 of host_find_byname() use the default settings of RES_DEFNAMES
3554 (qualify_single) but not RES_DNSRCH (search_parents).
3556 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3557 spool_read_header() do less string testing, by means of a preliminary
3558 switch on the second character of optional "-foo" lines. (This is
3559 overdue, caused by the large number of possibilities that now exist.
3560 Originally there were few.) While I was there, I also converted the
3561 str(n)cmp tests so they don't re-test the leading "-" and the first
3562 character, in the hope this might squeeze out yet more improvement.
3564 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3565 flag allowing group syntax was set by the header_syntax check but not
3566 turned off, possible causing trouble later; (2) The flag was not being
3567 set at all for the header_verify test, causing "group"-style headers to
3568 be rejected. I have now set it in this case, and also caused header_
3569 verify to ignore an empty address taken from a group. While doing this, I
3570 came across some other cases where the code for allowing group syntax
3571 while scanning a header line wasn't quite right (mostly, not resetting
3572 the flag correctly in the right place). These bugs could have caused
3573 trouble for malformed header lines. I hope it is now all correct.
3575 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3576 with the "reply" argument non-NULL. The code, however (which originally
3577 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3578 but it didn't always do it. This confused somebody who was copying the
3579 code for some other use. I have removed all the tests.
3581 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3582 feature that was used to support insecure browsers during the U.S. crypto
3583 embargo. It requires special client support, and Exim is probably the
3584 only MTA that supported it -- and would never use it because real RSA is
3585 always available. This code has been removed, because it had the bad
3586 effect of slowing Exim down by computing (never used) parameters for the
3587 RSA_EXPORT functionality.
3589 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3590 authenticator to fail if there's a tab character in the incoming data
3591 (there should never be unless someone is messing about, as it's supposed
3592 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3593 if the connection is using TLS or if the remote IP is the same as the
3594 local IP, and the "valid-client-cert option" if a client certificate has
3597 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3598 authenticators. This can be used for authorization after authentication
3599 succeeds. (In the case of plaintext, it servers for both authentication
3602 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3603 if any retry times were supplied.
3605 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3606 connection, where there is no client IP address to check. In this
3607 situation, the verify now always succeeds.
3609 PH/25 Applied John Jetmore's -Mset patch.
3611 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3613 PH/27 In a string expansion for a processed (not raw) header when multiple
3614 headers of the same name were present, leading whitespace was being
3615 removed from all of them, but trailing whitespace was being removed only
3616 from the last one. Now trailing whitespace is removed from each header
3617 before concatenation. Completely empty headers in a concatenation (as
3618 before) are ignored.
3620 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3621 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3623 PH/29 [Removed. This was a change that I later backed out, and forgot to
3624 correct the ChangeLog entry (that I had efficiently created) before
3625 committing the later change.]
3627 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3628 address errors (4xx response to RCPT) over the same connection as other
3629 messages routed to the same hosts. Such deliveries are always "forced",
3630 so retry times are not inspected. This resulted in far too many retries
3631 for the affected addresses. The effect occurred only when there were more
3632 hosts than the hosts_max_try setting in the smtp transport when it had
3633 the 4xx errors. Those hosts that it had tried were not added to the list
3634 of hosts for which the message was waiting, so if all were tried, there
3635 was no problem. Two fixes have been applied:
3637 (i) If there are any address or message errors in an SMTP delivery, none
3638 of the hosts (tried or untried) are now added to the list of hosts
3639 for which the message is waiting, so the message should not be a
3640 candidate for sending over the same connection that was used for a
3641 successful delivery of some other message. This seems entirely
3642 reasonable: after all the message is NOT "waiting for some host".
3643 This is so "obvious" that I'm not sure why it wasn't done
3644 previously. Hope I haven't missed anything, but it can't do any
3645 harm, as the worst effect is to miss an optimization.
3647 (ii) If, despite (i), such a delivery is accidentally attempted, the
3648 routing retry time is respected, so at least it doesn't keep
3649 hammering the server.
3651 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3652 in ${readsocket because some servers need this prod.
3654 PH/32 Added some extra debug output when updating a wait-xxx database.
3656 PH/33 The hint "could be header name not terminated by colon", which has been
3657 given for certain expansion errors for a long time, was not being given
3658 for the ${if def:h_colon_omitted{... case.
3660 PH/34 The spec says: "With one important exception, whenever a domain list is
3661 being scanned, $domain contains the subject domain." There was at least
3662 one case where this was not true.
3664 PH/35 The error "getsockname() failed: connection reset by peer" was being
3665 written to the panic log as well as the main log, but it isn't really
3666 panic-worthy as it just means the connection died rather early on. I have
3667 removed the panic log writing for the ECONNRESET error when getsockname()
3670 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3671 runs only) independently of the message's sender address. This meant
3672 that, if the 4xx error was in fact related to the sender, a different
3673 message to the same recipient with a different sender could confuse
3674 things. In particular, this can happen when sending to a greylisting
3675 server, but other circumstances could also provoke similar problems.
3676 I have changed the default so that the retry time for these errors is now
3677 based a combination of the sender and recipient addresses. This change
3678 can be overridden by setting address_retry_include_sender=false in the
3681 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3682 remote server are returned as part of bounce messages. This was not
3683 happening for LMTP over a pipe (the lmtp transport), but now it is the
3684 same for both kinds of LMTP.
3686 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3687 in header lines that were in fact CNAMEs. This is no longer the case.
3689 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3690 and queue runs started by the daemon processed all messages. This has
3691 been fixed so that -R and -S can now usefully be given with -q<time>.
3693 PH/40 Import PCRE release 6.7 (fixes some bugs).
3695 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3697 PH/42 Give an error if -q is specified more than once.
3699 PH/43 Renamed the variables $interface_address and $interface_port as
3700 $received_ip_address and $received_port, to make it clear that these
3701 values apply to message reception, and not to the outgoing interface when
3702 a message is delivered. (The old names remain recognized, of course.)
3704 PH/44 There was no timeout on the connect() call when using a Unix domain
3705 socket in the ${readsocket expansion. There now is.
3707 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3708 be meaningful with "accept".
3710 SC/01 Eximstats V1.43
3711 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3713 SC/02 Eximstats V1.44
3714 Use a glob alias rather than an array ref in the generated
3715 parser. This improves both readability and performance.
3717 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3718 Collect SpamAssassin and rejection statistics.
3719 Don't display local sender or destination tables unless
3720 there is data to show.
3721 Added average volumes into the top table text output.
3723 SC/04 Eximstats V1.46
3724 Collect data on the number of addresses (recipients)
3725 as well as the number of messages.
3727 SC/05 Eximstats V1.47
3728 Added 'Message too big' to the list of mail rejection
3729 reasons (thanks to Marco Gaiarin).
3731 SC/06 Eximstats V1.48
3732 Mainlog lines which have GMT offsets and are too short to
3733 have a flag are now skipped.
3735 SC/07 Eximstats V1.49 (Alain Williams)
3736 Added the -emptyok flag.
3738 SC/08 Eximstats V1.50
3739 Fixes for obtaining the IP address from reject messages.
3741 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3742 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3743 whitespace changes from 4.64-PH/27
3745 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3748 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3749 are found, allow negative numbers in numeric criteria)
3751 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3753 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3754 to match changes made in 4.64-PH/43
3756 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3758 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3759 "domain" is actually an IP address, and force a failure. This locks out
3760 those revolvers/nameservers that support "A-for-A" lookups, in
3761 contravention of the specifications.
3763 PH/48 When a host name was looked up from an IP address, and the subsequent
3764 forward lookup of the name timed out, the host name was left in
3765 $sender_host_name, contrary to the specification.
3767 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3768 restricted to single-key lookups, Exim was not diagnosing an error if
3769 * or *@ was used with a query-style lookup.
3771 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3773 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3774 long ago, but noone interested enough thought of it. Let's just say that
3775 the "1.1" means that there are some new functions that weren't there at
3776 some point in the past.
3778 PH/51 Error processing for expansion failure of helo_data from an smtp
3779 transport during callout processing was broken.
3781 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3782 tested/used via the -bh/-bhc/-bs options.
3784 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3785 bug, fixed in subsequent PCRE releases).
3787 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3788 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3790 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3796 SC/01 Use a glob alias rather than an array ref in eximstats generated
3797 parser. This improves both readability and performance.
3799 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3800 Don't display local sender or destination tables in eximstats unless
3801 there is data to show.
3802 Added average volumes into the eximstats top table text output.
3804 SC/03 Collect data on the number of addresses (recipients) as well
3805 as the number of messages in eximstats.
3807 TF/01 Correct an error in the documentation for the redirect router. Exim
3808 does (usually) call initgroups() when daemonizing.
3810 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3811 with consistent privilege compared to when running as a daemon.
3813 TF/03 Note in the spec that $authenticated_id is not set for local
3814 submissions from trusted users.
3816 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3817 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3819 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3820 by adding some example configuration directives to the default
3821 configuration file. A little bit of work is required to uncomment the
3822 directives and define how usernames and passwords are checked, but
3823 there is now a framework to start from.
3825 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3826 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3827 without this. I don't know how relevant this is to other LDAP libraries.
3829 PH/02 Add the verb name to the "unknown ACL verb" error.
3831 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3833 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3835 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3836 directory not expanded when it should be if an expanded home directory
3837 was set for the address (which is overridden by the transport).
3839 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3842 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3843 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3844 because it is too late at that time, and has no effect.
3846 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3847 security issue with \' (bugzilla #107). I could not use the
3848 PQescapeStringConn() function, because it needs a PGconn value as one of
3851 PH/08 When testing addresses using -bt, indicate those final addresses that
3852 are duplicates that would not cause an additional delivery. At least one
3853 person was confused, thinking that -bt output corresponded to deliveries.
3854 (Suppressing duplicates isn't a good idea as you lose the information
3855 about possibly different redirections that led to the duplicates.)
3857 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3858 systems where poll() doesn't work, in particular OS X.
3860 PH/10 Added more information to debugging output for retry time not reached.
3862 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3863 operations in malware.c.
3865 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3868 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3869 syslog_duplication set false, log lines that would normally be written
3870 both the the main log and to the reject log were not written to syslog at
3873 PH/14 In the default configuration, change the use of "message" in ACL warn
3874 statements to "add_header".
3876 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3877 not followed by a command (e.g. "seen endif").
3879 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3880 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3883 PH/17 Added extra conditions to the default value of delay_warning_condition
3887 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3888 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3889 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3892 The Auto-Submitted: and various List- headers are standardised, whereas I
3893 don't think Precedence: ever was.
3895 PH/18 Refactored debugging code in route_finduser() to show more information,
3896 in particular, the error code if getpwnam() issues one.
3898 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3899 This is apparently needed in addition to the PH/07 change above to avoid
3900 any possible encoding problems.
3902 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3903 but not after initializing Perl.
3905 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3906 output them only if debugging. By default they are written stderr,
3907 apparently, which is not desirable.
3909 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3912 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3915 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3917 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3918 authenticated or an ident call has been made. Suppress the default
3919 values for $authenticated_id and $authenticated_sender (but permit -oMai
3920 and -oMas) when testing with -bh.
3922 PH/24 Re-jigged the order of the tests in the default configuration so that the
3923 tests for valid domains and recipients precede the DNS black list and CSA
3924 tests, on the grounds that those ones are more expensive.
3926 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3927 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3928 This bug exists in every version of Exim that I still have, right back to
3931 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3932 However, an attempt to turn on case-sensitivity in a regex key by
3933 including (?-i) didn't work because the subject string was already
3934 lowercased, and the effects were non-intuitive. It turns out that a
3935 one-line patch can be used to allow (?-i) to work as expected.
3941 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3942 other effects) broke the use of negated acl sub-conditions.
3944 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3947 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3948 "Deny" causes Exim to reject the incoming connection with a 554 error.
3949 Unfortunately, if there is a major crisis, such as a disk failure,
3950 tcp-wrappers gives "deny", whereas what one would like would be some
3951 kind of temporary error. A kludge has been added to help with this.
3952 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3953 554 error is used if errno is still zero or contains ENOENT (which occurs
3954 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3957 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3959 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3960 errors. Otherwise a message that provokes a temporary error (when other
3961 messages do not) can cause a whole host to time out.
3963 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3964 addresses were routed directly to files or pipes from a redirect router.
3965 File deliveries just didn't batch; pipe deliveries might have suffered
3968 PH/06 A failure to get a lock for a hints database would erroneously always say
3969 "Failed to get write lock", even when it was really a read lock.
3971 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3972 of 0600. This has been changed to use the value of the lockfile_mode
3973 option (which defaults to 0600).
3975 PH/08 Applied small patch from the Sieve maintainer.
3977 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3978 folder from quota calculations, a direct delivery into this folder messed
3979 up the contents of the maildirsize file. This was because the regex was
3980 used only to exclude .Trash (or whatever) when the size of the mailbox
3981 was calculated. There was no check that a delivery was happening into an
3982 excluded directory. This bug has been fixed by ignoring all quota
3983 processing for deliveries into excluded directories.
3985 PH/10 Added the maildirfolder_create_regex option to appendfile.
3991 PH/01 The code for finding all the local interface addresses on a FreeBSD
3992 system running IPv6 was broken. This may well have applied to all BSD
3993 systems, as well as to others that have similar system calls. The broken
3994 code found IPv4 interfaces correctly, but gave incorrect values for the
3995 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3996 that it would not match correctly against @[] and not recognize the IPv6
3999 PH/02 The ipliteral router was not recognizing addresses of the form user@
4000 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4002 PH/03 Added disable_ipv6.
4004 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4005 decoded form, because it is most often used to construct To: headers
4006 lines in autoreplies, and the decoded form may well be syntactically
4007 invalid. However, $reply_address has leading white space removed, and all
4008 newlines turned into spaces so that the autoreply transport does not
4011 PH/05 If group was specified without a user on a router, and no group or user
4012 was specified on a transport, the group from the router was ignored.
4014 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4015 for visible compile-time settings that can be used to change these
4016 numbers, for those that want even more. Backwards compatibility with old
4017 spool files has been maintained. However, going back to a previous Exim
4018 release will lost any variables that are in spool files.
4020 PH/07 Two small changes when running in the test harness: increase delay when
4021 passing a TCP/IP connection to a new process, in case the original
4022 process has to generate a bounce, and remove special handling of
4023 127.0.0.2 (sic), which is no longer necessary.
4025 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4026 be the same on different OS.
4028 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4031 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4032 whether --show-vars was specified or not
4034 JJ/02 exipick: Added support for new ACL variable spool format introduced
4037 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4038 syntactically invalid From: or Reply-to: line, and a filter used this to
4039 generate an autoreply, and therefore failed to obtain an address for the
4040 autoreply, Exim could try to deliver to a non-existent relative file
4041 name, causing unrelated and misleading errors. What now happens is that
4042 it logs this as a hard delivery error, but does not attempt to create a
4045 PH/11 The exinext utility has a -C option for testing purposes, but although
4046 the given file was scanned by exinext itself; it wasn't being passed on
4047 when Exim was called.
4049 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4050 an end-of-file indication when reading a command response.
4052 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4053 compiled. In many other places in Exim, IPv6 addresses are always
4054 recognized, so I have changed this. It also means that IPv4 domain
4055 literals of the form [IPV4:n.n.n.n] are now always recognized.
4057 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4058 used if the router is not running as root, for example, when verifying at
4059 ACL time, or when using -bh. The debugging output from this situation was
4060 non-existent - all you got was a failure to exec. I have made two
4063 (a) Failures to set uid/gid, the current directory, or a process leader
4064 in a subprocess such as that created by queryprogram now generate
4065 suitable debugging output when -d is set.
4067 (b) The queryprogram router detects when it is not running as root,
4068 outputs suitable debugging information if -d is set, and then runs
4069 the subprocess without attempting to change uid/gid.
4071 PH/15 Minor change to Makefile for building test_host (undocumented testing
4074 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4075 additional section of a DNS packet that returns MX or SRV records.
4076 Instead, it always explicitly searches for A/AAAA records. This avoids
4077 major problems that occur when a DNS server includes only records of one
4078 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4079 fixed another bug: if SRV records were looked up and the corresponding
4080 address records were *not* found in the additional section, the port
4081 values from the SRV records were lost.
4083 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4084 using the correct key (the original address) when searching the retry
4085 rules in order to find which one to use for generating the retry hint.
4087 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4088 adding the default one. Similarly, if it contains a Reply-To: header, the
4089 errors_reply_to option, if set, is not used.
4091 PH/19 When calculating a retry time, Exim used to measure the "time since
4092 failure" by looking at the "first failed" field in the retry record. Now
4093 it does not use this if it is later than than the arrival time of the
4094 message. Instead it uses the arrival time. This makes for better
4095 behaviour in cases where some deliveries succeed, thus re-setting the
4096 "first failed" field. An example is a quota failure for a huge message
4097 when small messages continue to be delivered. Without this change, the
4098 "time since failure" will always be short, possible causing more frequent
4099 delivery attempts for the huge message than are intended.
4100 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4102 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4103 $1, $2, $3) because the numerical variables can be reset during some
4104 expansion items (e.g. "match"), thereby losing the authentication data.
4106 PH/21 Make -bV show the size of off_t variables so that the test suite can
4107 decide whether to run tests for quotas > 2G.
4109 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4110 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4111 filecount value is greater than 2G or if a quota value is greater than 2G
4112 on a system where the size of off_t is not greater than 4, a panic error
4115 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4116 never match. The debug and -bh output now contains an explicit error
4117 message indicating a malformed IPv4 address or mask.
4119 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4120 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4121 PH/23 above applies.
4123 PH/25 Do not write to syslog when running in the test harness. The only
4124 occasion when this arises is a failure to open the main or panic logs
4125 (for which there is an explicit test).
4127 PH/26 Added the /no_tell option to "control=freeze".
4129 PH/27 If a host name lookup failed very early in a connection, for example, if
4130 the IP address matched host_lookup and the reverse lookup yielded a name
4131 that did not have a forward lookup, an error message of the form "no IP
4132 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4133 could be logged. Now it outputs the IP address instead of "NULL".
4135 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4136 allows the sender and the authenticated sender to be set when
4137 submitting a message from within Exim. Since child_open_exim() is
4138 documented for local_scan(), the new function should be too.
4140 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4141 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4142 results in an empty string is now treated as unset.
4144 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4146 PH/31 Added sender_verify_fail logging option.
4148 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4149 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4150 tidied the source and removed it altogether.
4152 PH/33 When a queue run was abandoned because the load average was too high, a
4153 log line was always written; now it is written only if the queue_run log
4154 selector is set. In addition, the log line for abandonment now contains
4155 information about the queue run such as the pid. This is always present
4156 in "start" and "stop" lines but was omitted from the "abandon" line.
4158 PH/34 Omit spaces between a header name and the colon in the error message that
4159 is given when verify = headers_syntax fails (if there are lots of them,
4160 the message gets confusing).
4162 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4163 names, as there are now some PTR records that contain slashes. This check
4164 is only to protect against broken name servers that fall over on strange
4165 characters, so the fact that it applies to all lookups doesn't matter.
4167 PH/36 Now that the new test suite is complete, we can remove some of the
4168 special code in Exim that was needed for the old test suite. For example,
4169 sorting DNS records because real resolvers return them in an arbitrary
4170 order. The new test suite's fake resolver always returns records in the
4173 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4174 bounces) except when queue_only is set, to avoid logging races between
4175 the different processes.
4177 PH/38 Panic-die if .include specifies a non-absolute path.
4179 PH/39 A tweak to the "H" retry rule from its user.
4181 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4182 a label. They prevented compilation on older perls.
4184 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4185 a warning to be raised on newish perls.
4187 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4188 on queue. Changes to match documented behaviour of showing count of
4189 messages matching specified criteria.
4191 PH/40 Changed the default ident timeout from 30s to 5s.
4193 PH/41 Added support for the use of login_cap features, on those BSD systems
4194 that have them, for controlling the resources used by pipe deliveries.
4196 PH/42 The content-scanning code uses fopen() to create files in which to put
4197 message data. Previously it was not paying any attention to the mode of
4198 the files. Exim runs with umask(0) because the rest of the code creates
4199 files with open(), and sets the required mode explicitly. Thus, these
4200 files were ending up world-writeable. This was not a big issue, because,
4201 being within the spool directory, they were not world-accessible. I have
4202 created a function called modefopen, which takes an additional mode
4203 argument. It sets umask(777), creates the file, chmods it to the required
4204 mode, then resets the umask. All the relevant calls to fopen() in the
4205 content scanning code have been changed to use this function.
4207 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4208 to 24 hours. This avoids potential overflow problems when processing G
4209 and H retry rules. I suspect nobody ever tinkers with this value.
4211 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4213 PH/45 When the plaintext authenticator is running as a client, the server's
4214 challenges are checked to ensure they are valid base64 strings. By
4215 default, the authentication attempt is cancelled if an invalid string is
4216 received. Setting client_ignore_invalid_base64 true ignores these errors.
4217 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4218 they are received. Thus, the responses can be made to depend on the
4219 challenges. If an invalid string is ignored, an empty string is placed in
4222 PH/46 Messages that are created by the autoreply transport now contains a
4223 References: header, in accordance with RFCs 2822 and 3834.
4225 PH/47 Added authenticated_sender_force to the smtp transport.
4227 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4229 PH/49 Installed latest patch from the Sieve maintainer.
4231 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4232 was also set, the appendfile transport was unnecessarily scanning a
4233 directory of message files (e.g. for maildir delivery) to find the count
4234 of files (along with the size), even though it did not need this
4235 information. It now does the scan only if it needs to find either the
4236 size of the count of files.
4238 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4240 PH/52 Two bugs concerned with error handling when the smtp transport is
4243 (i) Exim was not creating retry information for temporary errors given
4244 for individual recipients after the DATA command when the smtp transport
4245 was used in LMTP mode. This meant that they could be retried too
4246 frequently, and not timed out correctly.
4248 (ii) Exim was setting the flag that allows error details to be returned
4249 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4250 recipients that were returned after the DATA command.
4252 PH/53 This is related to PH/52, but is more general: for any failing address,
4253 when detailed error information was permitted to be returned to the
4254 sender, but the error was temporary, then after the final timeout, only
4255 "retry timeout exceeded" was returned. Now it returns the full error as
4256 well as "retry timeout exceeded".
4258 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4259 do this, and (what is worse) MTAs that accept it.
4261 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4262 will now be deprecated.
4264 PH/56 New os.c-cygwin from the Cygwin maintainer.
4266 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4267 formats (previously only available in exim formats via -bpr, -bpru,
4268 and -bpra. Now also available in native and exiqgrep formats)
4270 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4271 with very large, slow to parse queues
4273 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4275 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4277 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4278 responses to authentication challenges, though it was showing the
4279 challenges; (ii) I've removed the CR characters from the debug output for
4282 PH/58 Allow for the insertion of a newline as well as a space when a string
4283 is turned into more than one encoded-word during RFC 2047 encoding. The
4284 Sieve code now uses this.
4286 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4287 data_4xx, lost_connection, tls_required.
4289 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4290 message was being sent as an SMTP response.
4292 PH/61 Add -l and -k options to exicyclog.
4294 PH/62 When verifying, if an address was redirected to one new address, so that
4295 verification continued, and the new address failed or deferred after
4296 having set something in $address_data, the value of $address_data was not
4297 passed back to the ACL. This was different to the case when no
4298 redirection occurred. The value is now passed back in both cases.
4300 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4301 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4302 use login_cap.h, so on its own it isn't the distinguishing feature. The
4303 new name refers directly to the setclassresources() function.
4305 PH/65 Added configuration files for NetBSD3.
4307 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4309 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4310 is preferred over IPv4.
4312 PH/68 The bounce_return_message and bounce_return_body options were not being
4313 honoured for bounces generated during the reception of non-SMTP messages.
4314 In particular, this applied to messages rejected by the ACL. This bug has
4315 been fixed. However, if bounce_return_message is true and bounce_return_
4316 body is false, the headers that are returned for a non-SMTP message
4317 include only those that have been read before the error was detected.
4318 (In the case of an ACL rejection, they have all been read.)
4320 PH/69 The HTML version of the specification is now built in a directory called
4321 spec_html instead of spec.html, because the latter looks like a path with
4322 a MIME-type, and this confuses some software.
4324 PH/70 Catch two compiler warnings in sieve.c.
4326 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4327 function verify_get_ident() calls ip_connect() to connect a socket, but
4328 if the "connect()" function timed out, ip_connect() used to close the
4329 socket. However, verify_get_ident() also closes the socket later, and in
4330 between Exim writes to the log, which may get opened at this point. When
4331 the socket was closed in ip_connect(), the log could get the same file
4332 descriptor number as the socket. This naturally causes chaos. The fix is
4333 not to close the socket in ip_connect(); the socket should be closed by
4334 the function that creates it. There was only one place in the code where
4335 this was missing, in the iplookup router, which I don't think anybody now
4336 uses, but I've fixed it anyway.
4338 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4339 well as to direct DNS lookups. Otherwise the handling of names in host
4340 lists is inconsistent and therefore confusing.
4346 PH/01 Two changes to the default runtime configuration:
4348 (1) Move the checks for relay_from_hosts and authenticated clients from
4349 after to before the (commented out) DNS black list checks.
4351 (2) Add control=submission to the relay_from_hosts and authenticated
4352 clients checks, on the grounds that messages accepted by these
4353 statements are most likely to be submissions.
4355 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4357 (1) Generate an error if the third argument for the ${prvs expansion is
4360 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4363 (3) Reset the variables that are obtained from the first argument of
4364 ${prvscheck and used in the second argument before leaving the code,
4365 because their memory is reclaimed, so using them afterwards may do
4368 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4369 one (it's much easier than Tom thought :-).
4371 (5) Because of (4), we can now allow for the use of $prvscheck_result
4372 inside the third argument.
4374 PH/03 For some reason, the default setting of PATH when running a command from
4375 a pipe transport was just "/usr/bin". I have changed it to
4378 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4379 anything to be listed in the output from -bV.
4381 PH/05 When a filter generated an autoreply, the entire To: header line was
4382 quoted in the delivery log line, like this:
4384 => >A.N.Other <ano@some.domain> <original@ddress> ...
4386 This has been changed so that it extracts the operative address. There
4387 may be more than one such address. If so, they are comma-separated, like
4390 => >ano@some.domain,ona@other.domain <original@ddress> ...
4392 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4393 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4394 not being looked up in the rDNS to get a host name, Exim was showing the
4395 IP address twice in Received: lines, even though the IP addresses were
4396 identical. For example:
4398 Received: from [1.2.3.4] (helo=[1.2.3.4])
4400 However, if the real host name was known, it was omitting the HELO data
4401 if it matched the actual IP address. This has been tidied up so that it
4402 doesn't show the same IP address twice.
4404 PH/07 When both +timestamp and +memory debugging was on, the value given by
4405 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4406 called by the debug printing, thereby overwriting the timestamp buffer.
4407 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4409 PH/08 When the original message was included in an autoreply transport, it
4410 always said "this is a copy of the message, including all the headers",
4411 even if body_only or headers_only was set. It now gives an appropriate
4414 PH/09 Applied a patch from the Sieve maintainer which:
4416 o fixes some comments
4417 o adds the (disabled) notify extension core
4418 o adds some debug output for the result of if/elsif tests
4419 o points to the current vacation draft in the documentation
4420 and documents the missing references header update
4424 o fixes a bug in processing the envelope test (when testing
4425 multiple envelope elements, the last element determined the
4428 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4429 Electronic Mail") by including:
4431 Auto-submitted: auto-generated
4433 in the messages that it generates (bounce messages and others, such as
4434 warnings). In the case of bounce messages for non-SMTP messages, there was
4435 also a typo: it was using "Auto_submitted" (underscore instead of
4436 hyphen). Since every message generated by Exim is necessarily in response
4437 to another message, thes have all been changed to:
4439 Auto-Submitted: auto-replied
4441 in accordance with these statements in the RFC:
4443 The auto-replied keyword:
4445 - SHOULD be used on messages sent in direct response to another
4446 message by an automatic process,
4448 - MUST NOT be used on manually-generated messages,
4450 - MAY be used on Delivery Status Notifications (DSNs) and Message
4451 Disposition Notifications (MDNs),
4453 - MUST NOT be used on messages generated by automatic or periodic
4454 processes, except for messages which are automatic responses to
4457 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4458 to the default Received: header definition.
4460 PH/12 Added log selector acl_warn_skipped (default on).
4462 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4463 variables because (a) they are in the wrong storage pool and (b) even if
4464 they were copied, it wouldn't work properly because of the caching.
4466 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4467 checking when decoding. Apparently there are clients that generate
4468 overlong encoded strings. Why am I not surprised?
4470 PH/15 If the first argument of "${if match_address" was not empty, but did not
4471 contain an "@" character, Exim crashed. Now it writes a panic log message
4472 and treats the condition as false.
4474 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4476 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4477 extension "envelope-auth". The code is finished and in agreement with
4478 other implementations, but there is no documentation so far and in fact,
4479 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4480 not changing the active code.
4482 Print executed "if" and "elsif" statements when debugging is used. This
4483 helps a great deal to understand what a filter does.
4485 Document more things not specified clearly in RFC3028. I had all this
4486 sorted out, when out of a sudden new issues came to my mind. Oops."
4488 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4491 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4492 canonical form (as documented). However, after a host name lookup from
4493 the IP address, check_host() was doing a simple string comparison with
4494 addresses acquired from the DNS when checking that the found name did
4495 have the original IP as one of its addresses. Since any found IPv6
4496 addresses are likely to be in abbreviated form, the comparison could
4497 fail. Luckily, there already exists a function for doing the comparison
4498 by converting both addresses to binary, so now that is used instead of
4499 the text comparison.
4501 PH/20 There was another similar case to PH/19, when a complete host name was
4502 given in a host list; looking up its IP address could give an abbreviated
4503 form, whereas the current host's name might or might not be abbreviated.
4504 The same fix has been applied.
4510 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4511 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4514 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4515 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4517 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4519 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4520 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4521 Also, for "Auto-Submitted", treat anything other than "no" as
4522 non-personal, in accordance with RFC 3834. (Previously it treated
4523 anything starting "auto-" as non-personal.)
4525 TF/01 The control=submission/name=... option had a problem with syntax
4526 errors if the name included a slash character. The /name= option
4527 now slurps the rest of the string, so it can include any characters
4528 but it must come last in the list of options (after /sender_retain
4531 PH/05 Some modifications to the interface to the fake nameserver for the new
4539 TK/01 Added the "success_on_redirect" address verification option. See
4540 NewStuff for rationale and an example.
4542 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4544 PH/02 Patch to exigrep to allow it to work on syslog lines.
4546 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4547 fread() to skip over the body file's header line, because in Cygwin the
4548 header line is locked and is inaccessible.
4550 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4551 co-exist for some time) to make it clear that it is the Exim ID that is
4552 referenced, not the Message-ID: header line.
4554 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4555 string_format() function, because snprintf() does not exist on all
4558 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4559 ${stat: expansion item.
4561 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4562 protocol synchronization error", to keep the pedants happy.
4564 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4565 well as for IRIX systems, when gcc is being used. See the host.c source
4568 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4570 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4573 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4574 transport and to the smtp transport in LMTP mode.
4576 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4578 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4579 run a filter in a subprocess. This could lead to confusion in subsequent
4580 lookups in the parent process. There should also be a search_tidyup() at
4581 the end of the subprocess.
4583 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4584 only if the host matched helo_try_verify_hosts, which caused the
4585 verification to occur when the EHLO/HELO command was issued. The ACL just
4586 tested the remembered result. Now, if a previous verification attempt has
4587 not happened, "verify = helo" does it there and then.
4589 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4591 TK/03 Fix log output including CR from clamd.
4593 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4594 exist provoked a memory error which could cause a segfault.
4596 PH/15 Installed PCRE 6.2
4598 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4600 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4601 of the problem. Specifically, suggested +O2 rather than +O1 for the
4604 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4606 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4607 record, the port was not correctly logged when the outgoing_port log
4608 selector was set (it logged the transort's default port).
4610 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4611 fallback_hosts, and "hosts" in the smtp transport.
4613 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4614 host errors such as "Connection refused".
4616 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4617 authentication with radiusclient 0.4.9:
4619 - Error returned from rc_read_config was caught wrongly
4620 - Username/password not passed on to radius server due to wrong length.
4622 The presumption is that some radiusclient API changes for 4.51/PH/17
4623 were not taken care of correctly. The code is still untested by me (my
4624 Linux distribution still has 0.3.2 of radiusclient), but it was
4625 contributed by a Radius user.
4627 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4628 expanding the "port" option of the smtp transport.
4630 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4631 while reading a MIME header. Thanks to Tom Hughes for a patch.
4633 PH/24 Include config.h inside local_scan.h so that configuration settings are
4636 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4637 that in an ACL for RCPT (for example), you can examine exactly what was
4640 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4641 commands, but it was not correctly comparing the address with the actual
4642 client host address. Thus, it would show the EHLO address in Received:
4643 header lines when this was not necessary.
4645 PH/27 Added the % operator to ${eval:}.
4647 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4648 it should be ignoring failures (because with -C, for example, it has lost
4649 privilege). It wasn't ignoring creation failures other than "already
4652 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4655 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4656 because an input line was too long, either on its own, or by virtue of
4657 too many continuations, the temporary file was not being removed, and the
4658 return code was incorrect.
4660 PH/31 Missing "BOOL" in function definition in filtertest.c.
4662 PH/32 Applied Sieve patches from the maintainer.
4664 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4666 PH/33 Added "verify = not_blind".
4668 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4669 Local/Makefile (with some defaults set). These are used in built scripts
4670 such as exicyclog, but they have never been used in the exim_install
4671 script (though there are many overriding facilities there). I have
4672 arranged that the exim_install script now takes note of these two
4675 PH/35 Installed configuration files for Dragonfly.
4677 PH/36 When a locally submitted message by a trusted user did not contain a
4678 From: header, and the sender address was obtained from -f or from an SMTP
4679 MAIL command, and the trusted user did not use -F to supply a sender
4680 name, $originator_name was incorrectly used when constructing a From:
4681 header. Furthermore, $originator_name was used for submission mode
4682 messages from external hosts without From: headers in a similar way,
4683 which is clearly wrong.
4685 PH/37 Added control=suppress_local_fixups.
4687 PH/38 When log_selector = +received_sender was set, and the addition of the
4688 sender made the log line's construction buffer exactly full, or one byte
4689 less than full, an overflow happened when the terminating "\n" was
4692 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4693 when the result of a list match is failure because a DNS lookup failed.
4695 PH/40 RM_COMMAND is now used in the building process.
4697 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4698 the "build-* directories that it finds.
4700 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4701 address in a domain literal was a prefix of an interface address.
4703 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4704 when verifying a sender address, unless rewrite_headers is false.
4706 PH/44 Wrote a long comment about why errors_to addresses are verified as
4707 recipients, not senders.
4709 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4710 the ratelimit ACL was added.
4712 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4714 PH/46 Added extra information about PostgreSQL errors to the error string.
4716 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4717 avoiding the need to install special zones in a real server. This is
4718 backwards compatible; if it can't find the fake resolver, it drops back.
4719 Thus, both old and new test suites can be run.
4721 TF/02 Added util/ratelimit.pl
4723 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4724 clock is set back in time.
4726 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4727 Candler <B.Candler@pobox.com>.
4729 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4730 OK for addresses that are the result of redirections.
4732 PH/48 A number of further additions for the benefit of the new test suite,
4733 including a fake gethostbyname() that interfaces to the fake DNS resolver
4736 TF/06 The fix for widen_domains has also been applied to qualify_single and
4737 search_parents which are the other dnslookup options that can cause
4740 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4743 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4745 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4746 service" if it encountered an MX record with an empty target hostname.
4747 The message is now "an MX or SRV record indicated no SMTP service".
4749 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4750 if the DNS of the sending site is misconfigured. This is quite a
4751 common situation. This change restores the behaviour of treating a
4752 helo verification defer as a failure.
4754 PH/51 If self=fail was set on a router, the bounce message did not include the
4755 actual error message.
4761 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4763 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4764 command itself ended in error, the underlying message about the transport
4765 filter timeout was being overwritten with the pipe command error. Now the
4766 underlying error message should be appended to the second error message.
4768 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4770 PH/02 Reduce the amount of output that "make" produces by default. Full output
4771 can still be requested.
4773 PH/03 The warning log line about a condition test deferring for a "warn" verb
4774 was being output only once per connection, rather than after each
4775 occurrence (because it was using the same function as for successful
4776 "warn" verbs). This seems wrong, so I have changed it.
4778 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4779 it should not have, which might have caused a crash in the right
4780 circumstances, but probably never did.
4782 PH/04 Installed a modified version of Tony Finch's patch to make submission
4783 mode fix the return path as well as the Sender: header line, and to
4784 add a /name= option so that you can make the user's friendly name appear
4787 TF/03 Added the control = fakedefer ACL modifier.
4789 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4790 Mark Lowes for thorough testing.
4792 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4794 TK/03 Merged latest SRS patch from Miles Wilton.
4796 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4797 with the definition in sysexits.h (which is #included earlier).
4798 Fortunately, Exim does not actually use EX_OK. The code used to try to
4799 preserve the sysexits.h value, by assuming that macro definitions were
4800 scanned for macro replacements. I have been disabused of this notion,
4801 so now the code just undefines EX_OK before #including unistd.h.
4803 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4804 in the smtp transport. When a block could not be written in a single
4805 write() function, the timeout was being re-applied to each part-write.
4806 This seems wrong - if the receiver was accepting one byte at a time it
4807 would take for ever. The timeout is now adjusted when this happens. It
4808 doesn't have to be particularly precise.
4810 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4811 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4813 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4814 without a domain if the check to <postmaster@domain> fails.
4816 SC/01 Eximstats: added -xls and the ability to specify output files
4817 (patch written by Frank Heydlauf).
4819 SC/02 Eximstats: use FileHandles for outputting results.
4821 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4823 SC/04 Eximstats: fixed display of large numbers with -nvr option
4825 SC/05 Eximstats: fixed merging of reports with empty tables.
4827 SC/06 Eximstats: added the -include_original_destination flag
4829 SC/07 Eximstats: removed tabs and trailing whitespace.
4831 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4833 TK/06 MBOX spool code: Add real "From " MBOX separator line
4834 so the .eml file is really in mbox format (even though
4835 most programs do not really care). Patch from Alex Miller.
4837 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4838 The latter is generated from $received_to and is only set if the
4839 message has one envelope recipient. SA can use these headers,
4840 obviously out-of-the-box. Patch from Alex Miller.
4842 PH/08 The ${def test on a variable was returning false if the variable's
4843 value was "0", contrary to what the specification has always said!
4844 The result should be true unless the variable is empty.
4846 PH/09 The syntax error of a character other than { following "${if
4847 def:variable_name" (after optional whitespace) was not being diagnosed.
4848 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4849 accidental colon was present, for example, could give incorrect results.
4851 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4852 result is used (not including appendfile, where other changes are about
4855 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4856 This involved changing a lot of size variables from int to off_t. It
4857 should work with maildirs and everything.
4859 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4860 spamd dying while we are connected to it.
4862 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4865 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4866 function for BDB 4.3.
4868 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4870 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4871 the sources for PCRE have been re-arranged and more files are now
4874 PH/15 The code I had for printing potentially long long variables in PH/11
4875 above was not the best (it lost precision). The length of off_t variables
4876 is now inspected at build time, and an appropriate printing format (%ld
4877 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4878 to be "long long int" or "long int". This is needed for the internal
4879 formatting function string_vformat().
4881 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4882 the configuration file to be ":syslog", then the script "guesses" where
4883 the logs files are, rather than using the compiled in default. In our
4884 case the guess is not the same as the compiled default, so the script
4885 suddenly stopped working when I started to use syslog. The patch checks
4886 to see if log_file_path is "". If so, it attempts to read it from exim
4887 with no configuration file to get the compiled in version, before it
4888 falls back to the previous guessing code."
4890 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4891 implementing BATV in an Exim configuration. See NewStuff for the gory
4894 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4895 Makefile that are specific to HP-UX.
4897 PH/18 If the "use_postmaster" option was set for a recipient callout together
4898 with the "random" option, the postmaster address was used as the MAIL
4899 FROM address for the random test, but not for the subsequent recipient
4900 test. It is now used for both.
4902 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4903 patch removes a few documentation additions to RFC 3028, because the
4904 latest draft now contains them. It adds the new en;ascii-case comparator
4905 and a new error check for 8bit text in MIME parts. Comparator and
4906 require names are now matched exactly. I enabled the subaddress
4907 extension, but it is not well tested yet (read: it works for me)."
4909 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4910 rework some of the code of TK/09 above to avoid the hardwired use of
4911 "%lld" and "long long". Replaced the call to snprintf() with a call to
4914 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4915 records point to non-existent hosts", "retry timeout exceeded", and
4916 "retry time not reached for any host after a long failure period".
4918 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4919 experimental DomainKeys support:
4921 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4922 (2) On an error such as an illegally used "control", the wrong name for
4923 the control was given.
4925 These problems did NOT occur unless DomainKeys support was compiled.
4927 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4929 PH/24 Added ${if match_ip condition.
4931 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4932 databases so that it will be absolutely obvious if a crash occurs in the
4933 DB library. This is a regular occurrence (often caused by mis-matched
4936 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4937 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4938 picked up on a user's system that detects such things. There doesn't seem
4939 to be a gcc warning option for this - only an attribute that has to be
4940 put on the function's prototype. It seems that in Fedora Core 4 they have
4941 set this on a number of new functions. No doubt there will be more in due
4944 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4945 specify a transport. However, if an address that was verified by such a
4946 router was the subject of a callout, Exim crashed because it tried to
4947 read the rcpt_include_affixes from the non-existent transport. Now it
4948 just assumes that the setting of that option is false. This bug was
4949 introduced by 4.51/PH/31.
4951 PH/28 Changed -d+all to exclude +memory, because that information is very
4952 rarely of interest, but it makes the output a lot bigger. People tend to
4953 do -d+all out of habit.
4955 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4956 code in os-type was giving problems when libc.so lives in lib64, like on
4959 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4960 aren't the modern standard, and it seems that some systems' include files
4961 don't always have them. Exim was already checking for some of the newer
4962 ones like T_AAAA, and defining it itself. I've added checks for all the
4963 record types that Exim uses.
4965 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4966 not automatically generating a new one, as it is supposed to. This
4967 prevented TLS from working. If the file did exist, but contained invalid
4968 data, a new version was generated, as expected. It was only the case of a
4969 non-existent file that was broken.
4971 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4972 with a change in libdomainkeys > 0.64.
4974 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4975 from DNS. If the selector record carries the flag, it now has
4976 precedence over the domain-wide flag.
4978 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4980 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4981 the use of a port name that isn't defined in /etc/services) occurred, the
4982 message was deferred as in a normal delivery, and thus remained on the
4983 spool, instead of being failed because of the mua_wrapper setting. This
4984 is now fixed, and I tidied up some of the mua_wrapper messages at the
4987 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4988 the messages in a hash of arrays rather than using individual hashes.
4989 This is a bit cleaner and results in dramatic memory savings, albeit
4990 at a slight CPU cost.
4992 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4993 as requested by Marc Sherman.
4995 SC/10 Eximstats: added histograms for user specified patterns as requested
4998 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5000 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5001 fopen() in the content-scanning modules that did not already have it.
5007 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5008 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5010 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5012 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5014 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5015 to test Sieve filters that use "vacation".
5017 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5018 that changes the way the GnuTLS parameters are stored in the cache file.
5019 The new format can be generated externally. For backward compatibility,
5020 if the data in the cache doesn't make sense, Exim assumes it has read an
5021 old-format file, and it generates new data and writes a new file. This
5022 means that you can't go back to an older release without removing the
5025 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5026 work if there are any delivery delays because "one_time" forces the
5027 parent to be marked "delivered", so its unseen clone is never tried
5028 again. For this reason, Exim now forbids the simultaneous setting of
5031 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5032 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5033 entry if you want to know the details. The fix, however, neglected to
5034 consider the case when local delivery batching is involved. The test for
5035 "previously delivered" was not happening when checking to see if an
5036 address could be batched with a previous (undelivered) one; under
5037 certain circumstances this could lead to multiple deliveries to the same
5040 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5041 in its include files, and this causes problems building Exim.
5043 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5044 header_syntax) but Exim was just ignoring anything given after a slash.
5045 In particular, this caused confusion with an attempt to use "verify =
5046 reverse_host_lookup/defer_ok". An error is now given when options are
5047 supplied for verify items that do not have them. (Maybe reverse_host_
5048 lookup should have a defer_ok option, but that's a different point.)
5050 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5051 defined by RFC 821) to 2048, because there were problems with some AUTH
5052 commands, and RFC 1869 says the size should be increased for extended
5053 SMTP commands that take arguments.
5055 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5058 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5059 "unknown" error; now it says that the functionality isn't in the binary.
5061 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5062 an address' error message when a string expansion fails (syntax or
5063 whatever). Otherwise the password may appear in the log. Following change
5064 PH/42 below, there is no longer a chance of it appearing in a bounce
5067 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5069 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5070 of its addresses was ever tried. (Bugzilla bug #2.)
5072 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5073 the result incorrectly in the debug output. (It correctly added a newline
5074 to what was transported.)
5076 TF/01 Added $received_time.
5078 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5079 commented out examples of how to interface to a virus scanner and to
5080 SpamAssassin. Also added commented examples of av_scanner and
5081 spamd_address settings.
5083 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5084 and controls are allowed in which ACLs. There were a couple of minor
5085 errors. Some of the entries in the conditions table (which is a table of
5086 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5087 negation of where the condition IS allowed.
5089 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5091 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5092 header file does not have a version number, so I've had to invent a new
5093 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5094 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5095 radiusclient), but it was contributed by a Radius user.
5097 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5098 files or directories, for OpenSSL.
5100 PH/19 When an Exim process that is running as root has to create an Exim log
5101 file, it does so in a subprocess that runs as exim:exim so as to get the
5102 ownership right at creation (otherwise, other Exim processes might see
5103 the file with the wrong ownership). There was no test for failure of this
5104 fork() call, which would lead to the process getting stuck as it waited
5105 for a non-existent subprocess. Forks do occasionally fail when resources
5106 run out. I reviewed all the other calls to fork(); they all seem to check
5109 PH/20 When checking for unexpected SMTP input at connect time (before writing
5110 the banner), Exim was not dealing correctly with a non-positive return
5111 from the read() function. If the client had disconnected by this time,
5112 the result was a log entry for a synchronization error with an empty
5113 string after "input=" when read() returned zero. If read() returned -1
5114 (an event I could not check), uninitialized data bytes were printed.
5115 There were reports of junk text (parts of files, etc) appearing after
5118 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5120 PH/22 Added support for macro redefinition, and (re)definition in between
5121 driver and ACL definitions.
5123 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5124 forgetting to use the resulting value; it was using the unexpanded value.
5126 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5127 hadn't been configured. The fix is from Juergen Kreileder, who
5128 understands it better than I do:
5130 "Here's what I see happening with three configured cyrus_sasl
5131 authenticators configured (plain, login, cram-md5):
5133 On startup auth_cyrus_sasl_init() gets called for each of these.
5134 This means three calls to sasl_listmech() without a specified mech_list.
5135 => SASL tests which mechs of all available mechs actually work
5136 => three warnings about OTP not working
5137 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5139 With the patch, sasl_listmech() also gets called three times. But now
5140 SASL's mech_list option is set to the server_mech specified in the the
5141 authenticator. Or in other words, the answer from sasl_listmech()
5142 gets limited to just the mech you're testing for (which is different
5144 => the return list contains just 'plain' or 'login', 'cram-md5' or
5145 nothing depending on the value of ob->server_mech.
5147 I've just tested the patch: Authentication still works fine,
5148 unavailable mechs specified in the exim configuration are still
5149 caught, and the auth.log warnings about OTP are gone."
5151 PH/25 When debugging is enabled, the contents of the command line are added
5152 to the debugging output, even when log_selector=+arguments is not
5155 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5156 answer is "GNU", and only if the return is "GNU/something" is the answer
5159 PH/27 $acl_verify_message is now set immediately after the failure of a
5160 verification in an ACL, and so is available in subsequent modifiers. In
5161 particular, the message can be preserved by coding like this:
5163 warn !verify = sender
5164 set acl_m0 = $acl_verify_message
5166 Previously, $acl_verify_message was set only while expanding "message"
5167 and "log_message" when a very denied access.
5169 PH/28 Modified OS/os.c-Linux with
5171 -#ifndef OS_LOAD_AVERAGE
5172 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5174 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5175 nomenclature these days.)
5177 PH/29 Installed patch from the Sieve maintainer that adds the options
5178 sieve_useraddress and sieve_subaddress to the redirect router.
5180 PH/30 In these circumstances:
5181 . Two addresses routed to the same list of hosts;
5182 . First host does not offer TLS;
5183 . First host accepts first address;
5184 . First host gives temporary error to second address;
5185 . Second host offers TLS and a TLS session is established;
5186 . Second host accepts second address.
5187 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5188 and peerdn, if requested) that were in fact used only for the second
5191 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5192 attention to any local part prefix or suffix that was matched by the
5193 router that accepted the address. It now behaves in the same way as it
5194 does for delivery: the affixes are removed from the local part unless
5195 rcpt_include_affixes is set on the transport.
5197 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5198 timeout during the DATA phase of an incoming message.
5200 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5201 applied a patch sanctioned by the Sieve maintainer.
5203 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5204 the uid or gid is negative. A case of a negative gid caused this to be
5205 noticed. The fix allows for either to be negative.
5207 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5208 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5211 PH/36 The change PH/12 above was broken. Fixed it.
5213 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5214 the grounds that routing the same address twice would always produce the
5215 same answer. This might have been true once, but it is certainly no
5216 longer true now. Routing a child address may depend on the previous
5217 routing that produced that child. Some complicated redirection strategies
5218 went wrong when messages had multiple recipients, and made Exim's
5219 behaviour dependent on the order in which the addresses were given.
5221 I have moved the duplicate checking until after the routing is complete.
5222 Exim scans the addresses that are assigned to local and remote
5223 transports, and removes any duplicates. This means that more work will be
5224 done, as duplicates will always all be routed, but duplicates are
5225 presumably rare, so I don't expect this is of any significance.
5227 For deliveries to pipes, files, and autoreplies, the duplicate checking
5228 still happens during the routing process, since they are not going to be
5231 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5232 It corrects a timeout issue with spamd. This is Ian's comment: "The
5233 background is that sometimes spamd either never reads data from a
5234 connection it has accepted, or it never writes response data. The exiscan
5235 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5236 blindly assumes that writes won't block so it may never time out."
5238 PH/39 Allow G after quota size as well as K and M.
5240 PH/40 The value set for $authenticated_id in an authenticator may not contain
5241 binary zeroes or newlines because the value is written to log lines and
5242 to spool files. There was no check on this. Now the value is run through
5243 the string_printing() function so that such characters are converted to
5244 printable escape sequences.
5246 PH/41 $message_linecount is a new variable that contains the total number of
5247 lines in the message. Compare $body_linecount, which is the count for the
5250 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5251 bounce and delay warning messages, except in certain special cases, which
5254 (a) An SMTP error message from a remote host;
5255 (b) A message specified in a :fail: redirection;
5256 (c) A message specified in a "fail" command in a system filter;
5257 (d) A message specified in a FAIL return from the queryprogram router;
5258 (e) A message specified by the cannot_route_message router option.
5260 In these cases only, Exim does include the error details in bounce and
5261 warning messages. There are also a few cases where bland messages such
5262 as "unrouteable address" or "local delivery error" are given.
5264 PH/43 $value is now also set for the "else" part of a ${run expansion.
5266 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5267 being worked on, but at least Exim now implements the latest version to
5270 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5271 process to complete was treated as a delivery failure, a timeout while
5272 writing the message to the pipe was logged, but erroneously treated as a
5273 successful delivery. Such timeouts include transport filter timeouts. For
5274 consistency with the overall process timeout, these timeouts are now
5275 treated as errors, giving rise to delivery failures by default. However,
5276 there is now a new Boolean option for the pipe transport called
5277 timeout_defer, which, if set TRUE, converts the failures into defers for
5278 both kinds of timeout. A transport filter timeout is now identified in
5281 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5282 systems where "make" and "gmake" are different, calling "gmake" at top
5283 level broke things. I've arranged for the value of $(MAKE) to be passed
5284 from the Makefile to this script so that it can call the same version of
5288 A note about Exim versions 4.44 and 4.50
5289 ----------------------------------------
5291 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5292 changes of various kinds. As a consequence, a big documentation update was
5293 needed. This delayed the release for rather longer than seemed good, especially
5294 in the light of a couple of (minor) security issues. Therefore, the changes
5295 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5296 release. So 4.44 and 4.50 are in effect two different branches that both start
5299 I have left the 4.50 change log unchanged; it contains all the changes since
5300 4.43. The change log for 4.44 is below; many of its items are identical to
5301 those for 4.50. This seems to be the most sensible way to preserve the
5302 historical information.
5308 1. Minor wording change to the doc/README.SIEVE file.
5310 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5311 computation of the current number of files was incorrect.
5313 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5314 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5317 4. Give more explanation in the error message when the command for a transport
5318 filter fails to execute.
5320 5. There are several places where Exim runs a non-Exim command in a
5321 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5322 was being done only for the command run by the queryprogram router. It is
5323 now done for all such subprocesses. The other cases are: ${run, transport
5324 filters, and the commands run by the lmtp and pipe transports.
5326 6. Added CONFIGURE_GROUP build-time option.
5328 7. Some older OS have a limit of 256 on the maximum number of file
5329 descriptors. Exim was using setrlimit() to set 1000 as a large value
5330 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5331 systems. I've change it so that if it can't get 1000, it tries for 256.
5333 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5334 was an oversight, and furthermore, ever since the addition of extra
5335 controls (e.g. 4.43/32), the checks on when to allow different forms of
5336 "control" were broken. There should now be diagnostics for all cases when a
5337 control that does not make sense is encountered.
5339 9. Added the /retain_sender option to "control=submission".
5341 10. $recipients is now available in the predata ACL (oversight).
5343 11. Tidy the search cache before the fork to do a delivery from a message
5344 received from the command line. Otherwise the child will trigger a lookup
5345 failure and thereby defer the delivery if it tries to use (for example) a
5346 cached ldap connection that the parent has called unbind on.
5348 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5349 of $address_data from the recipient verification was clobbered by the
5350 sender verification.
5352 13. The value of address_data from a sender verification is now available in
5353 $sender_address_data in subsequent conditions in the ACL statement.
5355 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5357 15. Added a new option "connect=<time>" to callout options, to set a different
5360 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5361 was its contents. (It was OK if the option was not defined at all.)
5363 17. A "Completed" log line is now written for messages that are removed from
5364 the spool by the -Mrm option.
5366 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5367 information about exactly what failed.
5369 19. Added -dd to debug only the daemon process.
5371 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5372 handles timeouts, both on the server side and network timeouts. Renamed the
5373 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5375 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5376 It is now set to "smtps".
5378 22. $host_address is now set to the target address during the checking of
5379 ignore_target_hosts.
5381 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5382 being passed; this would have caused $sender_host_name to have been used if
5383 matching the list had actually called for a host name (not very likely,
5384 since this list is usually IP addresses). A host name is now passed as
5387 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5388 code that specifies a non-restarting handler (typically sigaction() in
5389 modern systems) in an attempt to fix a rare and obscure crash bug.
5391 25. Narrowed the window for a race in the daemon that could cause it to ignore
5392 SIGCHLD signals. This is not a major problem, because they are used only to
5393 wake it up if nothing else does.
5395 26. A malformed maildirsize file could cause Exim to calculate negative values
5396 for the mailbox size or file count. Odd effects could occur as a result.
5397 The maildirsize information is now recalculated if the size or filecount
5400 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5401 support for a long time. Removed HAVE_SYS_VFS_H.
5403 28. Installed the latest version of exipick from John Jetmore.
5405 29. In an address list, if the pattern was not a regular expression, an empty
5406 subject address (from a bounce message) matched only if the pattern was an
5407 empty string. Non-empty patterns were not even tested. This was the wrong
5408 because it is perfectly reasonable to use an empty address as part of a
5409 database query. An empty address is now tested by patterns that are
5410 lookups. However, all the other forms of pattern expect the subject to
5411 contain a local part and a domain, and therefore, for them, an empty
5412 address still always fails if the pattern is not itself empty.
5414 30. Exim went into a mad DNS loop when attempting to do a callout where the
5415 host was specified on an smtp transport, and looking it up yielded more
5416 than one IP address.
5418 31. Re-factored the code for checking spool and log partition space into a
5419 function that finds that data and another that does the check. The former
5420 is then used to implement four new variables: $spool_space, $log_space,
5421 $spool_inodes, and $log_inodes.
5423 32. The RFC2047 encoding function was originally intended for short strings
5424 such as real names; it was not keeping to the 75-character limit for
5425 encoded words that the RFC imposes. It now respects the limit, and
5426 generates multiple encoded words if necessary. To be on the safe side, I
5427 have increased the buffer size for the ${rfc2047: expansion operator from
5430 33. It is now permitted to omit both strings after an "if" condition; if the
5431 condition is true, the result is "true". As before, when the second string
5432 is omitted, a false condition yields an empty string. This makes it less
5433 cumbersome to write custom ACL and router conditions.
5435 34. Failure to deliver a bounce message always caused it to be frozen, even if
5436 there was an errors_to setting on the router. The errors_to setting is now
5439 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5440 canonical form (fully expanded) before being placed in
5441 $sender_host_address.
5443 36. The table in the code that translates DNS record types into text (T_A to
5444 "A" for instance) was missing entries for NS and CNAME. It is just possible
5445 that this could have caused confusion if both these types were looked up
5446 for the same domain, because the text type is used as part of Exim's
5447 per-process caching. But the chance of anyone hitting this buglet seems
5450 37. The dnsdb lookup has been extended in a number of ways.
5452 (1) There is a new type, "zns", which walks up the domain tree until it
5453 finds some nameserver records. It should be used with care.
5455 (2) There is a new type, "mxh", which is like "mx" except that it returns
5456 just the host names, not the priorities.
5458 (3) It is now possible to give a list of domains (or IP addresses) to be
5459 looked up. The behaviour when one of the lookups defers can be
5460 controlled by a keyword.
5462 (4) It is now possible to specify the separator character for use when
5463 multiple records are returned.
5465 38. The dnslists ACL condition has been extended: it is now possible to supply
5466 a list of IP addresses and/or domains to be looked up in a particular DNS
5469 39. Added log_selector=+queue_time_overall.
5471 40. When running the queue in the test harness, wait just a tad after forking a
5472 delivery process, to get repeatability of debugging output.
5474 41. Include certificate and key file names in error message when GnuTLS fails
5475 to set them up, because the GnuTLS error message doesn't include the name
5476 of the failing file when there is a problem reading it.
5478 42. Allow both -bf and -bF in the same test run.
5480 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5482 44. The "Exiscan patch" is now merged into the mainline Exim source.
5484 45. Sometimes the final signoff response after QUIT could fail to get
5485 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5486 < 0 before doing a fflush(). This bug looks as though it goes back to the
5487 introduction of TLS in release 3.20, but "sometimes" must have been rare
5488 because the tests only now provoked it.
5490 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5491 (this can affect the format of dates).
5493 47. exim_tidydb, when checking for the continued existence of a message for
5494 which it has found a message-specific retry record, was not finding
5495 messages that were in split spool directories. Consequently, it was
5496 deleting retry records that should have stayed in existence.
5498 48. Steve fixed some bugs in eximstats.
5500 49. The SPA authentication driver was not abandoning authentication and moving
5501 on to the next authenticator when an expansion was forced to fail,
5502 contradicting the general specification for all authenticators. Instead it
5503 was generating a temporary error. It now behaves as specified.
5505 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5506 (the order specifies the preference for clients). The order is now AES256,
5507 AES128, 3DES, ARCFOUR128.
5509 51. Small patch to Sieve code - explicitly set From: when generating an
5512 52. Exim crashed if a remote delivery caused a very long error message to be
5513 recorded - for instance if somebody sent an entire SpamAssassin report back
5514 as a large number of 550 error lines. This bug was coincidentally fixed by
5515 increasing the size of one of Exim's internal buffers (big_buffer) that
5516 happened as part of the Exiscan merge. However, to be on the safe side, I
5517 have made the code more robust (and fixed the comments that describe what
5520 53. Now that there can be additional text after "Completed" in log lines (if
5521 the queue_time_overall log selector is set), a one-byte patch to exigrep
5522 was needed to allow it to recognize "Completed" as not the last thing in
5525 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5526 patch that reportedly fixes this has been added. I am not expert enough to
5527 create a test for it. This is what the patch creator wrote:
5529 "I found a little strange behaviour of ldap code when working with
5530 Windows 2003 AD Domain, where users was placed in more than one
5531 Organization Units. When I tried to give exim partial DN, the exit code
5532 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5533 But simultaneously result of request was absolutely normal ldap result,
5534 so I produce this patch..."
5536 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5537 REFERENCE, so I have modified the code to exclude the patch when that macro
5540 55. Some experimental protocols are using DNS PTR records for new purposes. The
5541 keys for these records are domain names, not reversed IP addresses. The
5542 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5543 leaves it alone. Component reversal etc. now happens only for IP addresses.
5546 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5548 57. Double the size of the debug message buffer (to 2048) so that more of very
5549 long debug lines gets shown.
5551 58. The exicyclog utility now does better if the number of log files to keep
5552 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5554 59. Two changes related to the smtp_active_hostname option:
5556 (1) $smtp_active_hostname is now available as a variable.
5557 (2) The default for smtp_banner uses $smtp_active_hostname instead
5558 of $primary_hostname.
5560 60. The host_aton() function is supposed to be passed a string that is known
5561 to be a valid IP address. However, in the case of IPv6 addresses, it was
5562 not checking this. This is a hostage to fortune. Exim now panics and dies
5563 if the condition is not met. A case was found where this could be provoked
5564 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5565 components; fortuitously, this particular loophole had already been fixed
5566 by change 4.50/55 above.
5568 If there are any other similar loopholes, the new check in host_aton()
5569 itself should stop them being exploited. The report I received stated that
5570 data on the command line could provoke the exploit when Exim was running as
5571 exim, but did not say which command line option was involved. All I could
5572 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5573 running as the user.
5576 61. There was a buffer overflow vulnerability in the SPA authentication code
5577 (which came originally from the Samba project). I have added a test to the
5578 spa_base64_to_bits() function which I hope fixes it.
5581 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5582 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5584 63. The daemon start-up calls getloadavg() while still root for those OS that
5585 need the first call to be done as root, but it missed one case: when
5586 deliver_queue_load_max is set with deliver_drop_privilege. This is
5587 necessary for the benefit of the queue runner, because there is no re-exec
5588 when deliver_drop_privilege is set.
5590 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5591 This has been fixed.
5593 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5594 was in use, was not putting the data itself into the right store pool;
5595 consequently, it could be overwritten for a subsequent message in the same
5596 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5599 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5601 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5602 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5603 the function were treating the return as a boolean value, which happened to
5604 work because 0=false and not-0=true, but is not correct code.
5606 68. The host_aton() function was not handling scoped IPv6 addresses (those
5607 with, for example, "%eth0" on the end) correctly.
5609 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5610 negated items (that is, ~something) in unsigned ints. Some compilers
5611 apparently mutter when there is no cast.
5613 70. If an address verification called from an ACL failed, and did not produce a
5614 user-specific message (i.e. there was only a "system" message), nothing was
5615 put in $acl_verify_message. In this situation, it now puts the system
5618 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5619 session; change 4.31/43 added the unwanted input to the log line - except
5620 that it did not do this in the start of session case. It now does.
5622 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5623 This is wrong and can cause the other end to generate a synchronization
5624 error if it is another Exim or anything else that does the synchronization
5625 check. A QUIT command is no longer sent after a timeout.
5627 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5628 during host lookups.
5630 74. The defer_ok option of callout verification was not working if it was used
5631 when verifying addresses in header lines, that is, for this case:
5633 verify = header_sender/callout=defer_ok
5635 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5636 those file descriptors could be used for SMTP connections. If anything
5637 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5638 could be sent to the SMTP client, causing chaos. The daemon now opens
5639 stdin, stdout, and stderr to /dev/null when it puts itself into the
5642 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5643 log by default. The user can override this with suitable Perl magic.
5645 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5646 the log message when discard triggers, was not working for the DATA ACL or
5647 for the non-SMTP ACL.
5649 78. Error message wording change in sieve.c.
5651 79. If smtp_accept_max_per_host was set, the number of connections could be
5652 restricted to fewer than expected, because the daemon was trying to set up
5653 a new connection before checking whether the processes handling previous
5654 connections had finished. The check for completed processes is now done
5655 earlier. On busy systems, this bug wouldn't be noticed because something
5656 else would have woken the daemon, and it would have reaped the completed
5659 80. If a message was submitted locally by a user whose login name contained one
5660 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5661 It caused a spool format error. I have fixed the spool reading code. A
5662 related problem was that the "from" clause in the Received: line became
5663 illegal because of the space(s). It is now covered by ${quote_local_part.
5665 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5668 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5670 83. Patch from Sieve maintainer to fix unterminated string problem in
5671 "vacation" handling.
5673 84. Some minor changes to the Linux configuration files to help with other
5674 OS variants using glibc.
5676 85. One more patch for Sieve to update vacation handling to latest spec.
5679 ----------------------------------------------------
5680 See the note above about the 4.44 and 4.50 releases.
5681 ----------------------------------------------------
5687 1. Change 4.43/35 introduced a bug that caused file counts to be
5688 incorrectly computed when quota_filecount was set in an appendfile
5691 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5692 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5695 3. Give more explanation in the error message when the command for a transport
5696 filter fails to execute.
5698 4. There are several places where Exim runs a non-Exim command in a
5699 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5700 was being done only for the command run by the queryprogram router. It is
5701 now done for all such subprocesses. The other cases are: ${run, transport
5702 filters, and the commands run by the lmtp and pipe transports.
5704 5. Some older OS have a limit of 256 on the maximum number of file
5705 descriptors. Exim was using setrlimit() to set 1000 as a large value
5706 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5707 systems. I've change it so that if it can't get 1000, it tries for 256.
5709 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5710 was an oversight, and furthermore, ever since the addition of extra
5711 controls (e.g. 4.43/32), the checks on when to allow different forms of
5712 "control" were broken. There should now be diagnostics for all cases when a
5713 control that does not make sense is encountered.
5715 7. $recipients is now available in the predata ACL (oversight).
5717 8. Tidy the search cache before the fork to do a delivery from a message
5718 received from the command line. Otherwise the child will trigger a lookup
5719 failure and thereby defer the delivery if it tries to use (for example) a
5720 cached ldap connection that the parent has called unbind on.
5722 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5723 of $address_data from the recipient verification was clobbered by the
5724 sender verification.
5726 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5727 was its contents. (It was OK if the option was not defined at all.)
5729 11. A "Completed" log line is now written for messages that are removed from
5730 the spool by the -Mrm option.
5732 12. $host_address is now set to the target address during the checking of
5733 ignore_target_hosts.
5735 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5736 being passed; this would have caused $sender_host_name to have been used if
5737 matching the list had actually called for a host name (not very likely,
5738 since this list is usually IP addresses). A host name is now passed as
5741 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5742 code that specifies a non-restarting handler (typically sigaction() in
5743 modern systems) in an attempt to fix a rare and obscure crash bug.
5745 15. Narrowed the window for a race in the daemon that could cause it to ignore
5746 SIGCHLD signals. This is not a major problem, because they are used only to
5747 wake it up if nothing else does.
5749 16. A malformed maildirsize file could cause Exim to calculate negative values
5750 for the mailbox size or file count. Odd effects could occur as a result.
5751 The maildirsize information is now recalculated if the size or filecount
5754 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5755 support for a long time. Removed HAVE_SYS_VFS_H.
5757 18. Updated exipick to current release from John Jetmore.
5759 19. Allow an empty sender to be matched against a lookup in an address list.
5760 Previously the only cases considered were a regular expression, or an
5763 20. Exim went into a mad DNS lookup loop when doing a callout where the
5764 host was specified on the transport, if the DNS lookup yielded more than
5767 21. The RFC2047 encoding function was originally intended for short strings
5768 such as real names; it was not keeping to the 75-character limit for
5769 encoded words that the RFC imposes. It now respects the limit, and
5770 generates multiple encoded words if necessary. To be on the safe side, I
5771 have increased the buffer size for the ${rfc2047: expansion operator from
5774 22. Failure to deliver a bounce message always caused it to be frozen, even if
5775 there was an errors_to setting on the router. The errors_to setting is now
5778 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5779 canonical form (fully expanded) before being placed in
5780 $sender_host_address.
5782 24. Updated eximstats to version 1.33
5784 25. Include certificate and key file names in error message when GnuTLS fails
5785 to set them up, because the GnuTLS error message doesn't include the name
5786 of the failing file when there is a problem reading it.
5788 26. Expand error message when OpenSSL has problems setting up cert/key files.
5791 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5792 (this can affect the format of dates).
5794 28. exim_tidydb, when checking for the continued existence of a message for
5795 which it has found a message-specific retry record, was not finding
5796 messages that were in split spool directories. Consequently, it was
5797 deleting retry records that should have stayed in existence.
5799 29. eximstats updated to version 1.35
5800 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5801 1.35 - bugfix such that pie charts by volume are generated correctly
5803 30. The SPA authentication driver was not abandoning authentication and moving
5804 on to the next authenticator when an expansion was forced to fail,
5805 contradicting the general specification for all authenticators. Instead it
5806 was generating a temporary error. It now behaves as specified.
5808 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5809 (the order specifies the preference for clients). The order is now AES256,
5810 AES128, 3DES, ARCFOUR128.
5812 31. Small patch to Sieve code - explicitly set From: when generating an
5815 32. Exim crashed if a remote delivery caused a very long error message to be
5816 recorded - for instance if somebody sent an entire SpamAssassin report back
5817 as a large number of 550 error lines. This bug was coincidentally fixed by
5818 increasing the size of one of Exim's internal buffers (big_buffer) that
5819 happened as part of the Exiscan merge. However, to be on the safe side, I
5820 have made the code more robust (and fixed the comments that describe what
5823 33. Some experimental protocols are using DNS PTR records for new purposes. The
5824 keys for these records are domain names, not reversed IP addresses. The
5825 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5826 leaves it alone. Component reversal etc. now happens only for IP addresses.
5829 34. The host_aton() function is supposed to be passed a string that is known
5830 to be a valid IP address. However, in the case of IPv6 addresses, it was
5831 not checking this. This is a hostage to fortune. Exim now panics and dies
5832 if the condition is not met. A case was found where this could be provoked
5833 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5834 components; fortuitously, this particular loophole had already been fixed
5835 by change 4.50/55 or 4.44/33 above.
5837 If there are any other similar loopholes, the new check in host_aton()
5838 itself should stop them being exploited. The report I received stated that
5839 data on the command line could provoke the exploit when Exim was running as
5840 exim, but did not say which command line option was involved. All I could
5841 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5842 running as the user.
5845 35. There was a buffer overflow vulnerability in the SPA authentication code
5846 (which came originally from the Samba project). I have added a test to the
5847 spa_base64_to_bits() function which I hope fixes it.
5850 36. The daemon start-up calls getloadavg() while still root for those OS that
5851 need the first call to be done as root, but it missed one case: when
5852 deliver_queue_load_max is set with deliver_drop_privilege. This is
5853 necessary for the benefit of the queue runner, because there is no re-exec
5854 when deliver_drop_privilege is set.
5856 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5857 was in use, was not putting the data itself into the right store pool;
5858 consequently, it could be overwritten for a subsequent message in the same
5859 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5862 38. Sometimes the final signoff response after QUIT could fail to get
5863 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5864 < 0 before doing a fflush(). This bug looks as though it goes back to the
5865 introduction of TLS in release 3.20, but "sometimes" must have been rare
5866 because the tests only now provoked it.
5872 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5873 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5874 Now it returns an integer. A number of calls were still expecting a T/F
5875 return. Fortuitously, in all cases, the tests worked in OK situations,
5876 which is the norm. However, things would have gone wrong on any write
5877 failures on the smtp file descriptor. This function is used when sending
5878 messages over SMTP and also when doing verify callouts.
5880 2. When Exim is called to do synchronous delivery of a locally submitted
5881 message (the -odf or -odi options), it no longer closes stderr before doing
5884 3. Implemented the mua_wrapper option.
5886 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5888 5. Implemented the functions header_remove(), header_testname(),
5889 header_add_at_position(), and receive_remove_recipient(), and exported them
5892 6. If an ACL "warn" statement specified the addition of headers, Exim already
5893 inserted X-ACL-Warn: at the start if there was no header name. However, it
5894 was not making this test for the second and subsequent header lines if
5895 there were newlines in the string. This meant that an invalid header could
5896 be inserted if Exim was badly configured.
5898 7. Allow an ACL "warn" statement to add header lines at the start or after all
5899 the Received: headers, as well as at the end.
5901 8. Added the rcpt_4xx retry error code.
5903 9. Added postmaster_mailfrom=xxx to callout verification option.
5905 10. Added mailfrom=xxxx to the callout verification option, for verify=
5908 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5909 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5911 12. Inserted some casts to stop certain compilers warning when using pointer
5912 differences as field lengths or precisions in printf-type calls (mostly
5913 affecting debugging statements).
5915 13. Added optional readline() support for -be (dynamically loaded).
5917 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5918 same clock tick as a message's arrival, so that its received time was the
5919 same as the "first fail" time on the retry record, and that message
5920 remained on the queue past the ultimate address timeout, every queue runner
5921 would try a delivery (because it was past the ultimate address timeout) but
5922 after another failure, the ultimate address timeout, which should have then
5923 bounced the address, did not kick in. This was a "< instead of <=" error;
5924 in most cases the first failure would have been in the next clock tick
5925 after the received time, and all would be well.
5927 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5928 being recognized when the domain list was tested by the match_domain
5929 condition in an expansion string.
5931 16. Added the ${str2b64: operator.
5933 17. Exim was always calling setrlimit() to set a large limit for the number of
5934 processes, without checking whether the existing limit was already
5935 adequate. (It did check for the limit on file descriptors.) Furthermore,
5936 errors from getrlimit() and setrlimit() were being ignored. Now they are
5937 logged to the main and panic logs, but Exim does carry on, to try to do its
5938 job under whatever limits there are.
5940 18. Imported PCRE 5.0.
5942 19. Trivial typo in log message " temporarily refused connection" (the leading
5945 20. If the log selector return_path_on_delivery was set and an address was
5946 redirected to /dev/null, the delivery process crashed because it assumed
5947 that a return path would always be set for a "successful" delivery. In this
5948 case, the whole delivery is bypassed as an optimization, and therefore no
5951 21. Internal re-arrangement: the function for sending a challenge and reading
5952 a response while authentication was assuming a zero-terminated challenge
5953 string. It's now changed to take a pointer and a length, to allow for
5954 binary data in such strings.
5956 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5958 23. Exim was not respecting finduser_retries when seeking the login of the
5959 uid under which it was called; it was always trying 10 times. (The default
5960 setting of finduser_retries is zero.) Also, it was sleeping after the final
5961 failure, which is pointless.
5963 24. Implemented tls_on_connect_ports.
5965 25. Implemented acl_smtp_predata.
5967 26. If the domain in control=submission is set empty, Exim assumes that the
5968 authenticated id is a complete email address when it generates From: or
5969 Sender: header lines.
5971 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5972 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5973 chown and chgrp in /bin and hostname in /usr/bin.
5975 28. Exim was keeping the "process log" file open after each use, just as it
5976 does for the main log. This opens the possibility of it remaining open for
5977 long periods when the USR1 signal hits a daemon. Occasional processlog
5978 errors were reported, that could have been caused by this. Anyway, it seems
5979 much more sensible not to leave this file open at all, so that is what now
5982 29. The long-running daemon process does not normally write to the log once it
5983 has entered its main loop, and it closes the log before doing so. This is
5984 so that log files can straightforwardly be renamed and moved. However,
5985 there are a couple of unusual error situations where the daemon does write
5986 log entries, and I had neglected to close the log afterwards.
5988 30. The text of an SMTP error response that was received during a remote
5989 delivery was being truncated at 512 bytes. This is too short for some of
5990 the long messages that one sometimes sees. I've increased the limit to
5993 31. It is now possible to make retry rules that apply only when a message has a
5994 specific sender, in particular, an empty sender.
5996 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5997 it possible to be selective about when SMTP synchronization is enforced.
5999 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6001 32. Implemented hosts_connection_nolog.
6003 33. Added an ACL for QUIT.
6005 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6008 35. Added mailbox_size and mailbox_filecount to appendfile.
6010 36. Added control = no_multiline_responses to ACLs.
6012 37. There was a bug in the logic of the code that waits for the clock to tick
6013 in the case where the clock went backwards by a substantial amount such
6014 that the microsecond fraction of "now" was more than the microsecond
6015 fraction of "then" (but the whole seconds number was less).
6017 38. Added support for the libradius Radius client library this is found on
6018 FreeBSD (previously only the radiusclient library was supported).
6024 1. When certain lookups returned multiple values in the form name=value, the
6025 quoting of the values was not always being done properly. Specifically:
6026 (a) If the value started with a double quote, but contained no whitespace,
6028 (b) If the value contained whitespace other than a space character (i.e.
6029 tabs or newlines or carriage returns) it was not quoted.
6030 This fix has been applied to the mysql and pgsql lookups by writing a
6031 separate quoting function and calling it from the lookup code. The fix
6032 should probably also be applied to nisplus, ibase and oracle lookups, but
6033 since I cannot test any of those, I have not disturbed their existing code.
6035 2. A hit in the callout cache for a specific address caused a log line with no
6036 reason for rejecting RCPT. Now it says "Previous (cached) callout
6037 verification failure".
6039 3. There was an off-by-one bug in the queryprogram router. An over-long
6040 return line was truncated at 256 instead of 255 characters, thereby
6041 overflowing its buffer with the terminating zero. As well as fixing this, I
6042 have increased the buffer size to 1024 (and made a note to document this).
6044 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6045 when Exim is waiting for an SMTP response from a remote server, Exim
6046 restarts its select() call on the socket, thereby resetting its timeout.
6047 This is not a problem when such interrupts are rare. Somebody set up a cron
6048 job to run exiwhat every 2 minutes, which is less than the normal select()
6049 timeout (5 or 10 minutes). This meant that the select() timeout never
6050 kicked in because it was always reset. I have fixed this by comparing the
6051 time when an interrupt arrives with the time at the start of the first call
6052 to select(). If more time than the timeout has elapsed, the interrupt is
6053 treated as a timeout.
6055 5. Some internal re-factoring in preparation for the addition of Sieve
6056 extensions (by MH). In particular, the "personal" test is moved to a
6057 separate function, and given an option for scanning Cc: and Bcc: (which is
6058 not set for Exim filters).
6060 6. When Exim created an email address using the login of the caller as the
6061 local part (e.g. when creating a From: or Sender: header line), it was not
6062 quoting the local part when it contained special characters such as @.
6064 7. Installed new OpenBSD configuration files.
6066 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6067 try to make them clearer.
6069 9. Callout options, other than the timeout value, were being ignored when
6070 verifying sender addresses in header lines. For example, when using
6072 verify = header_sender/callout=no_cache
6074 the cache was (incorrectly) being used.
6076 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6077 only the Cygwin environment.
6079 11. When return_path_on_delivery was set as a log selector, if different remote
6080 addresses in the same message used different return paths and parallel
6081 remote delivery occurred, the wrong values would sometimes be logged.
6082 (Whenever a remote delivery process finished, the return path value from
6083 the most recently started remote delivery process was logged.)
6085 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6086 lines when AUTH and/or TLS are in use. This is the "received protocol"
6087 field. Exim used to use "asmtp" for authenticated SMTP, without any
6088 indication (in the protocol name) for TLS use. Now it follows the RFC and
6089 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6090 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6091 names appear in log lines as well as in Received: header lines.
6093 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6094 extensions, and comparison tests, and to fix some bugs.
6096 14. Changes to the "personal" filter test:
6098 (1) The test was buggy in that it was just doing the equivalent of
6099 "contains" tests on header lines. For example, if a user's address was
6100 anne@some.where, the "personal" test would incorrectly be true for
6102 To: susanne@some.where
6104 This test is now done by extracting each address from the header in turn,
6105 and checking the entire address. Other tests that are part of "personal"
6106 are now done using regular expressions (for example, to check local parts
6107 of addresses in From: header lines).
6109 (2) The list of non-personal local parts in From: addresses has been
6110 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6111 taken from the Sieve specification recommendations.
6113 (3) If the message contains any header line starting with "List-" it is
6114 treated as non-personal.
6116 (4) The test for "circular" in the Subject: header line has been removed
6117 because it now seems ill-conceived.
6119 15. Minor typos in src/EDITME comments corrected.
6121 16. Installed latest exipick from John Jetmore.
6123 17. If headers_add on a router specified a text string that was too long for
6124 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6125 of string_sprintf() is now avoided.
6127 18. $message_body_size was not set (it was always zero) when running the DATA
6128 ACL and the local_scan() function.
6130 19. For the "mail" command in an Exim filter, no default was being set for
6131 the once_repeat time, causing a random time value to be used if "once" was
6132 specified. (If the value happened to be <= 0, no repeat happened.) The
6133 default is now 0s, meaning "never repeat". The "vacation" command was OK
6134 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6135 (I found it when inspecting the code).
6137 20. There is now an overall timeout for performing a callout verification. It
6138 defaults to 4 times the callout timeout, which applies to individual SMTP
6139 commands during the callout. The overall timeout applies when there is more
6140 than one host that can be tried. The timeout is checked before trying the
6141 next host. This prevents very long delays if there are a large number of
6142 hosts and all are timing out (e.g. when the network connections are timing
6143 out). The value of the overall timeout can be changed by specifying an
6144 additional sub-option for "callout", called "maxwait". For example:
6146 verify = sender/callout=5s,maxwait=20s
6148 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6149 to the end before writing, but this should make it even safer).
6151 22. Exim was forgetting that it had advertised PIPELINING for the second and
6152 subsequent messages on an SMTP connection. It was also not resetting its
6153 memory on STARTTLS and an internal HELO.
6155 23. When Exim logs an SMTP synchronization error within a session, it now
6156 records whether PIPELINING has been advertised or not.
6158 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6159 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6160 rather than long int.
6162 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6164 26. Added the never_mail option to autoreply.
6170 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6171 crash if the getsockname() call failed; this can happen if a connection is
6172 closed very soon after it is established. The problem was simply in the
6173 order in which certain operations were done, causing Exim to try to write
6174 to the SMTP stream before it had set up the file descriptor. The bug has
6175 been fixed by making things happen in the correct order.
6181 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6182 before the connection was closed, thus losing the rejection response.
6184 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6185 some early Solaris releases, but causes trouble in current releases where
6186 socklen_t is defined.
6188 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6191 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6194 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6195 that releasing the top of it at the end releases what was used for sub-
6196 expansions (unless the block got too big). However, discard this block if
6197 the first thing is a variable or header, so that we can use its block when
6198 it is dynamic (useful for very large $message_headers, for example).
6200 6. Lookups now cache *every* query, not just the most recent. A new, separate
6201 store pool is used for this. It can be recovered when all lookup caches are
6202 flushed. Lookups now release memory at the end of their result strings.
6203 This has involved some general refactoring of the lookup sources.
6205 7. Some code has been added to the store_xxx() functions to reduce the amount
6206 of flapping under certain conditions.
6208 8. log_incoming_interface used to affect only the <= reception log lines. Now
6209 it causes the local interface and port to be added to several more SMTP log
6210 lines, for example "SMTP connection from", and rejection lines.
6212 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6214 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6216 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6217 could be overwritten at the end of the current message (or the start of a
6218 new message if it was set in a HELO ACL). The value is now preserved for
6219 the duration of the SMTP connection.
6221 12. If a transport had a headers_rewrite setting, and a matching header line
6222 contained an unqualified address, that address was qualified, even if it
6223 did not match any rewriting rules. The underlying bug was that the values
6224 of the flags that permit the existence of unqualified sender and recipient
6225 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6226 non-local messages, and by -bnq for local messages) were not being
6227 preserved with the message after it was received.
6229 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6230 "next input=" as part of the text comprising the host identity instead of
6231 the correct text. The code was using the same buffer for two different
6232 strings. However, depending on which order the printing function evaluated
6233 its arguments, the bug did not always show up. Under Linux, for example, my
6234 test suite worked just fine.
6236 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6237 doesn't work with some older versions of Perl. It has been changed to "my",
6238 which in any case is probably the better facility to use.
6240 15. A really picky compiler found some instances of statements for creating
6241 error messages that either had too many or two few arguments for the format
6244 16. The size of the buffer for calls to the DNS resolver has been increased
6245 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6246 for addresses that have a lot of PTR records. This alleviates a problem; it
6247 does not fully solve it.
6249 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6250 buffer now truncates the list and logs the incident, which is the same
6251 action as happens when Exim is looking up a host name and its aliases.
6252 Previously in this situation something unpredictable would happen;
6253 sometimes it was "internal error: store_reset failed".
6255 18. If a server dropped the connection unexpectedly when an Exim client was
6256 using GnuTLS and trying to read a response, the client delivery process
6257 crashed while trying to generate an error log message.
6259 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6260 string, for example:
6262 warn message = H1: something\nH2: something
6264 the text was added as a single header line from Exim's point of view
6265 though it ended up OK in the delivered message. However, searching for the
6266 second and subsequent header lines using $h_h2: did not work. This has been
6267 fixed. Similarly, if a system filter added multiple headers in this way,
6268 the routers could not see them.
6270 20. Expanded the error message when iplsearch is called with an invalid key to
6271 suggest using net-iplsearch in a host list.
6273 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6274 ACLs are no longer actually imposed (and a message to that effect is
6277 22. If a "gecos" field in a passwd entry contained escaped characters, in
6278 particular, if it contained a \" sequence, Exim got it wrong when building
6279 a From: or a Sender: header from that name. A second bug also caused
6280 incorrect handling when an unquoted " was present following a character
6281 that needed quoting.
6283 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6284 was not being matched caselessly.
6286 24. Arranged for all hyphens in the exim.8 source to be escaped with
6289 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6290 properly. Recipient callout cache records were still being keyed to include
6291 the sender, even when use_sender was set false. This led to far more
6292 callouts that were necessary. The sender is no longer included in the key
6293 when use_sender is false.
6295 26. Added "control = submission" modifier to ACLs.
6297 27. Added the ${base62d: operator to decode base 62 numbers.
6299 28. dnsdb lookups can now access SRV records.
6301 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6302 the configuration file.
6304 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6305 (-v) mode. This makes the output for a verbose queue run more intelligible.
6307 31. Added a use_postmaster feature to recipient callouts.
6309 32. Added the $body_zerocount variable, containing the number of binary zero
6310 bytes in the message body.
6312 33. The time of last modification of the "new" subdirectory is now used as the
6313 "mailbox time last read" when there is a quota error for a maildir
6316 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6318 35. Added +ignore_unknown as a special item in host lists.
6320 36. Code for decoding IPv6 addresses in host lists is now included, even if
6321 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6322 address was recognized as an IP address, but was then not correctly decoded
6323 into binary, causing unexpected and incorrect effects when compared with
6330 1. Very minor rewording of debugging text in manualroute to say "list of
6331 hosts" instead of "hostlist".
6333 2. If verify=header_syntax was set, and a header line with an unqualified
6334 address (no domain) and a large number of spaces between the end of the
6335 name and the colon was received, the reception process suffered a buffer
6336 overflow, and (when I tested it) crashed. This was caused by some obsolete
6337 code that should have been removed. The fix is to remove it!
6339 3. When running in the test harness, delay a bit after writing a bounce
6340 message to get a bit more predictability in the log output.
6342 4. Added a call to search_tidyup() just before forking a reception process. In
6343 theory, someone could use a lookup in the expansion of smtp_accept_max_
6344 per_host which, without the tidyup, could leave open a database connection.
6346 5. Added the variables $recipient_data and $sender_data which get set from a
6347 lookup success in an ACL "recipients" or "senders" condition, or a router
6348 "senders" option, similar to $domain_data and $local_part_data.
6350 6. Moved the writing of debug_print from before to after the "senders" test
6353 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6354 problems for message scanning, either using a data ACL, or using
6355 local_scan() because the Received: header was not generated till after they
6356 were called (in order to set the time as the time of reception completion).
6357 I have revised the way this works. The header is now generated after the
6358 body is received, but before the ACL or local_scan() are called. After they
6359 are run, the timestamp in the header is updated.
6365 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6366 before starting a queue runner without re-exec. This happened only when
6367 deliver_drop_privilege was set or when the Exim user was set to root. The
6368 effect of the bug was that timeouts during subsequent deliveries caused
6369 crashes instead of being properly handled. The handler is now left at its
6370 default (and expected) setting.
6372 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6373 message, again when deliver_drop_privilege is set or Exim is run as root.
6374 The bug described in (1) was not present in this case, but the tidying up
6375 of the other signals was missing. I have made the two cases consistent.
6377 3. The ignore_target_hosts setting on a manualroute router was being ignored
6378 for hosts that were looked up using the /MX notation.
6380 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6383 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6384 operated on the sender address. After changing the $sender_address to <>
6385 for the sender address verify, Exim was re-instated it as the original
6386 (before rewriting) address, but remembering that it had rewritten it, so it
6387 wasn't rewriting it again. This bug also had the effect of breaking the
6388 sender address verification caching when the sender address was rewritten.
6390 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6391 This has been changed so that if the ip literal address matches
6392 ignore_target_hosts, the router declines.
6394 7. Added expansion conditions match_domain, match_address, and match_local_
6395 part (NOT match_host).
6397 8. The placeholder for the Received: header didn't have a length field set.
6399 9. Added code to Exim itself and to exim_lock to test for a specific race
6400 condition that could lead to file corruption when using MBX delivery. The
6401 issue is with the lockfile that is created in /tmp. If this file is removed
6402 after a process has opened it but before that process has acquired a lock,
6403 there is the potential for a second process to recreate the file and also
6404 acquire a lock. This could lead to two Exim processes writing to the file
6405 at the same time. The added code performs the same test as UW imapd; it
6406 checks after acquiring the lock that its file descriptor still refers to
6407 the same named file.
6409 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6410 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6413 11. Added the smtp_active_hostname option. If used, this will typically be made
6414 to depend on the incoming interface address. Because $interface_address is
6415 not set up until the daemon has forked a reception process, error responses
6416 that can happen earlier (such as "too many connections") no longer contain
6419 12. If an expansion in a condition on a "warn" statement fails because a lookup
6420 defers, the "warn" statement is abandoned, and the next ACL statement is
6421 processed. Previously this caused the whole ACL to be aborted.
6423 13. Added the iplsearch lookup type.
6425 14. Added ident_timeout as a log selector.
6427 15. Added tls_certificate_verified as a log selector.
6429 16. Added a global option tls_require_ciphers (compare the smtp transport
6430 option of the same name). This controls incoming TLS connections.
6432 17. I finally figured out how to make tls_require_ciphers do a similar thing
6433 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6434 before starting the TLS session.
6436 18. Tabs are now shown as \t in -bP output.
6438 19. If the log selector return_path_on_delivery was set, Exim crashed when
6439 bouncing a message because it had too many Received: header lines.
6441 20. If two routers both had headers_remove settings, and the first one included
6442 a superfluous trailing colon, the final name in the first list and the
6443 first name in the second list were incorrectly joined into one item (with a
6444 colon in the middle).
6450 1. Added -C and -D options to the exinext utility, mainly to make it easier
6451 to include in the automated testing, but these could be helpful when
6452 multiple configurations are in use.
6454 2. The exinext utility was not formatting the output nicely when there was
6455 an alternate port involved in the retry record key, nor when there was a
6456 message id as well (for retries that were specific to a specific message
6457 and a specific host). It was also confused by IPv6 addresses, because of
6458 the additional colons they contain. I have fixed the IPv4 problem, and
6459 patched it up to do a reasonable job for IPv6.
6461 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6462 delivery, the log line now contains "pipelined" if PIPELINING was used.
6464 4. An SMTP transport process used to panic and die if the bind() call to set
6465 an explicit outgoing interface failed. This has been changed; it is now
6466 treated in the same way as a connect() failure.
6468 5. A reference to $sender_host_name in the part of a conditional expansion
6469 that was being skipped was still causing a DNS lookup. This no longer
6472 6. The def: expansion condition was not recognizing references to header lines
6473 that used bh_ and bheader_.
6475 7. Added the _cache feature to named lists.
6477 8. The code for checking quota_filecount in the appendfile transport was
6478 allowing one more file than it should have been.
6480 9. For compatibility with Sendmail, the command line option
6488 and sets the incoming protocol and host name (for trusted callers). The
6489 host name and its colon can be omitted when only the protocol is to be set.
6490 Note the Exim already has two private options, -pd and -ps, that refer to
6491 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6492 "s", but I don't think that's a major issue.
6494 10. A number of refactoring changes to the code, none of which should affect
6497 (a) The number of logging options was getting close to filling up the
6498 32-bit word that was used as a bit map. I have split them into two classes:
6499 those that are passed in the argument to log_write(), and those that are
6500 only ever tested independently outside of that function. These are now in
6501 separate 32-bit words, so there is plenty of room for expansion again.
6502 There is no change in the user interface or the logging behaviour.
6504 (b) When building, for example, log lines, the code previously used a
6505 macro that called string_cat() twice, in order to add two strings. This is
6506 not really sufficiently general. Furthermore, there was one instance where
6507 it was actually wrong because one of the argument was used twice, and in
6508 one call a function was used. (As it happened, calling the function twice
6509 did not affect the overall behaviour.) The macro has been replaced by a
6510 function that can join an arbitrary number of extra strings onto a growing
6513 (c) The code for expansion conditions now uses a table and a binary chop
6514 instead of a serial search (which was left over from when there were very
6515 few conditions). Also, it now recognizes conditions like "pam" even when
6516 the relevant support is not compiled in: a suitably worded error message is
6517 given if an attempt is made to use such a condition.
6519 11. Added ${time_interval:xxxxx}.
6521 12. A bug was causing one of the ddress fields not to be passed back correctly
6522 from remote delivery subprocesses. The field in question was not being
6523 subsequently used, so this caused to problems in practice.
6525 13. Added new log selectors queue_time and deliver_time.
6527 14. Might have fixed a bug in maildirsizefile handling that threw up
6528 "unexpected character" debug warnings, and recalculated the data
6529 unnecessarily. In any case, I expanded the warning message to give more
6532 15. Added the message "Restricted characters in address" to the statements in
6533 the default ACL that block characters like @ and % in local parts.
6535 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6536 Three changes have been made:
6538 (a) There was a serious bug; a negative response to MAIL caused the whole
6539 recipient domain to be cached as invalid, thereby blocking all messages
6540 to all local parts at the same domain, from all senders. This bug has
6541 been fixed. The domain is no longer cached after a negative response to
6542 MAIL if the sender used is not empty.
6544 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6547 (c) A new callout option, "use_sender" has been added for people who want
6548 the modified behaviour.
6554 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6557 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6558 indeed breaks things for older releases.
6560 3. Added additional logging to the case where there is a problem reading data
6561 from a filter that is running in a subprocess using a pipe, in order to
6562 try to track down a specific problem.
6564 4. Testing facility fudge: when running in the test harness and attempting
6565 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6566 getting "No route to host". Convert this to a timeout.
6568 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6571 6. Some OS don't have socklen_t but use size_t instead. This affects the
6572 fifth argument of getsockopt() amongst other things. This is now
6573 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6574 can be set for individual OS. I have set it for SunOS5, OSF1, and
6575 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6576 some earlier ones do not.
6578 7. Change 4.30/15 was not doing the test caselessly.
6580 8. The standard form for an IPv6 address literal was being rejected by address
6581 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6582 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6583 this, as well as the form without the "IPv6" on the front (but only when
6584 address literals are enabled, of course).
6586 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6588 10. Exim crashed if a message with an empty sender address specified by -f
6589 encountered a router with an errors_to setting. This could be provoked only
6590 by a command such as
6594 where an empty string was supplied; "<>" did not hit this bug.
6596 11. Installed PCRE release 4.5.
6598 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6599 remained set. It is now erased.
6601 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6602 times from message ids (which are base 36 rather than the normal 62).
6604 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6605 were being counted as actual protocol errors, and logged if the log
6606 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6607 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6608 and DATA following a set of rejected RCPTs do not count as protocol errors.
6609 In other words, Exim assumes they were pipelined, though this may not
6610 actually be the case. Of course, in all cases the client gets an
6611 appropriate error code.
6613 15. If a lookup fails in an ACL condition, a message about the failure may
6614 be available; it is used if testing the ACL cannot continue, because most
6615 such messages specify what the cause of the deferral is. However, some
6616 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6617 that caused an old message to be retained and used if a later statement
6618 caused a defer, replacing the real cause of the deferral.
6620 16. If an IP address had so many PTR records that the DNS lookup buffer
6621 was not large enough to hold them, Exim could crash while trying to process
6622 the truncated data. It now detects and logs this case.
6624 17. Further to 4.21/58, another change has been made: if (and only if) the
6625 first line of a message (the first header line) ends with CRLF, a bare LF
6626 in a subsequent header line has a space inserted after it, so as not to
6627 terminate the header.
6629 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6630 unnecessarily, used atoi() instead of strtol(), and didn't check the
6631 termination when getting file sizes from file names by regex.
6633 19. Completely re-implemented the support for maildirsize files, in the light
6634 of a number of problems with the previous contributed implementation
6635 (4.30/29). In particular:
6637 . If the quota is zero, the maildirsize file is maintained, but no quota is
6640 . If the maildir directory does not exist, it is created before any attempt
6641 to write a maildirsize file.
6643 . The quota value in the file is just a cache; if the quota is changed in
6644 the transport, the new value overrides.
6646 . A regular expression is available for excluding directories from the
6649 20. The autoreply transport checks the characters in options that define the
6650 message's headers; it allows continued headers, but it was checking with
6651 isspace() after an embedded newline instead of explicitly looking for a
6654 21. If all the "regular" hosts to which an address was routed had passed their
6655 expiry times, and had not reached their retry times, the address was
6656 bounced, even if fallback hosts were defined. Now Exim should go on to try
6659 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6660 equivalent code in the SMTP transport. Some hosts send humungous responses
6661 to HELO/EHLO, more than 1024 it seems.
6663 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6664 gives compiler warnings in some environments. I've now done it "properly",
6667 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6668 (because of problems with the built-in one) was declared to return uschar *
6669 instead of char *, causing compiler failure.
6671 25. Fixed a file descriptor leak when processing alias/forward files.
6673 26. Fixed a minor format string issue in dbfn.c.
6675 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6677 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6678 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6681 29. When a sender address is verified, it is cached, to save repeating the test
6682 when there is more than one recipient in a message. However, when the
6683 verification involves a callout, it is possible for different callout
6684 options to be set for different recipients. It is too complicated to keep
6685 track of this in the cache, so now Exim always runs a verification when a
6686 callout is required, relying on the callout cache for the optimization.
6687 The overhead is duplication of the address routing, but this should not be
6690 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6691 to be verified with callout=postmaster, and the main callout worked but the
6692 postmaster check failed, the verification correctly failed. However, if a
6693 subsequent RCPT command asked for sender verification *without* the
6694 postmaster check, incorrect caching caused this verification also to fail,
6697 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6698 it was not caching the DNS options (qualify_single, search_parents) that
6699 were used when the lookup failed. A subsequent lookup with different
6700 options therefore always gave the same answer, though there were cases
6701 where it should not have. (Example: a "domains = !$mx_any" option on a
6702 dnslookup router: the "domains" option is always processed without any
6703 widening, but the router might have qualify_single set.) Now Exim uses the
6704 cached value only when the same options are set.
6706 32. Added John Jetmore's "exipick" utility to the distribution.
6708 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6709 than a timeout (e.g. a certificate is required, and is not provided), an
6710 Exim server now closes the connection immediately. Previously it waited for
6711 the client to close - but if the client is SSL, it seems that they each
6712 wait for each other, leading to a delay before one of them times out.
6714 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6715 maintained 0.8.x compatibility because I don't think many are using it, and
6716 it is clearly obsolete.
6718 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6721 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6722 client certificate was expired. A simple patch fixes this, though I don't
6723 understand the full logic of why the verify callback is called multiple
6726 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6727 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6728 which causes problems with some clients (such as the Certicom SSL Plus
6729 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6730 disables the coutermeasure allowing Eudora to connect."
6732 38. Exim was not checking that a write() to a log file succeeded. This could
6733 lead to Bad Things if a log got too big, in particular if it hit a file
6734 size limit. Exim now panics and dies if it cannot write to a log file, just
6735 as it does if it cannot open a log file.
6737 39. Modified OS/Makefile-Linux so that it now contains
6739 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6741 The two -D definitions ensure that Exim is compiled with large file
6742 support, which makes it possible to handle log files that are bigger than
6745 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6746 instance) a domain was checked against a named list that involved a lookup,
6747 causing $domain_data to be set, then another domain was checked against the
6748 same list, then the first domain was re-checked, the value of $domain_data
6749 after the final check could be wrong. In particular, if the second check
6750 failed, it could be set empty. This bug probably also applied to
6753 41. The strip_trailing_dot option was not being applied to the address given
6754 with the -f command-line option.
6756 42. The code for reading a message's header from the spool was incrementing
6757 $received_count, but never initializing it. This meant that the value was
6758 incorrect (doubled) while delivering a message in the same process in which
6759 it was received. In the most common configuration of Exim, this never
6760 happens - a fresh exec is done - but it can happen when
6761 deliver_drop_privilege is set.
6763 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6764 it now includes up to 150 characters of the unexpected data in the log
6767 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6768 and building data strings. The size of both of these buffers was 10 000
6769 bytes - far larger than anybody would *ever* want, thought I. Needless to
6770 say, somebody hit the limit. I have increased the maximum line length to
6771 20 000 and the maximum data length of concatenated lines to 100 000. I have
6772 also fixed two bugs, because there was no checking on these buffers. Tsk,
6773 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6774 buffer is too small.
6776 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6777 lsearch lookups. Now it does.
6779 46. When parsing a route_list item in a manualroute router, a fixed-length
6780 buffer was used for the list of hosts. I made this 1024 bytes long,
6781 thinking that nobody would ever have a list of hosts that long. Wrong.
6782 Somebody had a whole pile of complicated expansion conditions, and the
6783 string was silently truncated, leading to an expansion error. It turns out
6784 that it is easier to change to an unlimited length (owing to other changes
6785 that have happened since this code was originally written) than to build
6786 structure for giving a limitation error. The length of the item that
6787 expands into the list of hosts is now unlimited.
6789 47. The lsearch lookup could not handle data where the length of text line was
6790 more than 4095 characters. Such lines were truncated, leading to shortened
6791 data being returned. It should now handle lines of any length.
6793 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6794 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6797 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6798 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6799 no longer happens in comment lines. A list of replacements is now placed
6800 at the head of all of the source files, except those whose only change is
6801 to replace PERL_COMMAND in the very first #! line.
6803 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6804 messages on the queue, with a bottom-up merge sort, using code contributed
6805 by Michael Haardt. This should make operations like -bp somewhat faster on
6806 large queues. It won't affect queue runners, except when queue_run_in_order
6809 51. Installed eximstats 1.31 in the distribution.
6811 52. Added support for SRV lookups to the dnslookup router.
6813 53. If an ACL referred to $message_body or $message_body_end, the value was not
6814 reset for any messages that followed in the same SMTP session.
6816 54. The store-handling optimization for building very long strings was not
6817 differentiating between the different store pools. I don't think this
6818 actually made any difference in practice, but I've tidied it.
6820 55. While running the routers to verify a sender address, $sender_address
6821 was still set to the sender address. This is wrong, because when routing to
6822 send a bounce to the sender, it would be empty. Therefore, I have changed
6823 it so that, while verifying a sender address, $sender_address is set to <>.
6824 (There is no change to what happens when verifying a recipient address.)
6826 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6827 target A or AAAA records (if not already returned) without resetting the
6828 qualify_single or search_parents options of the DNS resolver. These are
6829 inappropriate in this case because the targets of MX and SRV records must
6830 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6831 target that, when qualified, matched something in the local domain. These
6832 two options are now turned off when doing these lookups.
6834 57. It seems that at least some releases of Reiserfs (which does not have the
6835 concept of a fixed number of inodes) returns zero and not -1 for the
6836 number of available inodes. This interacted badly with check_spool_inodes,
6837 which assumed that -1 was the "no such thing" setting. What I have done is
6838 to check that the total number of inodes is greater than zero before doing
6839 the test of how many are available.
6841 58. When a "warn" ACL statement has a log_message modifier, the message is
6842 remembered, and not repeated. This is to avoid a lot of repetition when a
6843 message has many recipients that cause the same warning to be written.
6844 However, Exim was preserving the list of already written lines for an
6845 entire SMTP session, which doesn't seem right. The memory is now reset if a
6846 new message is started.
6848 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6849 debugging output unless log_rewrite was also set.
6851 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6852 of (int)(handle) when we know that handle contains (void *)(-1).
6854 61. The Exim daemon panic-logs an error return when it closes the incoming
6855 connection. However "connection reset by peer" seems to be common, and
6856 isn't really an error worthy of noting specially, so that particular error
6859 62. When Exim is trying to find all the local interfaces, it used to panic and
6860 die if the ioctl to get the interface flags failed. However, it seems that
6861 on at least one OS (Solaris 9) it is possible to have an interface that is
6862 included in the list of interfaces, but for which you get a failure error
6863 for this call. This happens when the interface is not "plumbed" into a
6864 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6865 failure of the "get flags" call assumes that the interface is down.
6867 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6868 makes life easier for people who are doing arithmetic on fields extracted
6869 from dates, where you often get leading zeros that should not be
6870 interpreted as octal.
6872 64. Added qualify_domain to the redirect router, to override the global
6875 65. If a pathologically long header line contained very many addresses (the
6876 report of this problem mentioned 10 000) and each of them was rewritten,
6877 Exim could use up a very large amount of memory. (It kept on making new
6878 copies of the header line as it rewrote, and never released the old ones.)
6879 At the expense of a bit more processing, the header rewriting function has
6880 been changed so that it no longer eats memory in this way.
6882 66. The generation of the Received: header has been moved from the time that a
6883 message starts to be received, to the time that it finishes. The timestamp
6884 in the Received: header should now be very close to that of the <= log
6885 line. There are two side-effects of this change:
6887 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6888 logged header lines no longer include the local Received: line, because
6889 it has not yet been created. The same applies to a copy of the message
6890 that is returned to a non-SMTP sender when a message is rejected.
6892 (b) When a filter file is tested using -bf, no additional Received: header
6893 is added to the test message. After some thought, I decided that this
6896 This change does not affect the value of $received_for. It is still set
6897 after address rewriting, but before local_scan() is called.
6899 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6901 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6902 gave an unhelpful panic error message, and a defer error. I have managed to
6903 change this behaviour so that it now rejects any supplied certificate,
6904 which seems right, as the list of acceptable certificates is empty.
6906 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6907 gave an unhelpful defer error. I have not managed to make this reject any
6908 supplied certificates, but the error message it gives is "no certificate
6909 supplied", which is not helpful.
6911 70. exigrep's output now also includes lines that are not associated with any
6912 message, but which match the given pattern. Implemented by a patch from
6913 Martin Sluka, which also tidied up the Perl a bit.
6915 71. Recipient callout verification, like sender verification, was using <> in
6916 the MAIL FROM command. This isn't really the right thing, since the actual
6917 sender may affect whether the remote host accepts the recipient or not. I
6918 have changed it to use the actual sender in the callout; this means that
6919 the cache record is now keyed on a recipient/sender pair, not just the
6920 recipient address. There doesn't seem to be a real danger of callout loops,
6921 since a callout by the remote host to check the sender would use <>.
6922 [SEE ABOVE: changed after hitting problems.]
6924 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6925 temporary errors. However, in the case of such a code being given after
6926 the end of a data transmission (i.e. after ".") Exim was failing to write
6927 a retry record for the message. (Yes, there was some broken host that was
6928 actually sending 8xx at this point.)
6930 73. An unknown lookup type in a host list could cause Exim to panic-die when
6931 the list was checked. (An example that provoked this was putting <; in the
6932 middle of a list instead of at the start.) If this happened during a DATA
6933 ACL check, a -D file could be left lying around. This kind of configuration
6934 error no longer causes Exim to die; instead it causes a defer error. The
6935 incident is still logged to the main and panic logs.
6937 74. Buglet left over from Exim 3 conversion. The message "too many messages
6938 in one connection" was written to the rejectlog but not the mainlog, except
6939 when address rewriting (yes!) was being logged.
6941 75. Added write_rejectlog option.
6943 76. When a system filter was run not as root (that is, when system_filter_user
6944 was set), the values of the $n variables were not being returned to the
6945 main process; thus, they were not subsequently available in the $sn
6948 77. Added +return_path_on_delivery log selector.
6950 78. A connection timeout was being treated differently from recipients deferred
6951 when testing hosts_max_try with a message that was older than the host's
6952 retry timeout. (The host should not be counted, thus allowing all hosts to
6953 be tried at least once before bouncing.) This may have been the cause of an
6954 occasionally reported bug whereby a message would remain on the queue
6955 longer than the retry timeout, but would be bounced if a delivery was
6956 forced. I say "may" because I never totally pinned down the problem;
6957 setting up timeout/retry tests is difficult. See also the next item.
6959 79. The ultimate address timeout was not being applied to errors that involved
6960 a combination of host plus message (for example, a timeout on a MAIL
6961 command). When an address resolved to a number of possible hosts, and they
6962 were not all tried for each delivery (e.g. because of hosts_max_try), a
6963 message could remain on the queue longer than the retry timeout.
6965 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6968 81. Fixed an obscure SMTP outgoing bug which required at least the following
6969 conditions: (a) there was another message waiting for the same server;
6970 (b) the server returned 5xx to all RCPT commands in the first message so
6971 that the message was not completed; (c) the server dropped the connection
6972 or gave a negative response to the RSET that Exim sends to abort the
6973 transaction. The observed case was a dropped connection after DATA that had
6974 been sent in pipelining mode. That is, the server had advertised PIPELINING
6975 but was not implementing it correctly. The effect of the bug was incorrect
6976 behaviour, such as trying another host, and this could lead to a crash.
6982 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6983 and daemon.c were passed as pointers to ints; they should have been
6984 pointers to socklen_t variables (which are typically unsigned ints).
6986 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6989 3. Fixed a really odd bug that affected only the testing scheme; patching a
6990 certain fixed string in the binary changed the value of another string that
6991 happened to be identical to the end of the original first string.
6993 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6994 name", it returns that address as the IP address. On some operating
6995 systems (e.g. Solaris), it also passes back the IP address string as the
6996 "host name". However, on others (e.g. Linux), it passes back an empty
6997 string. Exim wasn't checking for this, and was changing the host name to an
6998 empty string, assuming it had been canonicalized.
7000 5. Although rare, it is permitted to have more than one PTR record for a given
7001 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7002 all the names associated with an address, because they do in Solaris.
7003 However, it seems that they do not in Linux for data that comes from the
7004 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7005 I found this out when I moved to a new Linux workstation and tried to run
7006 the Exim test suite.
7008 To get round this problem I have changed the code so that it now does its
7009 own call to the DNS to look up PTR records when searching for a host name.
7010 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7011 addresses that are only in /etc/hosts are still found.
7013 This behaviour is, however, controlled by an option called host_lookup_
7014 order, which defaults to "bydns:byaddr". If people want to use the other
7015 order, or indeed, just use one or the other means of lookup, they can
7016 specify it in this variable.
7018 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7019 some operating systems, this comes back from gethostbyaddr() as an empty
7020 string, and this is what Exim used to test for. However, it seems that in
7021 other systems, "." is yielded. Exim now tests for this case too.
7023 7. The values of check_spool_space and check_log_space are now held internally
7024 as a number of kilobytes instead of an absolute number of bytes. If a
7025 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7026 kilobyte. This means that much larger values can be stored.
7028 8. Exim monitor: an attempt to get the action menu when not actually pointing
7029 at a message produces an empty menu entitled "No message selected". This
7030 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7031 no entries in it ("Shell widget menu has zero width and/or height"). So I
7032 have added a single, blank menu entry in this case.
7034 9. Added ${quote_local_part.
7036 10. MIME decoding is now applied to the contents of Subject: header lines when
7039 11. Now that a reference to $sender_host_address automatically causes a reverse
7040 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7041 host lookup before query-style lookups in lists that might use this
7042 variable. This has therefore been abolished, and the "net-" prefix is no
7043 longer necessary for query-style lookups.
7045 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7046 have been a typo for LFLAGS, so it has been changed.
7048 13. The install script calls Exim with "-C /dev/null" in order to find the
7049 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7050 to be output. However, since Exim outputs its version number before the
7051 error, it didn't break the script. It just looked ugly. I fixed this by
7052 always allowing "-C /dev/null" if the caller is root.
7054 14. Ignore overlarge ACL variable number when reading spool file - insurance
7055 against a later release with more variables having written the file.
7057 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7058 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7059 this, as well as the form without the "IPv6" on the front.
7061 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7062 OS/Makefile-Darwin file.
7064 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7065 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7067 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7068 variable. However, some OS use ints for inodes. I've added cast to long int
7069 to get rid of the compiler warning.
7071 19. I had forgotten to lock out "/../" in configuration file names when
7072 ALT_CONFIG_PREFIX was set.
7074 20. Routers used for verification do not need to specify transports. However,
7075 if such a router generated a host list, and callout was configured, Exim
7076 crashed, because it could not find a port number from the (non-existent)
7077 transport. It now assumes port 25 in this circumstance.
7079 21. Added the -t option to exigrep.
7081 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7082 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7083 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7084 them is set, LOOKUP_LSEARCH is forced.
7086 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7088 24. Added sender and host information to the "rejected by local_scan()" log
7089 line; previously there was no indication of these.
7091 25. Added .include_if_exists.
7093 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7094 Linux. It turns out that not all file systems support this. Apparently some
7095 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7096 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7097 not supported on the file descriptor, is now ignored when Exim is trying to
7098 sync a directory. This applies only to Linux.
7100 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7102 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7103 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7106 29. Added support for maildirsize files from supplied patch (modified a bit).
7108 30. The use of :fail: followed by an empty string could lead Exim to respond to
7109 sender verification failures with (e.g.):
7111 550 Verification failed for <xxx>
7112 550 Sender verify failed
7114 where the first response line was missing the '-' that indicates it is not
7115 the final line of the response.
7117 31. The loop for finding the name of the user that called Exim had a hardwired
7118 limit of 10; it now uses the value of finduser_retries, which is used for
7119 all other user lookups.
7121 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7124 33. Exim was neglecting to zero errno before one call of strtol() when
7125 expanding a string and expecting an integer value. On some systems this
7126 resulted in spurious "integer overflow" errors. Also, it was casting the
7127 result into an int without checking.
7129 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7130 did not work. The code looks as if it has *never* worked, though it appears
7131 to have been documented since at least release 1.62. I have made it work.
7133 35. The "timeout_DNS" error in retry rules, also documented since at least
7134 1.62, also never worked. As it isn't clear exactly what this means, and
7135 clearly it isn't a major issue, I have abolished the feature by treating it
7136 as "timeout", and writing a warning to the main and panic logs.
7138 36. The display of retry rules for -brt wasn't always showing the error code
7141 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7142 timeout_connect_A, timeout_connect_MX.
7144 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7145 to the empty sender.
7147 39. The daemon was not analysing the content of -oX till after it had closed
7148 stderr and disconnected from the controlling terminal. This meant that any
7149 syntax errors were only noted on the panic log, and the return code from
7150 the command was 0. By re-arranging the code a little, I've made the
7151 decoding happen first, so such errors now appear on stderr, and the return
7152 code is 1. However, the actual setting up of the sockets still happens in
7153 the disconnected process, so errors there are still only recorded on the
7156 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7157 connections (as happens on some IP stacks) was logged at start up time as
7158 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7159 it from "IPv6 and IPv4", which means that two separate sockets are being
7162 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7163 says whether AF_INET or AF_INET6 was passed as an argument.
7165 42. Exiwhat output was messed up when time zones were included in log
7168 43. Exiwhat now gives more information about the daemon's listening ports,
7169 and whether -tls-on-connect was used.
7171 44. The "port" option of the smtp transport is now expanded.
7173 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7174 silently ignored. Now an error message is written to the main and panic
7177 46. There's a new ACL modifier called "logwrite" which writes to a log file
7178 as soon as it is encountered.
7180 47. Added $local_user_uid and $local_user_gid at routing time.
7182 48. Exim crashed when trying to verify a sender address that was being
7185 49. Exim was recognizing only a space character after ".include". It now also
7186 recognizes a tab character.
7188 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7189 extracting the relevant information from the specification. The man page no
7190 longer contains scrambled data for the -d option, and I've added a section
7191 at the front about calling Exim under different names.
7193 51. Added "extra_headers" argument to the "mail" command in filter files.
7195 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7198 53. Installed eximstats 1.29.
7200 54. Added transport_filter_timeout as a generic transport option.
7202 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7203 Cc: header lines. This was required by RFC 822, but it not required by RFC
7206 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7207 incoming messages that did not have them. Now it does so only if the
7208 message originates locally, that is, if there is no associated remote host
7209 address. When Resent- header lines are present, this applies to the Resent-
7210 lines rather than the non-Resent- lines.
7212 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7213 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7215 58. Messages for configuration errors now include the name of the main
7216 configuration file - useful now that there may be more than one file in a
7217 list (.included file names were always shown).
7219 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7220 for those rare installations that do not start the daemon as root or run it
7221 setuid root. I've cut out the call to initgroups() if the daemon is not
7224 60. The Exim user and group can now be bound into the binary as text strings
7225 that are looked up at the start of Exim's processing.
7227 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7229 62. Added $mailstore_basename variable.
7231 63. Installed patch to sieve.c from Michael Haardt.
7233 64. When Exim failed to open the panic log after failing to open the main log,
7234 the original message it was trying to log was written to stderr and debug
7235 output, but if they were not available (the usual case in production), it
7236 was lost. Now it is written to syslog before the two lines that record the
7237 failures to open the logs.
7239 65. Users' Exim filters run in subprocesses under the user's uid. It is
7240 possible for a "deliver" command or an alias in a "personal" command to
7241 provoke an address rewrite. If logging of address rewriting is configured,
7242 this fails because the process is not running as root or exim. There may be
7243 a better way of dealing with this, but for the moment (because 4.30 needs
7244 to be released), I have disabled address rewrite logging when running a
7245 filter in a non-root, non-exim process.
7251 1. The buildconfig auxiliary program wasn't quoting the value set for
7252 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7253 not defined. This bug was masked in 4.22 by the effect that was fixed in
7256 2. Some messages that were rejected after a message id was allocated were
7257 shown as "incomplete" by exigrep. It no longer does this for messages that
7258 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7260 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7261 have allow_domain_literals set, the ID did not get logged in the <= line.
7262 Domain literals are now always recognized in Message-ID: header lines.
7264 4. The first argument for a ${extract expansion item is the key name or field
7265 number. Leading and trailing spaces in this item were not being ignored,
7266 causing some misleading effects.
7268 5. When deliver_drop_privilege was set, single queue runner processes started
7269 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7270 same command in the process it spins off) were not dropping privilege.
7272 6. When the daemon running as "exim" started a queue runner, it always
7273 re-executed Exim in the spun-off process. This is a waste of effort when
7274 deliver_drop_privilege is set. The new process now just calls the
7275 queue-runner function directly.
7281 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7284 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7285 ignored. Though the use of -oP was forcing the writing of a pid file, it
7286 was always written to the default place.
7288 3. If the message "no IP address found for host xxxx" is generated during
7289 incoming verification, it is now followed by identification of the incoming
7290 connection (so you can more easily find what provoked it).
7292 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7294 5. Added some features to "harden" Exim a bit more against certain attacks:
7296 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7297 be put in Local/Makefile. This is like the never_users runtime option,
7298 but it cannot be overridden. The default setting is "root".
7300 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7301 prefix string with which any file named in a -C command line option
7304 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7305 is retained for -C and -D only if the caller of Exim is root. Without
7306 it, the exim user may also use -C and -D and retain privilege.
7308 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7309 command line option is disabled.
7311 6. Macro names set by the -D option must start with an upper case letter, just
7312 like macro names defined in the configuration file.
7314 7. Added "dereference=" facility to LDAP.
7316 8. Two instances of the typo "uknown" in the source files are fixed.
7318 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7319 the Configure-Makefile script screwed up while processing it.
7321 10. Incorporated PCRE 4.4.
7323 11. The SMTP synchronization check was not operating right at the start of an
7324 SMTP session. For example, it could not catch a HELO sent before the client
7325 waited for the greeting. There is now a check for outstanding input at the
7326 point when the greeting is written. Because of the duplex, asynchronous
7327 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7328 way, but not yet received, when the check is performed.
7330 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7331 on TCP/IP sockets, because this apparently causes some broken clients to
7334 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7335 unchanged) from the Cygwin maintainer.
7337 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7338 was supported instead of testing for mbx. Effectively a typo.
7340 15. The spa authenticator server code was not checking that the input it
7341 received was valid base64.
7343 16. The debug output line for the "set" modifier in ACLs was not showing the
7344 name of the variable that was being set.
7346 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7348 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7349 lookup on its own. Something else was needed to trigger the lookup. For
7350 example, a match in host_lookup or the need for a host name in a host list.
7351 Now, if $sender_host_name is referenced and the host name has not yet been
7352 looked up, a lookup is performed. If the lookup fails, the variable remains
7353 empty, and $host_lookup_failed is set to "1".
7355 19. Added "eqi" as a case-independent comparison operator.
7357 20. The saslauthd authentication condition could segfault if neither service
7358 nor realm was specified.
7360 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7361 error message that was logged was misleading, and incoming SMTP
7362 connections were dropped. The message is now more accurate, and temporary
7363 errors are given to SMTP connections.
7365 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7366 (except RSET) with a 421 error, until QUIT is received. However, it was
7367 failing to send a response to QUIT.
7369 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7370 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7371 ACL is now run after the helo_try_verify_hosts code.
7373 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7374 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7375 case-independent, so other case variants are also recognized). Apparently
7376 some systems use these upper case variants.
7378 25. If more than two messages were waiting for the same host, and a transport
7379 filter was specified for the transport, Exim sent two messages over the
7380 same TCP/IP connection, and then failed with "socket operation on non-
7381 socket" when it tried to send the third.
7383 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7385 27. The extern definition of crypt16() in expand.c was not being excluded when
7386 the OS had its own crypt16() function.
7388 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7389 as a preferred synonym for return_size_limit, both as an option and as an
7392 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7394 30. Changed the default configuration ACL to relax the local part checking rule
7395 for addresses that are not in any local domains. For these addresses,
7396 slashes and pipe symbols are allowed within local parts, but the sequence
7397 /../ is explicitly forbidden.
7399 31. SPA server authentication was not clearing the challenge buffer before
7402 32. log_message in a "warn" ACL statement was writing to the reject log as
7403 well as to the main log, which contradicts the documentation and doesn't
7404 seem right (because no rejection is happening). So I have stopped it.
7406 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7407 However, I am unable to do any testing of this.
7409 34. Fixed an infelicity in the appendfile transport. When checking directories
7410 for a mailbox, to see if any needed to be created, it was accidentally
7411 using path names with one or more superfluous leading slashes; tracing
7412 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7414 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7415 contents are added to the log line that is written for every discarded
7416 recipient. (Previously a log_message setting was ignored.)
7418 36. The ${quote: operator now quotes the string if it is empty.
7420 37. The install script runs exim in order to find its version number. If for
7421 some reason other than non-existence or emptiness, which it checks, it
7422 could not run './exim', it was installing it with an empty version number,
7423 i.e. as "exim-". This error state is now caught, and the installation is
7426 38. An argument was missing from the function that creates an error message
7427 when Exim fails to connect to the socket for saslauthd authentication.
7428 This could cause Exim to crash, or give a corrupted message.
7430 39. Added isip, isip4, and isip6 to ${if conditions.
7432 40. The ACL variables $acl_xx are now saved with the message, and can be
7433 accessed later in routers, transports, and filters.
7435 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7436 strings in the file are not string-expanded.
7438 42. If a MAIL command specified a SIZE value that was too large to fit into an
7439 int variable, the check against message_size_limit failed. Such values are
7440 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7441 day this will have to be increased, but I don't think I want to be around
7442 when emails are that large.
7449 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7450 iconv() is not standard in FreeBSD.
7452 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7453 IPv6 enabled. The observed symptom was a segmentation fault on return from
7454 the function os_common_find_running_interfaces() in src/os.c.
7456 3. In the check_special_case() function in daemon.c I had used "errno" as an
7457 argument name, which causes warnings on some systems. This was basically a
7458 typo, since it was named "eno" in the comments!
7460 4. The code that waits for the clock to tick (at a resolution of some fraction
7461 of a second) so as to ensure message-id uniqueness was always waiting for
7462 at least one whole tick, when it could have waited for less. [This is
7463 almost certainly not relevant at current processor speeds, where it is
7464 unlikely to ever wait at all. But we try to future-proof.]
7466 5. The function that sleeps for a time interval that includes fractions of a
7467 second contained a race. It did not block SIGALRM between setting the
7468 timer, and suspending (a couple of lines later). If the interval was short
7469 and the sigsuspend() was delayed until after it had expired, the suspension
7470 never ended. On busy systems this could lead to processes getting stuck for
7473 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7474 process, before it forks any delivery processes. The open lookup caching
7475 mechanism meant that the open file or database connection was passed into
7476 the delivery process. The problem was that delivery processes always tidy
7477 up cached lookup data. This could cause a problem for the next delivery
7478 process started by the queue runner, because the external queue runner
7479 process does not know about the closure. So the next delivery process
7480 still has data in the lookup cache. In the case of a file lookup, there was
7481 no problem because closing a file descriptor in a subprocess doesn't affect
7482 the parent. However, if the lookup was caching a connection to a database,
7483 the connection was closed, and the second delivery process was likely to
7484 see errors such as "PGSQL: query failed: server closed the connection
7485 unexpectedly". The problem has been fixed by closing all cached lookups
7486 in a queue runner before running a delivery process.
7488 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7489 seem to have the "const" qualifier which it has on other OS. I've
7492 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7493 *of the same type* (client or server) with the same public name that an
7494 error should be diagnosed.
7496 9. When Exim looked up a host name for an IP address, but failed to find the
7497 original IP address when looking up the host name (a safety check), it
7498 output the message "<ip address> does not match any IP for NULL", which was
7499 confusing, to say the least. The bug was that the host name should have
7500 appeared instead of "NULL".
7502 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7503 user that is built into the binary, and the -C or -D options is used, root
7504 privilege is dropped before the configuration file is read. In addition,
7505 logging is switched to stderr instead of the normal log files. If the
7506 configuration then re-defines the Exim user, the unprivileged environment
7507 is probably not what is expected, so Exim logs a panic warning message (but
7510 However, if deliver_drop_privilege is set, the unprivileged state may well
7511 be exactly what is intended, so the warning has been cut out in that case,
7512 and Exim is allowed to try to write to its normal log files.
7518 1. smtp_return_error_details was not giving details for temporary sender
7519 or receiver verification errors.
7521 2. Diagnose a configuration error if two authenticators have the same public
7524 3. Exim used not to create the message log file for a message until the first
7525 delivery attempt. This could be confusing when incoming messages were held
7526 for policy or load reasons. The message log file is now created at the time
7527 the message is received, and an initial "Received" line is written to it.
7529 4. The automatically generated man page for command line options had a minor
7530 bug that caused no ill effects; however, a more serious problem was that
7531 the procedure for building the man page automatically didn't always
7532 operate. Consequently, release 4.20 contains an out-of-date version. This
7533 shouldn't happen again.
7535 5. When building Exim with embedded Perl support, the script that builds the
7536 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7537 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7539 6. The freeze_tell option was not being used for messages that were frozen on
7540 arrival, either by an ACL or by local_scan().
7542 7. Added the smtp_incomplete_transaction log selector.
7544 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7545 was accepting AUTH without a new EHLO.
7547 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7548 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7551 10. Logging of TCP/IP connections (when configured) now happens in the main
7552 daemon process instead of the child process, so that the TCP/IP connection
7553 count is more accurate (but it can never be perfect).
7555 11. The use of "drop" in a nested ACL was not being handled correctly in the
7556 outer ACL. Now, if condition failure induced by the nested "drop" causes
7557 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7558 or "require"), the connection is dropped.
7560 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7561 that yield "discard" can now be used with an "accept" or a "discard" verb,
7562 but an error is generated for any others (because I can't see a useful way
7563 to define what should happen).
7565 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7566 are now processed in the same way as lines in the Exim configuration file.
7567 In particular, continuation lines are supported.
7569 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7571 15. Added -ti meaning -t -i.
7573 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7574 domains, and warn by logging if others are found.
7576 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7577 returned from GIFCONF when Exim is trying to find the list of interfaces on
7578 a host. The code in os.c has been modified to copy each ifreq to an aligned
7579 structure in all cases.
7581 Also, in some cases, the returned ifreq's were being copied to a 'struct
7582 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7583 means the last couple of bytes of an IPv6 address could be chopped if the
7584 ifreq contained only a normal sockaddr (14 bytes storage).
7586 18. Named domain lists were not supported in the hosts_treat_as_local option.
7587 An entry such as +xxxx was not recognized, and was treated as a literal
7590 19. Ensure that header lines added by a DATA ACL are included in the reject log
7591 if the ACL subsequently rejects the message.
7593 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7594 MD5 (which is deprecated).
7596 21. When testing a filter file using -bf, Exim was writing a message when it
7597 took the sender from a "From " line in the message, but it was not doing so
7598 when it took $return_path from a Return-Path: header line. It now does.
7600 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7601 with a valid header line field name (a series of printing characters
7602 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7604 23. Changed "disc" in the source to "disk" to conform to the documentation and
7605 the book and for uniformity.
7607 24. Ignore Sendmail's -Ooption=value command line item.
7609 25. When execve() failed while trying to run a command in a pipe transport,
7610 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7611 could be confused with a return value of 69 from the command itself. This
7612 has been changed to 127, the value the shell returns if it is asked to run
7613 a non-existent command. The wording for the related log line suggests a
7614 non-existent command as the problem.
7616 26. If received_header_text expands to an empty string, do not add a Received:
7617 header line to the message. (Well, it adds a token one on the spool, but
7618 marks it "old" so that it doesn't get used or transmitted.)
7620 27. Installed eximstats 1.28 (addition of -nt option).
7622 28. There was no check for failure on the call to getsockname() in the daemon
7623 code. This can fail if there is a shortage of resources on the system, with
7624 ENOMEM, for example. A temporary error is now given on failure.
7626 29. Contrary to the C standard, it seems that in some environments, the
7627 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7628 program. Exim now does this explicitly; it affects the formatting of
7629 timestamps using strftime().
7631 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7632 complaints. I've now initialized all the variables, to avoid this.
7634 32. Header lines added by a system filter were not being "seen" during
7635 transport-time rewrites.
7637 33. The info_callback() function passed to OpenSSL is set up with type void
7638 (*)(SSL *, int, int), as described somewhere. However, when calling the
7639 function (actually a macro) that sets it up, the type void(*)() is
7640 expected. I've put in a cast to prevent warnings from picky compilers.
7642 34. If a DNS black list lookup found a CNAME record, but there were no A
7643 records associated with the domain it pointed at, Exim crashed.
7645 35. If a DNS black list lookup returned more than one A record, Exim ignored
7646 all but the first. It now scans all returned addresses if a particular IP
7647 value is being sought. In this situation, the contents of the
7648 $dnslist_value variable are a list of all the addresses, separated by a
7651 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7652 to accept a host name and all its aliases if the forward lookup for any of
7653 them yielded the IP address of the incoming connection. Now it accepts only
7654 those names whose forward lookup yields the correct IP address. Any other
7655 names are discarded. This closes a loophole whereby a rogue DNS
7656 administrator could create reverse DNS records to break through a
7657 wildcarded host restriction in an ACL.
7659 37. If a user filter or a system filter that ran in a subprocess used any of
7660 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7661 the wrong values were passed to the pipe command ($thisaddress had the
7662 value of $0, $0 had the value of $1, etc). This bug was introduced by
7663 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7665 38. Improved the line breaking for long SMTP error messages from ACLs.
7666 Previously, if there was no break point between 40 and 75 characters, Exim
7667 left the rest of the message alone. Two changes have been made: (a) I've
7668 reduced the minimum length to 35 characters; (b) if it can't find a break
7669 point between 35 and 75 characters, it looks ahead and uses the first one
7670 that it finds. This may give the occasional overlong line, but at least the
7671 remaining text gets split now.
7673 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7674 file descriptors might be low, and that setting 1000 would always raise it.
7675 It turns out that in some environments, the limit is already over 1000 and
7676 that lowering it causes trouble. So now Exim takes care not to decrease it.
7678 40. When delivering a message, the value of $return_path is set to $sender_
7679 address at the start of routing (routers may change the value). By an
7680 oversight, this default was not being set up when an address was tested by
7681 -bt or -bv, which affected the outcome if any router or filter referred to
7684 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7685 log only when "message" or "log_message" are set. However, if one of the
7686 conditions was an address verification, or a call to a nested ACL, the
7687 messages generated by the underlying test were being passed through. This
7688 no longer happens. The underlying message is available in $acl_verify_
7689 message for both "message" and "log_message" expansions, so it can be
7690 passed through if needed.
7692 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7693 new expansion $bh_ to give the encoded byte string without charset
7694 translation. Translation happens only if iconv() is available; HAVE_ICONV
7695 indicates this at build time. HEADERS_CHARSET gives the charset to
7696 translate to; headers_charset can change it in the configuration, and
7697 "headers charset" can change it in an individual filter file.
7699 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7700 that creates RFC 2047 encoded "words" labels them as that charset instead
7701 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7702 expansion operator; (ii) when Exim creates a From: line for a local
7703 message; (iii) when a header line is rewritten to include a "phrase" part.
7705 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7706 buggy, causing it to skip the first lines of messages whose message ID
7707 ended in 'D'. This would not have bitten before Exim release 4.14, because
7708 message IDs were unlikely to end in 'D' before then. The effect was to have
7709 incorrect size information for certain domains.
7711 45. #include "config.h" was missing at the start of the crypt16.c module. This
7712 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7715 46. If there was a timeout during a "random" callout check, Exim treated it as
7716 a failure of the random address, and carried on sending RSET and the real
7717 address. If the delay was just some slowness somewhere, the response to the
7718 original RCPT would be taken as a response to RSET and so on, causing
7719 mayhem of various kinds.
7721 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7722 when I implemented it. It didn't allow for the fact that some option values
7723 may legitimately be negative (e.g. size_addition), and it didn't even do
7724 the right test for positive values.
7726 48. Domain names in DNS records are case-independent. Exim always looks them up
7727 in lower case. Some resolvers return domain names in exactly the case they
7728 appear in the zone file, that is, they may contain uppercase letters. Not
7729 all resolvers do this - some return always lower case. Exim was treating a
7730 change of case by a resolver as a change of domain, similar to a widening
7731 of a domain abbreviation. This triggered its re-routing code and so it was
7732 trying to route what was effectively the same domain again. This normally
7733 caused routing to fail (because the router wouldn't handle the domain
7734 twice). Now Exim checks for this case specially, and just changes the
7735 casing of the domain that it ultimately uses when it transmits the message
7738 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7741 50. If a filter generated a file delivery with a non-absolute name (possible if
7742 no home directory exists for the router), the forbid_file option was not
7745 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7746 the existing equality matching.
7748 52. Exim was using ints instead of ino_t variables in some places where it was
7749 dealing with inode numbers.
7751 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7752 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7753 called TMPDIR, and if it finds it is different, it changes its value.
7755 54. The smtp_printf() function is now made available to local_scan() so
7756 additional output lines can be written before returning. There is also an
7757 smtp_fflush() function to enable the detection of a dropped connection.
7758 The variables smtp_input and smtp_batched_input are exported to
7761 55. Changed the default runtime configuration: the message "Unknown user"
7762 has been removed from the ACL, and instead placed on the localuser router,
7763 using the cannot_route_message feature. This means that any verification
7764 failures that generate their own messages won't get overridden. Similarly,
7765 the "Unrouteable address" message that was in the ACL for unverifiable
7766 relay addresses has also been removed.
7768 56. Added hosts_avoid_esmtp to the smtp transport.
7770 57. The exicyclog script was not checking for the esoteric option
7771 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7772 will work only if exicyclog is run under the appropriate euid.
7774 58. Following a discussion on the list, the rules by which Exim recognises line
7775 endings on incoming messages have been changed. The -dropcr and drop_cr
7776 options are now no-ops, retained only for backwards compatibility. The
7777 following line terminators are recognized: LF CRLF CR. However, special
7778 processing applies to CR:
7780 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7781 nor a local message in the state where . is a terminator.
7783 (ii) If a bare CR is encountered in a header line, an extra space is added
7784 after the line terminator so as not to end the header. The reasoning
7785 behind this is that bare CRs in header lines are most likely either
7786 to be mistakes, or people trying to play silly games.
7788 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7789 was being incorrectly given as 18 bytes larger than it should have been.
7790 This is a VOB (very old bug).
7792 60. This may never have affected anything current, but just in case it has:
7793 When the local host is found other than at the start of a list of hosts,
7794 the local host, those with the same MX, and any that follow, are discarded.
7795 When the list in question was part of a longer list of hosts, the following
7796 hosts (not currently being processed) were also being discarded. This no
7797 longer happens. I'm not sure if this situation could ever has previously
7800 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7803 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7804 header. This is something that is recommended in a new Internet Draft, and
7805 is something that is documented as being done by Sendmail. There are two
7806 possible values. For messages generated by the autoreply transport, Exim
7809 Auto-Submitted: auto-replied
7811 whereas for all other generated messages (e.g. bounces) it adds
7813 Auto-Submitted: auto-generated
7815 63. The "personal" condition in filters now includes a test for the
7816 Auto-Submitted: header. If it contains the string "auto-" the message it
7817 not considered personal.
7819 64. Added rcpt_include_affixes as a generic transport option.
7821 65. Added queue_only_override (default true).
7823 66. Added the syslog_duplication option.
7825 67. If what should have been the first header line of a message consisted of
7826 a space followed by a colon, Exim was mis-interpreting it as a header line.
7827 It isn't of course - it is syntactically invalid and should therefore be
7828 treated as the start of the message body. The misbehaviour could have
7829 caused a number of strange effects, including loss of data in subsequent
7830 header lines, and spool format errors.
7832 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7833 client host had authenticated. This control can now be exercised by an ACL
7834 for more flexibility.
7836 69. By default, callouts do not happen when testing with -bh. There is now a
7837 variant, -bhc, which does actually run the callout code, including
7838 consulting and updating the callout cache.
7840 70. Added support for saslauthd authentication, courtesy of Alexander
7843 71. If statvfs() failed on the spool or log directories while checking their
7844 size for availability, Exim confusingly gave the error "space shortage".
7845 Furthermore, in debugging mode it crashed with a floating point exception.
7846 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7847 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7848 serious problem, Exim now writes to the main and panic logs when this
7849 happens, with details of the failure. It then refuses to accept the
7850 incoming message, giving the message "spool directory problem" or "log
7851 directory problem" with a 421 code for SMTP messages.
7853 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7854 0, 1, and 2 exist, because some OS complain for execs without them (see
7855 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7856 descriptors. However, the code omitted to check that the open succeeded,
7857 causing mysterious errors if for some reason the permissions on /dev/null
7858 got screwed. Now Exim writes a message to the main and panic logs, and
7859 bombs out if it can't open /dev/null.
7861 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7862 interact so that it is all more flexible. It is supposed to remain
7863 backwards compatible. Also added extra_local_interfaces.
7865 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7866 to bomb out with an assertion failure - to the client this appears as a
7867 connection drop. This problem occurs in the part of the code that was taken
7868 from the Samba project. Fortunately, the assertion is in a very simple
7869 function, so I have fixed this by reproducing the function inline in the
7870 one place where it is called, and arranging for authentication to fail
7871 instead of killing the process with assert().
7873 75. The SPA client code was not working when the server requested OEM rather
7874 than Unicode encoding.
7876 76. Added code to make require_files with a specific uid setting more usable in
7877 the case where statting the file as root fails - usually a non-root-mounted
7878 NFS file system. When this happens and the failure is EACCES, Exim now
7879 forks a subprocess and does the per-uid checking as the relevant uid.
7881 77. Added process_log_path.
7883 78. If log_file_path was not explicitly set, a setting of check_log_space or
7884 check_log_inodes was ignored.
7886 79. If a space check for the spool or log partitions fails, the incident is now
7887 logged. Of course, in the latter case the data may get lost...
7889 80. Added the %p formatting code to string_format() so that it can be used to
7890 print addresses in debug_print(). Adjusted all the address printing in the
7891 debugging in store.c to use %p rather than %d.
7893 81. There was a concern that a line of code in smtp_in.c could overflow a
7894 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7895 initially expressed, the concern was not well-founded, because trailing
7896 spaces are removed early. However, if the trailing spaces were followed by
7897 a NULL, they did not get removed, so the overflow was possible. Two fixes
7900 (a) I re-wrote the offending code in a cleaner fashion.
7901 (b) If an incoming SMTP command contains a NULL character, it is rejected
7904 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7905 runs initgroups(), so that if the Exim user is in any additional groups,
7906 they will be used during message reception.
7912 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob