1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
107 JH/01 Avoid costly startup code when not strictly needed. This reduces time
108 for some exim process initialisations. It does mean that the logging
109 of TLS configuration problems is only done for the daemon startup.
111 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
113 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
114 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
117 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
118 DKIM signing is being done. Previously a transport_filter would always
119 disable CHUNKING, falling back to traditional DATA.
121 JH/05 Regard command-line receipients as tainted.
123 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
125 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
126 PAM library frees one of the arguments given to it, despite the
127 documentation. Therefore a plain malloc must be used.
129 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
130 on-stack buffers were used, resulting in a taint trap when DSN information
131 copied from a received message was written into the buffer.
133 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
134 the ordering of its ARC headers. This caused a crash.
136 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
137 a new record was being constructed with information from the peer, a trap
140 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
141 installation would get error messages from DMARC verify, when it hit the
142 nonexistent file indicated by the default. Distros wanting DMARC enabled
143 should both provide the file and set the option.
144 Also enforce no DMARC verification for command-line sourced messages.
146 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
147 could, depending on the platform, hang at the STARTTLS response.
149 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
150 message on a connection. Previously if one message had ARC headers and
151 the following one did not, a crash could result when adding an
152 Authentication-Results: header.
154 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
155 Exim main code and Exim-related utities. The introduction of taint
156 tracking also did many adjustments to string handling. Since then, eximon
157 frequently terminated with an assert failure.
159 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
160 check for 452 responses. This slightly helps the inefficieny of doing
161 a large alias-expansion into a recipient-limited target. The max_rcpt
162 transport option still applies (and at the current default, will override
163 the new feature). The check is done for either cause of synch, and forces
164 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
165 connection. The new facility is not tunable at this time.
167 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
168 library live data was being used, so the results became garbage. Make
169 copies while it is still usable.
171 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
172 on delivery deferred (==) and failed (**) lines (if a delivery was
173 attemtped). Previously it was only on completion (=>) lines.
175 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
176 for the expansion of the server_scram_iter and server_scram_salt options.
178 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
179 are now specifically given a NO_DATA response without hitting the system
180 resolver. The library goes on to do the now-standard TXT lookup.
181 Use of dnsdb lookups is not affected.
183 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
184 only retrieve the errormessage once. Previously two calls to dlerror()
185 were used, and the second one (for mainlog/paniclog) retrieved null
188 JH/20 Taint checking: disallow use of tainted data for
189 - the appendfile transport file and directory options
190 - the pipe transport command
191 - the autoreply transport file, log and once options
192 - file names used by the redirect router (including filter files)
194 - paths used by single-key lookups
195 Previously this was permitted.
197 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
198 adjusted the size of a major service buffer; this failed because the
199 buffer was in use at the time. Change to a compile-time increase in the
200 buffer size, when this authenticator is compiled into exim.
202 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
203 previous fast-mode was untenable in the face of glibs using mmap to
204 support larger malloc requests.
206 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
207 New values supported, if defined on system where compiled:
208 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
209 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
211 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
212 running a limited number of proceses in parallel, a benefit is gained. The
213 amount varies with the platform hardware and load. The use of the option
214 queue_run_in_order means we cannot do this, as ordering becomes
217 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
218 had introduced a string-copy (for ensuring NUL-termination) which was not
219 appropriate for that case, which can include embedded NUL bytes in the
220 block of data. Investigation showed the copy to actually be needless, the
221 data being length-specified.
223 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
224 done during a receiving connection, and both used TLS, global info was
225 used rather than per-connection info for tracking the state of data
226 queued for transmission. This could result in a connection hang.
228 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
229 Previously, when delivering serveral messages down a single connection
230 only the first would provide a SIZE. This was due to the size information
231 not being properly tracked.
233 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
234 TAI (at 37 seconds currently), pretend to be in UTC for time-related
235 expansion and logging. Previously, spurious values such as a future
236 minute could be seen.
238 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
239 it could crash from a null-deref. This could also affect the
240 ${addresses: } operator and ${readsock } item.
242 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
243 message following a mime one, the variable was not reset.
245 JH/30 When an pipelined-connect fails at the first response, assume incorrect
246 cached capability (perhaps the peer reneged?) and immediately retry in
249 JH/31 Fix spurious detection of timeout while writing to transport filter.
251 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
252 an attempt to copy the string was made before checking it.
254 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
255 taint of the lookup key was maintained; we now regard the presence in the
256 filesystem as sufficient validation.
258 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
259 argument is supplied.
261 JH/35 The dsearch lookup now requires that the directory is an absolute path.
262 Previously this was not checked, and nonempty relative paths made an
263 access under Exim's current working directory.
265 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
266 Previously no event was raised.
268 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
269 parameter supplied by the sender MAIL FROM command. Previously it was
270 ignored, and only the check_spool_space option value for the required
273 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
274 the size of the signing public-key. Previously it was instead giving
275 the size of the signature hash.
277 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
278 the default. See the (new) dkim_verify_min_keysizes option.
280 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
281 and an ACL use a lookup for checking either the local_part or domain,
282 stale data could be accessed. Ensure that variable references are
283 dropped between messages.
285 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
286 by the client was not checked as pointing within response data before
287 being used. A malicious client could thus cause an out-of-bounds read and
288 possibly gain authentication. Fix by adding the check.
290 JH/42 Internationalisation: change the default for downconversion in the smtp
291 transport to be "if needed". Previously it was "as previously set" for
292 the message, which usually meant "if needed" for message-submission but
293 "no" for everything else. However, MTAs have been seen using SMTPUTF8
294 even when the envelope addresses did not need it, resulting in forwarding
295 failures to non-supporting MTAs. A downconvert in such cases will be
296 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
297 transport. The change does mean that addresses needing conversion will
298 be converted when previously a delivery failure would occur.
300 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
301 response was received it would be used unchecked in a fail-DSN, violating
302 standards on line-length limits. Truncate if needed.
304 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
305 doesn't work. (Additionally add a new main config option to configure the
312 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
313 side, to decode a TLS 1.3 packet capture.
315 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
316 Previously the default library behaviour applied, sending two, each in
319 JH/03 Debug output for ACL now gives the config file name and line number for
322 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
324 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
326 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
327 buffer overrun for (non-chunking) other transports.
329 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
330 TLS1.3, means that a server rejecting a client certificate is not visible
331 to the client until the first read of encrypted data (typically the
332 response to EHLO). Add detection for that case and treat it as a failed
333 TLS connection attempt, so that the normal retry-in-clear can work (if
334 suitably configured).
336 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
337 and/or domain. Found and fixed by Jason Betts.
339 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
340 configuration). If a CNAME target was not a wellformed name pattern, a
343 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
344 the OS reports them interleaved with other addresses.
346 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
347 used both for input and for a verify callout, both encrypted, SMTP
348 responses being sent by the server could be lost. This resulted in
349 dropped connections and sometimes bounces generated by a peer sending
352 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
353 string. Previously it was possible to cause undefined behaviour in a
354 library routine (usually a crash). Found by "zerons".
356 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
359 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
360 API was removed, so update to use the newer ones.
362 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
363 any timeout set, is taking a long time. Previously we would hang on to a
364 rotated logfile "forever" if the input was arriving with long gaps
365 (a previous attempt to fix addressed lack, for a long time, of initial
368 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
369 shared (NFS) environment. The length of the tempfile name is now
370 4 + 16 ("hdr.$message_exim_id") which might break on file
371 systems which restrict the file name length to lower values.
374 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
375 shared (NFS) environment.
377 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
378 did for all versions <4.90). Notably -M, -m, --invert, -I may be
381 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
382 on some platforms for bit 31.
384 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
385 to changes apparently associated with TLS1.3 handling some of the APIs
386 previously used were either nonfunctional or inappropriate. Strings
387 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
388 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
389 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
390 This affects log line X= elements, the $tls_{in,out}_cipher variables,
391 and the use of specific cipher names in the encrypted= ACL condition.
393 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
395 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
396 verification result was not updated unless hosts_require_ocsp applied.
398 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
399 queue_list_requires_admin set to false, non-admin users were denied the
402 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
403 directory-of-certs mode. Previously they were advertised despite the
406 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
407 A single TCP connection by a client will now hold a TLS connection open
408 for multiple message deliveries, by default. Previously the default was to
411 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
412 default. If built with the facility, DANE will be used. The facility
413 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
415 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
416 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
417 must be defined and you must still, unless you define DISABLE_TLS, manage
418 the the include-dir and library-file requirements that go with that
419 choice. Non-TLS builds are still supported.
421 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
424 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
425 default. If the platform supports and has the facility enabled, it will
426 be requested on all coneections.
428 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
429 controlled by the build-time option SUPPORT_PIPE_CONNECT.
431 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
433 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
434 success-DSN messages. Previously the From: header was always the default
435 one for these; the option was ignored.
437 JH/28 Fix the timeout on smtp response to apply to the whole response.
438 Previously it was reset for every read, so a teergrubing peer sending
439 single bytes within the time limit could extend the connection for a
440 long time. Credit to Qualsys Security Advisory Team for the discovery.
442 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
443 delivery address, which leaked information of the results of local
444 forwarding. Change to the original envelope recipient address, per
447 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
448 requested. Previously not bounce was generated and a log entry of
449 error ignored was made.
451 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
453 JH/32 Introduce a general tainting mechanism for values read from the input
454 channel, and values derived from them. Refuse to expand any tainted
455 values, to catch one form of exploit.
457 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
458 was unused and the unexpanded text used for the test. Found and
459 fixed by Ruben Jenster.
461 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
462 an attempt to use a TLS library read routine dereffed a nul pointer,
465 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
466 them in our smtp response.
468 JH/36 Have the general router option retry_use_local_part default to true when
469 any of the restrictive preconditions are set (to anything). Previously it
470 was only for check_local user. The change removes one item of manual
471 configuration which is required for proper retries when a remote router
472 handles a subset of addresses for a domain.
474 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
475 link count into consideration.
477 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
478 caused the extension of big_buffer, the following lines were ignored.
480 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
481 accordance with RFC 2308. Previously there was no expiry, so a longlived
482 receive process (eg. due to ACL delays) versus a short SOA value could
485 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
487 JH/39 Promote DMARC support to mainline.
489 JH/40 Bug 2452: Add a References: header to DSNs.
491 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
492 parameters. The relevant library call is documented as "Deprecated: This
493 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
494 3.6.0, DH parameters are negotiated following RFC7919."
496 HS/06 Change the default of dnssec_request_domains to "*"
498 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
499 carried on and emitted a BDAT command, even when PIPELINING was not
502 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
503 buffer was used for the filename, resulting in a trap when tainted
504 arguments (eg. $domain) were used.
506 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
507 recommended to avoid a possible server-load attack. The feature can be
508 re-enabled via the openssl_options main cofiguration option.
510 JH/45 local_scan API: documented the current smtp_printf() call. This changed
511 for version 4.90 - adding a "more data" boolean to the arguments.
512 Bumped the ABI version number also, this having been missed previously;
513 release versions 4.90 to 4.92.3 inclusive were effectively broken in
514 respect of usage of smtp_printf() by either local_scan code or libraries
515 accessed via the ${dlfunc } expansion item. Both will need coding
516 adjustment for any calls to smtp_printf() to match the new function
517 signature; a FALSE value for the new argument is always safe.
519 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
520 the file-offset (which the Linux syscall does, and exim expects); this
521 resulted in an indefinite loop.
523 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
524 to do ARC verification. The Authentication-Results: header line added
525 by the configuration then had no ARC item.
531 JH/01 Remove code calling the customisable local_scan function, unless a new
532 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
534 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
535 non-signal-safe functions being used.
537 JH/03 Bug 2269: When presented with a received message having a stupidly large
538 number of DKIM-Signature headers, disable DKIM verification to avoid
539 a resource-consumption attack. The limit is set at twenty.
541 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
542 report of oldest_pass in ${authres } in consequence, and separate out
543 some descriptions of reasons for verification fail.
545 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
546 files in the spool were present and unlocked. A queue-runner could spot
547 them, resulting in a duplicate delivery. Fix that by doing the unlock
548 after the unlink. Investigation by Tim Stewart. Take the opportunity to
549 add more error-checking on spoolfile handling while that code is being
552 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
553 No known attacks, no CVE, this is defensive hardening.
555 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
556 a queue-runner could start a delivery while other operations were ongoing.
557 Cutthrough delivery was a common victim, resulting in duplicate delivery.
558 Found and investigated by Tim Stewart. Fix by using the open message data
559 file handle rather than opening another, and not locally closing it (which
560 releases a lock) for that case, while creating the temporary .eml format
561 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
563 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
564 $sender_verify_failure/$recipient_verify_failure to "random".
566 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
569 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
570 Previously this would segfault.
572 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
575 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
576 like zero, since the resolver should be doing this for us, But we need one
577 as a CNAME but no MX presence gets the CNAME returned; we need to check
578 that doesn't point to an MX to declare it "no MX returned" rather than
579 "error, loop". A new main option is added so the older capability of
580 following some limited number of chain links is maintained.
582 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
584 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
585 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
586 now-deprecated earlier definitions used only specified the range up to TLS
587 1.2 (in the older-version library docs).
589 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
591 JH/15 Rework TLS client-side context management. Stop using a global, and
592 explicitly pass a context around. This enables future use of TLS for
593 connections to service-daemons (eg. malware scanning) while a client smtp
594 connection is using TLS; with cutthrough connections this is quite likely.
596 JH/16 Fix ARC verification to do AS checks in reverse order.
598 JH/17 Support a "tls" option on the ${readsocket } expansion item.
600 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
601 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
602 Previously the "utf8" would be re-prepended for every additional message.
604 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
605 Previously thery were accepted, resulting in issues when attempting to
606 forward messages to a non-supporting MTA.
608 PP/02 Let -n work with printing macros too, not just options.
610 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
611 one parent address was copied, and bogus data was used at delivery-logging
612 time. Either a crash (after delivery) or bogus log data could result.
613 Discovery and analysis by Tim Stewart.
615 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
616 Previously if the string ended mid-character, we did not insert the
617 promised '?' replacement.
619 PP/04 Documentation: current string operators work on bytes, not codepoints.
621 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
622 should pack well giving a smaller memory footprint so better caching and
623 therefore performance. Group the declarations where this can't be done so
624 that the byte-sized flag variables are not interspersed among pointer
625 variables, giving a better chance of good packing by the compiler.
627 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
628 non-null, to avoid issues with sites running BATV. Previously reports were
629 sent with an empty envelope sender so looked like bounces.
631 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
632 The ignore_error flag wasn't being returned from the filter subprocess so
633 was not set for later routers. Investigation and fix by Matthias Kurz.
635 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
636 and a msg:complete for the whole, when a message is manually removed using
637 -Mrm. Developement by Matthias Kurz, hacked on by JH.
639 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
640 a "Gnu special" function, asprintf() in the DB utility binary builds; I
641 hope that is portable enough.
643 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
644 requiring a known-CA anchor certificate; make it now rely entirely on the
645 TLSA as an anchor. Checking the name on the leaf cert against the name
646 on the A-record for the host is still done for TA (but not for EE mode).
648 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
649 included in delivery lines for non-proxied connections, when compiled with
650 SUPPORT_SOCKS and running with proxy logging enabled.
652 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
653 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
654 move the existing event to fire before the normal logging of message
655 failure so that custom logging is bracketed by normal logging.
657 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
658 msg:fail:internal event. Developement by Matthias Kurz.
660 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
661 far too small for todays use of crypto signatures stored there. Go all
662 the way to the max DNS message size of 64kB, even though this might be
663 overmuch for IOT constrained device use.
665 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
666 copy a string over itself. The library routine is documented as not
667 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
669 JH/32 For main options check_spool_space and check_inode_space, where the
670 platform supports 64b integers, support more than the previous 2^31 kB
671 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
672 the previous G, M, k.
674 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
675 $authenticated_fail_id variable on authentication failure. Previously
678 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
679 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
680 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
681 GNUTLS_SEC_PARAM_MEDIUM.
683 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
684 side. Previously we would continue as if no SNI had been received.
686 JH/36 Harden the handling of string-lists. When a list consisted of a sole
687 "<" character, which should be a list-separator specification, we walked
688 off past the nul-terimation.
690 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
691 causes) even when the retry time is not yet met. Previously they were
692 not, meaning that when (say) an account was over-quota and temp-rejecting,
693 and multiple senders' messages were queued, only one sender would get
694 notified on each configured delay_warning cycle.
696 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
698 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
699 cipher-suites, an error can be left on the stack even for a succeeding
700 accept; this results in impossible error messages when a later operation
703 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
704 return error codes indicating retry. Under TLS1.3 this becomes required.
706 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
707 it only wrote the new authenticators, resulting in a lack of tracking of
708 peer changes of ESMTP extensions until the next cache flush.
710 JH/41 Fix the loop reading a message header line to check for integer overflow,
711 and more-often against header_maxsize. Previously a crafted message could
712 induce a crash of the recive process; now the message is cleanly rejected.
714 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
715 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
721 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
722 When redis_servers is set to a list of > 1 element, and the Redis servers
723 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
724 case of MOVED into a DEFER case instead, thus moving the query onto the
725 next server in the list. For a cluster of N elements, all N servers must
726 be defined in redis_servers.
728 GF/02 Catch and remove uninitialized value warning in exiqsumm
729 Check for existence of @ARGV before looking at $ARGV[0]
731 JH/01 Replace the store_release() internal interface with store_newblock(),
732 which internalises the check required to safely use the old one, plus
733 the allocate and data copy operations duplicated in both (!) of the
734 extant use locations.
736 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
737 modifier. This matches the restriction on the commandline.
739 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
740 Previously only the last row was returned.
742 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
743 we assumed that tags in the header were well-formed, and parsed the
744 element content after inspecting only the first char of the tag.
745 Assumptions at that stage could crash the receive process on malformed
748 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
749 While running the DKIM ACL we operate on the Permanent memory pool so that
750 variables created with "set" persist to the DATA ACL. Also (at any time)
751 DNS lookups that fail create cache records using the Permanent pool. But
752 expansions release any allocations made on the current pool - so a dnsdb
753 lookup expansion done in the DKIM ACL releases the memory used for the
754 DNS negative-cache, and bad things result. Solution is to switch to the
755 Main pool for expansions.
756 While we're in that code, add checks on the DNS cache during store_reset,
757 active in the testsuite.
758 Problem spotted, and debugging aided, by Wolfgang Breyha.
760 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
761 When none of the hosts presented to a transport match an already-open
762 connection, close it and proceed with the list. Previously we would
763 queue the message. Spotted by Lena with Yahoo, probably involving
766 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
767 Previously a spurious "250 OK id=" response was appended to the proper
770 JH/08 The "support for" informational output now, which built with Content
771 Scanning support, has a line for the malware scanner interfaces compiled
772 in. Interface can be individually included or not at build time.
774 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
775 by the template makefile "src/EDITME". The "STREAM" support for an older
776 ClamAV interface method is removed.
778 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
779 rows affected is given instead).
781 JH/11 The runtime Berkeley DB library version is now additionally output by
782 "exim -d -bV". Previously only the compile-time version was shown.
784 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
785 SMTP connection. Previously, when one had more receipients than the
786 first, an abortive onward connection was made. Move to full support for
787 multiple onward connections in sequence, handling cutthrough connection
788 for all multi-message initiating connections.
790 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
791 routers. Previously, a multi-recipient message would fail to match the
792 onward-connection opened for the first recipient, and cause its closure.
794 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
795 a timeout on read on a GnuTLS initiating connection, resulting in the
796 initiating connection being dropped. This mattered most when the callout
797 was marked defer_ok. Fix to keep the two timeout-detection methods
800 JH/15 Relax results from ACL control request to enable cutthrough, in
801 unsupported situations, from error to silently (except under debug)
802 ignoring. This covers use with PRDR, frozen messages, queue-only and
805 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
807 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
808 metadata, resulting in a crash in free().
810 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
811 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
812 Broken also in d185889f4, with init system revamp.
814 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
815 Previously we abruptly closed the connection after reading a malware-
816 found indication; now we go on to read the "scan ok" response line,
819 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
820 ACL. Previously, a crash would result.
822 JH/19 Speed up macro lookups during configuration file read, by skipping non-
823 macro text after a replacement (previously it was only once per line) and
824 by skipping builtin macros when searching for an uppercase lead character.
826 JH/20 DANE support moved from Experimental to mainline. The Makefile control
827 for the build is renamed.
829 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
830 was allocated for every new TLS startup, meaning one per message. Fix
831 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
833 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
834 reported the original. Fix to report (as far as possible) the ACL
835 result replacing the original.
837 JH/23 Fix memory leak during multi-message connections using STARTTLS under
838 OpenSSL. Certificate information is loaded for every new TLS startup,
839 and the resources needed to be freed.
841 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
843 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
846 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
847 DATA response info to the (existing) per-recipient response info for
848 the "C=" log element. It can have useful tracking info from the
849 destination system. Patch from Simon Arlott.
851 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
852 length value. Previously this would segfault.
854 HS/02 Support Avast multiline protoocol, this allows passing flags to
855 newer versions of the scanner.
857 JH/28 Ensure that variables possibly set during message acceptance are marked
858 dead before release of memory in the daemon loop. This stops complaints
859 about them when the debug_store option is enabled. Discovered specifically
860 for sender_rate_period, but applies to a whole set of variables.
861 Do the same for the queue-runner and queue-list loops, for variables set
862 from spool message files. Do the same for the SMTP per-message loop, for
863 certain variables indirectly set in ACL operations.
865 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
866 as a multi-recipient message from a mailinglist manager). The coding had
867 an arbitrary cutoff number of characters while checking for more input;
868 enforced by writing a NUL into the buffer. This corrupted long / fast
869 input. The problem was exposed more widely when more pipelineing of SMTP
870 responses was introduced, and one Exim system was feeding another.
871 The symptom is log complaints of SMTP syntax error (NUL chars) on the
872 receiving system, and refused recipients seen by the sending system
873 (propating to people being dropped from mailing lists).
874 Discovered and pinpointed by David Carter.
876 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
877 replaced by the ${authresults } expansion.
879 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
881 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
882 allows proper process termination in container environments.
884 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
885 Previously the "final dot" had a newline after it; ensure it is CR,LF.
887 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
888 and "err_perm", deprecated since 4.83 when the RFC-defined words
889 "temperror" and "permerror" were introduced.
891 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
892 transport-filters or DKIM-signing. The restriction was lost in the
893 consolidation of verify-callout and delivery SMTP handling.
894 Extend the restriction to also cover ARC-signing.
896 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
897 in defer=pass mode supply a 450 to the initiator. Previously the message
900 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
901 tls_require_ciphers is used as before.
903 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
904 "pass_unscanned". Only tmpfails from the scanner are written to
905 the paniclog, as they may require admin intervention (permission
906 denied, license issues). Other scanner errors (like decompression
907 bombs) do not cause a paniclog entry.
909 JH/36 Fix reinitialisation of DKIM logging variable between messages.
910 Previously it was possible to log spurious information in receive log
913 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
914 triggered odd behaviour from Outlook Express clients.
916 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
919 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
920 since the IETF WG has not yet settled on that versus the original
921 "bare" representation.
923 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
924 Previously the millisecond value corrupted the output.
925 Fix also for syslog_pid=no and log_selector +pid, for which the pid
926 corrupted the output.
932 JH/01 Rework error string handling in TLS interface so that the caller in
933 more cases is responsible for logging. This permits library-sourced
934 string to be attached to addresses during delivery, and collapses
935 pairs of long lines into single ones.
937 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
938 during configuration. Wildcards are allowed and expanded.
940 JH/02 Rework error string handling in DKIM to pass more info back to callers.
941 This permits better logging.
943 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
944 do not close it down and have the child transport start it up again on
945 the passed-on TCP connection. Instead, proxy the child (and any
946 subsequent ones) for TLS via a unix-domain socket channel. Logging is
947 affected: the continued delivery log lines do not have any DNSSEC, TLS
948 Certificate or OCSP information. TLS cipher information is still logged.
950 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
951 identical IP addresses on different listening ports. Will also affect
954 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
955 add noisy ifdef guards to special-case this sillyness.
956 Patch from Bernd Kuhls.
958 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
959 than 255 are no longer allowed.
961 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
962 Disable the session-cache too, which might reduce our load. Since we
963 currrectly use a new context for every connection, both as server and
964 client, there is no benefit for these.
965 GnuTLS appears to not support tickets server-side by default (we don't
966 call gnutls_session_ticket_enable_server()) but client side is enabled
967 by default on recent versions (3.1.3 +) unless the PFS priority string
970 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
971 <https://reproducible-builds.org/specs/source-date-epoch/>.
973 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
974 the check for any unsuccessful recipients did not notice the limit, and
975 erroneously found still-pending ones.
977 JH/08 Pipeline CHUNKING command and data together, on kernels that support
978 MSG_MORE. Only in-clear (not on TLS connections).
980 JH/09 Avoid using a temporary file during transport using dkim. Unless a
981 transport-filter is involved we can buffer the headers in memory for
982 creating the signature, and read the spool data file once for the
983 signature and again for transmission.
985 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
986 4.77 as the kernel support then wasn't solid, having issues in 64bit
987 mode. Now, it's been long enough. Add support for FreeBSD also.
989 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
990 case where the routing stage had gathered several addresses to send to
991 a host before calling the transport for the first, we previously failed
992 to close down TLS in the old transport process before passing the TCP
993 connection to the new process. The new one sent a STARTTLS command
994 which naturally failed, giving a failed delivery and bloating the retry
995 database. Investigation and fix prototype from Wolfgang Breyha.
997 JH/12 Fix check on SMTP command input synchronisation. Previously there were
998 false-negatives in the check that the sender had not preempted a response
999 or prompt from Exim (running as a server), due to that code's lack of
1000 awareness of the SMTP input buffering.
1002 PP/04 Add commandline_checks_require_admin option.
1003 Exim drops privileges sanely, various checks such as -be aren't a
1004 security problem, as long as you trust local users with access to their
1005 own account. When invoked by services which pass untrusted data to
1006 Exim, this might be an issue. Set this option in main configuration
1007 AND make fixes to the calling application, such as using `--` to stop
1010 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1011 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1012 responses to those, into a single TLS record each way (this usually means
1013 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1016 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1017 affects you only if you're dancing at the edge of the param size limits.
1018 If you are, and this message makes sense to you, then: raise the
1019 configured limit or use OpenSSL 1.1. Nothing we can do for older
1022 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1023 cmdline element to get the documented default one. Previously it was
1026 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1027 are unsuable for non-message-specific reasons - eg. connection timeout,
1028 banner-time rejection.
1030 JH/16 Fix logging of delivery remote port, when specified by router, under
1033 PP/06 Repair manualroute's ability to take options in any order, even if one
1034 is the name of a transport.
1037 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1039 JH/17 Change the list-building routines interface to use the expanding-string
1040 triplet model, for better allocation and copying behaviour.
1042 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1043 Previously it was constructed the first time a possibly-matching string
1044 was met in the configuration file input during startup; now it is done
1047 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1048 compatible one, to avoid the (poorly documented) possibility of a config
1049 file in the working directory redirecting the DB files, possibly correpting
1050 some existing file. CVE-2017-10140 assigned for BDB.
1052 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1053 cache-hot. Previously, although the result was properly cached, the
1054 initial verify call returned a defer.
1056 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1057 the main verify for receipient in uncached-mode.
1059 JH/22 Retire historical build files to an "unsupported" subdir. These are
1060 defined as "ones for which we have no current evidence of testing".
1062 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1063 if present. Previously it was ignored.
1065 JH/24 Start using specified-initialisers in C structure init coding. This is
1066 a C99 feature (it's 2017, so now considered safe).
1068 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1069 if was a fixed-sized field and bitmask ops via macros; it is now more
1072 PP/07 GitHub PR 56: Apply MariaDB build fix.
1073 Patch provided by Jaroslav Škarvada.
1075 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1076 during Coverity cleanups [4.87 JH/47]
1077 Diagnosis and fix provided by Michael Fischer v. Mollard.
1079 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1080 the right size to place the terminating semicolon on its own folded
1081 line, the header hash was calculated to an incorrect value thanks to
1082 the (relaxed) space the fold became.
1084 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1085 and confused the parent.
1087 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1088 which could crash as a result. This could lead to undeliverable messages.
1090 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1093 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1094 matters on fast-turnover and PID-randomising systems, which were getting
1095 out-of-order delivery.
1097 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1098 a possibly-overlapping copy. The symptom was that "Remote host closed
1099 connection in response to HELO" was logged instead of the actual 4xx
1102 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1103 Previously only that bufferd was discarded, resulting in SYMTP command
1106 JH/32 DKIM: when a message has multiple signatures matching an identity given
1107 in dkim_verify_signers, run the dkim acl once for each. Previously only
1108 one run was done. Bug 2189.
1110 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1111 "panic the current process" to "deliberately defer". The panic log is
1112 still written with the problem list name; the mail and reject logs now
1113 get a temp-reject line for the message that was being handled, saying
1114 something like "domains check lookup or other defer". The SMTP 451
1115 message is still "Temporary local problem".
1117 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1118 A crafted sequence of BDAT commands could result in in-use memory beeing
1119 freed. CVE-2017-16943.
1121 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1122 from SMTP input. Previously it was always done; now only done for DATA
1123 and not BDAT commands. CVE-2017-16944.
1125 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1126 to the message (such as an overlong header line). Previously this was
1127 not done and we did not exit BDAT mode. Followon from the previous item
1128 though a different problem.
1134 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1135 than -2003 did; needs libidn2 in addition to libidn.
1137 JH/02 The path option on a pipe transport is now expanded before use.
1139 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1140 Patch provided by "Björn", documentation fix added too.
1142 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1143 missing a wire-to-host endian conversion.
1145 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1146 close after a BDAT command line could be taken as a following command,
1147 giving a synch failure. Fix by only checking for synch immediately
1148 before acknowledging the chunk.
1150 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1151 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1152 macro. Patches provided by Josh Soref.
1154 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1155 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1156 by IANA as a service extension. However, John Klensin suggests that we
1159 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1160 the dkim code may be unix-mode line endings rather than smtp wire-format
1161 CRLF, so prepend a CR to any bare LF.
1163 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1164 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1166 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1167 the first were themselves being wrongly included in the feed into dkim
1168 processing; with most chunk sizes in use this resulted in an incorrect
1169 body hash calculated value.
1171 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1172 DKIM signature block, for verification. Although advised against by
1173 standards it is specifically not ruled illegal.
1175 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1177 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1178 missing a body hash (the bh= tag).
1180 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1181 It seems that HAProxy sends the Proxy Protocol information in clear and
1182 only then does a TLS startup, so do the same.
1184 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1185 TCP connections (such as for Spamd) unless the daemon successfully set
1186 Fast Open mode on its listening sockets. This fixes breakage seen on
1187 too-old kernels or those not configured for Fast Open, at the cost of
1188 requiring both directions being enabled for TFO, and TFO never being used
1189 by non-daemon-related Exim processes.
1191 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1192 endings, at least on the first header line. Try to canonify any that get
1193 past that check, despite the cost.
1195 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1196 now limited to an arbitrary five deep, while parsing addresses with the
1197 strip_excess_angle_brackets option enabled.
1199 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1200 instead leave the unprompted TLS handshake in socket buffer for the
1201 TLS library to consume.
1203 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1205 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1207 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1208 region in one operation, for speed, and this leaves assigned pointers
1209 dangling. Add checks run only under the testsuite which checks all
1210 variables at a store-reset and panics on a dangling pointer; add code
1211 explicitly nulling out all the variables discovered. Fixes one known
1212 bug: a transport crash, where a dangling pointer for $sending_ip_address
1213 originally assigned in a verify callout, is re-used.
1215 PP/06 Drop '.' from @INC in various Perl scripts.
1217 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1219 PP/08 Reduce a number of compilation warnings under clang; building with
1220 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1221 should be warning-free.
1223 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1225 HS/01 Fix portability problems introduced by PP/08 for platforms where
1226 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1228 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1229 chunk. This allows us to accept broken chunked messages. We need a more
1230 general solution here.
1232 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1233 already-broken messages in the queue.
1235 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1237 JH/19 Fix reference counting bug in routing-generated-address tracking.
1243 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1244 supports it and a size is available (ie. the sending peer gave us one).
1246 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1247 years of being deprecated). The replacements are the ACLs
1248 acl_smtp_mime and acl_not_smtp_mime.
1250 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1251 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1252 or even an in-clear connection were permitted. Now, if the host lookup
1253 was dnssec and dane was requested then the host is only used if the
1254 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1255 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1256 if one fails this test.
1257 This means that a poorly-configured remote DNS will make it incommunicado;
1258 but it protects against a DNS-interception attack on it.
1260 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1261 non-noisy when a race steals the message being considered.
1263 JH/05 If main configuration option tls_certificate is unset, generate a
1264 self-signed certificate for inbound TLS connections.
1266 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1267 in rewrites and routers.
1269 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1270 and logged a warning sing 4.83; now they are a configuration file error.
1272 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1273 (lacking @domain). Apply the same qualification processing as RCPT.
1275 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1277 JH/10 Support ${sha256:} applied to a string (as well as the previous
1280 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1281 a cutthrough deliver is pending, as we always want to make a connection.
1282 This also avoids re-routing the message when later placing the cutthrough
1283 connection after a verify cache hit.
1284 Do not update it with the verify result either.
1286 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1287 when routing results in more than one destination address.
1289 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1290 signing (which inhibits the cutthrough capability). Previously only
1291 the presence of an option was tested; now an expansion evaluating as
1292 empty is permissible (obviously it should depend only on data available
1293 when the cutthrough connection is made).
1295 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1296 the relevant preceding SMTP command did not note the pipelining mode.
1298 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1299 Previously they were not counted.
1301 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1302 as one having no matching records. Previously we deferred the message
1303 that needed the lookup.
1305 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1306 distinguished as "(=".
1308 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1309 for missing MX records. Previously it only worked for missing A records.
1311 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1313 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1314 after the data-go-ahead and data-ack. Patch from Jason Betts.
1316 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1317 even for a "none" policy. Patch from Tony Meyer.
1319 JH/22 Fix continued use of a connection for further deliveries. If a port was
1320 specified by a router, it must also match for the delivery to be
1323 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1324 When one of the recipients of a message was unsuitable for the connection
1325 (has no matching addresses), we lost track of needing to mark it
1326 deferred. As a result mail would be lost.
1328 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1330 JH/25 Decoding ACL controls is now done using a binary search; the source code
1331 takes up less space and should be simpler to maintain. Merge the ACL
1332 condition decode tables also, with similar effect.
1334 JH/26 Fix problem with one_time used on a redirect router which returned the
1335 parent address unchanged. A retry would see the parent address marked as
1336 delivered, so not attempt the (identical) child. As a result mail would
1339 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1340 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1341 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1344 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1345 Main config options check_{log,spool}_{inodes,space} are now
1346 100 inodes, 10MB unless set otherwise in the configuration.
1348 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1349 Previously it only applied to the main-section connection policy
1352 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1354 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1355 by me. Added RFC7919 DH primes as an alternative.
1357 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1358 are not in the system include path.
1360 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1361 GnuTLS, when a session startup failed (eg because the client disconnected)
1362 Exim did stdio operations after fclose. This was exposed by a recent
1363 change which nulled out the file handle after the fclose.
1365 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1366 signed directly by the cert-signing cert, rather than an intermediate
1367 OCSP-signing cert. This is the model used by LetsEncrypt.
1369 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1371 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1372 an incoming connection.
1374 HS/02 Bug 1802: Do not half-close the connection after sending a request
1377 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1378 fallback to "prime256v1".
1380 JH/34 SECURITY: Use proper copy of DATA command in error message.
1381 Could leak key material. Remotely exploitable. CVE-2016-9963.
1387 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1388 and 3.4.4 - once the server is enabled to respond to an OCSP request
1389 it does even when not requested, resulting in a stapling non-aware
1390 client dropping the TLS connection.
1392 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1393 support variable-length bit vectors. No functional change.
1395 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1396 The I= interface field on outgoing lines is now after the H= remote
1397 host field, same as incoming lines. There is a separate
1398 outgoing_interface log selector which allows you to disable the
1401 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1402 If not running log_selector +smtp_connection the mainlog would be held
1403 open indefinitely after a "too many connections" event, including to a
1404 deleted file after a log rotate. Leave the per net connection logging
1405 leaving it open for efficiency as that will be quickly detected by the
1406 check on the next write.
1408 HS/01 Bug 1671: Fix post transport crash.
1409 Processing the wait-<transport> messages could crash the delivery
1410 process if the message IDs didn't exist for some reason. When
1411 using 'split_spool_directory=yes' the construction of the spool
1412 file name failed already, exposing the same netto behaviour.
1414 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1415 mime_regex ACL conditions.
1417 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1418 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1419 to HELO, local diagnostic string.
1421 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1422 log line to debug. Even when configured with a tls authenticator many
1423 client connections are expected to not authenticate in this way, so
1424 an authenticate fail is not an error.
1426 HS/02 Add the Exim version string to the process info. This way exiwhat
1427 gives some more detail about the running daemon.
1429 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1430 matter for fast-change records such as DNSBLs.
1432 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1433 retry record, even if constant. There may be multiple transports with
1434 different interface settings and the retry behaviour needs to be kept
1437 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1439 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1441 JH/10 Bug 840: fix log_defer_output option of pipe transport
1443 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1444 hosts_randomize. This matters a lot when combined with mua_wrapper.
1446 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1447 ${quote_pgsql:<string>} operator.
1449 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1451 JH/14 Change header file name for freeradius-client. Relevant if compiling
1452 with Radius support; from the Gentoo tree and checked under Fedora.
1454 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1456 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1457 option transport_filter, meaning no filtering.
1459 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1461 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1462 defaults to "*" (all hosts). The variable is now available when not built
1463 with TLS, default unset, mainly to enable keeping the testsuite sane.
1464 If a server certificate is not supplied (via tls_certificate) an error is
1465 logged, and clients will find TLS connections fail on startup. Presumably
1466 they will retry in-clear.
1467 Packagers of Exim are strongly encouraged to create a server certificate
1468 at installation time.
1470 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1471 with the $config_file variable.
1473 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1474 in transport context, after the attempt, and per-recipient. The latter type
1475 is per host attempted. The event data is the error message, and the errno
1476 information encodes the lookup type (A vs. MX) used for the (first) host,
1477 and the trailing two digits of the smtp 4xx response.
1479 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1480 to write to mainlog (or rejectlog, paniclog) in the window between file
1481 creation and permissions/ownership being changed. Particularly affects
1482 installations where exicyclog is run as root, rather than exim user;
1483 result is that the running daemon panics and dies.
1485 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1487 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1488 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1489 "pri" and "weight". Note that the previous implicit priority given by the
1490 list order is no longer honoured.
1492 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1493 for DKIM processing.
1495 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1496 by defining SUPPORT_SOCKS.
1498 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1499 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1500 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1501 variables are renamed to proxy_{local,external}_{address,port}.
1503 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1504 for a build by defining SUPPORT_I18N
1506 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1507 of the query string, and make ${quote_redis:} do that quoting.
1509 JH/29 Move Events support from Experimental to mainline, enabled by default
1510 and removable for a build by defining DISABLE_EVENT.
1512 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1514 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1515 cached by the daemon.
1517 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1518 by defining LOOKUP_REDIS. The libhiredis library is required.
1520 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1521 keys are given for lookup.
1523 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1524 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1525 only supported when built with TLS support. The PolarSSL SHA routines
1526 are still used when the TLS library is too old for convenient support.
1528 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1529 openssl_options), for security. OpenSSL forces this from version 1.1.0
1530 server-side so match that on older versions.
1532 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1533 allocation for $value could be released as the expansion processing
1534 concluded, but leaving the global pointer active for it.
1536 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1537 and to use the domains and local_parts ACL conditions.
1539 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1540 incorrectly not doubled on cutthrough transmission, hence seen as a
1541 body-termination at the receiving system - resulting in truncated mails.
1542 Commonly the sender saw a TCP-level error, and retransmitted the message
1543 via the normal store-and-forward channel. This could result in duplicates
1544 received - but deduplicating mailstores were liable to retain only the
1545 initial truncated version.
1547 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1549 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1551 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1552 we're in there, support oversigning also; bug 1309.
1554 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1556 HS/04 Add support for keep_environment and add_environment options.
1558 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1559 either intentional arithmetic overflow during PRNG, or testing config-
1562 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1563 delivery resulted in actual delivery. Cancel cutthrough before DATA
1566 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1567 rejects a recipient: pass the reject to the originator.
1569 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1570 Many were false-positives and ignorable, but it's worth fixing the
1573 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1574 for the new environment-manipulation done at startup. Move the routines
1575 from being local to tls.c to being global via the os.c file.
1577 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1578 an extract embedded as result-arg for a map, the first arg for extract
1579 is unavailable so we cannot tell if this is a numbered or keyed
1580 extraction. Accept either.
1586 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1589 JH/02 The smtp transport option "multi_domain" is now expanded.
1591 JH/03 The smtp transport now requests PRDR by default, if the server offers
1594 JH/04 Certificate name checking on server certificates, when exim is a client,
1595 is now done by default. The transport option tls_verify_cert_hostnames
1596 can be used to disable this per-host. The build option
1597 EXPERIMENTAL_CERTNAMES is withdrawn.
1599 JH/05 The value of the tls_verify_certificates smtp transport and main options
1600 default to the word "system" to access the system default CA bundle.
1601 For GnuTLS, only version 3.0.20 or later.
1603 JH/06 Verification of the server certificate for a TLS connection is now tried
1604 (but not required) by default. The verification status is now logged by
1605 default, for both outbound TLS and client-certificate supplying inbound
1608 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1611 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1612 Status Notification (bounce) messages are now MIME format per RFC 3464.
1613 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1614 under the control of the dsn_advertise_hosts option, and routers may
1615 have a dsn_lasthop option.
1617 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1618 default, modifiable by a malware= option. The list separator for
1619 the options can now be changed in the usual way. Bug 68.
1621 JH/10 The smtp_receive_timeout main option is now expanded before use.
1623 JH/11 The incoming_interface log option now also enables logging of the
1624 local interface on delivery outgoing connections.
1626 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1627 if the interface and destination host and port all match.
1629 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1632 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1633 Patch from Andrew Lewis.
1635 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1636 now supports optional time-restrictions, weighting, and priority
1637 modifiers per server. Patch originally by <rommer@active.by>.
1639 JH/16 The spamd_address main option now supports a mixed list of local
1640 and remote servers. Remote servers can be IPv6 addresses, and
1641 specify a port-range.
1643 JH/17 Bug 68: The spamd_address main option now supports an optional
1644 timeout value per server.
1646 JH/18 Bug 1581: Router and transport options headers_add/remove can
1647 now have the list separator specified.
1649 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1652 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1655 JH/21 Support for the A6 type of dns record is withdrawn.
1657 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1658 rather than the verbs used.
1660 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1661 from 255 to 1024 chars.
1663 JH/24 Verification callouts now attempt to use TLS by default.
1665 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1666 are generic router options now. The defaults didn't change.
1668 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1669 Original patch from Alexander Shikoff, worked over by JH.
1671 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1672 files if ZCAT_COMMAND is not executable.
1674 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1676 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1678 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1679 Normally benign, it bites when the pair was led to by a CNAME;
1680 modern usage is to not canonicalize the domain to a CNAME target
1681 (and we were inconsistent anyway for A-only vs AAAA+A).
1683 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1685 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1686 when evaluating $sender_host_dnssec.
1688 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1689 $sender_helo_dnssec variable.
1691 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1693 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1695 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1697 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1698 documented as working, but never had. Support all but $spam_report.
1700 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1701 added for tls authenticator.
1703 HS/03 Add perl_taintmode main config option
1709 TL/01 When running the test suite, the README says that variables such as
1710 no_msglog_check are global and can be placed anywhere in a specific
1711 test's script, however it was observed that placement needed to be near
1712 the beginning for it to behave that way. Changed the runtest perl
1713 script to read through the entire script once to detect and set these
1714 variables, reset to the beginning of the script, and then run through
1715 the script parsing/test process like normal.
1717 TL/02 The BSD's have an arc4random API. One of the functions to induce
1718 adding randomness was arc4random_stir(), but it has been removed in
1719 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1720 function when detected.
1722 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1723 cause callback expansion.
1725 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1726 syntax errors in an expansion can be treated as a string instead of
1727 logging or causing an error, due to the internal use of bool_lax
1728 instead of bool when processing it.
1730 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1731 server certificates when making smtp deliveries.
1733 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1735 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1737 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1739 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1740 Merged patch from Sebastian Wiedenroth.
1742 JH/05 Fix results-pipe from transport process. Several recipients, combined
1743 with certificate use, exposed issues where response data items split
1744 over buffer boundaries were not parsed properly. This eventually
1745 resulted in duplicates being sent. This issue only became common enough
1746 to notice due to the introduction of connection certificate information,
1747 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1749 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1750 size buffer was used, resulting in syntax errors when an expansion
1753 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1754 version 3.3.6 or later.
1756 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1757 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1758 both become "event_action", the variables become $event_name, $event_data
1759 and $event_defer_errno. There is a new variable $verify_mode, usable in
1760 routers, transports and related events. The tls:cert event is now also
1761 raised for inbound connections, if the main configuration event_action
1764 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1765 early OCSP support, but no stapling (appears to be less than 1.0.0).
1767 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1768 server certificate names available under the smtp transport option
1769 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1772 JH/10 Time-related extraction expansions from certificates now use the main
1773 option "timezone" setting for output formatting, and are consistent
1774 between OpenSSL and GnuTLS compilations. Bug 1541.
1776 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1777 encoded parameter in the incoming message. Bug 1558.
1779 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1780 include certificate info, eximon was claiming there were spoolfile
1783 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1785 JH/14 Log delivery-related information more consistently, using the sequence
1786 "H=<name> [<ip>]" wherever possible.
1788 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1789 are problematic for Debian distribution, omit them from the release
1792 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1794 JH/16 Fix string representation of time values on 64bit time_t architectures.
1797 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1798 output list separator was used.
1803 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1804 checkers that were complaining about end of non-void function with no
1807 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1808 This was a regression introduced in 4.83 by another bugfix.
1810 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1812 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1813 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1819 TF/01 Correctly close the server side of TLS when forking for delivery.
1821 When a message was received over SMTP with TLS, Exim failed to clear up
1822 the incoming connection properly after forking off the child process to
1823 deliver the message. In some situations the subsequent outgoing
1824 delivery connection happened to have the same fd number as the incoming
1825 connection previously had. Exim would try to use TLS and fail, logging
1826 a "Bad file descriptor" error.
1828 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1829 utilities have not been installed.
1831 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1832 temporary space as the ACL may create new global variables.
1834 TL/01 LDAP support uses per connection or global context settings, depending
1835 upon the detected version of the libraries at build time.
1837 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1838 to extract and use the src ip:port in logging and expansions as if it
1839 were a direct connection from the outside internet. PPv2 support was
1840 updated based on HAProxy spec change in May 2014.
1842 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1844 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1845 Properly escape header and check for NULL return.
1847 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1850 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1852 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1853 characters in header names, implemented as a verify condition.
1854 Contributed by Michael Fischer v. Mollard.
1856 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1857 results permerror and temperror. Previous values are deprecated but
1858 still accepted. In a future release, err_perm and err_temp will be
1859 completely removed, which will be a backward incompatibility if the
1860 ACL tests for either of these two old results. Patch contributed by
1861 user bes-internal on the mailing list.
1863 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1865 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1866 selectors, in both main and reject logs.
1868 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1871 JH/07 Add malware type "sock" for talking to simple daemon.
1873 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1875 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1876 routers/transports under cutthrough routing.
1878 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1879 numbers. Touch up "bool" conditional to keep the same definition.
1881 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1883 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1885 JH/12 Expand items in router/transport headers_add or headers_remove lists
1886 individually rather than the list as a whole. Bug 1452.
1888 Required for reasonable handling of multiple headers_ options when
1889 they may be empty; requires that headers_remove items with embedded
1890 colons must have them doubled (or the list-separator changed).
1892 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1893 view the policy declared in the DMARC record. Currently, $dmarc_status
1894 is a combined value of both the record presence and the result of the
1897 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1899 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1900 dnslookup router and the smtp transport (applying to the forward
1903 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1904 of ldap servers used for a specific lookup. Patch provided by Heiko
1907 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1908 New variable $lookup_dnssec_authenticated for observability.
1910 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1911 Patch submitted by Lars Timman.
1913 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1915 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1916 Requires trusted mode and valid format message id, aborts otherwise.
1917 Patch contributed by Heiko Schlichting.
1919 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1920 certextract with support for various fields. Bug 1358.
1922 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1923 is requested by default, modifiable by smtp transport option
1926 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1927 operate on certificate variables to give certificate fingerprints
1928 Also new ${sha256:cert_variable}.
1930 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1932 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1933 Christian Aistleitner.
1935 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1937 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1938 file. Patch from Wolfgang Breyha.
1940 JH/25 Expand the coverage of the delivery $host and $host_address to
1941 client authenticators run in verify callout. Bug 1476.
1943 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1944 align with daemon_smtp_ports. Bug 72.
1946 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1947 support and error reporting did not work properly.
1949 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1950 and is readable. Patch from Andrew Colin Kissa.
1952 TL/14 Enhance documentation of ${run expansion and how it parses the
1953 commandline after expansion, particularly in the case when an
1954 unquoted variable expansion results in an empty value.
1956 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1958 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1959 by renaming away from it. Addresses GH issue 15, reported by
1962 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1964 TL/15 SECURITY: prevent double expansion in math comparison functions
1965 (can expand unsanitized data). Not remotely exploitable.
1972 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1974 PP/02 Make -n do something, by making it not do something.
1975 When combined with -bP, the name of an option is not output.
1977 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1980 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1981 $sender_host_name and config options to manage this, and basic check
1984 PP/05 DSCP support for outbound connections and control modifier for inbound.
1986 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1987 (Only plugin which currently uses this is kerberos4, which nobody should
1988 be using, but we should make it available and other future plugins might
1989 conceivably use it, even though it would break NAT; stuff *should* be
1990 using channel bindings instead).
1992 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1993 name; added for Sendmail compatibility; requires admin caller.
1994 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1995 just ignore it); requires trusted caller.
1996 Also parse but ignore: -Ac -Am -X<logfile>
1999 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2001 TL/02 Add +smtp_confirmation as a default logging option.
2003 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2004 Patch by Magnus Holmgren from 2007-02-20.
2006 TL/04 Bugzilla 1281 - Spec typo.
2007 Bugzilla 1283 - Spec typo.
2008 Bugzilla 1290 - Spec grammar fixes.
2010 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2012 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2014 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2015 mailing list by by Dmitry Isaikin.
2017 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2019 JH/02 Support "G" suffix to numbers in ${if comparisons.
2021 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2023 NM/01 Bugzilla 1197 - Spec typo
2024 Bugzilla 1196 - Spec examples corrections
2026 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2028 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2029 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2032 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2033 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2035 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2036 "acl {{name}{arg}...}", and optional args on acl condition
2039 JH/05 Permit multiple router/transport headers_add/remove lines.
2041 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2043 JH/07 Avoid using a waiting database for a single-message-only transport.
2044 Performance patch from Paul Fisher. Bugzilla 1262.
2046 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2049 JH/09 Add $headers_added variable, with content from use of ACL modifier
2050 add_header (but not yet added to the message). Bugzilla 199.
2052 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2053 Pulled from Bugzilla 817 by Wolfgang Breyha.
2055 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2057 (nb: this is the same fix as in Exim 4.80.1)
2059 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2062 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2065 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2067 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2068 advertise SMTP AUTH mechanism to us, instead of a generic
2069 protocol violation error. Also, make Exim more robust to bad
2070 data from the Dovecot auth socket.
2072 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2074 When a queue runner is handling a message, Exim first routes the
2075 recipient addresses, during which it prunes them based on the retry
2076 hints database. After that it attempts to deliver the message to
2077 any remaining recipients. It then updates the hints database using
2080 So if a recipient address works intermittently, it can get repeatedly
2081 deferred at routing time. The retry hints record remains fresh so the
2082 address never reaches the final cutoff time.
2084 This is a fairly common occurrence when a user is bumping up against
2085 their storage quota. Exim had some logic in its local delivery code
2086 to deal with this. However it did not apply to per-recipient defers
2087 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2089 This change adds a proper retry rule check during routing so that the
2090 final cutoff time is checked against the message's age. We only do
2091 this check if there is an address retry record and there is not a
2092 domain retry record; this implies that previous attempts to handle
2093 the address had the retry_use_local_parts option turned on. We use
2094 this as an approximation for the destination being like a local
2095 delivery, as in LMTP.
2097 I suspect this new check makes the old local delivery cutoff check
2098 redundant, but I have not verified this so I left the code in place.
2100 TF/02 Correct gecos expansion when From: is a prefix of the username.
2102 Test 0254 submits a message to Exim with the header
2106 When I ran the test suite under the user fanf2, Exim expanded
2107 the header to contain my full name, whereas it should have added
2108 a Resent-Sender: header. It erroneously treats any prefix of the
2109 username as equal to the username.
2111 This change corrects that bug.
2113 GF/01 DCC debug and logging tidyup
2114 Error conditions log to paniclog rather than rejectlog.
2115 Debug lines prefixed by "DCC: " to remove any ambiguity.
2117 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2119 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2120 Bug spotted by Jeremy Harris; was flawed since initial commit.
2121 Would have resulted in OCSP responses post-SNI triggering an Exim
2122 NULL dereference and crash.
2124 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2126 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2127 Bug detection, analysis and fix by Samuel Thibault.
2128 Bugzilla 1331, Debian bug #698092.
2130 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2132 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2133 Server implementation by Todd Lyons, client by JH.
2134 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2135 config variable "prdr_enable" controls whether the server
2136 advertises the facility. If the client requests PRDR a new
2137 acl_data_smtp_prdr ACL is called once for each recipient, after
2138 the body content is received and before the acl_smtp_data ACL.
2139 The client is controlled by both of: a hosts_try_prdr option
2140 on the smtp transport, and the server advertisement.
2141 Default client logging of deliveries and rejections involving
2142 PRDR are flagged with the string "PRDR".
2144 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2145 fclose(). Diagnosis by Todd Lyons.
2147 PP/17 Update configure.default to handle IPv6 localhost better.
2148 Patch by Alain Williams (plus minor tweaks).
2151 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2152 This is now consistent with GnuTLS, and is now documented: the
2153 previous undocumented portable approach to treating the option as
2154 unset was to force an expansion failure. That still works, and
2155 an empty string is now equivalent.
2157 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2158 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2159 not performing validation itself.
2161 PP/20 Added force_command boolean option to pipe transport.
2162 Patch from Nick Koston, of cPanel Inc.
2164 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2167 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2169 PP/21 Fix eximon continuous updating with timestamped log-files.
2170 Broken in a format-string cleanup in 4.80, missed when I repaired the
2171 other false fix of the same issue.
2172 Report and fix from Heiko Schlichting.
2175 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2176 Report from Prashanth Katuri.
2178 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2179 It's SecureTransport, so affects any MacOS clients which use the
2180 system-integrated TLS libraries, including email clients.
2182 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2183 using a MIME ACL for non-SMTP local injection.
2184 Report and assistance in diagnosis by Warren Baker.
2186 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2188 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2190 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2191 last id that failed so it may be referenced in subsequent ACL's.
2193 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2196 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2197 ldap library initialization, allowing self-signed CA's to be
2198 used. Also properly sets require_cert option later in code by
2199 using NULL (global ldap config) instead of ldap handle (per
2200 session). Bug diagnosis and testing by alxgomz.
2202 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2203 the src/util/ subdirectory.
2205 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2206 renamed to Transport Post Delivery Action by Jeremy Harris, as
2209 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2210 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2211 redis_servers = needs to be configured which will be used by the redis
2212 lookup. Patch from Warren Baker, of The Packet Hub.
2214 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2216 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2217 hostname or reverse DNS when processing a host list. Used suggestions
2218 from multiple comments on this bug.
2220 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2222 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2223 Missed a few lines, added it to make the runtest require no keyboard
2226 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2227 contains upper case chars. Make router use caseful_local_part.
2229 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2230 support when GnuTLS has been built with p11-kit.
2236 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2238 This, or similar/improved, will also be change PP/11 of 4.82.
2244 PP/01 Handle short writes when writing local log-files.
2245 In practice, only affects FreeBSD (8 onwards).
2246 Bugzilla 1053, with thanks to Dmitry Isaikin.
2248 NM/01 Bugzilla 949 - Documentation tweak
2250 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2253 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2255 PP/02 Implemented gsasl authenticator.
2257 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2259 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2260 `pkg-config foo` for cflags/libs.
2262 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2263 with rest of GSASL and with heimdal_gssapi.
2265 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2266 `pkg-config foo` for cflags/libs for the TLS implementation.
2268 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2269 properties get this fed in as external SSF. A number of robustness
2270 and debugging improvements to the cyrus_sasl authenticator.
2272 PP/08 cyrus_sasl server now expands the server_realm option.
2274 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2275 Patch by Jeremy Harris.
2277 PP/10 Added dbmjz lookup type.
2279 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2281 PP/12 MAIL args handles TAB as well as SP, for better interop with
2282 non-compliant senders.
2283 Analysis and variant patch by Todd Lyons.
2285 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2286 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2287 Patch from Dirk Mueller <dmueller@suse.com>
2289 PP/13 tls_peerdn now print-escaped for spool files.
2290 Observed some $tls_peerdn in wild which contained \n, which resulted
2291 in spool file corruption.
2293 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2294 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2295 or write after TLS renegotiation, which otherwise led to messages
2298 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2299 as a tracking header (ie: a signed header comes before the signature).
2300 Patch from Wolfgang Breyha.
2302 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2303 comma-sep list; embedded commas doubled.
2305 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2307 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2309 Report and patch from Dmitry Banschikov.
2311 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2312 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2313 failures. We appear to now support TLS1.1+ with Exim.
2315 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2316 lets Exim select keys and certificates based upon TLS SNI from client.
2317 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2318 before an outbound SMTP session. New log_selector, +tls_sni.
2320 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2321 NULL dereference. Report and patch from Alun Jones.
2323 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2324 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2325 Not seeing resolver debug output on NetBSD, but suspect this is a
2326 resolver implementation change.
2328 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2329 Left warnings. Added "eximon gdb" invocation mode.
2331 PP/21 Defaulting "accept_8bitmime" to true, not false.
2333 PP/22 Added -bw for inetd wait mode support.
2335 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2336 locate the relevant includes and libraries. Made this the default.
2338 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2339 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2341 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2342 This may cause build issues on older platforms.
2344 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2345 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2346 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2347 Added SNI support via GnuTLS too.
2348 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2350 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2352 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2353 Applied second patch from Janne, implementing suggestion to default
2354 multiple-strings-in-record handling to match SPF spec.
2356 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2358 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2359 read-only, out of scope).
2360 Patch from Wolfgang Breyha, report from Stuart Northfield.
2362 PP/29 Fix three issues highlighted by clang analyser static analysis.
2363 Only crash-plausible issue would require the Cambridge-specific
2364 iplookup router and a misconfiguration.
2365 Report from Marcin Mirosław.
2367 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2369 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2370 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2371 As part of this, removing so much warning spew let me fix some minor
2372 real issues in debug logging.
2374 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2375 assignment on my part. Fixed.
2377 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2378 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2379 Janne Snabb (who went above and beyond: thank you).
2381 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2382 string otherwise requires a connection and a bunch more work and it's
2383 relatively easy to get wrong. Should also expose TLS library linkage
2386 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2387 64-bit ${eval} (JH/03).
2389 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2390 GNU libc to support some of the 64-bit stuff, should not lead to
2391 conflicts. Defined before os.h is pulled in, so if a given platform
2392 needs to override this, it can.
2394 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2395 protection layer was required, which is not implemented.
2396 Bugzilla 1254, patch from Wolfgang Breyha.
2398 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2399 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2400 tls_dhparam take prime identifiers. Also unbreak combination of
2401 OpenSSL+DH_params+TLSSNI.
2403 PP/39 Disable SSLv2 by default in OpenSSL support.
2409 PP/01 Solaris build fix for Oracle's LDAP libraries.
2410 Bugzilla 1109, patch from Stephen Usher.
2412 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2414 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2417 TF/02 Fix a couple more cases where we did not log the error message
2418 when unlink() failed. See also change 4.74-TF/03.
2420 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2421 lock up or crash if it happened to be inside a call to libc when it
2422 got a SIGUSR1 from exiwhat.
2424 The SIGUSR1 handler appends the current process status to the process
2425 log which is later printed by exiwhat. It used to use the general
2426 purpose logging code to do this, but several functions it calls are
2427 not safe for signals.
2429 The new output code in the SIGUSR1 handler is specific to the process
2430 log, and simple enough that it's easy to inspect for signal safety.
2431 Removing some special cases also simplifies the general logging code.
2432 Removing the spurious timestamps from the process log simplifies
2435 TF/04 Improved ratelimit ACL condition.
2437 The /noupdate option has been deprecated in favour of /readonly which
2438 has clearer semantics. The /leaky, /strict, and /readonly update modes
2439 are mutually exclusive. The update mode is no longer included in the
2440 database key; it just determines when the database is updated. (This
2441 means that when you upgrade Exim will forget old rate measurements.)
2443 Exim now checks that the per_* options are used with an update mode that
2444 makes sense for the current ACL. For example, when Exim is processing a
2445 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2446 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2447 must specify per_mail/readonly. If you omit the update mode it defaults to
2448 /leaky where that makes sense (as before) or /readonly where required.
2450 The /noupdate option is now undocumented but still supported for
2451 backwards compatibility. It is equivalent to /readonly except that in
2452 ACLs where /readonly is required you may specify /leaky/noupdate or
2453 /strict/noupdate which are treated the same as /readonly.
2455 A useful new feature is the /count= option. This is a generalization
2456 of the per_byte option, so that you can measure the throughput of other
2457 aggregate values. For example, the per_byte option is now equivalent
2458 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2460 The per_rcpt option has been generalized using the /count= mechanism
2461 (though it's more complicated than the per_byte equivalence). When it is
2462 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2463 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2464 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2465 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2466 Note that using per_rcpt with a non-readonly update mode in more than
2467 one ACL will cause the recipients to be double-counted. (The per_mail
2468 and per_byte options don't have this problem.)
2470 The handling of very low rates has changed slightly. If the computed rate
2471 is less than the event's count (usually one) then this event is the first
2472 after a long gap. In this case the rate is set to the same as this event's
2473 count, so that the first message of a spam run is counted properly.
2475 The major new feature is a mechanism for counting the rate of unique
2476 events. The new per_addr option counts the number of different
2477 recipients that someone has sent messages to in the last time period. It
2478 behaves like per_rcpt if all the recipient addresses are different, but
2479 duplicate recipient addresses do not increase the measured rate. Like
2480 the /count= option this is a general mechanism, so the per_addr option
2481 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2482 example, measure the rate that a client uses different sender addresses
2483 with the options per_mail/unique=$sender_address. There are further
2484 details in the main documentation.
2486 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2488 TF/06 Removed a few PCRE remnants.
2490 TF/07 Automatically extract Exim's version number from tags in the git
2491 repository when doing development or release builds.
2493 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2494 Bugzilla 879. Patch from Paul Fisher.
2496 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2497 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2500 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2502 PP/05 Variable $av_failed, true if the AV scanner deferred.
2503 Bugzilla 1078. Patch from John Horne.
2505 PP/06 Stop make process more reliably on build failure.
2506 Bugzilla 1087. Patch from Heiko Schlittermann.
2508 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2509 Bugzilla 1089. Patch from Heiko Schlittermann.
2511 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2512 Bugzilla 1131. Patch from Holger Weiß.
2514 PP/09 Handle IPv6 addresses with SPF.
2515 Bugzilla 860. Patch from Wolfgang Breyha.
2517 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2519 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2522 PP/11 match_* no longer expand right-hand-side by default.
2523 New compile-time build option, EXPAND_LISTMATCH_RHS.
2524 New expansion conditions, "inlist", "inlisti".
2526 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2528 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2530 PP/14 fix log_write() format string regression from TF/03.
2531 Bugzilla 1152. Patch from Dmitry Isaikin.
2537 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2539 PP/02 Harmonised TLS library version reporting; only show if debugging.
2540 Layout now matches that introduced for other libraries in 4.74 PP/03.
2542 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2544 PP/04 New "dns_use_edns0" global option.
2546 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2549 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2550 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2552 TK/01 Updated PolarSSL code to 0.14.2.
2553 Bugzilla 1097. Patch from Andreas Metzler.
2555 PP/07 Catch divide-by-zero in ${eval:...}.
2556 Fixes bugzilla 1102.
2558 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2561 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2562 format-string attack -- SECURITY: remote arbitrary code execution.
2564 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2565 time unintentionally subject to list matching rules, letting the header
2566 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2567 arbitrary string expansion). This allowed for information disclosure.
2569 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2570 INT_MIN/-1 -- value coerced to INT_MAX.
2576 NM/01 Workaround for PCRE version dependency in version reporting
2579 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2580 This fixes portability to compilers other than gcc, notably
2581 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2583 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2584 makefiles for portability to HP-UX and POSIX correctness.
2586 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2587 Also via indented variable definition in the Makefile.
2588 (Debugging by Oliver Heesakkers).
2590 PP/02 Restore caching of spamd results with expanded spamd_address.
2591 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2593 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2594 Improves build reliability. Fix from: Frank Elsner
2596 NM/02 Fix wide character breakage in the rfc2047 coding
2597 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2599 NM/03 Allow underscore in dnslist lookups
2600 Fixes bug 1026. Patch from Graeme Fowler
2602 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2603 Code patches from Adam Ciarcinski of NetBSD.
2605 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2608 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2609 is logged, to avoid truncation. Patch from John Horne.
2611 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2612 Patch from Jakob Hirsch.
2614 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2615 SQL string expansion failure details.
2616 Patch from Andrey Oktyabrski.
2618 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2619 Patch from Simon Arlott.
2621 PP/09 New lookups functionality failed to compile on old gcc which rejects
2622 extern declarations in function scope.
2623 Patch from Oliver Fleischmann
2625 PP/10 Use sig_atomic_t for flags set from signal handlers.
2626 Check getgroups() return and improve debugging.
2627 Fixed developed for diagnosis in bug 927 (which turned out to be
2630 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2631 Patch from Mark Zealey.
2633 PP/12 Bugzilla 1056: Improved spamd server selection.
2634 Patch from Mark Zealey.
2636 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2637 Based on patch from Heiko Schlittermann.
2639 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2640 Patch from Uwe Doering, sign-off by Michael Haardt.
2642 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2643 variable declaration deep within a block. Bug and patch from
2646 PP/15 lookups-Makefile IRIX compatibility coercion.
2648 PP/16 Make DISABLE_DKIM build knob functional.
2650 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2651 Patch by Simon Arlott
2653 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2654 variable argument macros. Our copy now differs from upstream.
2660 TF/01 Failure to get a lock on a hints database can have serious
2661 consequences so log it to the panic log.
2663 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2664 controlled using the smtp_confirmation log selector.
2666 TF/03 Include the error message when we fail to unlink a spool file.
2668 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2669 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2670 for maintaining out-of-tree patches for some time.
2672 PP/01 Bugzilla 139: Documentation and portability issues.
2673 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2674 Handle per-OS dynamic-module compilation flags.
2676 PP/02 Let /dev/null have normal permissions.
2677 The 4.73 fixes were a little too stringent and complained about the
2678 permissions on /dev/null. Exempt it from some checks.
2679 Reported by Andreas M. Kirchwitz.
2681 PP/03 Report version information for many libraries, including
2682 Exim version information for dynamically loaded libraries. Created
2683 version.h, now support a version extension string for distributors
2684 who patch heavily. Dynamic module ABI change.
2686 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2687 privilege escalation vulnerability whereby the Exim run-time user
2688 can cause root to append content of the attacker's choosing to
2691 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2694 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2695 If dropping privileges for untrusted macros, we disabled normal logging
2696 on the basis that it would fail; for the Exim run-time user, this is not
2697 the case, and it resulted in successful deliveries going unlogged.
2698 Fixed. Reported by Andreas Metzler.
2704 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2705 only prepend for the Resent-* case. Fixes regression introduced in
2706 Exim 4.70 by NM/22 for Bugzilla 607.
2708 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2709 increasing numbers of administrators be bitten by this.
2711 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2713 PP/03 Bugzilla 994: added openssl_options main configuration option.
2715 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2717 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2719 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2721 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2722 without a peer certificate, leading to a segfault because of an
2723 assumption that peers always have certificates. Be a little more
2724 paranoid. Problem reported by Martin Tscholak.
2726 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2727 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2728 NB: ClamAV planning to remove STREAM in "middle of 2010".
2729 CL also introduces -bmalware, various -d+acl logging additions and
2730 more caution in buffer sizes.
2732 PP/09 Implemented reverse_ip expansion operator.
2734 PP/10 Bugzilla 937: provide a "debug" ACL control.
2736 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2738 PP/12 Bugzilla 973: Implement --version.
2740 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2742 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2744 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2746 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2747 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2748 ignore trailing whitespace.
2750 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2752 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2755 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2756 Notification from Dr Andrew Aitchison.
2758 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2759 ExtendedDetectionInfo response format.
2760 Notification from John Horne.
2762 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2765 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2766 XSL and documented dependency on system catalogs, with examples of how
2769 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2772 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2773 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2774 configuration file which is writeable by the Exim user or group.
2776 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2777 of configuration files to cover files specified with the -C option if
2778 they are going to be used with root privileges, not just the default
2781 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2782 option (effectively making it always true).
2784 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2785 files to be used while preserving root privileges.
2787 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2788 that rogue child processes cannot use them.
2790 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2791 run-time user, instead of root.
2793 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2794 Exim run-time user without dropping privileges.
2796 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2797 result string, instead of calling string_vformat() twice with the same
2800 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2801 for other users. Others should always drop root privileges if they use
2802 -C on the command line, even for a whitelisted configure file.
2804 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2806 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2812 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2813 $data_path, and $header_path variables; fixed documentation bugs and
2816 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2817 exipick to access non-standard spools, including the "frozen" queue
2820 NM/01 Bugzilla 965: Support mysql stored procedures.
2821 Patch from Alain Williams
2823 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2825 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2826 Patch from Andreas Metzler
2828 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2829 Patch from Kirill Miazine
2831 NM/05 Bugzilla 671: Added umask to procmail example.
2833 JJ/03 installed exipick 20100323.0, fixing doc bug
2835 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2836 directory. Notification and patch from Dan Rosenberg.
2838 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2840 TK/02 Improve log output when DKIM signing operation fails.
2842 MH/01 Treat the transport option dkim_domain as a colon separated
2843 list, not as a single string, and sign the message with each element,
2844 omitting multiple occurences of the same signer.
2846 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2847 Bugzilla 985, 986. Patch by Simon Arlott
2849 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2850 Patch by Simon Arlott
2852 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2853 MBX locking. Notification from Dan Rosenberg.
2859 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2861 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2863 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2865 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2867 NM/04 Bugzilla 915: Flag broken perl installation during build.
2873 TK/01 Added patch by Johannes Berg that expands the main option
2874 "spamd_address" if it starts with a dollar sign.
2876 TK/02 Write list of recipients to X-Envelope-Sender header when building
2877 the mbox-format spool file for content scanning (suggested by Jakob
2880 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2881 (http://www.dcc-servers.net/) support via dccifd. Activated by
2882 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2884 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2885 by Mark Daniel Reidel <mr@df.eu>.
2887 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2888 When building exim an external PCRE library is now needed -
2889 PCRE is a system library on the majority of modern systems.
2890 See entry on PCRE_LIBS in EDITME file.
2892 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2893 conversation. Added nologin parameter to request.
2894 Patch contributed by Kirill Miazine.
2896 TF/01 Do not log submission mode rewrites if they do not change the address.
2898 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2900 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2901 log files in place. Contributed by Roberto Lima.
2903 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2905 TF/03 Bugzilla 615: When checking the local_parts router precondition
2906 after a local_part_suffix or local_part_prefix option, Exim now
2907 does not use the address's named list lookup cache, since this
2908 contains cached lookups for the whole local part.
2910 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2911 Robert Millan. Documentation is in experimental-spec.txt.
2913 TF/04 Bugzilla 668: Fix parallel build (make -j).
2915 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2917 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2918 Patch provided by Jan Srzednicki.
2920 TF/05 Leading white space used to be stripped from $spam_report which
2921 wrecked the formatting. Now it is preserved.
2923 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2924 that they are available at delivery time.
2926 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2928 TF/08 TLS error reporting now respects the incoming_interface and
2929 incoming_port log selectors.
2931 TF/09 Produce a more useful error message if an SMTP transport's hosts
2932 setting expands to an empty string.
2934 NM/06 Bugzilla 744: EXPN did not work under TLS.
2935 Patch provided by Phil Pennock.
2937 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2938 Patch provided by Richard Godbee.
2940 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2941 acl_smtp_notquit, added index entry.
2943 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2944 Patch provided by Eugene Bujak.
2946 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2947 accept(). Patch provided by Maxim Dounin.
2949 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2950 Patch provided by Phil Pennock.
2952 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2954 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2955 Patch provided by Brad "anomie" Jorsch.
2957 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2958 Patch provided by Dean Brooks.
2960 TK/05 Add native DKIM support (does not depend on external libraries).
2962 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2963 Patch provided by Graeme Fowler.
2965 NM/16 Bugzilla 851: Documentation example syntax fix.
2967 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2969 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2972 NM/19 Bugzilla 745: TLS version reporting.
2973 Patch provided by Phil Pennock.
2975 NM/20 Bugzilla 167: bool: condition support.
2976 Patch provided by Phil Pennock.
2978 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2979 clients. Patch provided by Phil Pennock.
2981 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2982 Patch provided by Brad "anomie" Jorsch.
2984 NM/23 Bugzilla 687: Fix misparses in eximstats.
2985 Patch provided by Heiko Schlittermann.
2987 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2988 Patch provided by Heiko Schlittermann.
2990 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2991 plus update to original patch.
2993 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2995 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2996 Patch provided by David Brownlee.
2998 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3000 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3002 NM/30 Bugzilla 888: TLS documentation bugfixes.
3004 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3006 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3007 Unlike the original bugzilla I have changed all shell scripts in src tree.
3009 NM/33 Bugzilla 898: Transport filter timeout fix.
3010 Patch by Todd Rinaldo.
3012 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3013 Patch by Serge Demonchaux.
3015 NM/35 Bugzilla 39: Base64 decode bug fixes.
3016 Patch by Jakob Hirsch.
3018 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3020 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3022 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3024 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3030 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3031 ALT-N's libdkim that I have put here:
3032 http://duncanthrax.net/exim-experimental/
3034 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3035 were called 'true' and it seems that C99 defines that as a reserved
3036 keyword to be used with 'bool' variable types. That means you could
3037 not include C99-style headers which use bools without triggering
3038 build errors in sieve.c.
3040 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3041 as mailq or other aliases. Changed the --help handling significantly
3042 to do whats expected. exim_usage() emits usage/help information.
3044 SC/01 Added the -bylocaldomain option to eximstats.
3046 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3048 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3050 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3056 PH/01 Another patch from the Sieve maintainer.
3058 PH/02 When an IPv6 address is converted to a string for single-key lookup
3059 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3060 dots are used instead of colons so that keys in lsearch files need not
3061 contain colons. This was done some time before quoting was made available
3062 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3063 (notated using the quote facility) so as to distinguish them from IPv4
3064 keys. This meant that lookups for IP addresses in host lists did not work
3065 for iplsearch lookups.
3067 This has been fixed by arranging for IPv6 addresses to be expressed with
3068 colons if the lookup type is iplsearch. This is not incompatible, because
3069 previously such lookups could never work.
3071 The situation is now rather anomalous, since one *can* have colons in
3072 ordinary lsearch keys. However, making the change in all cases is
3073 incompatible and would probably break a number of configurations.
3075 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3078 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3079 conversion specification without a maximum field width, thereby enabling
3080 a rogue spamd server to cause a buffer overflow. While nobody in their
3081 right mind would setup Exim to query an untrusted spamd server, an
3082 attacker that gains access to a server running spamd could potentially
3083 exploit this vulnerability to run arbitrary code as the Exim user.
3085 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3086 $primary_hostname instead of what libspf2 thinks the hosts name is.
3088 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3089 a directory entry by the name of the lookup key. Previously, if a
3090 symlink pointed to a non-existing file or a file in a directory that
3091 Exim lacked permissions to read, a lookup for a key matching that
3092 symlink would fail. Now it is enough that a matching directory entry
3093 exists, symlink or not. (Bugzilla 503.)
3095 PH/03 The body_linecount and body_zerocount variables are now exported in the
3098 PH/04 Added the $dnslist_matched variable.
3100 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3101 This means they are set thereafter only if the connection becomes
3104 PH/06 Added the client_condition to authenticators so that some can be skipped
3105 by clients under certain conditions.
3107 PH/07 The error message for a badly-placed control=no_multiline_responses left
3108 "_responses" off the end of the name.
3110 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3112 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3113 (without spaces) instead of just copying the configuration text.
3115 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3117 PH/11 Added $max_received_linelength.
3119 PH/12 Added +ignore_defer and +include_defer to host lists.
3121 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3122 way in which PCRE > 7.0 is built.
3124 PH/14 Implemented queue_only_load_latch.
3126 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3127 MAIL command. The effect was to mangle the value on 64-bit systems.
3129 PH/16 Another patch from the Sieve maintainer.
3131 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3133 PH/18 If a system quota error occurred while trying to create the file for
3134 a maildir delivery, the message "Mailbox is full" was not appended to the
3135 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3136 only to a quota excession during the actual writing of the file.
3138 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3139 characters?) which causes problems in log lines. The DN values are now
3140 passed through string_printing() before being added to log lines.
3142 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3143 and InterBase are left for another time.)
3145 PH/21 Added message_body_newlines option.
3147 PH/22 Guard against possible overflow in moan_check_errorcopy().
3149 PH/23 POSIX allows open() to be a macro; guard against that.
3151 PH/24 If the recipient of an error message contained an @ in the local part
3152 (suitably quoted, of course), incorrect values were put in $domain and
3153 $local_part during the evaluation of errors_copy.
3159 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3160 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3163 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3164 issue a MAIL command.
3166 PH/02 In an ACL statement such as
3168 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3170 if a client was not listed at all, or was listed with a value other than
3171 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3172 the condition was not true (as it should be), so access was not denied.
3173 The bug was that the ! inversion was incorrectly passed on to the second
3174 item. This has been fixed.
3176 PH/03 Added additional dnslists conditions == and =& which are different from
3177 = and & when the dns lookup returns more than one IP address.
3179 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3180 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3182 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3183 FSYNC, which compiles an option called disable_fsync that allows for
3184 bypassing fsync(). The documentation is heavily laced with warnings.
3186 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3188 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3189 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3190 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3191 including adding "make clean"; (3) Added -fPIC when compiling the test
3192 dynamically loaded module, to get rid of a warning.
3194 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3195 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3196 = 0s. The bug is otherwise harmless.
3198 PH/07 There was a bug in the dovecot authenticator such that the value of
3199 $auth1 could be overwritten, and so not correctly preserved, after a
3200 successful authentication. This usually meant that the value preserved by
3201 the server_setid option was incorrect.
3203 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3205 PH/09 Installed PCRE release 7.0.
3207 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3208 run for batched SMTP input. It is now run at the start of every message
3209 in the batch. While fixing this I discovered that the process information
3210 (output by running exiwhat) was not always getting set for -bs and -bS
3211 input. This is fixed, and it now also says "batched" for BSMTP.
3213 PH/11 Added control=no_pipelining.
3215 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3216 patch, slightly modified), and move the expansion of helo_data till after
3217 the connection is made in the smtp transport (so it can use these
3220 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3222 PH/14 Added log_selector = +pid.
3224 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3226 PH/16 Add ${if forany and ${if forall.
3228 PH/17 Added dsn_from option to vary the From: line in DSNs.
3230 PH/18 Flush SMTP output before performing a callout, unless control =
3231 no_callout_flush is set.
3233 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3234 was true (the default) a successful delivery failed to delete the retry
3235 item, thus causing premature timeout of the address. The bug is now
3238 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3240 PH/21 Long custom messages for fakedefer and fakereject are now split up
3241 into multiline responses in the same way that messages for "deny" and
3242 other ACL rejections are.
3244 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3245 with slight modification.
3247 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3248 draft, changing the syntax and factoring some duplicate code".
3250 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3251 for deliveries of the second and subsequent messages over the same SMTP
3254 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3255 ${reduce, with only minor "tidies".
3257 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3259 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3260 expansion side effects.
3262 PH/27 When a message times out after an over-quota error from an Exim-imposed
3263 quota, the bounce message says "mailbox is full". This message was not
3264 being given when it was a system quota that was exceeded. It now should
3267 MH/03 Made $recipients available in local_scan(). local_scan() already has
3268 better access to the recipient list through recipients_list[], but
3269 $recipients can be useful in postmaster-provided expansion strings.
3271 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3272 in the case of a MAIL command with additional options following the
3273 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3274 were accidentally chopped off.
3276 PH/29 SMTP synchronization checks are implemented when a command is read -
3277 there is a check that no more input is waiting when there shouldn't be
3278 any. However, for some commands, a delay in an ACL can mean that it is
3279 some time before the response is written. In this time, more input might
3280 arrive, invalidly. So now there are extra checks after an ACL has run for
3281 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3282 pipelining has not been advertised.
3284 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3286 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3287 returned in the SMTP response when smtp_return_error_details was false.
3288 This has been fixed.
3290 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3291 instead of the C I/O that was originally supplied, because problems were
3292 reported on Solaris.
3294 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3295 Exim which did not show up earlier: it was assuming that a call to
3296 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3297 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3298 was a macro that became an assignment, so it seemed to work. This has
3299 changed to a proper function call with a void return, hence the compile
3300 error. Exim's code has been fixed.
3302 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3305 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3307 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3309 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3310 because it turns out to be too big or there is a timeout) there is no
3311 "Completed" line in the log. When some messages of this type were
3312 selected by exigrep, they were listed as "not completed". Others were
3313 picked up by some special patterns. I have improved the selection
3314 criteria to be more general.
3316 PH/38 The host_find_failed option in the manualroute router can now be set
3317 to "ignore", to completely ignore a host whose IP address cannot be
3318 found. If all hosts are ignored, the behaviour is controlled by the new
3319 host_all_ignored option.
3321 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3322 homing or because of multiple MX records with /mx) generated more than
3323 one IP address, and the following item turned out to be the local host,
3324 all the secondary addresses of the first item were incorrectly removed
3325 from the list, along with the local host and any following hosts (which
3326 is what is supposed to happen).
3328 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3329 whoever called Exim into the -H file. In the case of the daemon it was
3330 behaving confusingly. When first started, it used values for whoever
3331 started the daemon, but after a SIGHUP it used the Exim user (because it
3332 calls itself on a restart). I have changed the code so that it now always
3335 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3336 message are rejected with the same error (e.g. no authentication or bad
3337 sender address), and a DATA command is nevertheless sent (as can happen
3338 with PIPELINING or a stupid MUA), the error message that was given to the
3339 RCPT commands is included in the rejection of the DATA command. This is
3340 intended to be helpful for MUAs that show only the final error to their
3343 PH/42 Another patch from the Sieve maintainer.
3345 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3346 Eximstats - Fixed some broken HTML links and added missing column headers
3348 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3349 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3351 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3353 PH/43 Yet another patch from the Sieve maintainer.
3355 PH/44 I found a way to check for a TCP/IP connection going away before sending
3356 the response to the final '.' that terminates a message, but only in the
3357 case where the client has not sent further data following the '.'
3358 (unfortunately, this is allowed). However, in many cases there won't be
3359 any further data because there won't be any more messages to send. A call
3360 to select() can be used: if it shows that the input is "ready", there is
3361 either input waiting, or the socket has been closed. An attempt to read
3362 the next input character can distinguish the two cases. Previously, Exim
3363 would have sent an OK response which the client would never have see.
3364 This could lead to message repetition. This fix should cure that, at
3365 least in a lot of common cases.
3367 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3368 advertised in response to EHLO.
3374 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3375 fixed by 4.65/MH/01 (is this a record?) are fixed:
3377 (i) An empty string was always treated as zero by the numeric comparison
3378 operators. This behaviour has been restored.
3380 (ii) It is documented that the numeric comparison operators always treat
3381 their arguments as decimal numbers. This was broken in that numbers
3382 starting with 0 were being interpreted as octal.
3384 While fixing these problems I realized that there was another issue that
3385 hadn't been noticed. Values of message_size_limit (both the global option
3386 and the transport option) were treated as octal if they started with 0.
3387 The documentation was vague. These values are now always treated as
3388 decimal, and I will make that clear in the documentation.
3394 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3395 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3398 MH/01 Don't check that the operands of numeric comparison operators are
3399 integers when their expansion is in "skipping" mode (fixes bug
3400 introduced by 4.64-PH/07).
3402 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3403 child addresses, Exim now panics and dies. Previously, because the count
3404 is held in a short int, deliveries were likely to be lost. As such a
3405 large number of recipients for a single message is ridiculous
3406 (performance will be very, very poor), I have chosen to impose a limit
3407 rather than extend the field.
3413 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3414 leftover -K file (the existence of which was triggered by #402).
3415 While we were at it, introduced process PID as part of the -K
3416 filename. This should rule out race conditions when creating
3419 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3420 processing considerably. Previous code took too long for large mails,
3421 triggering a timeout which in turn triggers #401.
3423 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3424 in the DK code in transports.c. sendfile() is not really portable,
3425 hence the _LINUX specificness.
3427 TF/01 In the add_headers option to the mail command in an Exim filter,
3428 there was a bug that Exim would claim a syntax error in any
3429 header after the first one which had an odd number of characters
3432 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3433 callout verification, Exim cached a "reject" for the entire domain. This
3434 is correct for most verifications, but it is not correct for a recipient
3435 verification with use_sender or use_postmaster set, because in that case
3436 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3437 case of MAIL FROM:<> rejection from other early rejections (e.g.
3438 rejection of HELO). When verifying a recipient using a non-null MAIL
3439 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3440 Whatever the result of the callout, the value of the domain cache is
3441 left unchanged (for any other kind of callout, getting as far as trying
3442 RCPT means that the domain itself is ok).
3444 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3447 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3448 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3449 session, the daemon ignores ECONNECT errors and logs others; it now
3450 ignores EPIPE as well.
3452 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3453 (quoted-printable decoding).
3455 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3456 later the small subsequent patch to fix an introduced bug.
3458 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3460 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3462 PH/08 An error is now given if message_size_limit is specified negative.
3464 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3465 to be given (somewhat) arbitrary names.
3467 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3470 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3471 miscellaneous code fixes
3473 PH/10 Added the log_reject_target ACL modifier to specify where to log
3476 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3477 hostname. This is wrong, because it relates to the incoming message (and
3478 probably the interface on which it is arriving) and not to the outgoing
3479 callout (which could be using a different interface). This has been
3480 changed to use the value of the helo_data option from the smtp transport
3481 instead - this is what is used when a message is actually being sent. If
3482 there is no remote transport (possible with a router that sets up host
3483 addresses), $smtp_active_hostname is used.
3485 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3486 tweaks were necessary in order to get it to work (see also 21 below):
3487 (a) The code assumed that strncpy() returns a negative number on buffer
3488 overflow, which isn't the case. Replaced with Exim's string_format()
3490 (b) There were several signed/unsigned issues. I just did the minimum
3491 hacking in of casts. There is scope for a larger refactoring.
3492 (c) The code used strcasecmp() which is not a standard C function.
3493 Replaced with Exim's strcmpic() function.
3494 (d) The code set only $1; it now sets $auth1 as well.
3495 (e) A simple test gave the error "authentication client didn't specify
3496 service in request". It would seem that Dovecot has changed its
3497 interface. Fortunately there's a specification; I followed it and
3498 changed what the client sends and it appears to be working now.
3500 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3503 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3504 address A is aliased to B and C, where B exists and C does not. Without
3505 -v the output is "A verified" because verification stops after a
3506 successful redirection if more than one address is generated. However,
3507 with -v the child addresses are also verified. Exim was outputting "A
3508 failed to verify" and then showing the successful verification for C,
3509 with its parentage. It now outputs "B failed to verify", showing B's
3510 parentage before showing the successful verification of C.
3512 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3513 look up a TXT record in a specific list after matching in a combined
3516 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3517 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3518 they consult the DNS. I had assumed they would set it the way they
3519 wanted; and indeed my experiments on Linux seem to show that in some
3520 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3521 To be on the safe side, however, I have now made the interface to
3522 host_find_byname() similar to host_find_bydns(), with an argument
3523 containing the DNS resolver options. The host_find_byname() function now
3524 sets these options at its start, just as host_find_bydns() does. The smtp
3525 transport options dns_qualify_single and dns_search_parents are passed to
3526 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3527 of host_find_byname() use the default settings of RES_DEFNAMES
3528 (qualify_single) but not RES_DNSRCH (search_parents).
3530 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3531 spool_read_header() do less string testing, by means of a preliminary
3532 switch on the second character of optional "-foo" lines. (This is
3533 overdue, caused by the large number of possibilities that now exist.
3534 Originally there were few.) While I was there, I also converted the
3535 str(n)cmp tests so they don't re-test the leading "-" and the first
3536 character, in the hope this might squeeze out yet more improvement.
3538 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3539 flag allowing group syntax was set by the header_syntax check but not
3540 turned off, possible causing trouble later; (2) The flag was not being
3541 set at all for the header_verify test, causing "group"-style headers to
3542 be rejected. I have now set it in this case, and also caused header_
3543 verify to ignore an empty address taken from a group. While doing this, I
3544 came across some other cases where the code for allowing group syntax
3545 while scanning a header line wasn't quite right (mostly, not resetting
3546 the flag correctly in the right place). These bugs could have caused
3547 trouble for malformed header lines. I hope it is now all correct.
3549 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3550 with the "reply" argument non-NULL. The code, however (which originally
3551 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3552 but it didn't always do it. This confused somebody who was copying the
3553 code for some other use. I have removed all the tests.
3555 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3556 feature that was used to support insecure browsers during the U.S. crypto
3557 embargo. It requires special client support, and Exim is probably the
3558 only MTA that supported it -- and would never use it because real RSA is
3559 always available. This code has been removed, because it had the bad
3560 effect of slowing Exim down by computing (never used) parameters for the
3561 RSA_EXPORT functionality.
3563 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3564 authenticator to fail if there's a tab character in the incoming data
3565 (there should never be unless someone is messing about, as it's supposed
3566 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3567 if the connection is using TLS or if the remote IP is the same as the
3568 local IP, and the "valid-client-cert option" if a client certificate has
3571 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3572 authenticators. This can be used for authorization after authentication
3573 succeeds. (In the case of plaintext, it servers for both authentication
3576 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3577 if any retry times were supplied.
3579 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3580 connection, where there is no client IP address to check. In this
3581 situation, the verify now always succeeds.
3583 PH/25 Applied John Jetmore's -Mset patch.
3585 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3587 PH/27 In a string expansion for a processed (not raw) header when multiple
3588 headers of the same name were present, leading whitespace was being
3589 removed from all of them, but trailing whitespace was being removed only
3590 from the last one. Now trailing whitespace is removed from each header
3591 before concatenation. Completely empty headers in a concatenation (as
3592 before) are ignored.
3594 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3595 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3597 PH/29 [Removed. This was a change that I later backed out, and forgot to
3598 correct the ChangeLog entry (that I had efficiently created) before
3599 committing the later change.]
3601 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3602 address errors (4xx response to RCPT) over the same connection as other
3603 messages routed to the same hosts. Such deliveries are always "forced",
3604 so retry times are not inspected. This resulted in far too many retries
3605 for the affected addresses. The effect occurred only when there were more
3606 hosts than the hosts_max_try setting in the smtp transport when it had
3607 the 4xx errors. Those hosts that it had tried were not added to the list
3608 of hosts for which the message was waiting, so if all were tried, there
3609 was no problem. Two fixes have been applied:
3611 (i) If there are any address or message errors in an SMTP delivery, none
3612 of the hosts (tried or untried) are now added to the list of hosts
3613 for which the message is waiting, so the message should not be a
3614 candidate for sending over the same connection that was used for a
3615 successful delivery of some other message. This seems entirely
3616 reasonable: after all the message is NOT "waiting for some host".
3617 This is so "obvious" that I'm not sure why it wasn't done
3618 previously. Hope I haven't missed anything, but it can't do any
3619 harm, as the worst effect is to miss an optimization.
3621 (ii) If, despite (i), such a delivery is accidentally attempted, the
3622 routing retry time is respected, so at least it doesn't keep
3623 hammering the server.
3625 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3626 in ${readsocket because some servers need this prod.
3628 PH/32 Added some extra debug output when updating a wait-xxx database.
3630 PH/33 The hint "could be header name not terminated by colon", which has been
3631 given for certain expansion errors for a long time, was not being given
3632 for the ${if def:h_colon_omitted{... case.
3634 PH/34 The spec says: "With one important exception, whenever a domain list is
3635 being scanned, $domain contains the subject domain." There was at least
3636 one case where this was not true.
3638 PH/35 The error "getsockname() failed: connection reset by peer" was being
3639 written to the panic log as well as the main log, but it isn't really
3640 panic-worthy as it just means the connection died rather early on. I have
3641 removed the panic log writing for the ECONNRESET error when getsockname()
3644 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3645 runs only) independently of the message's sender address. This meant
3646 that, if the 4xx error was in fact related to the sender, a different
3647 message to the same recipient with a different sender could confuse
3648 things. In particular, this can happen when sending to a greylisting
3649 server, but other circumstances could also provoke similar problems.
3650 I have changed the default so that the retry time for these errors is now
3651 based a combination of the sender and recipient addresses. This change
3652 can be overridden by setting address_retry_include_sender=false in the
3655 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3656 remote server are returned as part of bounce messages. This was not
3657 happening for LMTP over a pipe (the lmtp transport), but now it is the
3658 same for both kinds of LMTP.
3660 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3661 in header lines that were in fact CNAMEs. This is no longer the case.
3663 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3664 and queue runs started by the daemon processed all messages. This has
3665 been fixed so that -R and -S can now usefully be given with -q<time>.
3667 PH/40 Import PCRE release 6.7 (fixes some bugs).
3669 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3671 PH/42 Give an error if -q is specified more than once.
3673 PH/43 Renamed the variables $interface_address and $interface_port as
3674 $received_ip_address and $received_port, to make it clear that these
3675 values apply to message reception, and not to the outgoing interface when
3676 a message is delivered. (The old names remain recognized, of course.)
3678 PH/44 There was no timeout on the connect() call when using a Unix domain
3679 socket in the ${readsocket expansion. There now is.
3681 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3682 be meaningful with "accept".
3684 SC/01 Eximstats V1.43
3685 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3687 SC/02 Eximstats V1.44
3688 Use a glob alias rather than an array ref in the generated
3689 parser. This improves both readability and performance.
3691 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3692 Collect SpamAssassin and rejection statistics.
3693 Don't display local sender or destination tables unless
3694 there is data to show.
3695 Added average volumes into the top table text output.
3697 SC/04 Eximstats V1.46
3698 Collect data on the number of addresses (recipients)
3699 as well as the number of messages.
3701 SC/05 Eximstats V1.47
3702 Added 'Message too big' to the list of mail rejection
3703 reasons (thanks to Marco Gaiarin).
3705 SC/06 Eximstats V1.48
3706 Mainlog lines which have GMT offsets and are too short to
3707 have a flag are now skipped.
3709 SC/07 Eximstats V1.49 (Alain Williams)
3710 Added the -emptyok flag.
3712 SC/08 Eximstats V1.50
3713 Fixes for obtaining the IP address from reject messages.
3715 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3716 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3717 whitespace changes from 4.64-PH/27
3719 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3722 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3723 are found, allow negative numbers in numeric criteria)
3725 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3727 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3728 to match changes made in 4.64-PH/43
3730 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3732 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3733 "domain" is actually an IP address, and force a failure. This locks out
3734 those revolvers/nameservers that support "A-for-A" lookups, in
3735 contravention of the specifications.
3737 PH/48 When a host name was looked up from an IP address, and the subsequent
3738 forward lookup of the name timed out, the host name was left in
3739 $sender_host_name, contrary to the specification.
3741 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3742 restricted to single-key lookups, Exim was not diagnosing an error if
3743 * or *@ was used with a query-style lookup.
3745 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3747 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3748 long ago, but noone interested enough thought of it. Let's just say that
3749 the "1.1" means that there are some new functions that weren't there at
3750 some point in the past.
3752 PH/51 Error processing for expansion failure of helo_data from an smtp
3753 transport during callout processing was broken.
3755 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3756 tested/used via the -bh/-bhc/-bs options.
3758 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3759 bug, fixed in subsequent PCRE releases).
3761 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3762 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3764 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3770 SC/01 Use a glob alias rather than an array ref in eximstats generated
3771 parser. This improves both readability and performance.
3773 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3774 Don't display local sender or destination tables in eximstats unless
3775 there is data to show.
3776 Added average volumes into the eximstats top table text output.
3778 SC/03 Collect data on the number of addresses (recipients) as well
3779 as the number of messages in eximstats.
3781 TF/01 Correct an error in the documentation for the redirect router. Exim
3782 does (usually) call initgroups() when daemonizing.
3784 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3785 with consistent privilege compared to when running as a daemon.
3787 TF/03 Note in the spec that $authenticated_id is not set for local
3788 submissions from trusted users.
3790 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3791 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3793 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3794 by adding some example configuration directives to the default
3795 configuration file. A little bit of work is required to uncomment the
3796 directives and define how usernames and passwords are checked, but
3797 there is now a framework to start from.
3799 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3800 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3801 without this. I don't know how relevant this is to other LDAP libraries.
3803 PH/02 Add the verb name to the "unknown ACL verb" error.
3805 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3807 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3809 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3810 directory not expanded when it should be if an expanded home directory
3811 was set for the address (which is overridden by the transport).
3813 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3816 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3817 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3818 because it is too late at that time, and has no effect.
3820 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3821 security issue with \' (bugzilla #107). I could not use the
3822 PQescapeStringConn() function, because it needs a PGconn value as one of
3825 PH/08 When testing addresses using -bt, indicate those final addresses that
3826 are duplicates that would not cause an additional delivery. At least one
3827 person was confused, thinking that -bt output corresponded to deliveries.
3828 (Suppressing duplicates isn't a good idea as you lose the information
3829 about possibly different redirections that led to the duplicates.)
3831 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3832 systems where poll() doesn't work, in particular OS X.
3834 PH/10 Added more information to debugging output for retry time not reached.
3836 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3837 operations in malware.c.
3839 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3842 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3843 syslog_duplication set false, log lines that would normally be written
3844 both the the main log and to the reject log were not written to syslog at
3847 PH/14 In the default configuration, change the use of "message" in ACL warn
3848 statements to "add_header".
3850 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3851 not followed by a command (e.g. "seen endif").
3853 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3854 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3857 PH/17 Added extra conditions to the default value of delay_warning_condition
3861 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3862 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3863 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3866 The Auto-Submitted: and various List- headers are standardised, whereas I
3867 don't think Precedence: ever was.
3869 PH/18 Refactored debugging code in route_finduser() to show more information,
3870 in particular, the error code if getpwnam() issues one.
3872 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3873 This is apparently needed in addition to the PH/07 change above to avoid
3874 any possible encoding problems.
3876 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3877 but not after initializing Perl.
3879 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3880 output them only if debugging. By default they are written stderr,
3881 apparently, which is not desirable.
3883 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3886 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3889 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3891 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3892 authenticated or an ident call has been made. Suppress the default
3893 values for $authenticated_id and $authenticated_sender (but permit -oMai
3894 and -oMas) when testing with -bh.
3896 PH/24 Re-jigged the order of the tests in the default configuration so that the
3897 tests for valid domains and recipients precede the DNS black list and CSA
3898 tests, on the grounds that those ones are more expensive.
3900 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3901 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3902 This bug exists in every version of Exim that I still have, right back to
3905 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3906 However, an attempt to turn on case-sensitivity in a regex key by
3907 including (?-i) didn't work because the subject string was already
3908 lowercased, and the effects were non-intuitive. It turns out that a
3909 one-line patch can be used to allow (?-i) to work as expected.
3915 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3916 other effects) broke the use of negated acl sub-conditions.
3918 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3921 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3922 "Deny" causes Exim to reject the incoming connection with a 554 error.
3923 Unfortunately, if there is a major crisis, such as a disk failure,
3924 tcp-wrappers gives "deny", whereas what one would like would be some
3925 kind of temporary error. A kludge has been added to help with this.
3926 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3927 554 error is used if errno is still zero or contains ENOENT (which occurs
3928 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3931 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3933 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3934 errors. Otherwise a message that provokes a temporary error (when other
3935 messages do not) can cause a whole host to time out.
3937 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3938 addresses were routed directly to files or pipes from a redirect router.
3939 File deliveries just didn't batch; pipe deliveries might have suffered
3942 PH/06 A failure to get a lock for a hints database would erroneously always say
3943 "Failed to get write lock", even when it was really a read lock.
3945 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3946 of 0600. This has been changed to use the value of the lockfile_mode
3947 option (which defaults to 0600).
3949 PH/08 Applied small patch from the Sieve maintainer.
3951 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3952 folder from quota calculations, a direct delivery into this folder messed
3953 up the contents of the maildirsize file. This was because the regex was
3954 used only to exclude .Trash (or whatever) when the size of the mailbox
3955 was calculated. There was no check that a delivery was happening into an
3956 excluded directory. This bug has been fixed by ignoring all quota
3957 processing for deliveries into excluded directories.
3959 PH/10 Added the maildirfolder_create_regex option to appendfile.
3965 PH/01 The code for finding all the local interface addresses on a FreeBSD
3966 system running IPv6 was broken. This may well have applied to all BSD
3967 systems, as well as to others that have similar system calls. The broken
3968 code found IPv4 interfaces correctly, but gave incorrect values for the
3969 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3970 that it would not match correctly against @[] and not recognize the IPv6
3973 PH/02 The ipliteral router was not recognizing addresses of the form user@
3974 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3976 PH/03 Added disable_ipv6.
3978 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3979 decoded form, because it is most often used to construct To: headers
3980 lines in autoreplies, and the decoded form may well be syntactically
3981 invalid. However, $reply_address has leading white space removed, and all
3982 newlines turned into spaces so that the autoreply transport does not
3985 PH/05 If group was specified without a user on a router, and no group or user
3986 was specified on a transport, the group from the router was ignored.
3988 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3989 for visible compile-time settings that can be used to change these
3990 numbers, for those that want even more. Backwards compatibility with old
3991 spool files has been maintained. However, going back to a previous Exim
3992 release will lost any variables that are in spool files.
3994 PH/07 Two small changes when running in the test harness: increase delay when
3995 passing a TCP/IP connection to a new process, in case the original
3996 process has to generate a bounce, and remove special handling of
3997 127.0.0.2 (sic), which is no longer necessary.
3999 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4000 be the same on different OS.
4002 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4005 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4006 whether --show-vars was specified or not
4008 JJ/02 exipick: Added support for new ACL variable spool format introduced
4011 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4012 syntactically invalid From: or Reply-to: line, and a filter used this to
4013 generate an autoreply, and therefore failed to obtain an address for the
4014 autoreply, Exim could try to deliver to a non-existent relative file
4015 name, causing unrelated and misleading errors. What now happens is that
4016 it logs this as a hard delivery error, but does not attempt to create a
4019 PH/11 The exinext utility has a -C option for testing purposes, but although
4020 the given file was scanned by exinext itself; it wasn't being passed on
4021 when Exim was called.
4023 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4024 an end-of-file indication when reading a command response.
4026 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4027 compiled. In many other places in Exim, IPv6 addresses are always
4028 recognized, so I have changed this. It also means that IPv4 domain
4029 literals of the form [IPV4:n.n.n.n] are now always recognized.
4031 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4032 used if the router is not running as root, for example, when verifying at
4033 ACL time, or when using -bh. The debugging output from this situation was
4034 non-existent - all you got was a failure to exec. I have made two
4037 (a) Failures to set uid/gid, the current directory, or a process leader
4038 in a subprocess such as that created by queryprogram now generate
4039 suitable debugging output when -d is set.
4041 (b) The queryprogram router detects when it is not running as root,
4042 outputs suitable debugging information if -d is set, and then runs
4043 the subprocess without attempting to change uid/gid.
4045 PH/15 Minor change to Makefile for building test_host (undocumented testing
4048 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4049 additional section of a DNS packet that returns MX or SRV records.
4050 Instead, it always explicitly searches for A/AAAA records. This avoids
4051 major problems that occur when a DNS server includes only records of one
4052 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4053 fixed another bug: if SRV records were looked up and the corresponding
4054 address records were *not* found in the additional section, the port
4055 values from the SRV records were lost.
4057 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4058 using the correct key (the original address) when searching the retry
4059 rules in order to find which one to use for generating the retry hint.
4061 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4062 adding the default one. Similarly, if it contains a Reply-To: header, the
4063 errors_reply_to option, if set, is not used.
4065 PH/19 When calculating a retry time, Exim used to measure the "time since
4066 failure" by looking at the "first failed" field in the retry record. Now
4067 it does not use this if it is later than than the arrival time of the
4068 message. Instead it uses the arrival time. This makes for better
4069 behaviour in cases where some deliveries succeed, thus re-setting the
4070 "first failed" field. An example is a quota failure for a huge message
4071 when small messages continue to be delivered. Without this change, the
4072 "time since failure" will always be short, possible causing more frequent
4073 delivery attempts for the huge message than are intended.
4074 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4076 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4077 $1, $2, $3) because the numerical variables can be reset during some
4078 expansion items (e.g. "match"), thereby losing the authentication data.
4080 PH/21 Make -bV show the size of off_t variables so that the test suite can
4081 decide whether to run tests for quotas > 2G.
4083 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4084 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4085 filecount value is greater than 2G or if a quota value is greater than 2G
4086 on a system where the size of off_t is not greater than 4, a panic error
4089 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4090 never match. The debug and -bh output now contains an explicit error
4091 message indicating a malformed IPv4 address or mask.
4093 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4094 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4095 PH/23 above applies.
4097 PH/25 Do not write to syslog when running in the test harness. The only
4098 occasion when this arises is a failure to open the main or panic logs
4099 (for which there is an explicit test).
4101 PH/26 Added the /no_tell option to "control=freeze".
4103 PH/27 If a host name lookup failed very early in a connection, for example, if
4104 the IP address matched host_lookup and the reverse lookup yielded a name
4105 that did not have a forward lookup, an error message of the form "no IP
4106 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4107 could be logged. Now it outputs the IP address instead of "NULL".
4109 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4110 allows the sender and the authenticated sender to be set when
4111 submitting a message from within Exim. Since child_open_exim() is
4112 documented for local_scan(), the new function should be too.
4114 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4115 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4116 results in an empty string is now treated as unset.
4118 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4120 PH/31 Added sender_verify_fail logging option.
4122 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4123 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4124 tidied the source and removed it altogether.
4126 PH/33 When a queue run was abandoned because the load average was too high, a
4127 log line was always written; now it is written only if the queue_run log
4128 selector is set. In addition, the log line for abandonment now contains
4129 information about the queue run such as the pid. This is always present
4130 in "start" and "stop" lines but was omitted from the "abandon" line.
4132 PH/34 Omit spaces between a header name and the colon in the error message that
4133 is given when verify = headers_syntax fails (if there are lots of them,
4134 the message gets confusing).
4136 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4137 names, as there are now some PTR records that contain slashes. This check
4138 is only to protect against broken name servers that fall over on strange
4139 characters, so the fact that it applies to all lookups doesn't matter.
4141 PH/36 Now that the new test suite is complete, we can remove some of the
4142 special code in Exim that was needed for the old test suite. For example,
4143 sorting DNS records because real resolvers return them in an arbitrary
4144 order. The new test suite's fake resolver always returns records in the
4147 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4148 bounces) except when queue_only is set, to avoid logging races between
4149 the different processes.
4151 PH/38 Panic-die if .include specifies a non-absolute path.
4153 PH/39 A tweak to the "H" retry rule from its user.
4155 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4156 a label. They prevented compilation on older perls.
4158 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4159 a warning to be raised on newish perls.
4161 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4162 on queue. Changes to match documented behaviour of showing count of
4163 messages matching specified criteria.
4165 PH/40 Changed the default ident timeout from 30s to 5s.
4167 PH/41 Added support for the use of login_cap features, on those BSD systems
4168 that have them, for controlling the resources used by pipe deliveries.
4170 PH/42 The content-scanning code uses fopen() to create files in which to put
4171 message data. Previously it was not paying any attention to the mode of
4172 the files. Exim runs with umask(0) because the rest of the code creates
4173 files with open(), and sets the required mode explicitly. Thus, these
4174 files were ending up world-writeable. This was not a big issue, because,
4175 being within the spool directory, they were not world-accessible. I have
4176 created a function called modefopen, which takes an additional mode
4177 argument. It sets umask(777), creates the file, chmods it to the required
4178 mode, then resets the umask. All the relevant calls to fopen() in the
4179 content scanning code have been changed to use this function.
4181 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4182 to 24 hours. This avoids potential overflow problems when processing G
4183 and H retry rules. I suspect nobody ever tinkers with this value.
4185 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4187 PH/45 When the plaintext authenticator is running as a client, the server's
4188 challenges are checked to ensure they are valid base64 strings. By
4189 default, the authentication attempt is cancelled if an invalid string is
4190 received. Setting client_ignore_invalid_base64 true ignores these errors.
4191 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4192 they are received. Thus, the responses can be made to depend on the
4193 challenges. If an invalid string is ignored, an empty string is placed in
4196 PH/46 Messages that are created by the autoreply transport now contains a
4197 References: header, in accordance with RFCs 2822 and 3834.
4199 PH/47 Added authenticated_sender_force to the smtp transport.
4201 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4203 PH/49 Installed latest patch from the Sieve maintainer.
4205 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4206 was also set, the appendfile transport was unnecessarily scanning a
4207 directory of message files (e.g. for maildir delivery) to find the count
4208 of files (along with the size), even though it did not need this
4209 information. It now does the scan only if it needs to find either the
4210 size of the count of files.
4212 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4214 PH/52 Two bugs concerned with error handling when the smtp transport is
4217 (i) Exim was not creating retry information for temporary errors given
4218 for individual recipients after the DATA command when the smtp transport
4219 was used in LMTP mode. This meant that they could be retried too
4220 frequently, and not timed out correctly.
4222 (ii) Exim was setting the flag that allows error details to be returned
4223 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4224 recipients that were returned after the DATA command.
4226 PH/53 This is related to PH/52, but is more general: for any failing address,
4227 when detailed error information was permitted to be returned to the
4228 sender, but the error was temporary, then after the final timeout, only
4229 "retry timeout exceeded" was returned. Now it returns the full error as
4230 well as "retry timeout exceeded".
4232 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4233 do this, and (what is worse) MTAs that accept it.
4235 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4236 will now be deprecated.
4238 PH/56 New os.c-cygwin from the Cygwin maintainer.
4240 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4241 formats (previously only available in exim formats via -bpr, -bpru,
4242 and -bpra. Now also available in native and exiqgrep formats)
4244 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4245 with very large, slow to parse queues
4247 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4249 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4251 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4252 responses to authentication challenges, though it was showing the
4253 challenges; (ii) I've removed the CR characters from the debug output for
4256 PH/58 Allow for the insertion of a newline as well as a space when a string
4257 is turned into more than one encoded-word during RFC 2047 encoding. The
4258 Sieve code now uses this.
4260 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4261 data_4xx, lost_connection, tls_required.
4263 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4264 message was being sent as an SMTP response.
4266 PH/61 Add -l and -k options to exicyclog.
4268 PH/62 When verifying, if an address was redirected to one new address, so that
4269 verification continued, and the new address failed or deferred after
4270 having set something in $address_data, the value of $address_data was not
4271 passed back to the ACL. This was different to the case when no
4272 redirection occurred. The value is now passed back in both cases.
4274 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4275 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4276 use login_cap.h, so on its own it isn't the distinguishing feature. The
4277 new name refers directly to the setclassresources() function.
4279 PH/65 Added configuration files for NetBSD3.
4281 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4283 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4284 is preferred over IPv4.
4286 PH/68 The bounce_return_message and bounce_return_body options were not being
4287 honoured for bounces generated during the reception of non-SMTP messages.
4288 In particular, this applied to messages rejected by the ACL. This bug has
4289 been fixed. However, if bounce_return_message is true and bounce_return_
4290 body is false, the headers that are returned for a non-SMTP message
4291 include only those that have been read before the error was detected.
4292 (In the case of an ACL rejection, they have all been read.)
4294 PH/69 The HTML version of the specification is now built in a directory called
4295 spec_html instead of spec.html, because the latter looks like a path with
4296 a MIME-type, and this confuses some software.
4298 PH/70 Catch two compiler warnings in sieve.c.
4300 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4301 function verify_get_ident() calls ip_connect() to connect a socket, but
4302 if the "connect()" function timed out, ip_connect() used to close the
4303 socket. However, verify_get_ident() also closes the socket later, and in
4304 between Exim writes to the log, which may get opened at this point. When
4305 the socket was closed in ip_connect(), the log could get the same file
4306 descriptor number as the socket. This naturally causes chaos. The fix is
4307 not to close the socket in ip_connect(); the socket should be closed by
4308 the function that creates it. There was only one place in the code where
4309 this was missing, in the iplookup router, which I don't think anybody now
4310 uses, but I've fixed it anyway.
4312 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4313 well as to direct DNS lookups. Otherwise the handling of names in host
4314 lists is inconsistent and therefore confusing.
4320 PH/01 Two changes to the default runtime configuration:
4322 (1) Move the checks for relay_from_hosts and authenticated clients from
4323 after to before the (commented out) DNS black list checks.
4325 (2) Add control=submission to the relay_from_hosts and authenticated
4326 clients checks, on the grounds that messages accepted by these
4327 statements are most likely to be submissions.
4329 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4331 (1) Generate an error if the third argument for the ${prvs expansion is
4334 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4337 (3) Reset the variables that are obtained from the first argument of
4338 ${prvscheck and used in the second argument before leaving the code,
4339 because their memory is reclaimed, so using them afterwards may do
4342 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4343 one (it's much easier than Tom thought :-).
4345 (5) Because of (4), we can now allow for the use of $prvscheck_result
4346 inside the third argument.
4348 PH/03 For some reason, the default setting of PATH when running a command from
4349 a pipe transport was just "/usr/bin". I have changed it to
4352 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4353 anything to be listed in the output from -bV.
4355 PH/05 When a filter generated an autoreply, the entire To: header line was
4356 quoted in the delivery log line, like this:
4358 => >A.N.Other <ano@some.domain> <original@ddress> ...
4360 This has been changed so that it extracts the operative address. There
4361 may be more than one such address. If so, they are comma-separated, like
4364 => >ano@some.domain,ona@other.domain <original@ddress> ...
4366 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4367 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4368 not being looked up in the rDNS to get a host name, Exim was showing the
4369 IP address twice in Received: lines, even though the IP addresses were
4370 identical. For example:
4372 Received: from [1.2.3.4] (helo=[1.2.3.4])
4374 However, if the real host name was known, it was omitting the HELO data
4375 if it matched the actual IP address. This has been tidied up so that it
4376 doesn't show the same IP address twice.
4378 PH/07 When both +timestamp and +memory debugging was on, the value given by
4379 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4380 called by the debug printing, thereby overwriting the timestamp buffer.
4381 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4383 PH/08 When the original message was included in an autoreply transport, it
4384 always said "this is a copy of the message, including all the headers",
4385 even if body_only or headers_only was set. It now gives an appropriate
4388 PH/09 Applied a patch from the Sieve maintainer which:
4390 o fixes some comments
4391 o adds the (disabled) notify extension core
4392 o adds some debug output for the result of if/elsif tests
4393 o points to the current vacation draft in the documentation
4394 and documents the missing references header update
4398 o fixes a bug in processing the envelope test (when testing
4399 multiple envelope elements, the last element determined the
4402 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4403 Electronic Mail") by including:
4405 Auto-submitted: auto-generated
4407 in the messages that it generates (bounce messages and others, such as
4408 warnings). In the case of bounce messages for non-SMTP messages, there was
4409 also a typo: it was using "Auto_submitted" (underscore instead of
4410 hyphen). Since every message generated by Exim is necessarily in response
4411 to another message, thes have all been changed to:
4413 Auto-Submitted: auto-replied
4415 in accordance with these statements in the RFC:
4417 The auto-replied keyword:
4419 - SHOULD be used on messages sent in direct response to another
4420 message by an automatic process,
4422 - MUST NOT be used on manually-generated messages,
4424 - MAY be used on Delivery Status Notifications (DSNs) and Message
4425 Disposition Notifications (MDNs),
4427 - MUST NOT be used on messages generated by automatic or periodic
4428 processes, except for messages which are automatic responses to
4431 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4432 to the default Received: header definition.
4434 PH/12 Added log selector acl_warn_skipped (default on).
4436 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4437 variables because (a) they are in the wrong storage pool and (b) even if
4438 they were copied, it wouldn't work properly because of the caching.
4440 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4441 checking when decoding. Apparently there are clients that generate
4442 overlong encoded strings. Why am I not surprised?
4444 PH/15 If the first argument of "${if match_address" was not empty, but did not
4445 contain an "@" character, Exim crashed. Now it writes a panic log message
4446 and treats the condition as false.
4448 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4450 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4451 extension "envelope-auth". The code is finished and in agreement with
4452 other implementations, but there is no documentation so far and in fact,
4453 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4454 not changing the active code.
4456 Print executed "if" and "elsif" statements when debugging is used. This
4457 helps a great deal to understand what a filter does.
4459 Document more things not specified clearly in RFC3028. I had all this
4460 sorted out, when out of a sudden new issues came to my mind. Oops."
4462 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4465 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4466 canonical form (as documented). However, after a host name lookup from
4467 the IP address, check_host() was doing a simple string comparison with
4468 addresses acquired from the DNS when checking that the found name did
4469 have the original IP as one of its addresses. Since any found IPv6
4470 addresses are likely to be in abbreviated form, the comparison could
4471 fail. Luckily, there already exists a function for doing the comparison
4472 by converting both addresses to binary, so now that is used instead of
4473 the text comparison.
4475 PH/20 There was another similar case to PH/19, when a complete host name was
4476 given in a host list; looking up its IP address could give an abbreviated
4477 form, whereas the current host's name might or might not be abbreviated.
4478 The same fix has been applied.
4484 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4485 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4488 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4489 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4491 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4493 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4494 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4495 Also, for "Auto-Submitted", treat anything other than "no" as
4496 non-personal, in accordance with RFC 3834. (Previously it treated
4497 anything starting "auto-" as non-personal.)
4499 TF/01 The control=submission/name=... option had a problem with syntax
4500 errors if the name included a slash character. The /name= option
4501 now slurps the rest of the string, so it can include any characters
4502 but it must come last in the list of options (after /sender_retain
4505 PH/05 Some modifications to the interface to the fake nameserver for the new
4513 TK/01 Added the "success_on_redirect" address verification option. See
4514 NewStuff for rationale and an example.
4516 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4518 PH/02 Patch to exigrep to allow it to work on syslog lines.
4520 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4521 fread() to skip over the body file's header line, because in Cygwin the
4522 header line is locked and is inaccessible.
4524 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4525 co-exist for some time) to make it clear that it is the Exim ID that is
4526 referenced, not the Message-ID: header line.
4528 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4529 string_format() function, because snprintf() does not exist on all
4532 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4533 ${stat: expansion item.
4535 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4536 protocol synchronization error", to keep the pedants happy.
4538 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4539 well as for IRIX systems, when gcc is being used. See the host.c source
4542 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4544 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4547 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4548 transport and to the smtp transport in LMTP mode.
4550 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4552 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4553 run a filter in a subprocess. This could lead to confusion in subsequent
4554 lookups in the parent process. There should also be a search_tidyup() at
4555 the end of the subprocess.
4557 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4558 only if the host matched helo_try_verify_hosts, which caused the
4559 verification to occur when the EHLO/HELO command was issued. The ACL just
4560 tested the remembered result. Now, if a previous verification attempt has
4561 not happened, "verify = helo" does it there and then.
4563 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4565 TK/03 Fix log output including CR from clamd.
4567 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4568 exist provoked a memory error which could cause a segfault.
4570 PH/15 Installed PCRE 6.2
4572 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4574 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4575 of the problem. Specifically, suggested +O2 rather than +O1 for the
4578 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4580 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4581 record, the port was not correctly logged when the outgoing_port log
4582 selector was set (it logged the transort's default port).
4584 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4585 fallback_hosts, and "hosts" in the smtp transport.
4587 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4588 host errors such as "Connection refused".
4590 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4591 authentication with radiusclient 0.4.9:
4593 - Error returned from rc_read_config was caught wrongly
4594 - Username/password not passed on to radius server due to wrong length.
4596 The presumption is that some radiusclient API changes for 4.51/PH/17
4597 were not taken care of correctly. The code is still untested by me (my
4598 Linux distribution still has 0.3.2 of radiusclient), but it was
4599 contributed by a Radius user.
4601 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4602 expanding the "port" option of the smtp transport.
4604 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4605 while reading a MIME header. Thanks to Tom Hughes for a patch.
4607 PH/24 Include config.h inside local_scan.h so that configuration settings are
4610 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4611 that in an ACL for RCPT (for example), you can examine exactly what was
4614 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4615 commands, but it was not correctly comparing the address with the actual
4616 client host address. Thus, it would show the EHLO address in Received:
4617 header lines when this was not necessary.
4619 PH/27 Added the % operator to ${eval:}.
4621 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4622 it should be ignoring failures (because with -C, for example, it has lost
4623 privilege). It wasn't ignoring creation failures other than "already
4626 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4629 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4630 because an input line was too long, either on its own, or by virtue of
4631 too many continuations, the temporary file was not being removed, and the
4632 return code was incorrect.
4634 PH/31 Missing "BOOL" in function definition in filtertest.c.
4636 PH/32 Applied Sieve patches from the maintainer.
4638 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4640 PH/33 Added "verify = not_blind".
4642 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4643 Local/Makefile (with some defaults set). These are used in built scripts
4644 such as exicyclog, but they have never been used in the exim_install
4645 script (though there are many overriding facilities there). I have
4646 arranged that the exim_install script now takes note of these two
4649 PH/35 Installed configuration files for Dragonfly.
4651 PH/36 When a locally submitted message by a trusted user did not contain a
4652 From: header, and the sender address was obtained from -f or from an SMTP
4653 MAIL command, and the trusted user did not use -F to supply a sender
4654 name, $originator_name was incorrectly used when constructing a From:
4655 header. Furthermore, $originator_name was used for submission mode
4656 messages from external hosts without From: headers in a similar way,
4657 which is clearly wrong.
4659 PH/37 Added control=suppress_local_fixups.
4661 PH/38 When log_selector = +received_sender was set, and the addition of the
4662 sender made the log line's construction buffer exactly full, or one byte
4663 less than full, an overflow happened when the terminating "\n" was
4666 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4667 when the result of a list match is failure because a DNS lookup failed.
4669 PH/40 RM_COMMAND is now used in the building process.
4671 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4672 the "build-* directories that it finds.
4674 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4675 address in a domain literal was a prefix of an interface address.
4677 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4678 when verifying a sender address, unless rewrite_headers is false.
4680 PH/44 Wrote a long comment about why errors_to addresses are verified as
4681 recipients, not senders.
4683 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4684 the ratelimit ACL was added.
4686 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4688 PH/46 Added extra information about PostgreSQL errors to the error string.
4690 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4691 avoiding the need to install special zones in a real server. This is
4692 backwards compatible; if it can't find the fake resolver, it drops back.
4693 Thus, both old and new test suites can be run.
4695 TF/02 Added util/ratelimit.pl
4697 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4698 clock is set back in time.
4700 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4701 Candler <B.Candler@pobox.com>.
4703 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4704 OK for addresses that are the result of redirections.
4706 PH/48 A number of further additions for the benefit of the new test suite,
4707 including a fake gethostbyname() that interfaces to the fake DNS resolver
4710 TF/06 The fix for widen_domains has also been applied to qualify_single and
4711 search_parents which are the other dnslookup options that can cause
4714 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4717 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4719 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4720 service" if it encountered an MX record with an empty target hostname.
4721 The message is now "an MX or SRV record indicated no SMTP service".
4723 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4724 if the DNS of the sending site is misconfigured. This is quite a
4725 common situation. This change restores the behaviour of treating a
4726 helo verification defer as a failure.
4728 PH/51 If self=fail was set on a router, the bounce message did not include the
4729 actual error message.
4735 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4737 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4738 command itself ended in error, the underlying message about the transport
4739 filter timeout was being overwritten with the pipe command error. Now the
4740 underlying error message should be appended to the second error message.
4742 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4744 PH/02 Reduce the amount of output that "make" produces by default. Full output
4745 can still be requested.
4747 PH/03 The warning log line about a condition test deferring for a "warn" verb
4748 was being output only once per connection, rather than after each
4749 occurrence (because it was using the same function as for successful
4750 "warn" verbs). This seems wrong, so I have changed it.
4752 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4753 it should not have, which might have caused a crash in the right
4754 circumstances, but probably never did.
4756 PH/04 Installed a modified version of Tony Finch's patch to make submission
4757 mode fix the return path as well as the Sender: header line, and to
4758 add a /name= option so that you can make the user's friendly name appear
4761 TF/03 Added the control = fakedefer ACL modifier.
4763 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4764 Mark Lowes for thorough testing.
4766 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4768 TK/03 Merged latest SRS patch from Miles Wilton.
4770 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4771 with the definition in sysexits.h (which is #included earlier).
4772 Fortunately, Exim does not actually use EX_OK. The code used to try to
4773 preserve the sysexits.h value, by assuming that macro definitions were
4774 scanned for macro replacements. I have been disabused of this notion,
4775 so now the code just undefines EX_OK before #including unistd.h.
4777 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4778 in the smtp transport. When a block could not be written in a single
4779 write() function, the timeout was being re-applied to each part-write.
4780 This seems wrong - if the receiver was accepting one byte at a time it
4781 would take for ever. The timeout is now adjusted when this happens. It
4782 doesn't have to be particularly precise.
4784 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4785 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4787 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4788 without a domain if the check to <postmaster@domain> fails.
4790 SC/01 Eximstats: added -xls and the ability to specify output files
4791 (patch written by Frank Heydlauf).
4793 SC/02 Eximstats: use FileHandles for outputting results.
4795 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4797 SC/04 Eximstats: fixed display of large numbers with -nvr option
4799 SC/05 Eximstats: fixed merging of reports with empty tables.
4801 SC/06 Eximstats: added the -include_original_destination flag
4803 SC/07 Eximstats: removed tabs and trailing whitespace.
4805 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4807 TK/06 MBOX spool code: Add real "From " MBOX separator line
4808 so the .eml file is really in mbox format (even though
4809 most programs do not really care). Patch from Alex Miller.
4811 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4812 The latter is generated from $received_to and is only set if the
4813 message has one envelope recipient. SA can use these headers,
4814 obviously out-of-the-box. Patch from Alex Miller.
4816 PH/08 The ${def test on a variable was returning false if the variable's
4817 value was "0", contrary to what the specification has always said!
4818 The result should be true unless the variable is empty.
4820 PH/09 The syntax error of a character other than { following "${if
4821 def:variable_name" (after optional whitespace) was not being diagnosed.
4822 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4823 accidental colon was present, for example, could give incorrect results.
4825 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4826 result is used (not including appendfile, where other changes are about
4829 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4830 This involved changing a lot of size variables from int to off_t. It
4831 should work with maildirs and everything.
4833 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4834 spamd dying while we are connected to it.
4836 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4839 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4840 function for BDB 4.3.
4842 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4844 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4845 the sources for PCRE have been re-arranged and more files are now
4848 PH/15 The code I had for printing potentially long long variables in PH/11
4849 above was not the best (it lost precision). The length of off_t variables
4850 is now inspected at build time, and an appropriate printing format (%ld
4851 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4852 to be "long long int" or "long int". This is needed for the internal
4853 formatting function string_vformat().
4855 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4856 the configuration file to be ":syslog", then the script "guesses" where
4857 the logs files are, rather than using the compiled in default. In our
4858 case the guess is not the same as the compiled default, so the script
4859 suddenly stopped working when I started to use syslog. The patch checks
4860 to see if log_file_path is "". If so, it attempts to read it from exim
4861 with no configuration file to get the compiled in version, before it
4862 falls back to the previous guessing code."
4864 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4865 implementing BATV in an Exim configuration. See NewStuff for the gory
4868 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4869 Makefile that are specific to HP-UX.
4871 PH/18 If the "use_postmaster" option was set for a recipient callout together
4872 with the "random" option, the postmaster address was used as the MAIL
4873 FROM address for the random test, but not for the subsequent recipient
4874 test. It is now used for both.
4876 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4877 patch removes a few documentation additions to RFC 3028, because the
4878 latest draft now contains them. It adds the new en;ascii-case comparator
4879 and a new error check for 8bit text in MIME parts. Comparator and
4880 require names are now matched exactly. I enabled the subaddress
4881 extension, but it is not well tested yet (read: it works for me)."
4883 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4884 rework some of the code of TK/09 above to avoid the hardwired use of
4885 "%lld" and "long long". Replaced the call to snprintf() with a call to
4888 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4889 records point to non-existent hosts", "retry timeout exceeded", and
4890 "retry time not reached for any host after a long failure period".
4892 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4893 experimental DomainKeys support:
4895 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4896 (2) On an error such as an illegally used "control", the wrong name for
4897 the control was given.
4899 These problems did NOT occur unless DomainKeys support was compiled.
4901 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4903 PH/24 Added ${if match_ip condition.
4905 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4906 databases so that it will be absolutely obvious if a crash occurs in the
4907 DB library. This is a regular occurrence (often caused by mis-matched
4910 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4911 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4912 picked up on a user's system that detects such things. There doesn't seem
4913 to be a gcc warning option for this - only an attribute that has to be
4914 put on the function's prototype. It seems that in Fedora Core 4 they have
4915 set this on a number of new functions. No doubt there will be more in due
4918 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4919 specify a transport. However, if an address that was verified by such a
4920 router was the subject of a callout, Exim crashed because it tried to
4921 read the rcpt_include_affixes from the non-existent transport. Now it
4922 just assumes that the setting of that option is false. This bug was
4923 introduced by 4.51/PH/31.
4925 PH/28 Changed -d+all to exclude +memory, because that information is very
4926 rarely of interest, but it makes the output a lot bigger. People tend to
4927 do -d+all out of habit.
4929 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4930 code in os-type was giving problems when libc.so lives in lib64, like on
4933 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4934 aren't the modern standard, and it seems that some systems' include files
4935 don't always have them. Exim was already checking for some of the newer
4936 ones like T_AAAA, and defining it itself. I've added checks for all the
4937 record types that Exim uses.
4939 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4940 not automatically generating a new one, as it is supposed to. This
4941 prevented TLS from working. If the file did exist, but contained invalid
4942 data, a new version was generated, as expected. It was only the case of a
4943 non-existent file that was broken.
4945 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4946 with a change in libdomainkeys > 0.64.
4948 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4949 from DNS. If the selector record carries the flag, it now has
4950 precedence over the domain-wide flag.
4952 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4954 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4955 the use of a port name that isn't defined in /etc/services) occurred, the
4956 message was deferred as in a normal delivery, and thus remained on the
4957 spool, instead of being failed because of the mua_wrapper setting. This
4958 is now fixed, and I tidied up some of the mua_wrapper messages at the
4961 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4962 the messages in a hash of arrays rather than using individual hashes.
4963 This is a bit cleaner and results in dramatic memory savings, albeit
4964 at a slight CPU cost.
4966 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4967 as requested by Marc Sherman.
4969 SC/10 Eximstats: added histograms for user specified patterns as requested
4972 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4974 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4975 fopen() in the content-scanning modules that did not already have it.
4981 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4982 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4984 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4986 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4988 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4989 to test Sieve filters that use "vacation".
4991 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4992 that changes the way the GnuTLS parameters are stored in the cache file.
4993 The new format can be generated externally. For backward compatibility,
4994 if the data in the cache doesn't make sense, Exim assumes it has read an
4995 old-format file, and it generates new data and writes a new file. This
4996 means that you can't go back to an older release without removing the
4999 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5000 work if there are any delivery delays because "one_time" forces the
5001 parent to be marked "delivered", so its unseen clone is never tried
5002 again. For this reason, Exim now forbids the simultaneous setting of
5005 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5006 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5007 entry if you want to know the details. The fix, however, neglected to
5008 consider the case when local delivery batching is involved. The test for
5009 "previously delivered" was not happening when checking to see if an
5010 address could be batched with a previous (undelivered) one; under
5011 certain circumstances this could lead to multiple deliveries to the same
5014 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5015 in its include files, and this causes problems building Exim.
5017 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5018 header_syntax) but Exim was just ignoring anything given after a slash.
5019 In particular, this caused confusion with an attempt to use "verify =
5020 reverse_host_lookup/defer_ok". An error is now given when options are
5021 supplied for verify items that do not have them. (Maybe reverse_host_
5022 lookup should have a defer_ok option, but that's a different point.)
5024 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5025 defined by RFC 821) to 2048, because there were problems with some AUTH
5026 commands, and RFC 1869 says the size should be increased for extended
5027 SMTP commands that take arguments.
5029 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5032 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5033 "unknown" error; now it says that the functionality isn't in the binary.
5035 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5036 an address' error message when a string expansion fails (syntax or
5037 whatever). Otherwise the password may appear in the log. Following change
5038 PH/42 below, there is no longer a chance of it appearing in a bounce
5041 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5043 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5044 of its addresses was ever tried. (Bugzilla bug #2.)
5046 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5047 the result incorrectly in the debug output. (It correctly added a newline
5048 to what was transported.)
5050 TF/01 Added $received_time.
5052 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5053 commented out examples of how to interface to a virus scanner and to
5054 SpamAssassin. Also added commented examples of av_scanner and
5055 spamd_address settings.
5057 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5058 and controls are allowed in which ACLs. There were a couple of minor
5059 errors. Some of the entries in the conditions table (which is a table of
5060 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5061 negation of where the condition IS allowed.
5063 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5065 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5066 header file does not have a version number, so I've had to invent a new
5067 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5068 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5069 radiusclient), but it was contributed by a Radius user.
5071 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5072 files or directories, for OpenSSL.
5074 PH/19 When an Exim process that is running as root has to create an Exim log
5075 file, it does so in a subprocess that runs as exim:exim so as to get the
5076 ownership right at creation (otherwise, other Exim processes might see
5077 the file with the wrong ownership). There was no test for failure of this
5078 fork() call, which would lead to the process getting stuck as it waited
5079 for a non-existent subprocess. Forks do occasionally fail when resources
5080 run out. I reviewed all the other calls to fork(); they all seem to check
5083 PH/20 When checking for unexpected SMTP input at connect time (before writing
5084 the banner), Exim was not dealing correctly with a non-positive return
5085 from the read() function. If the client had disconnected by this time,
5086 the result was a log entry for a synchronization error with an empty
5087 string after "input=" when read() returned zero. If read() returned -1
5088 (an event I could not check), uninitialized data bytes were printed.
5089 There were reports of junk text (parts of files, etc) appearing after
5092 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5094 PH/22 Added support for macro redefinition, and (re)definition in between
5095 driver and ACL definitions.
5097 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5098 forgetting to use the resulting value; it was using the unexpanded value.
5100 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5101 hadn't been configured. The fix is from Juergen Kreileder, who
5102 understands it better than I do:
5104 "Here's what I see happening with three configured cyrus_sasl
5105 authenticators configured (plain, login, cram-md5):
5107 On startup auth_cyrus_sasl_init() gets called for each of these.
5108 This means three calls to sasl_listmech() without a specified mech_list.
5109 => SASL tests which mechs of all available mechs actually work
5110 => three warnings about OTP not working
5111 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5113 With the patch, sasl_listmech() also gets called three times. But now
5114 SASL's mech_list option is set to the server_mech specified in the the
5115 authenticator. Or in other words, the answer from sasl_listmech()
5116 gets limited to just the mech you're testing for (which is different
5118 => the return list contains just 'plain' or 'login', 'cram-md5' or
5119 nothing depending on the value of ob->server_mech.
5121 I've just tested the patch: Authentication still works fine,
5122 unavailable mechs specified in the exim configuration are still
5123 caught, and the auth.log warnings about OTP are gone."
5125 PH/25 When debugging is enabled, the contents of the command line are added
5126 to the debugging output, even when log_selector=+arguments is not
5129 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5130 answer is "GNU", and only if the return is "GNU/something" is the answer
5133 PH/27 $acl_verify_message is now set immediately after the failure of a
5134 verification in an ACL, and so is available in subsequent modifiers. In
5135 particular, the message can be preserved by coding like this:
5137 warn !verify = sender
5138 set acl_m0 = $acl_verify_message
5140 Previously, $acl_verify_message was set only while expanding "message"
5141 and "log_message" when a very denied access.
5143 PH/28 Modified OS/os.c-Linux with
5145 -#ifndef OS_LOAD_AVERAGE
5146 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5148 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5149 nomenclature these days.)
5151 PH/29 Installed patch from the Sieve maintainer that adds the options
5152 sieve_useraddress and sieve_subaddress to the redirect router.
5154 PH/30 In these circumstances:
5155 . Two addresses routed to the same list of hosts;
5156 . First host does not offer TLS;
5157 . First host accepts first address;
5158 . First host gives temporary error to second address;
5159 . Second host offers TLS and a TLS session is established;
5160 . Second host accepts second address.
5161 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5162 and peerdn, if requested) that were in fact used only for the second
5165 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5166 attention to any local part prefix or suffix that was matched by the
5167 router that accepted the address. It now behaves in the same way as it
5168 does for delivery: the affixes are removed from the local part unless
5169 rcpt_include_affixes is set on the transport.
5171 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5172 timeout during the DATA phase of an incoming message.
5174 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5175 applied a patch sanctioned by the Sieve maintainer.
5177 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5178 the uid or gid is negative. A case of a negative gid caused this to be
5179 noticed. The fix allows for either to be negative.
5181 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5182 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5185 PH/36 The change PH/12 above was broken. Fixed it.
5187 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5188 the grounds that routing the same address twice would always produce the
5189 same answer. This might have been true once, but it is certainly no
5190 longer true now. Routing a child address may depend on the previous
5191 routing that produced that child. Some complicated redirection strategies
5192 went wrong when messages had multiple recipients, and made Exim's
5193 behaviour dependent on the order in which the addresses were given.
5195 I have moved the duplicate checking until after the routing is complete.
5196 Exim scans the addresses that are assigned to local and remote
5197 transports, and removes any duplicates. This means that more work will be
5198 done, as duplicates will always all be routed, but duplicates are
5199 presumably rare, so I don't expect this is of any significance.
5201 For deliveries to pipes, files, and autoreplies, the duplicate checking
5202 still happens during the routing process, since they are not going to be
5205 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5206 It corrects a timeout issue with spamd. This is Ian's comment: "The
5207 background is that sometimes spamd either never reads data from a
5208 connection it has accepted, or it never writes response data. The exiscan
5209 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5210 blindly assumes that writes won't block so it may never time out."
5212 PH/39 Allow G after quota size as well as K and M.
5214 PH/40 The value set for $authenticated_id in an authenticator may not contain
5215 binary zeroes or newlines because the value is written to log lines and
5216 to spool files. There was no check on this. Now the value is run through
5217 the string_printing() function so that such characters are converted to
5218 printable escape sequences.
5220 PH/41 $message_linecount is a new variable that contains the total number of
5221 lines in the message. Compare $body_linecount, which is the count for the
5224 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5225 bounce and delay warning messages, except in certain special cases, which
5228 (a) An SMTP error message from a remote host;
5229 (b) A message specified in a :fail: redirection;
5230 (c) A message specified in a "fail" command in a system filter;
5231 (d) A message specified in a FAIL return from the queryprogram router;
5232 (e) A message specified by the cannot_route_message router option.
5234 In these cases only, Exim does include the error details in bounce and
5235 warning messages. There are also a few cases where bland messages such
5236 as "unrouteable address" or "local delivery error" are given.
5238 PH/43 $value is now also set for the "else" part of a ${run expansion.
5240 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5241 being worked on, but at least Exim now implements the latest version to
5244 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5245 process to complete was treated as a delivery failure, a timeout while
5246 writing the message to the pipe was logged, but erroneously treated as a
5247 successful delivery. Such timeouts include transport filter timeouts. For
5248 consistency with the overall process timeout, these timeouts are now
5249 treated as errors, giving rise to delivery failures by default. However,
5250 there is now a new Boolean option for the pipe transport called
5251 timeout_defer, which, if set TRUE, converts the failures into defers for
5252 both kinds of timeout. A transport filter timeout is now identified in
5255 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5256 systems where "make" and "gmake" are different, calling "gmake" at top
5257 level broke things. I've arranged for the value of $(MAKE) to be passed
5258 from the Makefile to this script so that it can call the same version of
5262 A note about Exim versions 4.44 and 4.50
5263 ----------------------------------------
5265 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5266 changes of various kinds. As a consequence, a big documentation update was
5267 needed. This delayed the release for rather longer than seemed good, especially
5268 in the light of a couple of (minor) security issues. Therefore, the changes
5269 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5270 release. So 4.44 and 4.50 are in effect two different branches that both start
5273 I have left the 4.50 change log unchanged; it contains all the changes since
5274 4.43. The change log for 4.44 is below; many of its items are identical to
5275 those for 4.50. This seems to be the most sensible way to preserve the
5276 historical information.
5282 1. Minor wording change to the doc/README.SIEVE file.
5284 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5285 computation of the current number of files was incorrect.
5287 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5288 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5291 4. Give more explanation in the error message when the command for a transport
5292 filter fails to execute.
5294 5. There are several places where Exim runs a non-Exim command in a
5295 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5296 was being done only for the command run by the queryprogram router. It is
5297 now done for all such subprocesses. The other cases are: ${run, transport
5298 filters, and the commands run by the lmtp and pipe transports.
5300 6. Added CONFIGURE_GROUP build-time option.
5302 7. Some older OS have a limit of 256 on the maximum number of file
5303 descriptors. Exim was using setrlimit() to set 1000 as a large value
5304 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5305 systems. I've change it so that if it can't get 1000, it tries for 256.
5307 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5308 was an oversight, and furthermore, ever since the addition of extra
5309 controls (e.g. 4.43/32), the checks on when to allow different forms of
5310 "control" were broken. There should now be diagnostics for all cases when a
5311 control that does not make sense is encountered.
5313 9. Added the /retain_sender option to "control=submission".
5315 10. $recipients is now available in the predata ACL (oversight).
5317 11. Tidy the search cache before the fork to do a delivery from a message
5318 received from the command line. Otherwise the child will trigger a lookup
5319 failure and thereby defer the delivery if it tries to use (for example) a
5320 cached ldap connection that the parent has called unbind on.
5322 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5323 of $address_data from the recipient verification was clobbered by the
5324 sender verification.
5326 13. The value of address_data from a sender verification is now available in
5327 $sender_address_data in subsequent conditions in the ACL statement.
5329 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5331 15. Added a new option "connect=<time>" to callout options, to set a different
5334 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5335 was its contents. (It was OK if the option was not defined at all.)
5337 17. A "Completed" log line is now written for messages that are removed from
5338 the spool by the -Mrm option.
5340 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5341 information about exactly what failed.
5343 19. Added -dd to debug only the daemon process.
5345 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5346 handles timeouts, both on the server side and network timeouts. Renamed the
5347 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5349 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5350 It is now set to "smtps".
5352 22. $host_address is now set to the target address during the checking of
5353 ignore_target_hosts.
5355 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5356 being passed; this would have caused $sender_host_name to have been used if
5357 matching the list had actually called for a host name (not very likely,
5358 since this list is usually IP addresses). A host name is now passed as
5361 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5362 code that specifies a non-restarting handler (typically sigaction() in
5363 modern systems) in an attempt to fix a rare and obscure crash bug.
5365 25. Narrowed the window for a race in the daemon that could cause it to ignore
5366 SIGCHLD signals. This is not a major problem, because they are used only to
5367 wake it up if nothing else does.
5369 26. A malformed maildirsize file could cause Exim to calculate negative values
5370 for the mailbox size or file count. Odd effects could occur as a result.
5371 The maildirsize information is now recalculated if the size or filecount
5374 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5375 support for a long time. Removed HAVE_SYS_VFS_H.
5377 28. Installed the latest version of exipick from John Jetmore.
5379 29. In an address list, if the pattern was not a regular expression, an empty
5380 subject address (from a bounce message) matched only if the pattern was an
5381 empty string. Non-empty patterns were not even tested. This was the wrong
5382 because it is perfectly reasonable to use an empty address as part of a
5383 database query. An empty address is now tested by patterns that are
5384 lookups. However, all the other forms of pattern expect the subject to
5385 contain a local part and a domain, and therefore, for them, an empty
5386 address still always fails if the pattern is not itself empty.
5388 30. Exim went into a mad DNS loop when attempting to do a callout where the
5389 host was specified on an smtp transport, and looking it up yielded more
5390 than one IP address.
5392 31. Re-factored the code for checking spool and log partition space into a
5393 function that finds that data and another that does the check. The former
5394 is then used to implement four new variables: $spool_space, $log_space,
5395 $spool_inodes, and $log_inodes.
5397 32. The RFC2047 encoding function was originally intended for short strings
5398 such as real names; it was not keeping to the 75-character limit for
5399 encoded words that the RFC imposes. It now respects the limit, and
5400 generates multiple encoded words if necessary. To be on the safe side, I
5401 have increased the buffer size for the ${rfc2047: expansion operator from
5404 33. It is now permitted to omit both strings after an "if" condition; if the
5405 condition is true, the result is "true". As before, when the second string
5406 is omitted, a false condition yields an empty string. This makes it less
5407 cumbersome to write custom ACL and router conditions.
5409 34. Failure to deliver a bounce message always caused it to be frozen, even if
5410 there was an errors_to setting on the router. The errors_to setting is now
5413 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5414 canonical form (fully expanded) before being placed in
5415 $sender_host_address.
5417 36. The table in the code that translates DNS record types into text (T_A to
5418 "A" for instance) was missing entries for NS and CNAME. It is just possible
5419 that this could have caused confusion if both these types were looked up
5420 for the same domain, because the text type is used as part of Exim's
5421 per-process caching. But the chance of anyone hitting this buglet seems
5424 37. The dnsdb lookup has been extended in a number of ways.
5426 (1) There is a new type, "zns", which walks up the domain tree until it
5427 finds some nameserver records. It should be used with care.
5429 (2) There is a new type, "mxh", which is like "mx" except that it returns
5430 just the host names, not the priorities.
5432 (3) It is now possible to give a list of domains (or IP addresses) to be
5433 looked up. The behaviour when one of the lookups defers can be
5434 controlled by a keyword.
5436 (4) It is now possible to specify the separator character for use when
5437 multiple records are returned.
5439 38. The dnslists ACL condition has been extended: it is now possible to supply
5440 a list of IP addresses and/or domains to be looked up in a particular DNS
5443 39. Added log_selector=+queue_time_overall.
5445 40. When running the queue in the test harness, wait just a tad after forking a
5446 delivery process, to get repeatability of debugging output.
5448 41. Include certificate and key file names in error message when GnuTLS fails
5449 to set them up, because the GnuTLS error message doesn't include the name
5450 of the failing file when there is a problem reading it.
5452 42. Allow both -bf and -bF in the same test run.
5454 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5456 44. The "Exiscan patch" is now merged into the mainline Exim source.
5458 45. Sometimes the final signoff response after QUIT could fail to get
5459 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5460 < 0 before doing a fflush(). This bug looks as though it goes back to the
5461 introduction of TLS in release 3.20, but "sometimes" must have been rare
5462 because the tests only now provoked it.
5464 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5465 (this can affect the format of dates).
5467 47. exim_tidydb, when checking for the continued existence of a message for
5468 which it has found a message-specific retry record, was not finding
5469 messages that were in split spool directories. Consequently, it was
5470 deleting retry records that should have stayed in existence.
5472 48. Steve fixed some bugs in eximstats.
5474 49. The SPA authentication driver was not abandoning authentication and moving
5475 on to the next authenticator when an expansion was forced to fail,
5476 contradicting the general specification for all authenticators. Instead it
5477 was generating a temporary error. It now behaves as specified.
5479 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5480 (the order specifies the preference for clients). The order is now AES256,
5481 AES128, 3DES, ARCFOUR128.
5483 51. Small patch to Sieve code - explicitly set From: when generating an
5486 52. Exim crashed if a remote delivery caused a very long error message to be
5487 recorded - for instance if somebody sent an entire SpamAssassin report back
5488 as a large number of 550 error lines. This bug was coincidentally fixed by
5489 increasing the size of one of Exim's internal buffers (big_buffer) that
5490 happened as part of the Exiscan merge. However, to be on the safe side, I
5491 have made the code more robust (and fixed the comments that describe what
5494 53. Now that there can be additional text after "Completed" in log lines (if
5495 the queue_time_overall log selector is set), a one-byte patch to exigrep
5496 was needed to allow it to recognize "Completed" as not the last thing in
5499 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5500 patch that reportedly fixes this has been added. I am not expert enough to
5501 create a test for it. This is what the patch creator wrote:
5503 "I found a little strange behaviour of ldap code when working with
5504 Windows 2003 AD Domain, where users was placed in more than one
5505 Organization Units. When I tried to give exim partial DN, the exit code
5506 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5507 But simultaneously result of request was absolutely normal ldap result,
5508 so I produce this patch..."
5510 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5511 REFERENCE, so I have modified the code to exclude the patch when that macro
5514 55. Some experimental protocols are using DNS PTR records for new purposes. The
5515 keys for these records are domain names, not reversed IP addresses. The
5516 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5517 leaves it alone. Component reversal etc. now happens only for IP addresses.
5520 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5522 57. Double the size of the debug message buffer (to 2048) so that more of very
5523 long debug lines gets shown.
5525 58. The exicyclog utility now does better if the number of log files to keep
5526 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5528 59. Two changes related to the smtp_active_hostname option:
5530 (1) $smtp_active_hostname is now available as a variable.
5531 (2) The default for smtp_banner uses $smtp_active_hostname instead
5532 of $primary_hostname.
5534 60. The host_aton() function is supposed to be passed a string that is known
5535 to be a valid IP address. However, in the case of IPv6 addresses, it was
5536 not checking this. This is a hostage to fortune. Exim now panics and dies
5537 if the condition is not met. A case was found where this could be provoked
5538 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5539 components; fortuitously, this particular loophole had already been fixed
5540 by change 4.50/55 above.
5542 If there are any other similar loopholes, the new check in host_aton()
5543 itself should stop them being exploited. The report I received stated that
5544 data on the command line could provoke the exploit when Exim was running as
5545 exim, but did not say which command line option was involved. All I could
5546 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5547 running as the user.
5550 61. There was a buffer overflow vulnerability in the SPA authentication code
5551 (which came originally from the Samba project). I have added a test to the
5552 spa_base64_to_bits() function which I hope fixes it.
5555 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5556 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5558 63. The daemon start-up calls getloadavg() while still root for those OS that
5559 need the first call to be done as root, but it missed one case: when
5560 deliver_queue_load_max is set with deliver_drop_privilege. This is
5561 necessary for the benefit of the queue runner, because there is no re-exec
5562 when deliver_drop_privilege is set.
5564 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5565 This has been fixed.
5567 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5568 was in use, was not putting the data itself into the right store pool;
5569 consequently, it could be overwritten for a subsequent message in the same
5570 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5573 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5575 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5576 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5577 the function were treating the return as a boolean value, which happened to
5578 work because 0=false and not-0=true, but is not correct code.
5580 68. The host_aton() function was not handling scoped IPv6 addresses (those
5581 with, for example, "%eth0" on the end) correctly.
5583 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5584 negated items (that is, ~something) in unsigned ints. Some compilers
5585 apparently mutter when there is no cast.
5587 70. If an address verification called from an ACL failed, and did not produce a
5588 user-specific message (i.e. there was only a "system" message), nothing was
5589 put in $acl_verify_message. In this situation, it now puts the system
5592 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5593 session; change 4.31/43 added the unwanted input to the log line - except
5594 that it did not do this in the start of session case. It now does.
5596 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5597 This is wrong and can cause the other end to generate a synchronization
5598 error if it is another Exim or anything else that does the synchronization
5599 check. A QUIT command is no longer sent after a timeout.
5601 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5602 during host lookups.
5604 74. The defer_ok option of callout verification was not working if it was used
5605 when verifying addresses in header lines, that is, for this case:
5607 verify = header_sender/callout=defer_ok
5609 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5610 those file descriptors could be used for SMTP connections. If anything
5611 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5612 could be sent to the SMTP client, causing chaos. The daemon now opens
5613 stdin, stdout, and stderr to /dev/null when it puts itself into the
5616 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5617 log by default. The user can override this with suitable Perl magic.
5619 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5620 the log message when discard triggers, was not working for the DATA ACL or
5621 for the non-SMTP ACL.
5623 78. Error message wording change in sieve.c.
5625 79. If smtp_accept_max_per_host was set, the number of connections could be
5626 restricted to fewer than expected, because the daemon was trying to set up
5627 a new connection before checking whether the processes handling previous
5628 connections had finished. The check for completed processes is now done
5629 earlier. On busy systems, this bug wouldn't be noticed because something
5630 else would have woken the daemon, and it would have reaped the completed
5633 80. If a message was submitted locally by a user whose login name contained one
5634 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5635 It caused a spool format error. I have fixed the spool reading code. A
5636 related problem was that the "from" clause in the Received: line became
5637 illegal because of the space(s). It is now covered by ${quote_local_part.
5639 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5642 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5644 83. Patch from Sieve maintainer to fix unterminated string problem in
5645 "vacation" handling.
5647 84. Some minor changes to the Linux configuration files to help with other
5648 OS variants using glibc.
5650 85. One more patch for Sieve to update vacation handling to latest spec.
5653 ----------------------------------------------------
5654 See the note above about the 4.44 and 4.50 releases.
5655 ----------------------------------------------------
5661 1. Change 4.43/35 introduced a bug that caused file counts to be
5662 incorrectly computed when quota_filecount was set in an appendfile
5665 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5666 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5669 3. Give more explanation in the error message when the command for a transport
5670 filter fails to execute.
5672 4. There are several places where Exim runs a non-Exim command in a
5673 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5674 was being done only for the command run by the queryprogram router. It is
5675 now done for all such subprocesses. The other cases are: ${run, transport
5676 filters, and the commands run by the lmtp and pipe transports.
5678 5. Some older OS have a limit of 256 on the maximum number of file
5679 descriptors. Exim was using setrlimit() to set 1000 as a large value
5680 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5681 systems. I've change it so that if it can't get 1000, it tries for 256.
5683 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5684 was an oversight, and furthermore, ever since the addition of extra
5685 controls (e.g. 4.43/32), the checks on when to allow different forms of
5686 "control" were broken. There should now be diagnostics for all cases when a
5687 control that does not make sense is encountered.
5689 7. $recipients is now available in the predata ACL (oversight).
5691 8. Tidy the search cache before the fork to do a delivery from a message
5692 received from the command line. Otherwise the child will trigger a lookup
5693 failure and thereby defer the delivery if it tries to use (for example) a
5694 cached ldap connection that the parent has called unbind on.
5696 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5697 of $address_data from the recipient verification was clobbered by the
5698 sender verification.
5700 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5701 was its contents. (It was OK if the option was not defined at all.)
5703 11. A "Completed" log line is now written for messages that are removed from
5704 the spool by the -Mrm option.
5706 12. $host_address is now set to the target address during the checking of
5707 ignore_target_hosts.
5709 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5710 being passed; this would have caused $sender_host_name to have been used if
5711 matching the list had actually called for a host name (not very likely,
5712 since this list is usually IP addresses). A host name is now passed as
5715 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5716 code that specifies a non-restarting handler (typically sigaction() in
5717 modern systems) in an attempt to fix a rare and obscure crash bug.
5719 15. Narrowed the window for a race in the daemon that could cause it to ignore
5720 SIGCHLD signals. This is not a major problem, because they are used only to
5721 wake it up if nothing else does.
5723 16. A malformed maildirsize file could cause Exim to calculate negative values
5724 for the mailbox size or file count. Odd effects could occur as a result.
5725 The maildirsize information is now recalculated if the size or filecount
5728 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5729 support for a long time. Removed HAVE_SYS_VFS_H.
5731 18. Updated exipick to current release from John Jetmore.
5733 19. Allow an empty sender to be matched against a lookup in an address list.
5734 Previously the only cases considered were a regular expression, or an
5737 20. Exim went into a mad DNS lookup loop when doing a callout where the
5738 host was specified on the transport, if the DNS lookup yielded more than
5741 21. The RFC2047 encoding function was originally intended for short strings
5742 such as real names; it was not keeping to the 75-character limit for
5743 encoded words that the RFC imposes. It now respects the limit, and
5744 generates multiple encoded words if necessary. To be on the safe side, I
5745 have increased the buffer size for the ${rfc2047: expansion operator from
5748 22. Failure to deliver a bounce message always caused it to be frozen, even if
5749 there was an errors_to setting on the router. The errors_to setting is now
5752 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5753 canonical form (fully expanded) before being placed in
5754 $sender_host_address.
5756 24. Updated eximstats to version 1.33
5758 25. Include certificate and key file names in error message when GnuTLS fails
5759 to set them up, because the GnuTLS error message doesn't include the name
5760 of the failing file when there is a problem reading it.
5762 26. Expand error message when OpenSSL has problems setting up cert/key files.
5765 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5766 (this can affect the format of dates).
5768 28. exim_tidydb, when checking for the continued existence of a message for
5769 which it has found a message-specific retry record, was not finding
5770 messages that were in split spool directories. Consequently, it was
5771 deleting retry records that should have stayed in existence.
5773 29. eximstats updated to version 1.35
5774 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5775 1.35 - bugfix such that pie charts by volume are generated correctly
5777 30. The SPA authentication driver was not abandoning authentication and moving
5778 on to the next authenticator when an expansion was forced to fail,
5779 contradicting the general specification for all authenticators. Instead it
5780 was generating a temporary error. It now behaves as specified.
5782 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5783 (the order specifies the preference for clients). The order is now AES256,
5784 AES128, 3DES, ARCFOUR128.
5786 31. Small patch to Sieve code - explicitly set From: when generating an
5789 32. Exim crashed if a remote delivery caused a very long error message to be
5790 recorded - for instance if somebody sent an entire SpamAssassin report back
5791 as a large number of 550 error lines. This bug was coincidentally fixed by
5792 increasing the size of one of Exim's internal buffers (big_buffer) that
5793 happened as part of the Exiscan merge. However, to be on the safe side, I
5794 have made the code more robust (and fixed the comments that describe what
5797 33. Some experimental protocols are using DNS PTR records for new purposes. The
5798 keys for these records are domain names, not reversed IP addresses. The
5799 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5800 leaves it alone. Component reversal etc. now happens only for IP addresses.
5803 34. The host_aton() function is supposed to be passed a string that is known
5804 to be a valid IP address. However, in the case of IPv6 addresses, it was
5805 not checking this. This is a hostage to fortune. Exim now panics and dies
5806 if the condition is not met. A case was found where this could be provoked
5807 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5808 components; fortuitously, this particular loophole had already been fixed
5809 by change 4.50/55 or 4.44/33 above.
5811 If there are any other similar loopholes, the new check in host_aton()
5812 itself should stop them being exploited. The report I received stated that
5813 data on the command line could provoke the exploit when Exim was running as
5814 exim, but did not say which command line option was involved. All I could
5815 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5816 running as the user.
5819 35. There was a buffer overflow vulnerability in the SPA authentication code
5820 (which came originally from the Samba project). I have added a test to the
5821 spa_base64_to_bits() function which I hope fixes it.
5824 36. The daemon start-up calls getloadavg() while still root for those OS that
5825 need the first call to be done as root, but it missed one case: when
5826 deliver_queue_load_max is set with deliver_drop_privilege. This is
5827 necessary for the benefit of the queue runner, because there is no re-exec
5828 when deliver_drop_privilege is set.
5830 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5831 was in use, was not putting the data itself into the right store pool;
5832 consequently, it could be overwritten for a subsequent message in the same
5833 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5836 38. Sometimes the final signoff response after QUIT could fail to get
5837 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5838 < 0 before doing a fflush(). This bug looks as though it goes back to the
5839 introduction of TLS in release 3.20, but "sometimes" must have been rare
5840 because the tests only now provoked it.
5846 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5847 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5848 Now it returns an integer. A number of calls were still expecting a T/F
5849 return. Fortuitously, in all cases, the tests worked in OK situations,
5850 which is the norm. However, things would have gone wrong on any write
5851 failures on the smtp file descriptor. This function is used when sending
5852 messages over SMTP and also when doing verify callouts.
5854 2. When Exim is called to do synchronous delivery of a locally submitted
5855 message (the -odf or -odi options), it no longer closes stderr before doing
5858 3. Implemented the mua_wrapper option.
5860 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5862 5. Implemented the functions header_remove(), header_testname(),
5863 header_add_at_position(), and receive_remove_recipient(), and exported them
5866 6. If an ACL "warn" statement specified the addition of headers, Exim already
5867 inserted X-ACL-Warn: at the start if there was no header name. However, it
5868 was not making this test for the second and subsequent header lines if
5869 there were newlines in the string. This meant that an invalid header could
5870 be inserted if Exim was badly configured.
5872 7. Allow an ACL "warn" statement to add header lines at the start or after all
5873 the Received: headers, as well as at the end.
5875 8. Added the rcpt_4xx retry error code.
5877 9. Added postmaster_mailfrom=xxx to callout verification option.
5879 10. Added mailfrom=xxxx to the callout verification option, for verify=
5882 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5883 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5885 12. Inserted some casts to stop certain compilers warning when using pointer
5886 differences as field lengths or precisions in printf-type calls (mostly
5887 affecting debugging statements).
5889 13. Added optional readline() support for -be (dynamically loaded).
5891 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5892 same clock tick as a message's arrival, so that its received time was the
5893 same as the "first fail" time on the retry record, and that message
5894 remained on the queue past the ultimate address timeout, every queue runner
5895 would try a delivery (because it was past the ultimate address timeout) but
5896 after another failure, the ultimate address timeout, which should have then
5897 bounced the address, did not kick in. This was a "< instead of <=" error;
5898 in most cases the first failure would have been in the next clock tick
5899 after the received time, and all would be well.
5901 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5902 being recognized when the domain list was tested by the match_domain
5903 condition in an expansion string.
5905 16. Added the ${str2b64: operator.
5907 17. Exim was always calling setrlimit() to set a large limit for the number of
5908 processes, without checking whether the existing limit was already
5909 adequate. (It did check for the limit on file descriptors.) Furthermore,
5910 errors from getrlimit() and setrlimit() were being ignored. Now they are
5911 logged to the main and panic logs, but Exim does carry on, to try to do its
5912 job under whatever limits there are.
5914 18. Imported PCRE 5.0.
5916 19. Trivial typo in log message " temporarily refused connection" (the leading
5919 20. If the log selector return_path_on_delivery was set and an address was
5920 redirected to /dev/null, the delivery process crashed because it assumed
5921 that a return path would always be set for a "successful" delivery. In this
5922 case, the whole delivery is bypassed as an optimization, and therefore no
5925 21. Internal re-arrangement: the function for sending a challenge and reading
5926 a response while authentication was assuming a zero-terminated challenge
5927 string. It's now changed to take a pointer and a length, to allow for
5928 binary data in such strings.
5930 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5932 23. Exim was not respecting finduser_retries when seeking the login of the
5933 uid under which it was called; it was always trying 10 times. (The default
5934 setting of finduser_retries is zero.) Also, it was sleeping after the final
5935 failure, which is pointless.
5937 24. Implemented tls_on_connect_ports.
5939 25. Implemented acl_smtp_predata.
5941 26. If the domain in control=submission is set empty, Exim assumes that the
5942 authenticated id is a complete email address when it generates From: or
5943 Sender: header lines.
5945 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5946 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5947 chown and chgrp in /bin and hostname in /usr/bin.
5949 28. Exim was keeping the "process log" file open after each use, just as it
5950 does for the main log. This opens the possibility of it remaining open for
5951 long periods when the USR1 signal hits a daemon. Occasional processlog
5952 errors were reported, that could have been caused by this. Anyway, it seems
5953 much more sensible not to leave this file open at all, so that is what now
5956 29. The long-running daemon process does not normally write to the log once it
5957 has entered its main loop, and it closes the log before doing so. This is
5958 so that log files can straightforwardly be renamed and moved. However,
5959 there are a couple of unusual error situations where the daemon does write
5960 log entries, and I had neglected to close the log afterwards.
5962 30. The text of an SMTP error response that was received during a remote
5963 delivery was being truncated at 512 bytes. This is too short for some of
5964 the long messages that one sometimes sees. I've increased the limit to
5967 31. It is now possible to make retry rules that apply only when a message has a
5968 specific sender, in particular, an empty sender.
5970 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5971 it possible to be selective about when SMTP synchronization is enforced.
5973 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5975 32. Implemented hosts_connection_nolog.
5977 33. Added an ACL for QUIT.
5979 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5982 35. Added mailbox_size and mailbox_filecount to appendfile.
5984 36. Added control = no_multiline_responses to ACLs.
5986 37. There was a bug in the logic of the code that waits for the clock to tick
5987 in the case where the clock went backwards by a substantial amount such
5988 that the microsecond fraction of "now" was more than the microsecond
5989 fraction of "then" (but the whole seconds number was less).
5991 38. Added support for the libradius Radius client library this is found on
5992 FreeBSD (previously only the radiusclient library was supported).
5998 1. When certain lookups returned multiple values in the form name=value, the
5999 quoting of the values was not always being done properly. Specifically:
6000 (a) If the value started with a double quote, but contained no whitespace,
6002 (b) If the value contained whitespace other than a space character (i.e.
6003 tabs or newlines or carriage returns) it was not quoted.
6004 This fix has been applied to the mysql and pgsql lookups by writing a
6005 separate quoting function and calling it from the lookup code. The fix
6006 should probably also be applied to nisplus, ibase and oracle lookups, but
6007 since I cannot test any of those, I have not disturbed their existing code.
6009 2. A hit in the callout cache for a specific address caused a log line with no
6010 reason for rejecting RCPT. Now it says "Previous (cached) callout
6011 verification failure".
6013 3. There was an off-by-one bug in the queryprogram router. An over-long
6014 return line was truncated at 256 instead of 255 characters, thereby
6015 overflowing its buffer with the terminating zero. As well as fixing this, I
6016 have increased the buffer size to 1024 (and made a note to document this).
6018 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6019 when Exim is waiting for an SMTP response from a remote server, Exim
6020 restarts its select() call on the socket, thereby resetting its timeout.
6021 This is not a problem when such interrupts are rare. Somebody set up a cron
6022 job to run exiwhat every 2 minutes, which is less than the normal select()
6023 timeout (5 or 10 minutes). This meant that the select() timeout never
6024 kicked in because it was always reset. I have fixed this by comparing the
6025 time when an interrupt arrives with the time at the start of the first call
6026 to select(). If more time than the timeout has elapsed, the interrupt is
6027 treated as a timeout.
6029 5. Some internal re-factoring in preparation for the addition of Sieve
6030 extensions (by MH). In particular, the "personal" test is moved to a
6031 separate function, and given an option for scanning Cc: and Bcc: (which is
6032 not set for Exim filters).
6034 6. When Exim created an email address using the login of the caller as the
6035 local part (e.g. when creating a From: or Sender: header line), it was not
6036 quoting the local part when it contained special characters such as @.
6038 7. Installed new OpenBSD configuration files.
6040 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6041 try to make them clearer.
6043 9. Callout options, other than the timeout value, were being ignored when
6044 verifying sender addresses in header lines. For example, when using
6046 verify = header_sender/callout=no_cache
6048 the cache was (incorrectly) being used.
6050 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6051 only the Cygwin environment.
6053 11. When return_path_on_delivery was set as a log selector, if different remote
6054 addresses in the same message used different return paths and parallel
6055 remote delivery occurred, the wrong values would sometimes be logged.
6056 (Whenever a remote delivery process finished, the return path value from
6057 the most recently started remote delivery process was logged.)
6059 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6060 lines when AUTH and/or TLS are in use. This is the "received protocol"
6061 field. Exim used to use "asmtp" for authenticated SMTP, without any
6062 indication (in the protocol name) for TLS use. Now it follows the RFC and
6063 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6064 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6065 names appear in log lines as well as in Received: header lines.
6067 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6068 extensions, and comparison tests, and to fix some bugs.
6070 14. Changes to the "personal" filter test:
6072 (1) The test was buggy in that it was just doing the equivalent of
6073 "contains" tests on header lines. For example, if a user's address was
6074 anne@some.where, the "personal" test would incorrectly be true for
6076 To: susanne@some.where
6078 This test is now done by extracting each address from the header in turn,
6079 and checking the entire address. Other tests that are part of "personal"
6080 are now done using regular expressions (for example, to check local parts
6081 of addresses in From: header lines).
6083 (2) The list of non-personal local parts in From: addresses has been
6084 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6085 taken from the Sieve specification recommendations.
6087 (3) If the message contains any header line starting with "List-" it is
6088 treated as non-personal.
6090 (4) The test for "circular" in the Subject: header line has been removed
6091 because it now seems ill-conceived.
6093 15. Minor typos in src/EDITME comments corrected.
6095 16. Installed latest exipick from John Jetmore.
6097 17. If headers_add on a router specified a text string that was too long for
6098 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6099 of string_sprintf() is now avoided.
6101 18. $message_body_size was not set (it was always zero) when running the DATA
6102 ACL and the local_scan() function.
6104 19. For the "mail" command in an Exim filter, no default was being set for
6105 the once_repeat time, causing a random time value to be used if "once" was
6106 specified. (If the value happened to be <= 0, no repeat happened.) The
6107 default is now 0s, meaning "never repeat". The "vacation" command was OK
6108 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6109 (I found it when inspecting the code).
6111 20. There is now an overall timeout for performing a callout verification. It
6112 defaults to 4 times the callout timeout, which applies to individual SMTP
6113 commands during the callout. The overall timeout applies when there is more
6114 than one host that can be tried. The timeout is checked before trying the
6115 next host. This prevents very long delays if there are a large number of
6116 hosts and all are timing out (e.g. when the network connections are timing
6117 out). The value of the overall timeout can be changed by specifying an
6118 additional sub-option for "callout", called "maxwait". For example:
6120 verify = sender/callout=5s,maxwait=20s
6122 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6123 to the end before writing, but this should make it even safer).
6125 22. Exim was forgetting that it had advertised PIPELINING for the second and
6126 subsequent messages on an SMTP connection. It was also not resetting its
6127 memory on STARTTLS and an internal HELO.
6129 23. When Exim logs an SMTP synchronization error within a session, it now
6130 records whether PIPELINING has been advertised or not.
6132 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6133 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6134 rather than long int.
6136 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6138 26. Added the never_mail option to autoreply.
6144 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6145 crash if the getsockname() call failed; this can happen if a connection is
6146 closed very soon after it is established. The problem was simply in the
6147 order in which certain operations were done, causing Exim to try to write
6148 to the SMTP stream before it had set up the file descriptor. The bug has
6149 been fixed by making things happen in the correct order.
6155 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6156 before the connection was closed, thus losing the rejection response.
6158 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6159 some early Solaris releases, but causes trouble in current releases where
6160 socklen_t is defined.
6162 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6165 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6168 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6169 that releasing the top of it at the end releases what was used for sub-
6170 expansions (unless the block got too big). However, discard this block if
6171 the first thing is a variable or header, so that we can use its block when
6172 it is dynamic (useful for very large $message_headers, for example).
6174 6. Lookups now cache *every* query, not just the most recent. A new, separate
6175 store pool is used for this. It can be recovered when all lookup caches are
6176 flushed. Lookups now release memory at the end of their result strings.
6177 This has involved some general refactoring of the lookup sources.
6179 7. Some code has been added to the store_xxx() functions to reduce the amount
6180 of flapping under certain conditions.
6182 8. log_incoming_interface used to affect only the <= reception log lines. Now
6183 it causes the local interface and port to be added to several more SMTP log
6184 lines, for example "SMTP connection from", and rejection lines.
6186 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6188 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6190 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6191 could be overwritten at the end of the current message (or the start of a
6192 new message if it was set in a HELO ACL). The value is now preserved for
6193 the duration of the SMTP connection.
6195 12. If a transport had a headers_rewrite setting, and a matching header line
6196 contained an unqualified address, that address was qualified, even if it
6197 did not match any rewriting rules. The underlying bug was that the values
6198 of the flags that permit the existence of unqualified sender and recipient
6199 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6200 non-local messages, and by -bnq for local messages) were not being
6201 preserved with the message after it was received.
6203 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6204 "next input=" as part of the text comprising the host identity instead of
6205 the correct text. The code was using the same buffer for two different
6206 strings. However, depending on which order the printing function evaluated
6207 its arguments, the bug did not always show up. Under Linux, for example, my
6208 test suite worked just fine.
6210 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6211 doesn't work with some older versions of Perl. It has been changed to "my",
6212 which in any case is probably the better facility to use.
6214 15. A really picky compiler found some instances of statements for creating
6215 error messages that either had too many or two few arguments for the format
6218 16. The size of the buffer for calls to the DNS resolver has been increased
6219 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6220 for addresses that have a lot of PTR records. This alleviates a problem; it
6221 does not fully solve it.
6223 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6224 buffer now truncates the list and logs the incident, which is the same
6225 action as happens when Exim is looking up a host name and its aliases.
6226 Previously in this situation something unpredictable would happen;
6227 sometimes it was "internal error: store_reset failed".
6229 18. If a server dropped the connection unexpectedly when an Exim client was
6230 using GnuTLS and trying to read a response, the client delivery process
6231 crashed while trying to generate an error log message.
6233 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6234 string, for example:
6236 warn message = H1: something\nH2: something
6238 the text was added as a single header line from Exim's point of view
6239 though it ended up OK in the delivered message. However, searching for the
6240 second and subsequent header lines using $h_h2: did not work. This has been
6241 fixed. Similarly, if a system filter added multiple headers in this way,
6242 the routers could not see them.
6244 20. Expanded the error message when iplsearch is called with an invalid key to
6245 suggest using net-iplsearch in a host list.
6247 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6248 ACLs are no longer actually imposed (and a message to that effect is
6251 22. If a "gecos" field in a passwd entry contained escaped characters, in
6252 particular, if it contained a \" sequence, Exim got it wrong when building
6253 a From: or a Sender: header from that name. A second bug also caused
6254 incorrect handling when an unquoted " was present following a character
6255 that needed quoting.
6257 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6258 was not being matched caselessly.
6260 24. Arranged for all hyphens in the exim.8 source to be escaped with
6263 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6264 properly. Recipient callout cache records were still being keyed to include
6265 the sender, even when use_sender was set false. This led to far more
6266 callouts that were necessary. The sender is no longer included in the key
6267 when use_sender is false.
6269 26. Added "control = submission" modifier to ACLs.
6271 27. Added the ${base62d: operator to decode base 62 numbers.
6273 28. dnsdb lookups can now access SRV records.
6275 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6276 the configuration file.
6278 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6279 (-v) mode. This makes the output for a verbose queue run more intelligible.
6281 31. Added a use_postmaster feature to recipient callouts.
6283 32. Added the $body_zerocount variable, containing the number of binary zero
6284 bytes in the message body.
6286 33. The time of last modification of the "new" subdirectory is now used as the
6287 "mailbox time last read" when there is a quota error for a maildir
6290 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6292 35. Added +ignore_unknown as a special item in host lists.
6294 36. Code for decoding IPv6 addresses in host lists is now included, even if
6295 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6296 address was recognized as an IP address, but was then not correctly decoded
6297 into binary, causing unexpected and incorrect effects when compared with
6304 1. Very minor rewording of debugging text in manualroute to say "list of
6305 hosts" instead of "hostlist".
6307 2. If verify=header_syntax was set, and a header line with an unqualified
6308 address (no domain) and a large number of spaces between the end of the
6309 name and the colon was received, the reception process suffered a buffer
6310 overflow, and (when I tested it) crashed. This was caused by some obsolete
6311 code that should have been removed. The fix is to remove it!
6313 3. When running in the test harness, delay a bit after writing a bounce
6314 message to get a bit more predictability in the log output.
6316 4. Added a call to search_tidyup() just before forking a reception process. In
6317 theory, someone could use a lookup in the expansion of smtp_accept_max_
6318 per_host which, without the tidyup, could leave open a database connection.
6320 5. Added the variables $recipient_data and $sender_data which get set from a
6321 lookup success in an ACL "recipients" or "senders" condition, or a router
6322 "senders" option, similar to $domain_data and $local_part_data.
6324 6. Moved the writing of debug_print from before to after the "senders" test
6327 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6328 problems for message scanning, either using a data ACL, or using
6329 local_scan() because the Received: header was not generated till after they
6330 were called (in order to set the time as the time of reception completion).
6331 I have revised the way this works. The header is now generated after the
6332 body is received, but before the ACL or local_scan() are called. After they
6333 are run, the timestamp in the header is updated.
6339 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6340 before starting a queue runner without re-exec. This happened only when
6341 deliver_drop_privilege was set or when the Exim user was set to root. The
6342 effect of the bug was that timeouts during subsequent deliveries caused
6343 crashes instead of being properly handled. The handler is now left at its
6344 default (and expected) setting.
6346 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6347 message, again when deliver_drop_privilege is set or Exim is run as root.
6348 The bug described in (1) was not present in this case, but the tidying up
6349 of the other signals was missing. I have made the two cases consistent.
6351 3. The ignore_target_hosts setting on a manualroute router was being ignored
6352 for hosts that were looked up using the /MX notation.
6354 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6357 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6358 operated on the sender address. After changing the $sender_address to <>
6359 for the sender address verify, Exim was re-instated it as the original
6360 (before rewriting) address, but remembering that it had rewritten it, so it
6361 wasn't rewriting it again. This bug also had the effect of breaking the
6362 sender address verification caching when the sender address was rewritten.
6364 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6365 This has been changed so that if the ip literal address matches
6366 ignore_target_hosts, the router declines.
6368 7. Added expansion conditions match_domain, match_address, and match_local_
6369 part (NOT match_host).
6371 8. The placeholder for the Received: header didn't have a length field set.
6373 9. Added code to Exim itself and to exim_lock to test for a specific race
6374 condition that could lead to file corruption when using MBX delivery. The
6375 issue is with the lockfile that is created in /tmp. If this file is removed
6376 after a process has opened it but before that process has acquired a lock,
6377 there is the potential for a second process to recreate the file and also
6378 acquire a lock. This could lead to two Exim processes writing to the file
6379 at the same time. The added code performs the same test as UW imapd; it
6380 checks after acquiring the lock that its file descriptor still refers to
6381 the same named file.
6383 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6384 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6387 11. Added the smtp_active_hostname option. If used, this will typically be made
6388 to depend on the incoming interface address. Because $interface_address is
6389 not set up until the daemon has forked a reception process, error responses
6390 that can happen earlier (such as "too many connections") no longer contain
6393 12. If an expansion in a condition on a "warn" statement fails because a lookup
6394 defers, the "warn" statement is abandoned, and the next ACL statement is
6395 processed. Previously this caused the whole ACL to be aborted.
6397 13. Added the iplsearch lookup type.
6399 14. Added ident_timeout as a log selector.
6401 15. Added tls_certificate_verified as a log selector.
6403 16. Added a global option tls_require_ciphers (compare the smtp transport
6404 option of the same name). This controls incoming TLS connections.
6406 17. I finally figured out how to make tls_require_ciphers do a similar thing
6407 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6408 before starting the TLS session.
6410 18. Tabs are now shown as \t in -bP output.
6412 19. If the log selector return_path_on_delivery was set, Exim crashed when
6413 bouncing a message because it had too many Received: header lines.
6415 20. If two routers both had headers_remove settings, and the first one included
6416 a superfluous trailing colon, the final name in the first list and the
6417 first name in the second list were incorrectly joined into one item (with a
6418 colon in the middle).
6424 1. Added -C and -D options to the exinext utility, mainly to make it easier
6425 to include in the automated testing, but these could be helpful when
6426 multiple configurations are in use.
6428 2. The exinext utility was not formatting the output nicely when there was
6429 an alternate port involved in the retry record key, nor when there was a
6430 message id as well (for retries that were specific to a specific message
6431 and a specific host). It was also confused by IPv6 addresses, because of
6432 the additional colons they contain. I have fixed the IPv4 problem, and
6433 patched it up to do a reasonable job for IPv6.
6435 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6436 delivery, the log line now contains "pipelined" if PIPELINING was used.
6438 4. An SMTP transport process used to panic and die if the bind() call to set
6439 an explicit outgoing interface failed. This has been changed; it is now
6440 treated in the same way as a connect() failure.
6442 5. A reference to $sender_host_name in the part of a conditional expansion
6443 that was being skipped was still causing a DNS lookup. This no longer
6446 6. The def: expansion condition was not recognizing references to header lines
6447 that used bh_ and bheader_.
6449 7. Added the _cache feature to named lists.
6451 8. The code for checking quota_filecount in the appendfile transport was
6452 allowing one more file than it should have been.
6454 9. For compatibility with Sendmail, the command line option
6462 and sets the incoming protocol and host name (for trusted callers). The
6463 host name and its colon can be omitted when only the protocol is to be set.
6464 Note the Exim already has two private options, -pd and -ps, that refer to
6465 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6466 "s", but I don't think that's a major issue.
6468 10. A number of refactoring changes to the code, none of which should affect
6471 (a) The number of logging options was getting close to filling up the
6472 32-bit word that was used as a bit map. I have split them into two classes:
6473 those that are passed in the argument to log_write(), and those that are
6474 only ever tested independently outside of that function. These are now in
6475 separate 32-bit words, so there is plenty of room for expansion again.
6476 There is no change in the user interface or the logging behaviour.
6478 (b) When building, for example, log lines, the code previously used a
6479 macro that called string_cat() twice, in order to add two strings. This is
6480 not really sufficiently general. Furthermore, there was one instance where
6481 it was actually wrong because one of the argument was used twice, and in
6482 one call a function was used. (As it happened, calling the function twice
6483 did not affect the overall behaviour.) The macro has been replaced by a
6484 function that can join an arbitrary number of extra strings onto a growing
6487 (c) The code for expansion conditions now uses a table and a binary chop
6488 instead of a serial search (which was left over from when there were very
6489 few conditions). Also, it now recognizes conditions like "pam" even when
6490 the relevant support is not compiled in: a suitably worded error message is
6491 given if an attempt is made to use such a condition.
6493 11. Added ${time_interval:xxxxx}.
6495 12. A bug was causing one of the ddress fields not to be passed back correctly
6496 from remote delivery subprocesses. The field in question was not being
6497 subsequently used, so this caused to problems in practice.
6499 13. Added new log selectors queue_time and deliver_time.
6501 14. Might have fixed a bug in maildirsizefile handling that threw up
6502 "unexpected character" debug warnings, and recalculated the data
6503 unnecessarily. In any case, I expanded the warning message to give more
6506 15. Added the message "Restricted characters in address" to the statements in
6507 the default ACL that block characters like @ and % in local parts.
6509 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6510 Three changes have been made:
6512 (a) There was a serious bug; a negative response to MAIL caused the whole
6513 recipient domain to be cached as invalid, thereby blocking all messages
6514 to all local parts at the same domain, from all senders. This bug has
6515 been fixed. The domain is no longer cached after a negative response to
6516 MAIL if the sender used is not empty.
6518 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6521 (c) A new callout option, "use_sender" has been added for people who want
6522 the modified behaviour.
6528 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6531 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6532 indeed breaks things for older releases.
6534 3. Added additional logging to the case where there is a problem reading data
6535 from a filter that is running in a subprocess using a pipe, in order to
6536 try to track down a specific problem.
6538 4. Testing facility fudge: when running in the test harness and attempting
6539 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6540 getting "No route to host". Convert this to a timeout.
6542 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6545 6. Some OS don't have socklen_t but use size_t instead. This affects the
6546 fifth argument of getsockopt() amongst other things. This is now
6547 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6548 can be set for individual OS. I have set it for SunOS5, OSF1, and
6549 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6550 some earlier ones do not.
6552 7. Change 4.30/15 was not doing the test caselessly.
6554 8. The standard form for an IPv6 address literal was being rejected by address
6555 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6556 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6557 this, as well as the form without the "IPv6" on the front (but only when
6558 address literals are enabled, of course).
6560 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6562 10. Exim crashed if a message with an empty sender address specified by -f
6563 encountered a router with an errors_to setting. This could be provoked only
6564 by a command such as
6568 where an empty string was supplied; "<>" did not hit this bug.
6570 11. Installed PCRE release 4.5.
6572 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6573 remained set. It is now erased.
6575 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6576 times from message ids (which are base 36 rather than the normal 62).
6578 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6579 were being counted as actual protocol errors, and logged if the log
6580 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6581 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6582 and DATA following a set of rejected RCPTs do not count as protocol errors.
6583 In other words, Exim assumes they were pipelined, though this may not
6584 actually be the case. Of course, in all cases the client gets an
6585 appropriate error code.
6587 15. If a lookup fails in an ACL condition, a message about the failure may
6588 be available; it is used if testing the ACL cannot continue, because most
6589 such messages specify what the cause of the deferral is. However, some
6590 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6591 that caused an old message to be retained and used if a later statement
6592 caused a defer, replacing the real cause of the deferral.
6594 16. If an IP address had so many PTR records that the DNS lookup buffer
6595 was not large enough to hold them, Exim could crash while trying to process
6596 the truncated data. It now detects and logs this case.
6598 17. Further to 4.21/58, another change has been made: if (and only if) the
6599 first line of a message (the first header line) ends with CRLF, a bare LF
6600 in a subsequent header line has a space inserted after it, so as not to
6601 terminate the header.
6603 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6604 unnecessarily, used atoi() instead of strtol(), and didn't check the
6605 termination when getting file sizes from file names by regex.
6607 19. Completely re-implemented the support for maildirsize files, in the light
6608 of a number of problems with the previous contributed implementation
6609 (4.30/29). In particular:
6611 . If the quota is zero, the maildirsize file is maintained, but no quota is
6614 . If the maildir directory does not exist, it is created before any attempt
6615 to write a maildirsize file.
6617 . The quota value in the file is just a cache; if the quota is changed in
6618 the transport, the new value overrides.
6620 . A regular expression is available for excluding directories from the
6623 20. The autoreply transport checks the characters in options that define the
6624 message's headers; it allows continued headers, but it was checking with
6625 isspace() after an embedded newline instead of explicitly looking for a
6628 21. If all the "regular" hosts to which an address was routed had passed their
6629 expiry times, and had not reached their retry times, the address was
6630 bounced, even if fallback hosts were defined. Now Exim should go on to try
6633 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6634 equivalent code in the SMTP transport. Some hosts send humungous responses
6635 to HELO/EHLO, more than 1024 it seems.
6637 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6638 gives compiler warnings in some environments. I've now done it "properly",
6641 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6642 (because of problems with the built-in one) was declared to return uschar *
6643 instead of char *, causing compiler failure.
6645 25. Fixed a file descriptor leak when processing alias/forward files.
6647 26. Fixed a minor format string issue in dbfn.c.
6649 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6651 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6652 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6655 29. When a sender address is verified, it is cached, to save repeating the test
6656 when there is more than one recipient in a message. However, when the
6657 verification involves a callout, it is possible for different callout
6658 options to be set for different recipients. It is too complicated to keep
6659 track of this in the cache, so now Exim always runs a verification when a
6660 callout is required, relying on the callout cache for the optimization.
6661 The overhead is duplication of the address routing, but this should not be
6664 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6665 to be verified with callout=postmaster, and the main callout worked but the
6666 postmaster check failed, the verification correctly failed. However, if a
6667 subsequent RCPT command asked for sender verification *without* the
6668 postmaster check, incorrect caching caused this verification also to fail,
6671 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6672 it was not caching the DNS options (qualify_single, search_parents) that
6673 were used when the lookup failed. A subsequent lookup with different
6674 options therefore always gave the same answer, though there were cases
6675 where it should not have. (Example: a "domains = !$mx_any" option on a
6676 dnslookup router: the "domains" option is always processed without any
6677 widening, but the router might have qualify_single set.) Now Exim uses the
6678 cached value only when the same options are set.
6680 32. Added John Jetmore's "exipick" utility to the distribution.
6682 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6683 than a timeout (e.g. a certificate is required, and is not provided), an
6684 Exim server now closes the connection immediately. Previously it waited for
6685 the client to close - but if the client is SSL, it seems that they each
6686 wait for each other, leading to a delay before one of them times out.
6688 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6689 maintained 0.8.x compatibility because I don't think many are using it, and
6690 it is clearly obsolete.
6692 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6695 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6696 client certificate was expired. A simple patch fixes this, though I don't
6697 understand the full logic of why the verify callback is called multiple
6700 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6701 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6702 which causes problems with some clients (such as the Certicom SSL Plus
6703 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6704 disables the coutermeasure allowing Eudora to connect."
6706 38. Exim was not checking that a write() to a log file succeeded. This could
6707 lead to Bad Things if a log got too big, in particular if it hit a file
6708 size limit. Exim now panics and dies if it cannot write to a log file, just
6709 as it does if it cannot open a log file.
6711 39. Modified OS/Makefile-Linux so that it now contains
6713 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6715 The two -D definitions ensure that Exim is compiled with large file
6716 support, which makes it possible to handle log files that are bigger than
6719 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6720 instance) a domain was checked against a named list that involved a lookup,
6721 causing $domain_data to be set, then another domain was checked against the
6722 same list, then the first domain was re-checked, the value of $domain_data
6723 after the final check could be wrong. In particular, if the second check
6724 failed, it could be set empty. This bug probably also applied to
6727 41. The strip_trailing_dot option was not being applied to the address given
6728 with the -f command-line option.
6730 42. The code for reading a message's header from the spool was incrementing
6731 $received_count, but never initializing it. This meant that the value was
6732 incorrect (doubled) while delivering a message in the same process in which
6733 it was received. In the most common configuration of Exim, this never
6734 happens - a fresh exec is done - but it can happen when
6735 deliver_drop_privilege is set.
6737 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6738 it now includes up to 150 characters of the unexpected data in the log
6741 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6742 and building data strings. The size of both of these buffers was 10 000
6743 bytes - far larger than anybody would *ever* want, thought I. Needless to
6744 say, somebody hit the limit. I have increased the maximum line length to
6745 20 000 and the maximum data length of concatenated lines to 100 000. I have
6746 also fixed two bugs, because there was no checking on these buffers. Tsk,
6747 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6748 buffer is too small.
6750 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6751 lsearch lookups. Now it does.
6753 46. When parsing a route_list item in a manualroute router, a fixed-length
6754 buffer was used for the list of hosts. I made this 1024 bytes long,
6755 thinking that nobody would ever have a list of hosts that long. Wrong.
6756 Somebody had a whole pile of complicated expansion conditions, and the
6757 string was silently truncated, leading to an expansion error. It turns out
6758 that it is easier to change to an unlimited length (owing to other changes
6759 that have happened since this code was originally written) than to build
6760 structure for giving a limitation error. The length of the item that
6761 expands into the list of hosts is now unlimited.
6763 47. The lsearch lookup could not handle data where the length of text line was
6764 more than 4095 characters. Such lines were truncated, leading to shortened
6765 data being returned. It should now handle lines of any length.
6767 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6768 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6771 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6772 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6773 no longer happens in comment lines. A list of replacements is now placed
6774 at the head of all of the source files, except those whose only change is
6775 to replace PERL_COMMAND in the very first #! line.
6777 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6778 messages on the queue, with a bottom-up merge sort, using code contributed
6779 by Michael Haardt. This should make operations like -bp somewhat faster on
6780 large queues. It won't affect queue runners, except when queue_run_in_order
6783 51. Installed eximstats 1.31 in the distribution.
6785 52. Added support for SRV lookups to the dnslookup router.
6787 53. If an ACL referred to $message_body or $message_body_end, the value was not
6788 reset for any messages that followed in the same SMTP session.
6790 54. The store-handling optimization for building very long strings was not
6791 differentiating between the different store pools. I don't think this
6792 actually made any difference in practice, but I've tidied it.
6794 55. While running the routers to verify a sender address, $sender_address
6795 was still set to the sender address. This is wrong, because when routing to
6796 send a bounce to the sender, it would be empty. Therefore, I have changed
6797 it so that, while verifying a sender address, $sender_address is set to <>.
6798 (There is no change to what happens when verifying a recipient address.)
6800 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6801 target A or AAAA records (if not already returned) without resetting the
6802 qualify_single or search_parents options of the DNS resolver. These are
6803 inappropriate in this case because the targets of MX and SRV records must
6804 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6805 target that, when qualified, matched something in the local domain. These
6806 two options are now turned off when doing these lookups.
6808 57. It seems that at least some releases of Reiserfs (which does not have the
6809 concept of a fixed number of inodes) returns zero and not -1 for the
6810 number of available inodes. This interacted badly with check_spool_inodes,
6811 which assumed that -1 was the "no such thing" setting. What I have done is
6812 to check that the total number of inodes is greater than zero before doing
6813 the test of how many are available.
6815 58. When a "warn" ACL statement has a log_message modifier, the message is
6816 remembered, and not repeated. This is to avoid a lot of repetition when a
6817 message has many recipients that cause the same warning to be written.
6818 However, Exim was preserving the list of already written lines for an
6819 entire SMTP session, which doesn't seem right. The memory is now reset if a
6820 new message is started.
6822 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6823 debugging output unless log_rewrite was also set.
6825 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6826 of (int)(handle) when we know that handle contains (void *)(-1).
6828 61. The Exim daemon panic-logs an error return when it closes the incoming
6829 connection. However "connection reset by peer" seems to be common, and
6830 isn't really an error worthy of noting specially, so that particular error
6833 62. When Exim is trying to find all the local interfaces, it used to panic and
6834 die if the ioctl to get the interface flags failed. However, it seems that
6835 on at least one OS (Solaris 9) it is possible to have an interface that is
6836 included in the list of interfaces, but for which you get a failure error
6837 for this call. This happens when the interface is not "plumbed" into a
6838 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6839 failure of the "get flags" call assumes that the interface is down.
6841 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6842 makes life easier for people who are doing arithmetic on fields extracted
6843 from dates, where you often get leading zeros that should not be
6844 interpreted as octal.
6846 64. Added qualify_domain to the redirect router, to override the global
6849 65. If a pathologically long header line contained very many addresses (the
6850 report of this problem mentioned 10 000) and each of them was rewritten,
6851 Exim could use up a very large amount of memory. (It kept on making new
6852 copies of the header line as it rewrote, and never released the old ones.)
6853 At the expense of a bit more processing, the header rewriting function has
6854 been changed so that it no longer eats memory in this way.
6856 66. The generation of the Received: header has been moved from the time that a
6857 message starts to be received, to the time that it finishes. The timestamp
6858 in the Received: header should now be very close to that of the <= log
6859 line. There are two side-effects of this change:
6861 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6862 logged header lines no longer include the local Received: line, because
6863 it has not yet been created. The same applies to a copy of the message
6864 that is returned to a non-SMTP sender when a message is rejected.
6866 (b) When a filter file is tested using -bf, no additional Received: header
6867 is added to the test message. After some thought, I decided that this
6870 This change does not affect the value of $received_for. It is still set
6871 after address rewriting, but before local_scan() is called.
6873 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6875 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6876 gave an unhelpful panic error message, and a defer error. I have managed to
6877 change this behaviour so that it now rejects any supplied certificate,
6878 which seems right, as the list of acceptable certificates is empty.
6880 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6881 gave an unhelpful defer error. I have not managed to make this reject any
6882 supplied certificates, but the error message it gives is "no certificate
6883 supplied", which is not helpful.
6885 70. exigrep's output now also includes lines that are not associated with any
6886 message, but which match the given pattern. Implemented by a patch from
6887 Martin Sluka, which also tidied up the Perl a bit.
6889 71. Recipient callout verification, like sender verification, was using <> in
6890 the MAIL FROM command. This isn't really the right thing, since the actual
6891 sender may affect whether the remote host accepts the recipient or not. I
6892 have changed it to use the actual sender in the callout; this means that
6893 the cache record is now keyed on a recipient/sender pair, not just the
6894 recipient address. There doesn't seem to be a real danger of callout loops,
6895 since a callout by the remote host to check the sender would use <>.
6896 [SEE ABOVE: changed after hitting problems.]
6898 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6899 temporary errors. However, in the case of such a code being given after
6900 the end of a data transmission (i.e. after ".") Exim was failing to write
6901 a retry record for the message. (Yes, there was some broken host that was
6902 actually sending 8xx at this point.)
6904 73. An unknown lookup type in a host list could cause Exim to panic-die when
6905 the list was checked. (An example that provoked this was putting <; in the
6906 middle of a list instead of at the start.) If this happened during a DATA
6907 ACL check, a -D file could be left lying around. This kind of configuration
6908 error no longer causes Exim to die; instead it causes a defer error. The
6909 incident is still logged to the main and panic logs.
6911 74. Buglet left over from Exim 3 conversion. The message "too many messages
6912 in one connection" was written to the rejectlog but not the mainlog, except
6913 when address rewriting (yes!) was being logged.
6915 75. Added write_rejectlog option.
6917 76. When a system filter was run not as root (that is, when system_filter_user
6918 was set), the values of the $n variables were not being returned to the
6919 main process; thus, they were not subsequently available in the $sn
6922 77. Added +return_path_on_delivery log selector.
6924 78. A connection timeout was being treated differently from recipients deferred
6925 when testing hosts_max_try with a message that was older than the host's
6926 retry timeout. (The host should not be counted, thus allowing all hosts to
6927 be tried at least once before bouncing.) This may have been the cause of an
6928 occasionally reported bug whereby a message would remain on the queue
6929 longer than the retry timeout, but would be bounced if a delivery was
6930 forced. I say "may" because I never totally pinned down the problem;
6931 setting up timeout/retry tests is difficult. See also the next item.
6933 79. The ultimate address timeout was not being applied to errors that involved
6934 a combination of host plus message (for example, a timeout on a MAIL
6935 command). When an address resolved to a number of possible hosts, and they
6936 were not all tried for each delivery (e.g. because of hosts_max_try), a
6937 message could remain on the queue longer than the retry timeout.
6939 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6942 81. Fixed an obscure SMTP outgoing bug which required at least the following
6943 conditions: (a) there was another message waiting for the same server;
6944 (b) the server returned 5xx to all RCPT commands in the first message so
6945 that the message was not completed; (c) the server dropped the connection
6946 or gave a negative response to the RSET that Exim sends to abort the
6947 transaction. The observed case was a dropped connection after DATA that had
6948 been sent in pipelining mode. That is, the server had advertised PIPELINING
6949 but was not implementing it correctly. The effect of the bug was incorrect
6950 behaviour, such as trying another host, and this could lead to a crash.
6956 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6957 and daemon.c were passed as pointers to ints; they should have been
6958 pointers to socklen_t variables (which are typically unsigned ints).
6960 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6963 3. Fixed a really odd bug that affected only the testing scheme; patching a
6964 certain fixed string in the binary changed the value of another string that
6965 happened to be identical to the end of the original first string.
6967 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6968 name", it returns that address as the IP address. On some operating
6969 systems (e.g. Solaris), it also passes back the IP address string as the
6970 "host name". However, on others (e.g. Linux), it passes back an empty
6971 string. Exim wasn't checking for this, and was changing the host name to an
6972 empty string, assuming it had been canonicalized.
6974 5. Although rare, it is permitted to have more than one PTR record for a given
6975 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6976 all the names associated with an address, because they do in Solaris.
6977 However, it seems that they do not in Linux for data that comes from the
6978 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6979 I found this out when I moved to a new Linux workstation and tried to run
6980 the Exim test suite.
6982 To get round this problem I have changed the code so that it now does its
6983 own call to the DNS to look up PTR records when searching for a host name.
6984 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6985 addresses that are only in /etc/hosts are still found.
6987 This behaviour is, however, controlled by an option called host_lookup_
6988 order, which defaults to "bydns:byaddr". If people want to use the other
6989 order, or indeed, just use one or the other means of lookup, they can
6990 specify it in this variable.
6992 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6993 some operating systems, this comes back from gethostbyaddr() as an empty
6994 string, and this is what Exim used to test for. However, it seems that in
6995 other systems, "." is yielded. Exim now tests for this case too.
6997 7. The values of check_spool_space and check_log_space are now held internally
6998 as a number of kilobytes instead of an absolute number of bytes. If a
6999 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7000 kilobyte. This means that much larger values can be stored.
7002 8. Exim monitor: an attempt to get the action menu when not actually pointing
7003 at a message produces an empty menu entitled "No message selected". This
7004 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7005 no entries in it ("Shell widget menu has zero width and/or height"). So I
7006 have added a single, blank menu entry in this case.
7008 9. Added ${quote_local_part.
7010 10. MIME decoding is now applied to the contents of Subject: header lines when
7013 11. Now that a reference to $sender_host_address automatically causes a reverse
7014 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7015 host lookup before query-style lookups in lists that might use this
7016 variable. This has therefore been abolished, and the "net-" prefix is no
7017 longer necessary for query-style lookups.
7019 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7020 have been a typo for LFLAGS, so it has been changed.
7022 13. The install script calls Exim with "-C /dev/null" in order to find the
7023 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7024 to be output. However, since Exim outputs its version number before the
7025 error, it didn't break the script. It just looked ugly. I fixed this by
7026 always allowing "-C /dev/null" if the caller is root.
7028 14. Ignore overlarge ACL variable number when reading spool file - insurance
7029 against a later release with more variables having written the file.
7031 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7032 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7033 this, as well as the form without the "IPv6" on the front.
7035 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7036 OS/Makefile-Darwin file.
7038 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7039 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7041 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7042 variable. However, some OS use ints for inodes. I've added cast to long int
7043 to get rid of the compiler warning.
7045 19. I had forgotten to lock out "/../" in configuration file names when
7046 ALT_CONFIG_PREFIX was set.
7048 20. Routers used for verification do not need to specify transports. However,
7049 if such a router generated a host list, and callout was configured, Exim
7050 crashed, because it could not find a port number from the (non-existent)
7051 transport. It now assumes port 25 in this circumstance.
7053 21. Added the -t option to exigrep.
7055 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7056 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7057 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7058 them is set, LOOKUP_LSEARCH is forced.
7060 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7062 24. Added sender and host information to the "rejected by local_scan()" log
7063 line; previously there was no indication of these.
7065 25. Added .include_if_exists.
7067 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7068 Linux. It turns out that not all file systems support this. Apparently some
7069 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7070 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7071 not supported on the file descriptor, is now ignored when Exim is trying to
7072 sync a directory. This applies only to Linux.
7074 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7076 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7077 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7080 29. Added support for maildirsize files from supplied patch (modified a bit).
7082 30. The use of :fail: followed by an empty string could lead Exim to respond to
7083 sender verification failures with (e.g.):
7085 550 Verification failed for <xxx>
7086 550 Sender verify failed
7088 where the first response line was missing the '-' that indicates it is not
7089 the final line of the response.
7091 31. The loop for finding the name of the user that called Exim had a hardwired
7092 limit of 10; it now uses the value of finduser_retries, which is used for
7093 all other user lookups.
7095 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7098 33. Exim was neglecting to zero errno before one call of strtol() when
7099 expanding a string and expecting an integer value. On some systems this
7100 resulted in spurious "integer overflow" errors. Also, it was casting the
7101 result into an int without checking.
7103 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7104 did not work. The code looks as if it has *never* worked, though it appears
7105 to have been documented since at least release 1.62. I have made it work.
7107 35. The "timeout_DNS" error in retry rules, also documented since at least
7108 1.62, also never worked. As it isn't clear exactly what this means, and
7109 clearly it isn't a major issue, I have abolished the feature by treating it
7110 as "timeout", and writing a warning to the main and panic logs.
7112 36. The display of retry rules for -brt wasn't always showing the error code
7115 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7116 timeout_connect_A, timeout_connect_MX.
7118 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7119 to the empty sender.
7121 39. The daemon was not analysing the content of -oX till after it had closed
7122 stderr and disconnected from the controlling terminal. This meant that any
7123 syntax errors were only noted on the panic log, and the return code from
7124 the command was 0. By re-arranging the code a little, I've made the
7125 decoding happen first, so such errors now appear on stderr, and the return
7126 code is 1. However, the actual setting up of the sockets still happens in
7127 the disconnected process, so errors there are still only recorded on the
7130 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7131 connections (as happens on some IP stacks) was logged at start up time as
7132 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7133 it from "IPv6 and IPv4", which means that two separate sockets are being
7136 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7137 says whether AF_INET or AF_INET6 was passed as an argument.
7139 42. Exiwhat output was messed up when time zones were included in log
7142 43. Exiwhat now gives more information about the daemon's listening ports,
7143 and whether -tls-on-connect was used.
7145 44. The "port" option of the smtp transport is now expanded.
7147 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7148 silently ignored. Now an error message is written to the main and panic
7151 46. There's a new ACL modifier called "logwrite" which writes to a log file
7152 as soon as it is encountered.
7154 47. Added $local_user_uid and $local_user_gid at routing time.
7156 48. Exim crashed when trying to verify a sender address that was being
7159 49. Exim was recognizing only a space character after ".include". It now also
7160 recognizes a tab character.
7162 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7163 extracting the relevant information from the specification. The man page no
7164 longer contains scrambled data for the -d option, and I've added a section
7165 at the front about calling Exim under different names.
7167 51. Added "extra_headers" argument to the "mail" command in filter files.
7169 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7172 53. Installed eximstats 1.29.
7174 54. Added transport_filter_timeout as a generic transport option.
7176 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7177 Cc: header lines. This was required by RFC 822, but it not required by RFC
7180 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7181 incoming messages that did not have them. Now it does so only if the
7182 message originates locally, that is, if there is no associated remote host
7183 address. When Resent- header lines are present, this applies to the Resent-
7184 lines rather than the non-Resent- lines.
7186 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7187 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7189 58. Messages for configuration errors now include the name of the main
7190 configuration file - useful now that there may be more than one file in a
7191 list (.included file names were always shown).
7193 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7194 for those rare installations that do not start the daemon as root or run it
7195 setuid root. I've cut out the call to initgroups() if the daemon is not
7198 60. The Exim user and group can now be bound into the binary as text strings
7199 that are looked up at the start of Exim's processing.
7201 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7203 62. Added $mailstore_basename variable.
7205 63. Installed patch to sieve.c from Michael Haardt.
7207 64. When Exim failed to open the panic log after failing to open the main log,
7208 the original message it was trying to log was written to stderr and debug
7209 output, but if they were not available (the usual case in production), it
7210 was lost. Now it is written to syslog before the two lines that record the
7211 failures to open the logs.
7213 65. Users' Exim filters run in subprocesses under the user's uid. It is
7214 possible for a "deliver" command or an alias in a "personal" command to
7215 provoke an address rewrite. If logging of address rewriting is configured,
7216 this fails because the process is not running as root or exim. There may be
7217 a better way of dealing with this, but for the moment (because 4.30 needs
7218 to be released), I have disabled address rewrite logging when running a
7219 filter in a non-root, non-exim process.
7225 1. The buildconfig auxiliary program wasn't quoting the value set for
7226 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7227 not defined. This bug was masked in 4.22 by the effect that was fixed in
7230 2. Some messages that were rejected after a message id was allocated were
7231 shown as "incomplete" by exigrep. It no longer does this for messages that
7232 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7234 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7235 have allow_domain_literals set, the ID did not get logged in the <= line.
7236 Domain literals are now always recognized in Message-ID: header lines.
7238 4. The first argument for a ${extract expansion item is the key name or field
7239 number. Leading and trailing spaces in this item were not being ignored,
7240 causing some misleading effects.
7242 5. When deliver_drop_privilege was set, single queue runner processes started
7243 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7244 same command in the process it spins off) were not dropping privilege.
7246 6. When the daemon running as "exim" started a queue runner, it always
7247 re-executed Exim in the spun-off process. This is a waste of effort when
7248 deliver_drop_privilege is set. The new process now just calls the
7249 queue-runner function directly.
7255 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7258 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7259 ignored. Though the use of -oP was forcing the writing of a pid file, it
7260 was always written to the default place.
7262 3. If the message "no IP address found for host xxxx" is generated during
7263 incoming verification, it is now followed by identification of the incoming
7264 connection (so you can more easily find what provoked it).
7266 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7268 5. Added some features to "harden" Exim a bit more against certain attacks:
7270 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7271 be put in Local/Makefile. This is like the never_users runtime option,
7272 but it cannot be overridden. The default setting is "root".
7274 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7275 prefix string with which any file named in a -C command line option
7278 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7279 is retained for -C and -D only if the caller of Exim is root. Without
7280 it, the exim user may also use -C and -D and retain privilege.
7282 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7283 command line option is disabled.
7285 6. Macro names set by the -D option must start with an upper case letter, just
7286 like macro names defined in the configuration file.
7288 7. Added "dereference=" facility to LDAP.
7290 8. Two instances of the typo "uknown" in the source files are fixed.
7292 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7293 the Configure-Makefile script screwed up while processing it.
7295 10. Incorporated PCRE 4.4.
7297 11. The SMTP synchronization check was not operating right at the start of an
7298 SMTP session. For example, it could not catch a HELO sent before the client
7299 waited for the greeting. There is now a check for outstanding input at the
7300 point when the greeting is written. Because of the duplex, asynchronous
7301 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7302 way, but not yet received, when the check is performed.
7304 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7305 on TCP/IP sockets, because this apparently causes some broken clients to
7308 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7309 unchanged) from the Cygwin maintainer.
7311 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7312 was supported instead of testing for mbx. Effectively a typo.
7314 15. The spa authenticator server code was not checking that the input it
7315 received was valid base64.
7317 16. The debug output line for the "set" modifier in ACLs was not showing the
7318 name of the variable that was being set.
7320 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7322 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7323 lookup on its own. Something else was needed to trigger the lookup. For
7324 example, a match in host_lookup or the need for a host name in a host list.
7325 Now, if $sender_host_name is referenced and the host name has not yet been
7326 looked up, a lookup is performed. If the lookup fails, the variable remains
7327 empty, and $host_lookup_failed is set to "1".
7329 19. Added "eqi" as a case-independent comparison operator.
7331 20. The saslauthd authentication condition could segfault if neither service
7332 nor realm was specified.
7334 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7335 error message that was logged was misleading, and incoming SMTP
7336 connections were dropped. The message is now more accurate, and temporary
7337 errors are given to SMTP connections.
7339 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7340 (except RSET) with a 421 error, until QUIT is received. However, it was
7341 failing to send a response to QUIT.
7343 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7344 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7345 ACL is now run after the helo_try_verify_hosts code.
7347 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7348 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7349 case-independent, so other case variants are also recognized). Apparently
7350 some systems use these upper case variants.
7352 25. If more than two messages were waiting for the same host, and a transport
7353 filter was specified for the transport, Exim sent two messages over the
7354 same TCP/IP connection, and then failed with "socket operation on non-
7355 socket" when it tried to send the third.
7357 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7359 27. The extern definition of crypt16() in expand.c was not being excluded when
7360 the OS had its own crypt16() function.
7362 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7363 as a preferred synonym for return_size_limit, both as an option and as an
7366 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7368 30. Changed the default configuration ACL to relax the local part checking rule
7369 for addresses that are not in any local domains. For these addresses,
7370 slashes and pipe symbols are allowed within local parts, but the sequence
7371 /../ is explicitly forbidden.
7373 31. SPA server authentication was not clearing the challenge buffer before
7376 32. log_message in a "warn" ACL statement was writing to the reject log as
7377 well as to the main log, which contradicts the documentation and doesn't
7378 seem right (because no rejection is happening). So I have stopped it.
7380 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7381 However, I am unable to do any testing of this.
7383 34. Fixed an infelicity in the appendfile transport. When checking directories
7384 for a mailbox, to see if any needed to be created, it was accidentally
7385 using path names with one or more superfluous leading slashes; tracing
7386 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7388 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7389 contents are added to the log line that is written for every discarded
7390 recipient. (Previously a log_message setting was ignored.)
7392 36. The ${quote: operator now quotes the string if it is empty.
7394 37. The install script runs exim in order to find its version number. If for
7395 some reason other than non-existence or emptiness, which it checks, it
7396 could not run './exim', it was installing it with an empty version number,
7397 i.e. as "exim-". This error state is now caught, and the installation is
7400 38. An argument was missing from the function that creates an error message
7401 when Exim fails to connect to the socket for saslauthd authentication.
7402 This could cause Exim to crash, or give a corrupted message.
7404 39. Added isip, isip4, and isip6 to ${if conditions.
7406 40. The ACL variables $acl_xx are now saved with the message, and can be
7407 accessed later in routers, transports, and filters.
7409 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7410 strings in the file are not string-expanded.
7412 42. If a MAIL command specified a SIZE value that was too large to fit into an
7413 int variable, the check against message_size_limit failed. Such values are
7414 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7415 day this will have to be increased, but I don't think I want to be around
7416 when emails are that large.
7423 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7424 iconv() is not standard in FreeBSD.
7426 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7427 IPv6 enabled. The observed symptom was a segmentation fault on return from
7428 the function os_common_find_running_interfaces() in src/os.c.
7430 3. In the check_special_case() function in daemon.c I had used "errno" as an
7431 argument name, which causes warnings on some systems. This was basically a
7432 typo, since it was named "eno" in the comments!
7434 4. The code that waits for the clock to tick (at a resolution of some fraction
7435 of a second) so as to ensure message-id uniqueness was always waiting for
7436 at least one whole tick, when it could have waited for less. [This is
7437 almost certainly not relevant at current processor speeds, where it is
7438 unlikely to ever wait at all. But we try to future-proof.]
7440 5. The function that sleeps for a time interval that includes fractions of a
7441 second contained a race. It did not block SIGALRM between setting the
7442 timer, and suspending (a couple of lines later). If the interval was short
7443 and the sigsuspend() was delayed until after it had expired, the suspension
7444 never ended. On busy systems this could lead to processes getting stuck for
7447 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7448 process, before it forks any delivery processes. The open lookup caching
7449 mechanism meant that the open file or database connection was passed into
7450 the delivery process. The problem was that delivery processes always tidy
7451 up cached lookup data. This could cause a problem for the next delivery
7452 process started by the queue runner, because the external queue runner
7453 process does not know about the closure. So the next delivery process
7454 still has data in the lookup cache. In the case of a file lookup, there was
7455 no problem because closing a file descriptor in a subprocess doesn't affect
7456 the parent. However, if the lookup was caching a connection to a database,
7457 the connection was closed, and the second delivery process was likely to
7458 see errors such as "PGSQL: query failed: server closed the connection
7459 unexpectedly". The problem has been fixed by closing all cached lookups
7460 in a queue runner before running a delivery process.
7462 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7463 seem to have the "const" qualifier which it has on other OS. I've
7466 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7467 *of the same type* (client or server) with the same public name that an
7468 error should be diagnosed.
7470 9. When Exim looked up a host name for an IP address, but failed to find the
7471 original IP address when looking up the host name (a safety check), it
7472 output the message "<ip address> does not match any IP for NULL", which was
7473 confusing, to say the least. The bug was that the host name should have
7474 appeared instead of "NULL".
7476 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7477 user that is built into the binary, and the -C or -D options is used, root
7478 privilege is dropped before the configuration file is read. In addition,
7479 logging is switched to stderr instead of the normal log files. If the
7480 configuration then re-defines the Exim user, the unprivileged environment
7481 is probably not what is expected, so Exim logs a panic warning message (but
7484 However, if deliver_drop_privilege is set, the unprivileged state may well
7485 be exactly what is intended, so the warning has been cut out in that case,
7486 and Exim is allowed to try to write to its normal log files.
7492 1. smtp_return_error_details was not giving details for temporary sender
7493 or receiver verification errors.
7495 2. Diagnose a configuration error if two authenticators have the same public
7498 3. Exim used not to create the message log file for a message until the first
7499 delivery attempt. This could be confusing when incoming messages were held
7500 for policy or load reasons. The message log file is now created at the time
7501 the message is received, and an initial "Received" line is written to it.
7503 4. The automatically generated man page for command line options had a minor
7504 bug that caused no ill effects; however, a more serious problem was that
7505 the procedure for building the man page automatically didn't always
7506 operate. Consequently, release 4.20 contains an out-of-date version. This
7507 shouldn't happen again.
7509 5. When building Exim with embedded Perl support, the script that builds the
7510 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7511 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7513 6. The freeze_tell option was not being used for messages that were frozen on
7514 arrival, either by an ACL or by local_scan().
7516 7. Added the smtp_incomplete_transaction log selector.
7518 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7519 was accepting AUTH without a new EHLO.
7521 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7522 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7525 10. Logging of TCP/IP connections (when configured) now happens in the main
7526 daemon process instead of the child process, so that the TCP/IP connection
7527 count is more accurate (but it can never be perfect).
7529 11. The use of "drop" in a nested ACL was not being handled correctly in the
7530 outer ACL. Now, if condition failure induced by the nested "drop" causes
7531 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7532 or "require"), the connection is dropped.
7534 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7535 that yield "discard" can now be used with an "accept" or a "discard" verb,
7536 but an error is generated for any others (because I can't see a useful way
7537 to define what should happen).
7539 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7540 are now processed in the same way as lines in the Exim configuration file.
7541 In particular, continuation lines are supported.
7543 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7545 15. Added -ti meaning -t -i.
7547 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7548 domains, and warn by logging if others are found.
7550 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7551 returned from GIFCONF when Exim is trying to find the list of interfaces on
7552 a host. The code in os.c has been modified to copy each ifreq to an aligned
7553 structure in all cases.
7555 Also, in some cases, the returned ifreq's were being copied to a 'struct
7556 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7557 means the last couple of bytes of an IPv6 address could be chopped if the
7558 ifreq contained only a normal sockaddr (14 bytes storage).
7560 18. Named domain lists were not supported in the hosts_treat_as_local option.
7561 An entry such as +xxxx was not recognized, and was treated as a literal
7564 19. Ensure that header lines added by a DATA ACL are included in the reject log
7565 if the ACL subsequently rejects the message.
7567 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7568 MD5 (which is deprecated).
7570 21. When testing a filter file using -bf, Exim was writing a message when it
7571 took the sender from a "From " line in the message, but it was not doing so
7572 when it took $return_path from a Return-Path: header line. It now does.
7574 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7575 with a valid header line field name (a series of printing characters
7576 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7578 23. Changed "disc" in the source to "disk" to conform to the documentation and
7579 the book and for uniformity.
7581 24. Ignore Sendmail's -Ooption=value command line item.
7583 25. When execve() failed while trying to run a command in a pipe transport,
7584 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7585 could be confused with a return value of 69 from the command itself. This
7586 has been changed to 127, the value the shell returns if it is asked to run
7587 a non-existent command. The wording for the related log line suggests a
7588 non-existent command as the problem.
7590 26. If received_header_text expands to an empty string, do not add a Received:
7591 header line to the message. (Well, it adds a token one on the spool, but
7592 marks it "old" so that it doesn't get used or transmitted.)
7594 27. Installed eximstats 1.28 (addition of -nt option).
7596 28. There was no check for failure on the call to getsockname() in the daemon
7597 code. This can fail if there is a shortage of resources on the system, with
7598 ENOMEM, for example. A temporary error is now given on failure.
7600 29. Contrary to the C standard, it seems that in some environments, the
7601 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7602 program. Exim now does this explicitly; it affects the formatting of
7603 timestamps using strftime().
7605 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7606 complaints. I've now initialized all the variables, to avoid this.
7608 32. Header lines added by a system filter were not being "seen" during
7609 transport-time rewrites.
7611 33. The info_callback() function passed to OpenSSL is set up with type void
7612 (*)(SSL *, int, int), as described somewhere. However, when calling the
7613 function (actually a macro) that sets it up, the type void(*)() is
7614 expected. I've put in a cast to prevent warnings from picky compilers.
7616 34. If a DNS black list lookup found a CNAME record, but there were no A
7617 records associated with the domain it pointed at, Exim crashed.
7619 35. If a DNS black list lookup returned more than one A record, Exim ignored
7620 all but the first. It now scans all returned addresses if a particular IP
7621 value is being sought. In this situation, the contents of the
7622 $dnslist_value variable are a list of all the addresses, separated by a
7625 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7626 to accept a host name and all its aliases if the forward lookup for any of
7627 them yielded the IP address of the incoming connection. Now it accepts only
7628 those names whose forward lookup yields the correct IP address. Any other
7629 names are discarded. This closes a loophole whereby a rogue DNS
7630 administrator could create reverse DNS records to break through a
7631 wildcarded host restriction in an ACL.
7633 37. If a user filter or a system filter that ran in a subprocess used any of
7634 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7635 the wrong values were passed to the pipe command ($thisaddress had the
7636 value of $0, $0 had the value of $1, etc). This bug was introduced by
7637 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7639 38. Improved the line breaking for long SMTP error messages from ACLs.
7640 Previously, if there was no break point between 40 and 75 characters, Exim
7641 left the rest of the message alone. Two changes have been made: (a) I've
7642 reduced the minimum length to 35 characters; (b) if it can't find a break
7643 point between 35 and 75 characters, it looks ahead and uses the first one
7644 that it finds. This may give the occasional overlong line, but at least the
7645 remaining text gets split now.
7647 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7648 file descriptors might be low, and that setting 1000 would always raise it.
7649 It turns out that in some environments, the limit is already over 1000 and
7650 that lowering it causes trouble. So now Exim takes care not to decrease it.
7652 40. When delivering a message, the value of $return_path is set to $sender_
7653 address at the start of routing (routers may change the value). By an
7654 oversight, this default was not being set up when an address was tested by
7655 -bt or -bv, which affected the outcome if any router or filter referred to
7658 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7659 log only when "message" or "log_message" are set. However, if one of the
7660 conditions was an address verification, or a call to a nested ACL, the
7661 messages generated by the underlying test were being passed through. This
7662 no longer happens. The underlying message is available in $acl_verify_
7663 message for both "message" and "log_message" expansions, so it can be
7664 passed through if needed.
7666 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7667 new expansion $bh_ to give the encoded byte string without charset
7668 translation. Translation happens only if iconv() is available; HAVE_ICONV
7669 indicates this at build time. HEADERS_CHARSET gives the charset to
7670 translate to; headers_charset can change it in the configuration, and
7671 "headers charset" can change it in an individual filter file.
7673 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7674 that creates RFC 2047 encoded "words" labels them as that charset instead
7675 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7676 expansion operator; (ii) when Exim creates a From: line for a local
7677 message; (iii) when a header line is rewritten to include a "phrase" part.
7679 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7680 buggy, causing it to skip the first lines of messages whose message ID
7681 ended in 'D'. This would not have bitten before Exim release 4.14, because
7682 message IDs were unlikely to end in 'D' before then. The effect was to have
7683 incorrect size information for certain domains.
7685 45. #include "config.h" was missing at the start of the crypt16.c module. This
7686 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7689 46. If there was a timeout during a "random" callout check, Exim treated it as
7690 a failure of the random address, and carried on sending RSET and the real
7691 address. If the delay was just some slowness somewhere, the response to the
7692 original RCPT would be taken as a response to RSET and so on, causing
7693 mayhem of various kinds.
7695 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7696 when I implemented it. It didn't allow for the fact that some option values
7697 may legitimately be negative (e.g. size_addition), and it didn't even do
7698 the right test for positive values.
7700 48. Domain names in DNS records are case-independent. Exim always looks them up
7701 in lower case. Some resolvers return domain names in exactly the case they
7702 appear in the zone file, that is, they may contain uppercase letters. Not
7703 all resolvers do this - some return always lower case. Exim was treating a
7704 change of case by a resolver as a change of domain, similar to a widening
7705 of a domain abbreviation. This triggered its re-routing code and so it was
7706 trying to route what was effectively the same domain again. This normally
7707 caused routing to fail (because the router wouldn't handle the domain
7708 twice). Now Exim checks for this case specially, and just changes the
7709 casing of the domain that it ultimately uses when it transmits the message
7712 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7715 50. If a filter generated a file delivery with a non-absolute name (possible if
7716 no home directory exists for the router), the forbid_file option was not
7719 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7720 the existing equality matching.
7722 52. Exim was using ints instead of ino_t variables in some places where it was
7723 dealing with inode numbers.
7725 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7726 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7727 called TMPDIR, and if it finds it is different, it changes its value.
7729 54. The smtp_printf() function is now made available to local_scan() so
7730 additional output lines can be written before returning. There is also an
7731 smtp_fflush() function to enable the detection of a dropped connection.
7732 The variables smtp_input and smtp_batched_input are exported to
7735 55. Changed the default runtime configuration: the message "Unknown user"
7736 has been removed from the ACL, and instead placed on the localuser router,
7737 using the cannot_route_message feature. This means that any verification
7738 failures that generate their own messages won't get overridden. Similarly,
7739 the "Unrouteable address" message that was in the ACL for unverifiable
7740 relay addresses has also been removed.
7742 56. Added hosts_avoid_esmtp to the smtp transport.
7744 57. The exicyclog script was not checking for the esoteric option
7745 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7746 will work only if exicyclog is run under the appropriate euid.
7748 58. Following a discussion on the list, the rules by which Exim recognises line
7749 endings on incoming messages have been changed. The -dropcr and drop_cr
7750 options are now no-ops, retained only for backwards compatibility. The
7751 following line terminators are recognized: LF CRLF CR. However, special
7752 processing applies to CR:
7754 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7755 nor a local message in the state where . is a terminator.
7757 (ii) If a bare CR is encountered in a header line, an extra space is added
7758 after the line terminator so as not to end the header. The reasoning
7759 behind this is that bare CRs in header lines are most likely either
7760 to be mistakes, or people trying to play silly games.
7762 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7763 was being incorrectly given as 18 bytes larger than it should have been.
7764 This is a VOB (very old bug).
7766 60. This may never have affected anything current, but just in case it has:
7767 When the local host is found other than at the start of a list of hosts,
7768 the local host, those with the same MX, and any that follow, are discarded.
7769 When the list in question was part of a longer list of hosts, the following
7770 hosts (not currently being processed) were also being discarded. This no
7771 longer happens. I'm not sure if this situation could ever has previously
7774 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7777 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7778 header. This is something that is recommended in a new Internet Draft, and
7779 is something that is documented as being done by Sendmail. There are two
7780 possible values. For messages generated by the autoreply transport, Exim
7783 Auto-Submitted: auto-replied
7785 whereas for all other generated messages (e.g. bounces) it adds
7787 Auto-Submitted: auto-generated
7789 63. The "personal" condition in filters now includes a test for the
7790 Auto-Submitted: header. If it contains the string "auto-" the message it
7791 not considered personal.
7793 64. Added rcpt_include_affixes as a generic transport option.
7795 65. Added queue_only_override (default true).
7797 66. Added the syslog_duplication option.
7799 67. If what should have been the first header line of a message consisted of
7800 a space followed by a colon, Exim was mis-interpreting it as a header line.
7801 It isn't of course - it is syntactically invalid and should therefore be
7802 treated as the start of the message body. The misbehaviour could have
7803 caused a number of strange effects, including loss of data in subsequent
7804 header lines, and spool format errors.
7806 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7807 client host had authenticated. This control can now be exercised by an ACL
7808 for more flexibility.
7810 69. By default, callouts do not happen when testing with -bh. There is now a
7811 variant, -bhc, which does actually run the callout code, including
7812 consulting and updating the callout cache.
7814 70. Added support for saslauthd authentication, courtesy of Alexander
7817 71. If statvfs() failed on the spool or log directories while checking their
7818 size for availability, Exim confusingly gave the error "space shortage".
7819 Furthermore, in debugging mode it crashed with a floating point exception.
7820 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7821 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7822 serious problem, Exim now writes to the main and panic logs when this
7823 happens, with details of the failure. It then refuses to accept the
7824 incoming message, giving the message "spool directory problem" or "log
7825 directory problem" with a 421 code for SMTP messages.
7827 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7828 0, 1, and 2 exist, because some OS complain for execs without them (see
7829 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7830 descriptors. However, the code omitted to check that the open succeeded,
7831 causing mysterious errors if for some reason the permissions on /dev/null
7832 got screwed. Now Exim writes a message to the main and panic logs, and
7833 bombs out if it can't open /dev/null.
7835 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7836 interact so that it is all more flexible. It is supposed to remain
7837 backwards compatible. Also added extra_local_interfaces.
7839 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7840 to bomb out with an assertion failure - to the client this appears as a
7841 connection drop. This problem occurs in the part of the code that was taken
7842 from the Samba project. Fortunately, the assertion is in a very simple
7843 function, so I have fixed this by reproducing the function inline in the
7844 one place where it is called, and arranging for authentication to fail
7845 instead of killing the process with assert().
7847 75. The SPA client code was not working when the server requested OEM rather
7848 than Unicode encoding.
7850 76. Added code to make require_files with a specific uid setting more usable in
7851 the case where statting the file as root fails - usually a non-root-mounted
7852 NFS file system. When this happens and the failure is EACCES, Exim now
7853 forks a subprocess and does the per-uid checking as the relevant uid.
7855 77. Added process_log_path.
7857 78. If log_file_path was not explicitly set, a setting of check_log_space or
7858 check_log_inodes was ignored.
7860 79. If a space check for the spool or log partitions fails, the incident is now
7861 logged. Of course, in the latter case the data may get lost...
7863 80. Added the %p formatting code to string_format() so that it can be used to
7864 print addresses in debug_print(). Adjusted all the address printing in the
7865 debugging in store.c to use %p rather than %d.
7867 81. There was a concern that a line of code in smtp_in.c could overflow a
7868 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7869 initially expressed, the concern was not well-founded, because trailing
7870 spaces are removed early. However, if the trailing spaces were followed by
7871 a NULL, they did not get removed, so the overflow was possible. Two fixes
7874 (a) I re-wrote the offending code in a cleaner fashion.
7875 (b) If an incoming SMTP command contains a NULL character, it is rejected
7878 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7879 runs initgroups(), so that if the Exim user is in any additional groups,
7880 they will be used during message reception.
7886 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob