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 Fix "spam" ACL condition. Previously, tainted values for the "name"
73 argument resulted in a trap. There is no reason to disallow such; this
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.
88 JH/01 Avoid costly startup code when not strictly needed. This reduces time
89 for some exim process initialisations. It does mean that the logging
90 of TLS configuration problems is only done for the daemon startup.
92 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
94 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
95 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
98 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
99 DKIM signing is being done. Previously a transport_filter would always
100 disable CHUNKING, falling back to traditional DATA.
102 JH/05 Regard command-line receipients as tainted.
104 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
106 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
107 PAM library frees one of the arguments given to it, despite the
108 documentation. Therefore a plain malloc must be used.
110 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
111 on-stack buffers were used, resulting in a taint trap when DSN information
112 copied from a received message was written into the buffer.
114 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
115 the ordering of its ARC headers. This caused a crash.
117 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
118 a new record was being constructed with information from the peer, a trap
121 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
122 installation would get error messages from DMARC verify, when it hit the
123 nonexistent file indicated by the default. Distros wanting DMARC enabled
124 should both provide the file and set the option.
125 Also enforce no DMARC verification for command-line sourced messages.
127 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
128 could, depending on the platform, hang at the STARTTLS response.
130 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
131 message on a connection. Previously if one message had ARC headers and
132 the following one did not, a crash could result when adding an
133 Authentication-Results: header.
135 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
136 Exim main code and Exim-related utities. The introduction of taint
137 tracking also did many adjustments to string handling. Since then, eximon
138 frequently terminated with an assert failure.
140 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
141 check for 452 responses. This slightly helps the inefficieny of doing
142 a large alias-expansion into a recipient-limited target. The max_rcpt
143 transport option still applies (and at the current default, will override
144 the new feature). The check is done for either cause of synch, and forces
145 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
146 connection. The new facility is not tunable at this time.
148 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
149 library live data was being used, so the results became garbage. Make
150 copies while it is still usable.
152 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
153 on delivery deferred (==) and failed (**) lines (if a delivery was
154 attemtped). Previously it was only on completion (=>) lines.
156 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
157 for the expansion of the server_scram_iter and server_scram_salt options.
159 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
160 are now specifically given a NO_DATA response without hitting the system
161 resolver. The library goes on to do the now-standard TXT lookup.
162 Use of dnsdb lookups is not affected.
164 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
165 only retrieve the errormessage once. Previously two calls to dlerror()
166 were used, and the second one (for mainlog/paniclog) retrieved null
169 JH/20 Taint checking: disallow use of tainted data for
170 - the appendfile transport file and directory options
171 - the pipe transport command
172 - the autoreply transport file, log and once options
173 - file names used by the redirect router (including filter files)
175 - paths used by single-key lookups
176 Previously this was permitted.
178 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
179 adjusted the size of a major service buffer; this failed because the
180 buffer was in use at the time. Change to a compile-time increase in the
181 buffer size, when this authenticator is compiled into exim.
183 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
184 previous fast-mode was untenable in the face of glibs using mmap to
185 support larger malloc requests.
187 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
188 New values supported, if defined on system where compiled:
189 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
190 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
192 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
193 running a limited number of proceses in parallel, a benefit is gained. The
194 amount varies with the platform hardware and load. The use of the option
195 queue_run_in_order means we cannot do this, as ordering becomes
198 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
199 had introduced a string-copy (for ensuring NUL-termination) which was not
200 appropriate for that case, which can include embedded NUL bytes in the
201 block of data. Investigation showed the copy to actually be needless, the
202 data being length-specified.
204 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
205 done during a receiving connection, and both used TLS, global info was
206 used rather than per-connection info for tracking the state of data
207 queued for transmission. This could result in a connection hang.
209 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
210 Previously, when delivering serveral messages down a single connection
211 only the first would provide a SIZE. This was due to the size information
212 not being properly tracked.
214 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
215 TAI (at 37 seconds currently), pretend to be in UTC for time-related
216 expansion and logging. Previously, spurious values such as a future
217 minute could be seen.
219 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
220 it could crash from a null-deref. This could also affect the
221 ${addresses: } operator and ${readsock } item.
223 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
224 message following a mime one, the variable was not reset.
226 JH/30 When an pipelined-connect fails at the first response, assume incorrect
227 cached capability (perhaps the peer reneged?) and immediately retry in
230 JH/31 Fix spurious detection of timeout while writing to transport filter.
232 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
233 an attempt to copy the string was made before checking it.
235 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
236 taint of the lookup key was maintained; we now regard the presence in the
237 filesystem as sufficient validation.
239 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
240 argument is supplied.
242 JH/35 The dsearch lookup now requires that the directory is an absolute path.
243 Previously this was not checked, and nonempty relative paths made an
244 access under Exim's current working directory.
246 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
247 Previously no event was raised.
249 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
250 parameter supplied by the sender MAIL FROM command. Previously it was
251 ignored, and only the check_spool_space option value for the required
254 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
255 the size of the signing public-key. Previously it was instead giving
256 the size of the signature hash.
258 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
259 the default. See the (new) dkim_verify_min_keysizes option.
261 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
262 and an ACL use a lookup for checking either the local_part or domain,
263 stale data could be accessed. Ensure that variable references are
264 dropped between messages.
266 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
267 by the client was not checked as pointing within response data before
268 being used. A malicious client could thus cause an out-of-bounds read and
269 possibly gain authentication. Fix by adding the check.
271 JH/42 Internationalisation: change the default for downconversion in the smtp
272 transport to be "if needed". Previously it was "as previously set" for
273 the message, which usually meant "if needed" for message-submission but
274 "no" for everything else. However, MTAs have been seen using SMTPUTF8
275 even when the envelope addresses did not need it, resulting in forwarding
276 failures to non-supporting MTAs. A downconvert in such cases will be
277 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
278 transport. The change does mean that addresses needing conversion will
279 be converted when previously a delivery failure would occur.
281 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
282 response was received it would be used unchecked in a fail-DSN, violating
283 standards on line-length limits. Truncate if needed.
285 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
286 doesn't work. (Additionally add a new main config option to configure the
293 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
294 side, to decode a TLS 1.3 packet capture.
296 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
297 Previously the default library behaviour applied, sending two, each in
300 JH/03 Debug output for ACL now gives the config file name and line number for
303 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
305 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
307 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
308 buffer overrun for (non-chunking) other transports.
310 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
311 TLS1.3, means that a server rejecting a client certificate is not visible
312 to the client until the first read of encrypted data (typically the
313 response to EHLO). Add detection for that case and treat it as a failed
314 TLS connection attempt, so that the normal retry-in-clear can work (if
315 suitably configured).
317 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
318 and/or domain. Found and fixed by Jason Betts.
320 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
321 configuration). If a CNAME target was not a wellformed name pattern, a
324 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
325 the OS reports them interleaved with other addresses.
327 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
328 used both for input and for a verify callout, both encrypted, SMTP
329 responses being sent by the server could be lost. This resulted in
330 dropped connections and sometimes bounces generated by a peer sending
333 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
334 string. Previously it was possible to cause undefined behaviour in a
335 library routine (usually a crash). Found by "zerons".
337 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
340 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
341 API was removed, so update to use the newer ones.
343 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
344 any timeout set, is taking a long time. Previously we would hang on to a
345 rotated logfile "forever" if the input was arriving with long gaps
346 (a previous attempt to fix addressed lack, for a long time, of initial
349 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
350 shared (NFS) environment. The length of the tempfile name is now
351 4 + 16 ("hdr.$message_exim_id") which might break on file
352 systems which restrict the file name length to lower values.
355 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
356 shared (NFS) environment.
358 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
359 did for all versions <4.90). Notably -M, -m, --invert, -I may be
362 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
363 on some platforms for bit 31.
365 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
366 to changes apparently associated with TLS1.3 handling some of the APIs
367 previously used were either nonfunctional or inappropriate. Strings
368 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
369 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
370 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
371 This affects log line X= elements, the $tls_{in,out}_cipher variables,
372 and the use of specific cipher names in the encrypted= ACL condition.
374 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
376 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
377 verification result was not updated unless hosts_require_ocsp applied.
379 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
380 queue_list_requires_admin set to false, non-admin users were denied the
383 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
384 directory-of-certs mode. Previously they were advertised despite the
387 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
388 A single TCP connection by a client will now hold a TLS connection open
389 for multiple message deliveries, by default. Previously the default was to
392 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
393 default. If built with the facility, DANE will be used. The facility
394 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
396 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
397 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
398 must be defined and you must still, unless you define DISABLE_TLS, manage
399 the the include-dir and library-file requirements that go with that
400 choice. Non-TLS builds are still supported.
402 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
405 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
406 default. If the platform supports and has the facility enabled, it will
407 be requested on all coneections.
409 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
410 controlled by the build-time option SUPPORT_PIPE_CONNECT.
412 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
414 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
415 success-DSN messages. Previously the From: header was always the default
416 one for these; the option was ignored.
418 JH/28 Fix the timeout on smtp response to apply to the whole response.
419 Previously it was reset for every read, so a teergrubing peer sending
420 single bytes within the time limit could extend the connection for a
421 long time. Credit to Qualsys Security Advisory Team for the discovery.
423 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
424 delivery address, which leaked information of the results of local
425 forwarding. Change to the original envelope recipient address, per
428 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
429 requested. Previously not bounce was generated and a log entry of
430 error ignored was made.
432 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
434 JH/32 Introduce a general tainting mechanism for values read from the input
435 channel, and values derived from them. Refuse to expand any tainted
436 values, to catch one form of exploit.
438 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
439 was unused and the unexpanded text used for the test. Found and
440 fixed by Ruben Jenster.
442 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
443 an attempt to use a TLS library read routine dereffed a nul pointer,
446 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
447 them in our smtp response.
449 JH/36 Have the general router option retry_use_local_part default to true when
450 any of the restrictive preconditions are set (to anything). Previously it
451 was only for check_local user. The change removes one item of manual
452 configuration which is required for proper retries when a remote router
453 handles a subset of addresses for a domain.
455 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
456 link count into consideration.
458 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
459 caused the extension of big_buffer, the following lines were ignored.
461 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
462 accordance with RFC 2308. Previously there was no expiry, so a longlived
463 receive process (eg. due to ACL delays) versus a short SOA value could
466 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
468 JH/39 Promote DMARC support to mainline.
470 JH/40 Bug 2452: Add a References: header to DSNs.
472 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
473 parameters. The relevant library call is documented as "Deprecated: This
474 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
475 3.6.0, DH parameters are negotiated following RFC7919."
477 HS/06 Change the default of dnssec_request_domains to "*"
479 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
480 carried on and emitted a BDAT command, even when PIPELINING was not
483 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
484 buffer was used for the filename, resulting in a trap when tainted
485 arguments (eg. $domain) were used.
487 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
488 recommended to avoid a possible server-load attack. The feature can be
489 re-enabled via the openssl_options main cofiguration option.
491 JH/45 local_scan API: documented the current smtp_printf() call. This changed
492 for version 4.90 - adding a "more data" boolean to the arguments.
493 Bumped the ABI version number also, this having been missed previously;
494 release versions 4.90 to 4.92.3 inclusive were effectively broken in
495 respect of usage of smtp_printf() by either local_scan code or libraries
496 accessed via the ${dlfunc } expansion item. Both will need coding
497 adjustment for any calls to smtp_printf() to match the new function
498 signature; a FALSE value for the new argument is always safe.
500 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
501 the file-offset (which the Linux syscall does, and exim expects); this
502 resulted in an indefinite loop.
504 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
505 to do ARC verification. The Authentication-Results: header line added
506 by the configuration then had no ARC item.
512 JH/01 Remove code calling the customisable local_scan function, unless a new
513 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
515 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
516 non-signal-safe functions being used.
518 JH/03 Bug 2269: When presented with a received message having a stupidly large
519 number of DKIM-Signature headers, disable DKIM verification to avoid
520 a resource-consumption attack. The limit is set at twenty.
522 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
523 report of oldest_pass in ${authres } in consequence, and separate out
524 some descriptions of reasons for verification fail.
526 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
527 files in the spool were present and unlocked. A queue-runner could spot
528 them, resulting in a duplicate delivery. Fix that by doing the unlock
529 after the unlink. Investigation by Tim Stewart. Take the opportunity to
530 add more error-checking on spoolfile handling while that code is being
533 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
534 No known attacks, no CVE, this is defensive hardening.
536 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
537 a queue-runner could start a delivery while other operations were ongoing.
538 Cutthrough delivery was a common victim, resulting in duplicate delivery.
539 Found and investigated by Tim Stewart. Fix by using the open message data
540 file handle rather than opening another, and not locally closing it (which
541 releases a lock) for that case, while creating the temporary .eml format
542 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
544 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
545 $sender_verify_failure/$recipient_verify_failure to "random".
547 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
550 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
551 Previously this would segfault.
553 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
556 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
557 like zero, since the resolver should be doing this for us, But we need one
558 as a CNAME but no MX presence gets the CNAME returned; we need to check
559 that doesn't point to an MX to declare it "no MX returned" rather than
560 "error, loop". A new main option is added so the older capability of
561 following some limited number of chain links is maintained.
563 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
565 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
566 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
567 now-deprecated earlier definitions used only specified the range up to TLS
568 1.2 (in the older-version library docs).
570 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
572 JH/15 Rework TLS client-side context management. Stop using a global, and
573 explicitly pass a context around. This enables future use of TLS for
574 connections to service-daemons (eg. malware scanning) while a client smtp
575 connection is using TLS; with cutthrough connections this is quite likely.
577 JH/16 Fix ARC verification to do AS checks in reverse order.
579 JH/17 Support a "tls" option on the ${readsocket } expansion item.
581 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
582 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
583 Previously the "utf8" would be re-prepended for every additional message.
585 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
586 Previously thery were accepted, resulting in issues when attempting to
587 forward messages to a non-supporting MTA.
589 PP/02 Let -n work with printing macros too, not just options.
591 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
592 one parent address was copied, and bogus data was used at delivery-logging
593 time. Either a crash (after delivery) or bogus log data could result.
594 Discovery and analysis by Tim Stewart.
596 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
597 Previously if the string ended mid-character, we did not insert the
598 promised '?' replacement.
600 PP/04 Documentation: current string operators work on bytes, not codepoints.
602 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
603 should pack well giving a smaller memory footprint so better caching and
604 therefore performance. Group the declarations where this can't be done so
605 that the byte-sized flag variables are not interspersed among pointer
606 variables, giving a better chance of good packing by the compiler.
608 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
609 non-null, to avoid issues with sites running BATV. Previously reports were
610 sent with an empty envelope sender so looked like bounces.
612 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
613 The ignore_error flag wasn't being returned from the filter subprocess so
614 was not set for later routers. Investigation and fix by Matthias Kurz.
616 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
617 and a msg:complete for the whole, when a message is manually removed using
618 -Mrm. Developement by Matthias Kurz, hacked on by JH.
620 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
621 a "Gnu special" function, asprintf() in the DB utility binary builds; I
622 hope that is portable enough.
624 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
625 requiring a known-CA anchor certificate; make it now rely entirely on the
626 TLSA as an anchor. Checking the name on the leaf cert against the name
627 on the A-record for the host is still done for TA (but not for EE mode).
629 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
630 included in delivery lines for non-proxied connections, when compiled with
631 SUPPORT_SOCKS and running with proxy logging enabled.
633 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
634 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
635 move the existing event to fire before the normal logging of message
636 failure so that custom logging is bracketed by normal logging.
638 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
639 msg:fail:internal event. Developement by Matthias Kurz.
641 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
642 far too small for todays use of crypto signatures stored there. Go all
643 the way to the max DNS message size of 64kB, even though this might be
644 overmuch for IOT constrained device use.
646 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
647 copy a string over itself. The library routine is documented as not
648 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
650 JH/32 For main options check_spool_space and check_inode_space, where the
651 platform supports 64b integers, support more than the previous 2^31 kB
652 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
653 the previous G, M, k.
655 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
656 $authenticated_fail_id variable on authentication failure. Previously
659 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
660 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
661 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
662 GNUTLS_SEC_PARAM_MEDIUM.
664 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
665 side. Previously we would continue as if no SNI had been received.
667 JH/36 Harden the handling of string-lists. When a list consisted of a sole
668 "<" character, which should be a list-separator specification, we walked
669 off past the nul-terimation.
671 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
672 causes) even when the retry time is not yet met. Previously they were
673 not, meaning that when (say) an account was over-quota and temp-rejecting,
674 and multiple senders' messages were queued, only one sender would get
675 notified on each configured delay_warning cycle.
677 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
679 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
680 cipher-suites, an error can be left on the stack even for a succeeding
681 accept; this results in impossible error messages when a later operation
684 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
685 return error codes indicating retry. Under TLS1.3 this becomes required.
687 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
688 it only wrote the new authenticators, resulting in a lack of tracking of
689 peer changes of ESMTP extensions until the next cache flush.
691 JH/41 Fix the loop reading a message header line to check for integer overflow,
692 and more-often against header_maxsize. Previously a crafted message could
693 induce a crash of the recive process; now the message is cleanly rejected.
695 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
696 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
702 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
703 When redis_servers is set to a list of > 1 element, and the Redis servers
704 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
705 case of MOVED into a DEFER case instead, thus moving the query onto the
706 next server in the list. For a cluster of N elements, all N servers must
707 be defined in redis_servers.
709 GF/02 Catch and remove uninitialized value warning in exiqsumm
710 Check for existence of @ARGV before looking at $ARGV[0]
712 JH/01 Replace the store_release() internal interface with store_newblock(),
713 which internalises the check required to safely use the old one, plus
714 the allocate and data copy operations duplicated in both (!) of the
715 extant use locations.
717 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
718 modifier. This matches the restriction on the commandline.
720 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
721 Previously only the last row was returned.
723 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
724 we assumed that tags in the header were well-formed, and parsed the
725 element content after inspecting only the first char of the tag.
726 Assumptions at that stage could crash the receive process on malformed
729 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
730 While running the DKIM ACL we operate on the Permanent memory pool so that
731 variables created with "set" persist to the DATA ACL. Also (at any time)
732 DNS lookups that fail create cache records using the Permanent pool. But
733 expansions release any allocations made on the current pool - so a dnsdb
734 lookup expansion done in the DKIM ACL releases the memory used for the
735 DNS negative-cache, and bad things result. Solution is to switch to the
736 Main pool for expansions.
737 While we're in that code, add checks on the DNS cache during store_reset,
738 active in the testsuite.
739 Problem spotted, and debugging aided, by Wolfgang Breyha.
741 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
742 When none of the hosts presented to a transport match an already-open
743 connection, close it and proceed with the list. Previously we would
744 queue the message. Spotted by Lena with Yahoo, probably involving
747 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
748 Previously a spurious "250 OK id=" response was appended to the proper
751 JH/08 The "support for" informational output now, which built with Content
752 Scanning support, has a line for the malware scanner interfaces compiled
753 in. Interface can be individually included or not at build time.
755 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
756 by the template makefile "src/EDITME". The "STREAM" support for an older
757 ClamAV interface method is removed.
759 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
760 rows affected is given instead).
762 JH/11 The runtime Berkeley DB library version is now additionally output by
763 "exim -d -bV". Previously only the compile-time version was shown.
765 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
766 SMTP connection. Previously, when one had more receipients than the
767 first, an abortive onward connection was made. Move to full support for
768 multiple onward connections in sequence, handling cutthrough connection
769 for all multi-message initiating connections.
771 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
772 routers. Previously, a multi-recipient message would fail to match the
773 onward-connection opened for the first recipient, and cause its closure.
775 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
776 a timeout on read on a GnuTLS initiating connection, resulting in the
777 initiating connection being dropped. This mattered most when the callout
778 was marked defer_ok. Fix to keep the two timeout-detection methods
781 JH/15 Relax results from ACL control request to enable cutthrough, in
782 unsupported situations, from error to silently (except under debug)
783 ignoring. This covers use with PRDR, frozen messages, queue-only and
786 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
788 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
789 metadata, resulting in a crash in free().
791 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
792 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
793 Broken also in d185889f4, with init system revamp.
795 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
796 Previously we abruptly closed the connection after reading a malware-
797 found indication; now we go on to read the "scan ok" response line,
800 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
801 ACL. Previously, a crash would result.
803 JH/19 Speed up macro lookups during configuration file read, by skipping non-
804 macro text after a replacement (previously it was only once per line) and
805 by skipping builtin macros when searching for an uppercase lead character.
807 JH/20 DANE support moved from Experimental to mainline. The Makefile control
808 for the build is renamed.
810 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
811 was allocated for every new TLS startup, meaning one per message. Fix
812 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
814 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
815 reported the original. Fix to report (as far as possible) the ACL
816 result replacing the original.
818 JH/23 Fix memory leak during multi-message connections using STARTTLS under
819 OpenSSL. Certificate information is loaded for every new TLS startup,
820 and the resources needed to be freed.
822 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
824 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
827 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
828 DATA response info to the (existing) per-recipient response info for
829 the "C=" log element. It can have useful tracking info from the
830 destination system. Patch from Simon Arlott.
832 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
833 length value. Previously this would segfault.
835 HS/02 Support Avast multiline protoocol, this allows passing flags to
836 newer versions of the scanner.
838 JH/28 Ensure that variables possibly set during message acceptance are marked
839 dead before release of memory in the daemon loop. This stops complaints
840 about them when the debug_store option is enabled. Discovered specifically
841 for sender_rate_period, but applies to a whole set of variables.
842 Do the same for the queue-runner and queue-list loops, for variables set
843 from spool message files. Do the same for the SMTP per-message loop, for
844 certain variables indirectly set in ACL operations.
846 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
847 as a multi-recipient message from a mailinglist manager). The coding had
848 an arbitrary cutoff number of characters while checking for more input;
849 enforced by writing a NUL into the buffer. This corrupted long / fast
850 input. The problem was exposed more widely when more pipelineing of SMTP
851 responses was introduced, and one Exim system was feeding another.
852 The symptom is log complaints of SMTP syntax error (NUL chars) on the
853 receiving system, and refused recipients seen by the sending system
854 (propating to people being dropped from mailing lists).
855 Discovered and pinpointed by David Carter.
857 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
858 replaced by the ${authresults } expansion.
860 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
862 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
863 allows proper process termination in container environments.
865 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
866 Previously the "final dot" had a newline after it; ensure it is CR,LF.
868 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
869 and "err_perm", deprecated since 4.83 when the RFC-defined words
870 "temperror" and "permerror" were introduced.
872 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
873 transport-filters or DKIM-signing. The restriction was lost in the
874 consolidation of verify-callout and delivery SMTP handling.
875 Extend the restriction to also cover ARC-signing.
877 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
878 in defer=pass mode supply a 450 to the initiator. Previously the message
881 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
882 tls_require_ciphers is used as before.
884 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
885 "pass_unscanned". Only tmpfails from the scanner are written to
886 the paniclog, as they may require admin intervention (permission
887 denied, license issues). Other scanner errors (like decompression
888 bombs) do not cause a paniclog entry.
890 JH/36 Fix reinitialisation of DKIM logging variable between messages.
891 Previously it was possible to log spurious information in receive log
894 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
895 triggered odd behaviour from Outlook Express clients.
897 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
900 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
901 since the IETF WG has not yet settled on that versus the original
902 "bare" representation.
904 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
905 Previously the millisecond value corrupted the output.
906 Fix also for syslog_pid=no and log_selector +pid, for which the pid
907 corrupted the output.
913 JH/01 Rework error string handling in TLS interface so that the caller in
914 more cases is responsible for logging. This permits library-sourced
915 string to be attached to addresses during delivery, and collapses
916 pairs of long lines into single ones.
918 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
919 during configuration. Wildcards are allowed and expanded.
921 JH/02 Rework error string handling in DKIM to pass more info back to callers.
922 This permits better logging.
924 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
925 do not close it down and have the child transport start it up again on
926 the passed-on TCP connection. Instead, proxy the child (and any
927 subsequent ones) for TLS via a unix-domain socket channel. Logging is
928 affected: the continued delivery log lines do not have any DNSSEC, TLS
929 Certificate or OCSP information. TLS cipher information is still logged.
931 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
932 identical IP addresses on different listening ports. Will also affect
935 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
936 add noisy ifdef guards to special-case this sillyness.
937 Patch from Bernd Kuhls.
939 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
940 than 255 are no longer allowed.
942 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
943 Disable the session-cache too, which might reduce our load. Since we
944 currrectly use a new context for every connection, both as server and
945 client, there is no benefit for these.
946 GnuTLS appears to not support tickets server-side by default (we don't
947 call gnutls_session_ticket_enable_server()) but client side is enabled
948 by default on recent versions (3.1.3 +) unless the PFS priority string
951 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
952 <https://reproducible-builds.org/specs/source-date-epoch/>.
954 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
955 the check for any unsuccessful recipients did not notice the limit, and
956 erroneously found still-pending ones.
958 JH/08 Pipeline CHUNKING command and data together, on kernels that support
959 MSG_MORE. Only in-clear (not on TLS connections).
961 JH/09 Avoid using a temporary file during transport using dkim. Unless a
962 transport-filter is involved we can buffer the headers in memory for
963 creating the signature, and read the spool data file once for the
964 signature and again for transmission.
966 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
967 4.77 as the kernel support then wasn't solid, having issues in 64bit
968 mode. Now, it's been long enough. Add support for FreeBSD also.
970 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
971 case where the routing stage had gathered several addresses to send to
972 a host before calling the transport for the first, we previously failed
973 to close down TLS in the old transport process before passing the TCP
974 connection to the new process. The new one sent a STARTTLS command
975 which naturally failed, giving a failed delivery and bloating the retry
976 database. Investigation and fix prototype from Wolfgang Breyha.
978 JH/12 Fix check on SMTP command input synchronisation. Previously there were
979 false-negatives in the check that the sender had not preempted a response
980 or prompt from Exim (running as a server), due to that code's lack of
981 awareness of the SMTP input buffering.
983 PP/04 Add commandline_checks_require_admin option.
984 Exim drops privileges sanely, various checks such as -be aren't a
985 security problem, as long as you trust local users with access to their
986 own account. When invoked by services which pass untrusted data to
987 Exim, this might be an issue. Set this option in main configuration
988 AND make fixes to the calling application, such as using `--` to stop
991 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
992 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
993 responses to those, into a single TLS record each way (this usually means
994 a single packet). As a side issue, smtp_enforce_sync now works on TLS
997 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
998 affects you only if you're dancing at the edge of the param size limits.
999 If you are, and this message makes sense to you, then: raise the
1000 configured limit or use OpenSSL 1.1. Nothing we can do for older
1003 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1004 cmdline element to get the documented default one. Previously it was
1007 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1008 are unsuable for non-message-specific reasons - eg. connection timeout,
1009 banner-time rejection.
1011 JH/16 Fix logging of delivery remote port, when specified by router, under
1014 PP/06 Repair manualroute's ability to take options in any order, even if one
1015 is the name of a transport.
1018 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1020 JH/17 Change the list-building routines interface to use the expanding-string
1021 triplet model, for better allocation and copying behaviour.
1023 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1024 Previously it was constructed the first time a possibly-matching string
1025 was met in the configuration file input during startup; now it is done
1028 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1029 compatible one, to avoid the (poorly documented) possibility of a config
1030 file in the working directory redirecting the DB files, possibly correpting
1031 some existing file. CVE-2017-10140 assigned for BDB.
1033 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1034 cache-hot. Previously, although the result was properly cached, the
1035 initial verify call returned a defer.
1037 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1038 the main verify for receipient in uncached-mode.
1040 JH/22 Retire historical build files to an "unsupported" subdir. These are
1041 defined as "ones for which we have no current evidence of testing".
1043 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1044 if present. Previously it was ignored.
1046 JH/24 Start using specified-initialisers in C structure init coding. This is
1047 a C99 feature (it's 2017, so now considered safe).
1049 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1050 if was a fixed-sized field and bitmask ops via macros; it is now more
1053 PP/07 GitHub PR 56: Apply MariaDB build fix.
1054 Patch provided by Jaroslav Škarvada.
1056 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1057 during Coverity cleanups [4.87 JH/47]
1058 Diagnosis and fix provided by Michael Fischer v. Mollard.
1060 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1061 the right size to place the terminating semicolon on its own folded
1062 line, the header hash was calculated to an incorrect value thanks to
1063 the (relaxed) space the fold became.
1065 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1066 and confused the parent.
1068 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1069 which could crash as a result. This could lead to undeliverable messages.
1071 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1074 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1075 matters on fast-turnover and PID-randomising systems, which were getting
1076 out-of-order delivery.
1078 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1079 a possibly-overlapping copy. The symptom was that "Remote host closed
1080 connection in response to HELO" was logged instead of the actual 4xx
1083 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1084 Previously only that bufferd was discarded, resulting in SYMTP command
1087 JH/32 DKIM: when a message has multiple signatures matching an identity given
1088 in dkim_verify_signers, run the dkim acl once for each. Previously only
1089 one run was done. Bug 2189.
1091 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1092 "panic the current process" to "deliberately defer". The panic log is
1093 still written with the problem list name; the mail and reject logs now
1094 get a temp-reject line for the message that was being handled, saying
1095 something like "domains check lookup or other defer". The SMTP 451
1096 message is still "Temporary local problem".
1098 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1099 A crafted sequence of BDAT commands could result in in-use memory beeing
1100 freed. CVE-2017-16943.
1102 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1103 from SMTP input. Previously it was always done; now only done for DATA
1104 and not BDAT commands. CVE-2017-16944.
1106 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1107 to the message (such as an overlong header line). Previously this was
1108 not done and we did not exit BDAT mode. Followon from the previous item
1109 though a different problem.
1115 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1116 than -2003 did; needs libidn2 in addition to libidn.
1118 JH/02 The path option on a pipe transport is now expanded before use.
1120 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1121 Patch provided by "Björn", documentation fix added too.
1123 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1124 missing a wire-to-host endian conversion.
1126 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1127 close after a BDAT command line could be taken as a following command,
1128 giving a synch failure. Fix by only checking for synch immediately
1129 before acknowledging the chunk.
1131 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1132 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1133 macro. Patches provided by Josh Soref.
1135 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1136 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1137 by IANA as a service extension. However, John Klensin suggests that we
1140 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1141 the dkim code may be unix-mode line endings rather than smtp wire-format
1142 CRLF, so prepend a CR to any bare LF.
1144 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1145 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1147 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1148 the first were themselves being wrongly included in the feed into dkim
1149 processing; with most chunk sizes in use this resulted in an incorrect
1150 body hash calculated value.
1152 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1153 DKIM signature block, for verification. Although advised against by
1154 standards it is specifically not ruled illegal.
1156 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1158 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1159 missing a body hash (the bh= tag).
1161 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1162 It seems that HAProxy sends the Proxy Protocol information in clear and
1163 only then does a TLS startup, so do the same.
1165 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1166 TCP connections (such as for Spamd) unless the daemon successfully set
1167 Fast Open mode on its listening sockets. This fixes breakage seen on
1168 too-old kernels or those not configured for Fast Open, at the cost of
1169 requiring both directions being enabled for TFO, and TFO never being used
1170 by non-daemon-related Exim processes.
1172 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1173 endings, at least on the first header line. Try to canonify any that get
1174 past that check, despite the cost.
1176 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1177 now limited to an arbitrary five deep, while parsing addresses with the
1178 strip_excess_angle_brackets option enabled.
1180 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1181 instead leave the unprompted TLS handshake in socket buffer for the
1182 TLS library to consume.
1184 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1186 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1188 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1189 region in one operation, for speed, and this leaves assigned pointers
1190 dangling. Add checks run only under the testsuite which checks all
1191 variables at a store-reset and panics on a dangling pointer; add code
1192 explicitly nulling out all the variables discovered. Fixes one known
1193 bug: a transport crash, where a dangling pointer for $sending_ip_address
1194 originally assigned in a verify callout, is re-used.
1196 PP/06 Drop '.' from @INC in various Perl scripts.
1198 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1200 PP/08 Reduce a number of compilation warnings under clang; building with
1201 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1202 should be warning-free.
1204 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1206 HS/01 Fix portability problems introduced by PP/08 for platforms where
1207 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1209 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1210 chunk. This allows us to accept broken chunked messages. We need a more
1211 general solution here.
1213 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1214 already-broken messages in the queue.
1216 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1218 JH/19 Fix reference counting bug in routing-generated-address tracking.
1224 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1225 supports it and a size is available (ie. the sending peer gave us one).
1227 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1228 years of being deprecated). The replacements are the ACLs
1229 acl_smtp_mime and acl_not_smtp_mime.
1231 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1232 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1233 or even an in-clear connection were permitted. Now, if the host lookup
1234 was dnssec and dane was requested then the host is only used if the
1235 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1236 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1237 if one fails this test.
1238 This means that a poorly-configured remote DNS will make it incommunicado;
1239 but it protects against a DNS-interception attack on it.
1241 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1242 non-noisy when a race steals the message being considered.
1244 JH/05 If main configuration option tls_certificate is unset, generate a
1245 self-signed certificate for inbound TLS connections.
1247 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1248 in rewrites and routers.
1250 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1251 and logged a warning sing 4.83; now they are a configuration file error.
1253 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1254 (lacking @domain). Apply the same qualification processing as RCPT.
1256 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1258 JH/10 Support ${sha256:} applied to a string (as well as the previous
1261 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1262 a cutthrough deliver is pending, as we always want to make a connection.
1263 This also avoids re-routing the message when later placing the cutthrough
1264 connection after a verify cache hit.
1265 Do not update it with the verify result either.
1267 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1268 when routing results in more than one destination address.
1270 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1271 signing (which inhibits the cutthrough capability). Previously only
1272 the presence of an option was tested; now an expansion evaluating as
1273 empty is permissible (obviously it should depend only on data available
1274 when the cutthrough connection is made).
1276 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1277 the relevant preceding SMTP command did not note the pipelining mode.
1279 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1280 Previously they were not counted.
1282 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1283 as one having no matching records. Previously we deferred the message
1284 that needed the lookup.
1286 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1287 distinguished as "(=".
1289 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1290 for missing MX records. Previously it only worked for missing A records.
1292 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1294 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1295 after the data-go-ahead and data-ack. Patch from Jason Betts.
1297 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1298 even for a "none" policy. Patch from Tony Meyer.
1300 JH/22 Fix continued use of a connection for further deliveries. If a port was
1301 specified by a router, it must also match for the delivery to be
1304 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1305 When one of the recipients of a message was unsuitable for the connection
1306 (has no matching addresses), we lost track of needing to mark it
1307 deferred. As a result mail would be lost.
1309 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1311 JH/25 Decoding ACL controls is now done using a binary search; the source code
1312 takes up less space and should be simpler to maintain. Merge the ACL
1313 condition decode tables also, with similar effect.
1315 JH/26 Fix problem with one_time used on a redirect router which returned the
1316 parent address unchanged. A retry would see the parent address marked as
1317 delivered, so not attempt the (identical) child. As a result mail would
1320 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1321 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1322 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1325 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1326 Main config options check_{log,spool}_{inodes,space} are now
1327 100 inodes, 10MB unless set otherwise in the configuration.
1329 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1330 Previously it only applied to the main-section connection policy
1333 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1335 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1336 by me. Added RFC7919 DH primes as an alternative.
1338 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1339 are not in the system include path.
1341 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1342 GnuTLS, when a session startup failed (eg because the client disconnected)
1343 Exim did stdio operations after fclose. This was exposed by a recent
1344 change which nulled out the file handle after the fclose.
1346 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1347 signed directly by the cert-signing cert, rather than an intermediate
1348 OCSP-signing cert. This is the model used by LetsEncrypt.
1350 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1352 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1353 an incoming connection.
1355 HS/02 Bug 1802: Do not half-close the connection after sending a request
1358 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1359 fallback to "prime256v1".
1361 JH/34 SECURITY: Use proper copy of DATA command in error message.
1362 Could leak key material. Remotely exploitable. CVE-2016-9963.
1368 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1369 and 3.4.4 - once the server is enabled to respond to an OCSP request
1370 it does even when not requested, resulting in a stapling non-aware
1371 client dropping the TLS connection.
1373 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1374 support variable-length bit vectors. No functional change.
1376 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1377 The I= interface field on outgoing lines is now after the H= remote
1378 host field, same as incoming lines. There is a separate
1379 outgoing_interface log selector which allows you to disable the
1382 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1383 If not running log_selector +smtp_connection the mainlog would be held
1384 open indefinitely after a "too many connections" event, including to a
1385 deleted file after a log rotate. Leave the per net connection logging
1386 leaving it open for efficiency as that will be quickly detected by the
1387 check on the next write.
1389 HS/01 Bug 1671: Fix post transport crash.
1390 Processing the wait-<transport> messages could crash the delivery
1391 process if the message IDs didn't exist for some reason. When
1392 using 'split_spool_directory=yes' the construction of the spool
1393 file name failed already, exposing the same netto behaviour.
1395 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1396 mime_regex ACL conditions.
1398 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1399 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1400 to HELO, local diagnostic string.
1402 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1403 log line to debug. Even when configured with a tls authenticator many
1404 client connections are expected to not authenticate in this way, so
1405 an authenticate fail is not an error.
1407 HS/02 Add the Exim version string to the process info. This way exiwhat
1408 gives some more detail about the running daemon.
1410 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1411 matter for fast-change records such as DNSBLs.
1413 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1414 retry record, even if constant. There may be multiple transports with
1415 different interface settings and the retry behaviour needs to be kept
1418 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1420 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1422 JH/10 Bug 840: fix log_defer_output option of pipe transport
1424 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1425 hosts_randomize. This matters a lot when combined with mua_wrapper.
1427 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1428 ${quote_pgsql:<string>} operator.
1430 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1432 JH/14 Change header file name for freeradius-client. Relevant if compiling
1433 with Radius support; from the Gentoo tree and checked under Fedora.
1435 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1437 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1438 option transport_filter, meaning no filtering.
1440 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1442 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1443 defaults to "*" (all hosts). The variable is now available when not built
1444 with TLS, default unset, mainly to enable keeping the testsuite sane.
1445 If a server certificate is not supplied (via tls_certificate) an error is
1446 logged, and clients will find TLS connections fail on startup. Presumably
1447 they will retry in-clear.
1448 Packagers of Exim are strongly encouraged to create a server certificate
1449 at installation time.
1451 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1452 with the $config_file variable.
1454 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1455 in transport context, after the attempt, and per-recipient. The latter type
1456 is per host attempted. The event data is the error message, and the errno
1457 information encodes the lookup type (A vs. MX) used for the (first) host,
1458 and the trailing two digits of the smtp 4xx response.
1460 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1461 to write to mainlog (or rejectlog, paniclog) in the window between file
1462 creation and permissions/ownership being changed. Particularly affects
1463 installations where exicyclog is run as root, rather than exim user;
1464 result is that the running daemon panics and dies.
1466 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1468 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1469 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1470 "pri" and "weight". Note that the previous implicit priority given by the
1471 list order is no longer honoured.
1473 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1474 for DKIM processing.
1476 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1477 by defining SUPPORT_SOCKS.
1479 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1480 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1481 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1482 variables are renamed to proxy_{local,external}_{address,port}.
1484 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1485 for a build by defining SUPPORT_I18N
1487 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1488 of the query string, and make ${quote_redis:} do that quoting.
1490 JH/29 Move Events support from Experimental to mainline, enabled by default
1491 and removable for a build by defining DISABLE_EVENT.
1493 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1495 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1496 cached by the daemon.
1498 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1499 by defining LOOKUP_REDIS. The libhiredis library is required.
1501 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1502 keys are given for lookup.
1504 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1505 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1506 only supported when built with TLS support. The PolarSSL SHA routines
1507 are still used when the TLS library is too old for convenient support.
1509 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1510 openssl_options), for security. OpenSSL forces this from version 1.1.0
1511 server-side so match that on older versions.
1513 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1514 allocation for $value could be released as the expansion processing
1515 concluded, but leaving the global pointer active for it.
1517 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1518 and to use the domains and local_parts ACL conditions.
1520 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1521 incorrectly not doubled on cutthrough transmission, hence seen as a
1522 body-termination at the receiving system - resulting in truncated mails.
1523 Commonly the sender saw a TCP-level error, and retransmitted the message
1524 via the normal store-and-forward channel. This could result in duplicates
1525 received - but deduplicating mailstores were liable to retain only the
1526 initial truncated version.
1528 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1530 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1532 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1533 we're in there, support oversigning also; bug 1309.
1535 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1537 HS/04 Add support for keep_environment and add_environment options.
1539 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1540 either intentional arithmetic overflow during PRNG, or testing config-
1543 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1544 delivery resulted in actual delivery. Cancel cutthrough before DATA
1547 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1548 rejects a recipient: pass the reject to the originator.
1550 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1551 Many were false-positives and ignorable, but it's worth fixing the
1554 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1555 for the new environment-manipulation done at startup. Move the routines
1556 from being local to tls.c to being global via the os.c file.
1558 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1559 an extract embedded as result-arg for a map, the first arg for extract
1560 is unavailable so we cannot tell if this is a numbered or keyed
1561 extraction. Accept either.
1567 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1570 JH/02 The smtp transport option "multi_domain" is now expanded.
1572 JH/03 The smtp transport now requests PRDR by default, if the server offers
1575 JH/04 Certificate name checking on server certificates, when exim is a client,
1576 is now done by default. The transport option tls_verify_cert_hostnames
1577 can be used to disable this per-host. The build option
1578 EXPERIMENTAL_CERTNAMES is withdrawn.
1580 JH/05 The value of the tls_verify_certificates smtp transport and main options
1581 default to the word "system" to access the system default CA bundle.
1582 For GnuTLS, only version 3.0.20 or later.
1584 JH/06 Verification of the server certificate for a TLS connection is now tried
1585 (but not required) by default. The verification status is now logged by
1586 default, for both outbound TLS and client-certificate supplying inbound
1589 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1592 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1593 Status Notification (bounce) messages are now MIME format per RFC 3464.
1594 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1595 under the control of the dsn_advertise_hosts option, and routers may
1596 have a dsn_lasthop option.
1598 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1599 default, modifiable by a malware= option. The list separator for
1600 the options can now be changed in the usual way. Bug 68.
1602 JH/10 The smtp_receive_timeout main option is now expanded before use.
1604 JH/11 The incoming_interface log option now also enables logging of the
1605 local interface on delivery outgoing connections.
1607 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1608 if the interface and destination host and port all match.
1610 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1613 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1614 Patch from Andrew Lewis.
1616 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1617 now supports optional time-restrictions, weighting, and priority
1618 modifiers per server. Patch originally by <rommer@active.by>.
1620 JH/16 The spamd_address main option now supports a mixed list of local
1621 and remote servers. Remote servers can be IPv6 addresses, and
1622 specify a port-range.
1624 JH/17 Bug 68: The spamd_address main option now supports an optional
1625 timeout value per server.
1627 JH/18 Bug 1581: Router and transport options headers_add/remove can
1628 now have the list separator specified.
1630 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1633 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1636 JH/21 Support for the A6 type of dns record is withdrawn.
1638 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1639 rather than the verbs used.
1641 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1642 from 255 to 1024 chars.
1644 JH/24 Verification callouts now attempt to use TLS by default.
1646 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1647 are generic router options now. The defaults didn't change.
1649 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1650 Original patch from Alexander Shikoff, worked over by JH.
1652 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1653 files if ZCAT_COMMAND is not executable.
1655 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1657 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1659 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1660 Normally benign, it bites when the pair was led to by a CNAME;
1661 modern usage is to not canonicalize the domain to a CNAME target
1662 (and we were inconsistent anyway for A-only vs AAAA+A).
1664 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1666 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1667 when evaluating $sender_host_dnssec.
1669 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1670 $sender_helo_dnssec variable.
1672 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1674 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1676 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1678 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1679 documented as working, but never had. Support all but $spam_report.
1681 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1682 added for tls authenticator.
1684 HS/03 Add perl_taintmode main config option
1690 TL/01 When running the test suite, the README says that variables such as
1691 no_msglog_check are global and can be placed anywhere in a specific
1692 test's script, however it was observed that placement needed to be near
1693 the beginning for it to behave that way. Changed the runtest perl
1694 script to read through the entire script once to detect and set these
1695 variables, reset to the beginning of the script, and then run through
1696 the script parsing/test process like normal.
1698 TL/02 The BSD's have an arc4random API. One of the functions to induce
1699 adding randomness was arc4random_stir(), but it has been removed in
1700 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1701 function when detected.
1703 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1704 cause callback expansion.
1706 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1707 syntax errors in an expansion can be treated as a string instead of
1708 logging or causing an error, due to the internal use of bool_lax
1709 instead of bool when processing it.
1711 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1712 server certificates when making smtp deliveries.
1714 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1716 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1718 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1720 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1721 Merged patch from Sebastian Wiedenroth.
1723 JH/05 Fix results-pipe from transport process. Several recipients, combined
1724 with certificate use, exposed issues where response data items split
1725 over buffer boundaries were not parsed properly. This eventually
1726 resulted in duplicates being sent. This issue only became common enough
1727 to notice due to the introduction of connection certificate information,
1728 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1730 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1731 size buffer was used, resulting in syntax errors when an expansion
1734 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1735 version 3.3.6 or later.
1737 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1738 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1739 both become "event_action", the variables become $event_name, $event_data
1740 and $event_defer_errno. There is a new variable $verify_mode, usable in
1741 routers, transports and related events. The tls:cert event is now also
1742 raised for inbound connections, if the main configuration event_action
1745 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1746 early OCSP support, but no stapling (appears to be less than 1.0.0).
1748 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1749 server certificate names available under the smtp transport option
1750 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1753 JH/10 Time-related extraction expansions from certificates now use the main
1754 option "timezone" setting for output formatting, and are consistent
1755 between OpenSSL and GnuTLS compilations. Bug 1541.
1757 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1758 encoded parameter in the incoming message. Bug 1558.
1760 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1761 include certificate info, eximon was claiming there were spoolfile
1764 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1766 JH/14 Log delivery-related information more consistently, using the sequence
1767 "H=<name> [<ip>]" wherever possible.
1769 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1770 are problematic for Debian distribution, omit them from the release
1773 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1775 JH/16 Fix string representation of time values on 64bit time_t architectures.
1778 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1779 output list separator was used.
1784 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1785 checkers that were complaining about end of non-void function with no
1788 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1789 This was a regression introduced in 4.83 by another bugfix.
1791 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1793 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1794 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1800 TF/01 Correctly close the server side of TLS when forking for delivery.
1802 When a message was received over SMTP with TLS, Exim failed to clear up
1803 the incoming connection properly after forking off the child process to
1804 deliver the message. In some situations the subsequent outgoing
1805 delivery connection happened to have the same fd number as the incoming
1806 connection previously had. Exim would try to use TLS and fail, logging
1807 a "Bad file descriptor" error.
1809 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1810 utilities have not been installed.
1812 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1813 temporary space as the ACL may create new global variables.
1815 TL/01 LDAP support uses per connection or global context settings, depending
1816 upon the detected version of the libraries at build time.
1818 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1819 to extract and use the src ip:port in logging and expansions as if it
1820 were a direct connection from the outside internet. PPv2 support was
1821 updated based on HAProxy spec change in May 2014.
1823 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1825 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1826 Properly escape header and check for NULL return.
1828 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1831 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1833 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1834 characters in header names, implemented as a verify condition.
1835 Contributed by Michael Fischer v. Mollard.
1837 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1838 results permerror and temperror. Previous values are deprecated but
1839 still accepted. In a future release, err_perm and err_temp will be
1840 completely removed, which will be a backward incompatibility if the
1841 ACL tests for either of these two old results. Patch contributed by
1842 user bes-internal on the mailing list.
1844 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1846 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1847 selectors, in both main and reject logs.
1849 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1852 JH/07 Add malware type "sock" for talking to simple daemon.
1854 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1856 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1857 routers/transports under cutthrough routing.
1859 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1860 numbers. Touch up "bool" conditional to keep the same definition.
1862 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1864 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1866 JH/12 Expand items in router/transport headers_add or headers_remove lists
1867 individually rather than the list as a whole. Bug 1452.
1869 Required for reasonable handling of multiple headers_ options when
1870 they may be empty; requires that headers_remove items with embedded
1871 colons must have them doubled (or the list-separator changed).
1873 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1874 view the policy declared in the DMARC record. Currently, $dmarc_status
1875 is a combined value of both the record presence and the result of the
1878 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1880 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1881 dnslookup router and the smtp transport (applying to the forward
1884 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1885 of ldap servers used for a specific lookup. Patch provided by Heiko
1888 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1889 New variable $lookup_dnssec_authenticated for observability.
1891 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1892 Patch submitted by Lars Timman.
1894 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1896 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1897 Requires trusted mode and valid format message id, aborts otherwise.
1898 Patch contributed by Heiko Schlichting.
1900 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1901 certextract with support for various fields. Bug 1358.
1903 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1904 is requested by default, modifiable by smtp transport option
1907 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1908 operate on certificate variables to give certificate fingerprints
1909 Also new ${sha256:cert_variable}.
1911 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1913 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1914 Christian Aistleitner.
1916 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1918 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1919 file. Patch from Wolfgang Breyha.
1921 JH/25 Expand the coverage of the delivery $host and $host_address to
1922 client authenticators run in verify callout. Bug 1476.
1924 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1925 align with daemon_smtp_ports. Bug 72.
1927 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1928 support and error reporting did not work properly.
1930 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1931 and is readable. Patch from Andrew Colin Kissa.
1933 TL/14 Enhance documentation of ${run expansion and how it parses the
1934 commandline after expansion, particularly in the case when an
1935 unquoted variable expansion results in an empty value.
1937 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1939 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1940 by renaming away from it. Addresses GH issue 15, reported by
1943 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1945 TL/15 SECURITY: prevent double expansion in math comparison functions
1946 (can expand unsanitized data). Not remotely exploitable.
1953 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1955 PP/02 Make -n do something, by making it not do something.
1956 When combined with -bP, the name of an option is not output.
1958 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1961 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1962 $sender_host_name and config options to manage this, and basic check
1965 PP/05 DSCP support for outbound connections and control modifier for inbound.
1967 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1968 (Only plugin which currently uses this is kerberos4, which nobody should
1969 be using, but we should make it available and other future plugins might
1970 conceivably use it, even though it would break NAT; stuff *should* be
1971 using channel bindings instead).
1973 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1974 name; added for Sendmail compatibility; requires admin caller.
1975 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1976 just ignore it); requires trusted caller.
1977 Also parse but ignore: -Ac -Am -X<logfile>
1980 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1982 TL/02 Add +smtp_confirmation as a default logging option.
1984 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1985 Patch by Magnus Holmgren from 2007-02-20.
1987 TL/04 Bugzilla 1281 - Spec typo.
1988 Bugzilla 1283 - Spec typo.
1989 Bugzilla 1290 - Spec grammar fixes.
1991 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1993 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1995 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1996 mailing list by by Dmitry Isaikin.
1998 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2000 JH/02 Support "G" suffix to numbers in ${if comparisons.
2002 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2004 NM/01 Bugzilla 1197 - Spec typo
2005 Bugzilla 1196 - Spec examples corrections
2007 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2009 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2010 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2013 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2014 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2016 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2017 "acl {{name}{arg}...}", and optional args on acl condition
2020 JH/05 Permit multiple router/transport headers_add/remove lines.
2022 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2024 JH/07 Avoid using a waiting database for a single-message-only transport.
2025 Performance patch from Paul Fisher. Bugzilla 1262.
2027 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2030 JH/09 Add $headers_added variable, with content from use of ACL modifier
2031 add_header (but not yet added to the message). Bugzilla 199.
2033 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2034 Pulled from Bugzilla 817 by Wolfgang Breyha.
2036 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2038 (nb: this is the same fix as in Exim 4.80.1)
2040 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2043 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2046 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2048 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2049 advertise SMTP AUTH mechanism to us, instead of a generic
2050 protocol violation error. Also, make Exim more robust to bad
2051 data from the Dovecot auth socket.
2053 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2055 When a queue runner is handling a message, Exim first routes the
2056 recipient addresses, during which it prunes them based on the retry
2057 hints database. After that it attempts to deliver the message to
2058 any remaining recipients. It then updates the hints database using
2061 So if a recipient address works intermittently, it can get repeatedly
2062 deferred at routing time. The retry hints record remains fresh so the
2063 address never reaches the final cutoff time.
2065 This is a fairly common occurrence when a user is bumping up against
2066 their storage quota. Exim had some logic in its local delivery code
2067 to deal with this. However it did not apply to per-recipient defers
2068 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2070 This change adds a proper retry rule check during routing so that the
2071 final cutoff time is checked against the message's age. We only do
2072 this check if there is an address retry record and there is not a
2073 domain retry record; this implies that previous attempts to handle
2074 the address had the retry_use_local_parts option turned on. We use
2075 this as an approximation for the destination being like a local
2076 delivery, as in LMTP.
2078 I suspect this new check makes the old local delivery cutoff check
2079 redundant, but I have not verified this so I left the code in place.
2081 TF/02 Correct gecos expansion when From: is a prefix of the username.
2083 Test 0254 submits a message to Exim with the header
2087 When I ran the test suite under the user fanf2, Exim expanded
2088 the header to contain my full name, whereas it should have added
2089 a Resent-Sender: header. It erroneously treats any prefix of the
2090 username as equal to the username.
2092 This change corrects that bug.
2094 GF/01 DCC debug and logging tidyup
2095 Error conditions log to paniclog rather than rejectlog.
2096 Debug lines prefixed by "DCC: " to remove any ambiguity.
2098 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2100 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2101 Bug spotted by Jeremy Harris; was flawed since initial commit.
2102 Would have resulted in OCSP responses post-SNI triggering an Exim
2103 NULL dereference and crash.
2105 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2107 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2108 Bug detection, analysis and fix by Samuel Thibault.
2109 Bugzilla 1331, Debian bug #698092.
2111 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2113 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2114 Server implementation by Todd Lyons, client by JH.
2115 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2116 config variable "prdr_enable" controls whether the server
2117 advertises the facility. If the client requests PRDR a new
2118 acl_data_smtp_prdr ACL is called once for each recipient, after
2119 the body content is received and before the acl_smtp_data ACL.
2120 The client is controlled by both of: a hosts_try_prdr option
2121 on the smtp transport, and the server advertisement.
2122 Default client logging of deliveries and rejections involving
2123 PRDR are flagged with the string "PRDR".
2125 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2126 fclose(). Diagnosis by Todd Lyons.
2128 PP/17 Update configure.default to handle IPv6 localhost better.
2129 Patch by Alain Williams (plus minor tweaks).
2132 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2133 This is now consistent with GnuTLS, and is now documented: the
2134 previous undocumented portable approach to treating the option as
2135 unset was to force an expansion failure. That still works, and
2136 an empty string is now equivalent.
2138 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2139 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2140 not performing validation itself.
2142 PP/20 Added force_command boolean option to pipe transport.
2143 Patch from Nick Koston, of cPanel Inc.
2145 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2148 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2150 PP/21 Fix eximon continuous updating with timestamped log-files.
2151 Broken in a format-string cleanup in 4.80, missed when I repaired the
2152 other false fix of the same issue.
2153 Report and fix from Heiko Schlichting.
2156 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2157 Report from Prashanth Katuri.
2159 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2160 It's SecureTransport, so affects any MacOS clients which use the
2161 system-integrated TLS libraries, including email clients.
2163 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2164 using a MIME ACL for non-SMTP local injection.
2165 Report and assistance in diagnosis by Warren Baker.
2167 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2169 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2171 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2172 last id that failed so it may be referenced in subsequent ACL's.
2174 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2177 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2178 ldap library initialization, allowing self-signed CA's to be
2179 used. Also properly sets require_cert option later in code by
2180 using NULL (global ldap config) instead of ldap handle (per
2181 session). Bug diagnosis and testing by alxgomz.
2183 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2184 the src/util/ subdirectory.
2186 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2187 renamed to Transport Post Delivery Action by Jeremy Harris, as
2190 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2191 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2192 redis_servers = needs to be configured which will be used by the redis
2193 lookup. Patch from Warren Baker, of The Packet Hub.
2195 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2197 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2198 hostname or reverse DNS when processing a host list. Used suggestions
2199 from multiple comments on this bug.
2201 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2203 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2204 Missed a few lines, added it to make the runtest require no keyboard
2207 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2208 contains upper case chars. Make router use caseful_local_part.
2210 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2211 support when GnuTLS has been built with p11-kit.
2217 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2219 This, or similar/improved, will also be change PP/11 of 4.82.
2225 PP/01 Handle short writes when writing local log-files.
2226 In practice, only affects FreeBSD (8 onwards).
2227 Bugzilla 1053, with thanks to Dmitry Isaikin.
2229 NM/01 Bugzilla 949 - Documentation tweak
2231 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2234 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2236 PP/02 Implemented gsasl authenticator.
2238 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2240 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2241 `pkg-config foo` for cflags/libs.
2243 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2244 with rest of GSASL and with heimdal_gssapi.
2246 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2247 `pkg-config foo` for cflags/libs for the TLS implementation.
2249 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2250 properties get this fed in as external SSF. A number of robustness
2251 and debugging improvements to the cyrus_sasl authenticator.
2253 PP/08 cyrus_sasl server now expands the server_realm option.
2255 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2256 Patch by Jeremy Harris.
2258 PP/10 Added dbmjz lookup type.
2260 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2262 PP/12 MAIL args handles TAB as well as SP, for better interop with
2263 non-compliant senders.
2264 Analysis and variant patch by Todd Lyons.
2266 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2267 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2268 Patch from Dirk Mueller <dmueller@suse.com>
2270 PP/13 tls_peerdn now print-escaped for spool files.
2271 Observed some $tls_peerdn in wild which contained \n, which resulted
2272 in spool file corruption.
2274 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2275 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2276 or write after TLS renegotiation, which otherwise led to messages
2279 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2280 as a tracking header (ie: a signed header comes before the signature).
2281 Patch from Wolfgang Breyha.
2283 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2284 comma-sep list; embedded commas doubled.
2286 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2288 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2290 Report and patch from Dmitry Banschikov.
2292 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2293 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2294 failures. We appear to now support TLS1.1+ with Exim.
2296 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2297 lets Exim select keys and certificates based upon TLS SNI from client.
2298 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2299 before an outbound SMTP session. New log_selector, +tls_sni.
2301 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2302 NULL dereference. Report and patch from Alun Jones.
2304 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2305 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2306 Not seeing resolver debug output on NetBSD, but suspect this is a
2307 resolver implementation change.
2309 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2310 Left warnings. Added "eximon gdb" invocation mode.
2312 PP/21 Defaulting "accept_8bitmime" to true, not false.
2314 PP/22 Added -bw for inetd wait mode support.
2316 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2317 locate the relevant includes and libraries. Made this the default.
2319 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2320 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2322 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2323 This may cause build issues on older platforms.
2325 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2326 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2327 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2328 Added SNI support via GnuTLS too.
2329 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2331 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2333 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2334 Applied second patch from Janne, implementing suggestion to default
2335 multiple-strings-in-record handling to match SPF spec.
2337 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2339 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2340 read-only, out of scope).
2341 Patch from Wolfgang Breyha, report from Stuart Northfield.
2343 PP/29 Fix three issues highlighted by clang analyser static analysis.
2344 Only crash-plausible issue would require the Cambridge-specific
2345 iplookup router and a misconfiguration.
2346 Report from Marcin Mirosław.
2348 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2350 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2351 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2352 As part of this, removing so much warning spew let me fix some minor
2353 real issues in debug logging.
2355 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2356 assignment on my part. Fixed.
2358 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2359 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2360 Janne Snabb (who went above and beyond: thank you).
2362 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2363 string otherwise requires a connection and a bunch more work and it's
2364 relatively easy to get wrong. Should also expose TLS library linkage
2367 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2368 64-bit ${eval} (JH/03).
2370 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2371 GNU libc to support some of the 64-bit stuff, should not lead to
2372 conflicts. Defined before os.h is pulled in, so if a given platform
2373 needs to override this, it can.
2375 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2376 protection layer was required, which is not implemented.
2377 Bugzilla 1254, patch from Wolfgang Breyha.
2379 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2380 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2381 tls_dhparam take prime identifiers. Also unbreak combination of
2382 OpenSSL+DH_params+TLSSNI.
2384 PP/39 Disable SSLv2 by default in OpenSSL support.
2390 PP/01 Solaris build fix for Oracle's LDAP libraries.
2391 Bugzilla 1109, patch from Stephen Usher.
2393 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2395 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2398 TF/02 Fix a couple more cases where we did not log the error message
2399 when unlink() failed. See also change 4.74-TF/03.
2401 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2402 lock up or crash if it happened to be inside a call to libc when it
2403 got a SIGUSR1 from exiwhat.
2405 The SIGUSR1 handler appends the current process status to the process
2406 log which is later printed by exiwhat. It used to use the general
2407 purpose logging code to do this, but several functions it calls are
2408 not safe for signals.
2410 The new output code in the SIGUSR1 handler is specific to the process
2411 log, and simple enough that it's easy to inspect for signal safety.
2412 Removing some special cases also simplifies the general logging code.
2413 Removing the spurious timestamps from the process log simplifies
2416 TF/04 Improved ratelimit ACL condition.
2418 The /noupdate option has been deprecated in favour of /readonly which
2419 has clearer semantics. The /leaky, /strict, and /readonly update modes
2420 are mutually exclusive. The update mode is no longer included in the
2421 database key; it just determines when the database is updated. (This
2422 means that when you upgrade Exim will forget old rate measurements.)
2424 Exim now checks that the per_* options are used with an update mode that
2425 makes sense for the current ACL. For example, when Exim is processing a
2426 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2427 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2428 must specify per_mail/readonly. If you omit the update mode it defaults to
2429 /leaky where that makes sense (as before) or /readonly where required.
2431 The /noupdate option is now undocumented but still supported for
2432 backwards compatibility. It is equivalent to /readonly except that in
2433 ACLs where /readonly is required you may specify /leaky/noupdate or
2434 /strict/noupdate which are treated the same as /readonly.
2436 A useful new feature is the /count= option. This is a generalization
2437 of the per_byte option, so that you can measure the throughput of other
2438 aggregate values. For example, the per_byte option is now equivalent
2439 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2441 The per_rcpt option has been generalized using the /count= mechanism
2442 (though it's more complicated than the per_byte equivalence). When it is
2443 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2444 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2445 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2446 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2447 Note that using per_rcpt with a non-readonly update mode in more than
2448 one ACL will cause the recipients to be double-counted. (The per_mail
2449 and per_byte options don't have this problem.)
2451 The handling of very low rates has changed slightly. If the computed rate
2452 is less than the event's count (usually one) then this event is the first
2453 after a long gap. In this case the rate is set to the same as this event's
2454 count, so that the first message of a spam run is counted properly.
2456 The major new feature is a mechanism for counting the rate of unique
2457 events. The new per_addr option counts the number of different
2458 recipients that someone has sent messages to in the last time period. It
2459 behaves like per_rcpt if all the recipient addresses are different, but
2460 duplicate recipient addresses do not increase the measured rate. Like
2461 the /count= option this is a general mechanism, so the per_addr option
2462 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2463 example, measure the rate that a client uses different sender addresses
2464 with the options per_mail/unique=$sender_address. There are further
2465 details in the main documentation.
2467 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2469 TF/06 Removed a few PCRE remnants.
2471 TF/07 Automatically extract Exim's version number from tags in the git
2472 repository when doing development or release builds.
2474 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2475 Bugzilla 879. Patch from Paul Fisher.
2477 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2478 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2481 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2483 PP/05 Variable $av_failed, true if the AV scanner deferred.
2484 Bugzilla 1078. Patch from John Horne.
2486 PP/06 Stop make process more reliably on build failure.
2487 Bugzilla 1087. Patch from Heiko Schlittermann.
2489 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2490 Bugzilla 1089. Patch from Heiko Schlittermann.
2492 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2493 Bugzilla 1131. Patch from Holger Weiß.
2495 PP/09 Handle IPv6 addresses with SPF.
2496 Bugzilla 860. Patch from Wolfgang Breyha.
2498 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2500 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2503 PP/11 match_* no longer expand right-hand-side by default.
2504 New compile-time build option, EXPAND_LISTMATCH_RHS.
2505 New expansion conditions, "inlist", "inlisti".
2507 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2509 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2511 PP/14 fix log_write() format string regression from TF/03.
2512 Bugzilla 1152. Patch from Dmitry Isaikin.
2518 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2520 PP/02 Harmonised TLS library version reporting; only show if debugging.
2521 Layout now matches that introduced for other libraries in 4.74 PP/03.
2523 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2525 PP/04 New "dns_use_edns0" global option.
2527 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2530 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2531 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2533 TK/01 Updated PolarSSL code to 0.14.2.
2534 Bugzilla 1097. Patch from Andreas Metzler.
2536 PP/07 Catch divide-by-zero in ${eval:...}.
2537 Fixes bugzilla 1102.
2539 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2542 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2543 format-string attack -- SECURITY: remote arbitrary code execution.
2545 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2546 time unintentionally subject to list matching rules, letting the header
2547 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2548 arbitrary string expansion). This allowed for information disclosure.
2550 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2551 INT_MIN/-1 -- value coerced to INT_MAX.
2557 NM/01 Workaround for PCRE version dependency in version reporting
2560 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2561 This fixes portability to compilers other than gcc, notably
2562 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2564 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2565 makefiles for portability to HP-UX and POSIX correctness.
2567 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2568 Also via indented variable definition in the Makefile.
2569 (Debugging by Oliver Heesakkers).
2571 PP/02 Restore caching of spamd results with expanded spamd_address.
2572 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2574 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2575 Improves build reliability. Fix from: Frank Elsner
2577 NM/02 Fix wide character breakage in the rfc2047 coding
2578 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2580 NM/03 Allow underscore in dnslist lookups
2581 Fixes bug 1026. Patch from Graeme Fowler
2583 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2584 Code patches from Adam Ciarcinski of NetBSD.
2586 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2589 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2590 is logged, to avoid truncation. Patch from John Horne.
2592 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2593 Patch from Jakob Hirsch.
2595 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2596 SQL string expansion failure details.
2597 Patch from Andrey Oktyabrski.
2599 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2600 Patch from Simon Arlott.
2602 PP/09 New lookups functionality failed to compile on old gcc which rejects
2603 extern declarations in function scope.
2604 Patch from Oliver Fleischmann
2606 PP/10 Use sig_atomic_t for flags set from signal handlers.
2607 Check getgroups() return and improve debugging.
2608 Fixed developed for diagnosis in bug 927 (which turned out to be
2611 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2612 Patch from Mark Zealey.
2614 PP/12 Bugzilla 1056: Improved spamd server selection.
2615 Patch from Mark Zealey.
2617 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2618 Based on patch from Heiko Schlittermann.
2620 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2621 Patch from Uwe Doering, sign-off by Michael Haardt.
2623 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2624 variable declaration deep within a block. Bug and patch from
2627 PP/15 lookups-Makefile IRIX compatibility coercion.
2629 PP/16 Make DISABLE_DKIM build knob functional.
2631 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2632 Patch by Simon Arlott
2634 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2635 variable argument macros. Our copy now differs from upstream.
2641 TF/01 Failure to get a lock on a hints database can have serious
2642 consequences so log it to the panic log.
2644 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2645 controlled using the smtp_confirmation log selector.
2647 TF/03 Include the error message when we fail to unlink a spool file.
2649 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2650 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2651 for maintaining out-of-tree patches for some time.
2653 PP/01 Bugzilla 139: Documentation and portability issues.
2654 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2655 Handle per-OS dynamic-module compilation flags.
2657 PP/02 Let /dev/null have normal permissions.
2658 The 4.73 fixes were a little too stringent and complained about the
2659 permissions on /dev/null. Exempt it from some checks.
2660 Reported by Andreas M. Kirchwitz.
2662 PP/03 Report version information for many libraries, including
2663 Exim version information for dynamically loaded libraries. Created
2664 version.h, now support a version extension string for distributors
2665 who patch heavily. Dynamic module ABI change.
2667 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2668 privilege escalation vulnerability whereby the Exim run-time user
2669 can cause root to append content of the attacker's choosing to
2672 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2675 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2676 If dropping privileges for untrusted macros, we disabled normal logging
2677 on the basis that it would fail; for the Exim run-time user, this is not
2678 the case, and it resulted in successful deliveries going unlogged.
2679 Fixed. Reported by Andreas Metzler.
2685 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2686 only prepend for the Resent-* case. Fixes regression introduced in
2687 Exim 4.70 by NM/22 for Bugzilla 607.
2689 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2690 increasing numbers of administrators be bitten by this.
2692 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2694 PP/03 Bugzilla 994: added openssl_options main configuration option.
2696 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2698 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2700 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2702 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2703 without a peer certificate, leading to a segfault because of an
2704 assumption that peers always have certificates. Be a little more
2705 paranoid. Problem reported by Martin Tscholak.
2707 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2708 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2709 NB: ClamAV planning to remove STREAM in "middle of 2010".
2710 CL also introduces -bmalware, various -d+acl logging additions and
2711 more caution in buffer sizes.
2713 PP/09 Implemented reverse_ip expansion operator.
2715 PP/10 Bugzilla 937: provide a "debug" ACL control.
2717 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2719 PP/12 Bugzilla 973: Implement --version.
2721 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2723 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2725 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2727 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2728 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2729 ignore trailing whitespace.
2731 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2733 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2736 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2737 Notification from Dr Andrew Aitchison.
2739 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2740 ExtendedDetectionInfo response format.
2741 Notification from John Horne.
2743 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2746 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2747 XSL and documented dependency on system catalogs, with examples of how
2750 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2753 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2754 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2755 configuration file which is writeable by the Exim user or group.
2757 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2758 of configuration files to cover files specified with the -C option if
2759 they are going to be used with root privileges, not just the default
2762 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2763 option (effectively making it always true).
2765 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2766 files to be used while preserving root privileges.
2768 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2769 that rogue child processes cannot use them.
2771 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2772 run-time user, instead of root.
2774 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2775 Exim run-time user without dropping privileges.
2777 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2778 result string, instead of calling string_vformat() twice with the same
2781 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2782 for other users. Others should always drop root privileges if they use
2783 -C on the command line, even for a whitelisted configure file.
2785 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2787 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2793 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2794 $data_path, and $header_path variables; fixed documentation bugs and
2797 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2798 exipick to access non-standard spools, including the "frozen" queue
2801 NM/01 Bugzilla 965: Support mysql stored procedures.
2802 Patch from Alain Williams
2804 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2806 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2807 Patch from Andreas Metzler
2809 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2810 Patch from Kirill Miazine
2812 NM/05 Bugzilla 671: Added umask to procmail example.
2814 JJ/03 installed exipick 20100323.0, fixing doc bug
2816 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2817 directory. Notification and patch from Dan Rosenberg.
2819 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2821 TK/02 Improve log output when DKIM signing operation fails.
2823 MH/01 Treat the transport option dkim_domain as a colon separated
2824 list, not as a single string, and sign the message with each element,
2825 omitting multiple occurences of the same signer.
2827 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2828 Bugzilla 985, 986. Patch by Simon Arlott
2830 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2831 Patch by Simon Arlott
2833 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2834 MBX locking. Notification from Dan Rosenberg.
2840 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2842 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2844 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2846 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2848 NM/04 Bugzilla 915: Flag broken perl installation during build.
2854 TK/01 Added patch by Johannes Berg that expands the main option
2855 "spamd_address" if it starts with a dollar sign.
2857 TK/02 Write list of recipients to X-Envelope-Sender header when building
2858 the mbox-format spool file for content scanning (suggested by Jakob
2861 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2862 (http://www.dcc-servers.net/) support via dccifd. Activated by
2863 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2865 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2866 by Mark Daniel Reidel <mr@df.eu>.
2868 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2869 When building exim an external PCRE library is now needed -
2870 PCRE is a system library on the majority of modern systems.
2871 See entry on PCRE_LIBS in EDITME file.
2873 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2874 conversation. Added nologin parameter to request.
2875 Patch contributed by Kirill Miazine.
2877 TF/01 Do not log submission mode rewrites if they do not change the address.
2879 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2881 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2882 log files in place. Contributed by Roberto Lima.
2884 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2886 TF/03 Bugzilla 615: When checking the local_parts router precondition
2887 after a local_part_suffix or local_part_prefix option, Exim now
2888 does not use the address's named list lookup cache, since this
2889 contains cached lookups for the whole local part.
2891 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2892 Robert Millan. Documentation is in experimental-spec.txt.
2894 TF/04 Bugzilla 668: Fix parallel build (make -j).
2896 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2898 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2899 Patch provided by Jan Srzednicki.
2901 TF/05 Leading white space used to be stripped from $spam_report which
2902 wrecked the formatting. Now it is preserved.
2904 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2905 that they are available at delivery time.
2907 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2909 TF/08 TLS error reporting now respects the incoming_interface and
2910 incoming_port log selectors.
2912 TF/09 Produce a more useful error message if an SMTP transport's hosts
2913 setting expands to an empty string.
2915 NM/06 Bugzilla 744: EXPN did not work under TLS.
2916 Patch provided by Phil Pennock.
2918 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2919 Patch provided by Richard Godbee.
2921 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2922 acl_smtp_notquit, added index entry.
2924 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2925 Patch provided by Eugene Bujak.
2927 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2928 accept(). Patch provided by Maxim Dounin.
2930 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2931 Patch provided by Phil Pennock.
2933 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2935 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2936 Patch provided by Brad "anomie" Jorsch.
2938 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2939 Patch provided by Dean Brooks.
2941 TK/05 Add native DKIM support (does not depend on external libraries).
2943 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2944 Patch provided by Graeme Fowler.
2946 NM/16 Bugzilla 851: Documentation example syntax fix.
2948 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2950 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2953 NM/19 Bugzilla 745: TLS version reporting.
2954 Patch provided by Phil Pennock.
2956 NM/20 Bugzilla 167: bool: condition support.
2957 Patch provided by Phil Pennock.
2959 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2960 clients. Patch provided by Phil Pennock.
2962 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2963 Patch provided by Brad "anomie" Jorsch.
2965 NM/23 Bugzilla 687: Fix misparses in eximstats.
2966 Patch provided by Heiko Schlittermann.
2968 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2969 Patch provided by Heiko Schlittermann.
2971 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2972 plus update to original patch.
2974 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2976 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2977 Patch provided by David Brownlee.
2979 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2981 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2983 NM/30 Bugzilla 888: TLS documentation bugfixes.
2985 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2987 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2988 Unlike the original bugzilla I have changed all shell scripts in src tree.
2990 NM/33 Bugzilla 898: Transport filter timeout fix.
2991 Patch by Todd Rinaldo.
2993 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2994 Patch by Serge Demonchaux.
2996 NM/35 Bugzilla 39: Base64 decode bug fixes.
2997 Patch by Jakob Hirsch.
2999 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3001 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3003 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3005 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3011 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3012 ALT-N's libdkim that I have put here:
3013 http://duncanthrax.net/exim-experimental/
3015 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3016 were called 'true' and it seems that C99 defines that as a reserved
3017 keyword to be used with 'bool' variable types. That means you could
3018 not include C99-style headers which use bools without triggering
3019 build errors in sieve.c.
3021 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3022 as mailq or other aliases. Changed the --help handling significantly
3023 to do whats expected. exim_usage() emits usage/help information.
3025 SC/01 Added the -bylocaldomain option to eximstats.
3027 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3029 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3031 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3037 PH/01 Another patch from the Sieve maintainer.
3039 PH/02 When an IPv6 address is converted to a string for single-key lookup
3040 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3041 dots are used instead of colons so that keys in lsearch files need not
3042 contain colons. This was done some time before quoting was made available
3043 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3044 (notated using the quote facility) so as to distinguish them from IPv4
3045 keys. This meant that lookups for IP addresses in host lists did not work
3046 for iplsearch lookups.
3048 This has been fixed by arranging for IPv6 addresses to be expressed with
3049 colons if the lookup type is iplsearch. This is not incompatible, because
3050 previously such lookups could never work.
3052 The situation is now rather anomalous, since one *can* have colons in
3053 ordinary lsearch keys. However, making the change in all cases is
3054 incompatible and would probably break a number of configurations.
3056 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3059 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3060 conversion specification without a maximum field width, thereby enabling
3061 a rogue spamd server to cause a buffer overflow. While nobody in their
3062 right mind would setup Exim to query an untrusted spamd server, an
3063 attacker that gains access to a server running spamd could potentially
3064 exploit this vulnerability to run arbitrary code as the Exim user.
3066 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3067 $primary_hostname instead of what libspf2 thinks the hosts name is.
3069 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3070 a directory entry by the name of the lookup key. Previously, if a
3071 symlink pointed to a non-existing file or a file in a directory that
3072 Exim lacked permissions to read, a lookup for a key matching that
3073 symlink would fail. Now it is enough that a matching directory entry
3074 exists, symlink or not. (Bugzilla 503.)
3076 PH/03 The body_linecount and body_zerocount variables are now exported in the
3079 PH/04 Added the $dnslist_matched variable.
3081 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3082 This means they are set thereafter only if the connection becomes
3085 PH/06 Added the client_condition to authenticators so that some can be skipped
3086 by clients under certain conditions.
3088 PH/07 The error message for a badly-placed control=no_multiline_responses left
3089 "_responses" off the end of the name.
3091 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3093 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3094 (without spaces) instead of just copying the configuration text.
3096 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3098 PH/11 Added $max_received_linelength.
3100 PH/12 Added +ignore_defer and +include_defer to host lists.
3102 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3103 way in which PCRE > 7.0 is built.
3105 PH/14 Implemented queue_only_load_latch.
3107 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3108 MAIL command. The effect was to mangle the value on 64-bit systems.
3110 PH/16 Another patch from the Sieve maintainer.
3112 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3114 PH/18 If a system quota error occurred while trying to create the file for
3115 a maildir delivery, the message "Mailbox is full" was not appended to the
3116 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3117 only to a quota excession during the actual writing of the file.
3119 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3120 characters?) which causes problems in log lines. The DN values are now
3121 passed through string_printing() before being added to log lines.
3123 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3124 and InterBase are left for another time.)
3126 PH/21 Added message_body_newlines option.
3128 PH/22 Guard against possible overflow in moan_check_errorcopy().
3130 PH/23 POSIX allows open() to be a macro; guard against that.
3132 PH/24 If the recipient of an error message contained an @ in the local part
3133 (suitably quoted, of course), incorrect values were put in $domain and
3134 $local_part during the evaluation of errors_copy.
3140 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3141 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3144 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3145 issue a MAIL command.
3147 PH/02 In an ACL statement such as
3149 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3151 if a client was not listed at all, or was listed with a value other than
3152 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3153 the condition was not true (as it should be), so access was not denied.
3154 The bug was that the ! inversion was incorrectly passed on to the second
3155 item. This has been fixed.
3157 PH/03 Added additional dnslists conditions == and =& which are different from
3158 = and & when the dns lookup returns more than one IP address.
3160 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3161 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3163 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3164 FSYNC, which compiles an option called disable_fsync that allows for
3165 bypassing fsync(). The documentation is heavily laced with warnings.
3167 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3169 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3170 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3171 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3172 including adding "make clean"; (3) Added -fPIC when compiling the test
3173 dynamically loaded module, to get rid of a warning.
3175 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3176 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3177 = 0s. The bug is otherwise harmless.
3179 PH/07 There was a bug in the dovecot authenticator such that the value of
3180 $auth1 could be overwritten, and so not correctly preserved, after a
3181 successful authentication. This usually meant that the value preserved by
3182 the server_setid option was incorrect.
3184 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3186 PH/09 Installed PCRE release 7.0.
3188 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3189 run for batched SMTP input. It is now run at the start of every message
3190 in the batch. While fixing this I discovered that the process information
3191 (output by running exiwhat) was not always getting set for -bs and -bS
3192 input. This is fixed, and it now also says "batched" for BSMTP.
3194 PH/11 Added control=no_pipelining.
3196 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3197 patch, slightly modified), and move the expansion of helo_data till after
3198 the connection is made in the smtp transport (so it can use these
3201 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3203 PH/14 Added log_selector = +pid.
3205 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3207 PH/16 Add ${if forany and ${if forall.
3209 PH/17 Added dsn_from option to vary the From: line in DSNs.
3211 PH/18 Flush SMTP output before performing a callout, unless control =
3212 no_callout_flush is set.
3214 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3215 was true (the default) a successful delivery failed to delete the retry
3216 item, thus causing premature timeout of the address. The bug is now
3219 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3221 PH/21 Long custom messages for fakedefer and fakereject are now split up
3222 into multiline responses in the same way that messages for "deny" and
3223 other ACL rejections are.
3225 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3226 with slight modification.
3228 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3229 draft, changing the syntax and factoring some duplicate code".
3231 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3232 for deliveries of the second and subsequent messages over the same SMTP
3235 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3236 ${reduce, with only minor "tidies".
3238 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3240 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3241 expansion side effects.
3243 PH/27 When a message times out after an over-quota error from an Exim-imposed
3244 quota, the bounce message says "mailbox is full". This message was not
3245 being given when it was a system quota that was exceeded. It now should
3248 MH/03 Made $recipients available in local_scan(). local_scan() already has
3249 better access to the recipient list through recipients_list[], but
3250 $recipients can be useful in postmaster-provided expansion strings.
3252 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3253 in the case of a MAIL command with additional options following the
3254 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3255 were accidentally chopped off.
3257 PH/29 SMTP synchronization checks are implemented when a command is read -
3258 there is a check that no more input is waiting when there shouldn't be
3259 any. However, for some commands, a delay in an ACL can mean that it is
3260 some time before the response is written. In this time, more input might
3261 arrive, invalidly. So now there are extra checks after an ACL has run for
3262 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3263 pipelining has not been advertised.
3265 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3267 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3268 returned in the SMTP response when smtp_return_error_details was false.
3269 This has been fixed.
3271 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3272 instead of the C I/O that was originally supplied, because problems were
3273 reported on Solaris.
3275 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3276 Exim which did not show up earlier: it was assuming that a call to
3277 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3278 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3279 was a macro that became an assignment, so it seemed to work. This has
3280 changed to a proper function call with a void return, hence the compile
3281 error. Exim's code has been fixed.
3283 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3286 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3288 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3290 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3291 because it turns out to be too big or there is a timeout) there is no
3292 "Completed" line in the log. When some messages of this type were
3293 selected by exigrep, they were listed as "not completed". Others were
3294 picked up by some special patterns. I have improved the selection
3295 criteria to be more general.
3297 PH/38 The host_find_failed option in the manualroute router can now be set
3298 to "ignore", to completely ignore a host whose IP address cannot be
3299 found. If all hosts are ignored, the behaviour is controlled by the new
3300 host_all_ignored option.
3302 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3303 homing or because of multiple MX records with /mx) generated more than
3304 one IP address, and the following item turned out to be the local host,
3305 all the secondary addresses of the first item were incorrectly removed
3306 from the list, along with the local host and any following hosts (which
3307 is what is supposed to happen).
3309 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3310 whoever called Exim into the -H file. In the case of the daemon it was
3311 behaving confusingly. When first started, it used values for whoever
3312 started the daemon, but after a SIGHUP it used the Exim user (because it
3313 calls itself on a restart). I have changed the code so that it now always
3316 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3317 message are rejected with the same error (e.g. no authentication or bad
3318 sender address), and a DATA command is nevertheless sent (as can happen
3319 with PIPELINING or a stupid MUA), the error message that was given to the
3320 RCPT commands is included in the rejection of the DATA command. This is
3321 intended to be helpful for MUAs that show only the final error to their
3324 PH/42 Another patch from the Sieve maintainer.
3326 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3327 Eximstats - Fixed some broken HTML links and added missing column headers
3329 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3330 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3332 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3334 PH/43 Yet another patch from the Sieve maintainer.
3336 PH/44 I found a way to check for a TCP/IP connection going away before sending
3337 the response to the final '.' that terminates a message, but only in the
3338 case where the client has not sent further data following the '.'
3339 (unfortunately, this is allowed). However, in many cases there won't be
3340 any further data because there won't be any more messages to send. A call
3341 to select() can be used: if it shows that the input is "ready", there is
3342 either input waiting, or the socket has been closed. An attempt to read
3343 the next input character can distinguish the two cases. Previously, Exim
3344 would have sent an OK response which the client would never have see.
3345 This could lead to message repetition. This fix should cure that, at
3346 least in a lot of common cases.
3348 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3349 advertised in response to EHLO.
3355 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3356 fixed by 4.65/MH/01 (is this a record?) are fixed:
3358 (i) An empty string was always treated as zero by the numeric comparison
3359 operators. This behaviour has been restored.
3361 (ii) It is documented that the numeric comparison operators always treat
3362 their arguments as decimal numbers. This was broken in that numbers
3363 starting with 0 were being interpreted as octal.
3365 While fixing these problems I realized that there was another issue that
3366 hadn't been noticed. Values of message_size_limit (both the global option
3367 and the transport option) were treated as octal if they started with 0.
3368 The documentation was vague. These values are now always treated as
3369 decimal, and I will make that clear in the documentation.
3375 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3376 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3379 MH/01 Don't check that the operands of numeric comparison operators are
3380 integers when their expansion is in "skipping" mode (fixes bug
3381 introduced by 4.64-PH/07).
3383 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3384 child addresses, Exim now panics and dies. Previously, because the count
3385 is held in a short int, deliveries were likely to be lost. As such a
3386 large number of recipients for a single message is ridiculous
3387 (performance will be very, very poor), I have chosen to impose a limit
3388 rather than extend the field.
3394 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3395 leftover -K file (the existence of which was triggered by #402).
3396 While we were at it, introduced process PID as part of the -K
3397 filename. This should rule out race conditions when creating
3400 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3401 processing considerably. Previous code took too long for large mails,
3402 triggering a timeout which in turn triggers #401.
3404 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3405 in the DK code in transports.c. sendfile() is not really portable,
3406 hence the _LINUX specificness.
3408 TF/01 In the add_headers option to the mail command in an Exim filter,
3409 there was a bug that Exim would claim a syntax error in any
3410 header after the first one which had an odd number of characters
3413 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3414 callout verification, Exim cached a "reject" for the entire domain. This
3415 is correct for most verifications, but it is not correct for a recipient
3416 verification with use_sender or use_postmaster set, because in that case
3417 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3418 case of MAIL FROM:<> rejection from other early rejections (e.g.
3419 rejection of HELO). When verifying a recipient using a non-null MAIL
3420 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3421 Whatever the result of the callout, the value of the domain cache is
3422 left unchanged (for any other kind of callout, getting as far as trying
3423 RCPT means that the domain itself is ok).
3425 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3428 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3429 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3430 session, the daemon ignores ECONNECT errors and logs others; it now
3431 ignores EPIPE as well.
3433 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3434 (quoted-printable decoding).
3436 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3437 later the small subsequent patch to fix an introduced bug.
3439 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3441 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3443 PH/08 An error is now given if message_size_limit is specified negative.
3445 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3446 to be given (somewhat) arbitrary names.
3448 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3451 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3452 miscellaneous code fixes
3454 PH/10 Added the log_reject_target ACL modifier to specify where to log
3457 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3458 hostname. This is wrong, because it relates to the incoming message (and
3459 probably the interface on which it is arriving) and not to the outgoing
3460 callout (which could be using a different interface). This has been
3461 changed to use the value of the helo_data option from the smtp transport
3462 instead - this is what is used when a message is actually being sent. If
3463 there is no remote transport (possible with a router that sets up host
3464 addresses), $smtp_active_hostname is used.
3466 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3467 tweaks were necessary in order to get it to work (see also 21 below):
3468 (a) The code assumed that strncpy() returns a negative number on buffer
3469 overflow, which isn't the case. Replaced with Exim's string_format()
3471 (b) There were several signed/unsigned issues. I just did the minimum
3472 hacking in of casts. There is scope for a larger refactoring.
3473 (c) The code used strcasecmp() which is not a standard C function.
3474 Replaced with Exim's strcmpic() function.
3475 (d) The code set only $1; it now sets $auth1 as well.
3476 (e) A simple test gave the error "authentication client didn't specify
3477 service in request". It would seem that Dovecot has changed its
3478 interface. Fortunately there's a specification; I followed it and
3479 changed what the client sends and it appears to be working now.
3481 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3484 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3485 address A is aliased to B and C, where B exists and C does not. Without
3486 -v the output is "A verified" because verification stops after a
3487 successful redirection if more than one address is generated. However,
3488 with -v the child addresses are also verified. Exim was outputting "A
3489 failed to verify" and then showing the successful verification for C,
3490 with its parentage. It now outputs "B failed to verify", showing B's
3491 parentage before showing the successful verification of C.
3493 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3494 look up a TXT record in a specific list after matching in a combined
3497 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3498 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3499 they consult the DNS. I had assumed they would set it the way they
3500 wanted; and indeed my experiments on Linux seem to show that in some
3501 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3502 To be on the safe side, however, I have now made the interface to
3503 host_find_byname() similar to host_find_bydns(), with an argument
3504 containing the DNS resolver options. The host_find_byname() function now
3505 sets these options at its start, just as host_find_bydns() does. The smtp
3506 transport options dns_qualify_single and dns_search_parents are passed to
3507 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3508 of host_find_byname() use the default settings of RES_DEFNAMES
3509 (qualify_single) but not RES_DNSRCH (search_parents).
3511 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3512 spool_read_header() do less string testing, by means of a preliminary
3513 switch on the second character of optional "-foo" lines. (This is
3514 overdue, caused by the large number of possibilities that now exist.
3515 Originally there were few.) While I was there, I also converted the
3516 str(n)cmp tests so they don't re-test the leading "-" and the first
3517 character, in the hope this might squeeze out yet more improvement.
3519 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3520 flag allowing group syntax was set by the header_syntax check but not
3521 turned off, possible causing trouble later; (2) The flag was not being
3522 set at all for the header_verify test, causing "group"-style headers to
3523 be rejected. I have now set it in this case, and also caused header_
3524 verify to ignore an empty address taken from a group. While doing this, I
3525 came across some other cases where the code for allowing group syntax
3526 while scanning a header line wasn't quite right (mostly, not resetting
3527 the flag correctly in the right place). These bugs could have caused
3528 trouble for malformed header lines. I hope it is now all correct.
3530 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3531 with the "reply" argument non-NULL. The code, however (which originally
3532 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3533 but it didn't always do it. This confused somebody who was copying the
3534 code for some other use. I have removed all the tests.
3536 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3537 feature that was used to support insecure browsers during the U.S. crypto
3538 embargo. It requires special client support, and Exim is probably the
3539 only MTA that supported it -- and would never use it because real RSA is
3540 always available. This code has been removed, because it had the bad
3541 effect of slowing Exim down by computing (never used) parameters for the
3542 RSA_EXPORT functionality.
3544 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3545 authenticator to fail if there's a tab character in the incoming data
3546 (there should never be unless someone is messing about, as it's supposed
3547 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3548 if the connection is using TLS or if the remote IP is the same as the
3549 local IP, and the "valid-client-cert option" if a client certificate has
3552 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3553 authenticators. This can be used for authorization after authentication
3554 succeeds. (In the case of plaintext, it servers for both authentication
3557 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3558 if any retry times were supplied.
3560 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3561 connection, where there is no client IP address to check. In this
3562 situation, the verify now always succeeds.
3564 PH/25 Applied John Jetmore's -Mset patch.
3566 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3568 PH/27 In a string expansion for a processed (not raw) header when multiple
3569 headers of the same name were present, leading whitespace was being
3570 removed from all of them, but trailing whitespace was being removed only
3571 from the last one. Now trailing whitespace is removed from each header
3572 before concatenation. Completely empty headers in a concatenation (as
3573 before) are ignored.
3575 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3576 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3578 PH/29 [Removed. This was a change that I later backed out, and forgot to
3579 correct the ChangeLog entry (that I had efficiently created) before
3580 committing the later change.]
3582 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3583 address errors (4xx response to RCPT) over the same connection as other
3584 messages routed to the same hosts. Such deliveries are always "forced",
3585 so retry times are not inspected. This resulted in far too many retries
3586 for the affected addresses. The effect occurred only when there were more
3587 hosts than the hosts_max_try setting in the smtp transport when it had
3588 the 4xx errors. Those hosts that it had tried were not added to the list
3589 of hosts for which the message was waiting, so if all were tried, there
3590 was no problem. Two fixes have been applied:
3592 (i) If there are any address or message errors in an SMTP delivery, none
3593 of the hosts (tried or untried) are now added to the list of hosts
3594 for which the message is waiting, so the message should not be a
3595 candidate for sending over the same connection that was used for a
3596 successful delivery of some other message. This seems entirely
3597 reasonable: after all the message is NOT "waiting for some host".
3598 This is so "obvious" that I'm not sure why it wasn't done
3599 previously. Hope I haven't missed anything, but it can't do any
3600 harm, as the worst effect is to miss an optimization.
3602 (ii) If, despite (i), such a delivery is accidentally attempted, the
3603 routing retry time is respected, so at least it doesn't keep
3604 hammering the server.
3606 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3607 in ${readsocket because some servers need this prod.
3609 PH/32 Added some extra debug output when updating a wait-xxx database.
3611 PH/33 The hint "could be header name not terminated by colon", which has been
3612 given for certain expansion errors for a long time, was not being given
3613 for the ${if def:h_colon_omitted{... case.
3615 PH/34 The spec says: "With one important exception, whenever a domain list is
3616 being scanned, $domain contains the subject domain." There was at least
3617 one case where this was not true.
3619 PH/35 The error "getsockname() failed: connection reset by peer" was being
3620 written to the panic log as well as the main log, but it isn't really
3621 panic-worthy as it just means the connection died rather early on. I have
3622 removed the panic log writing for the ECONNRESET error when getsockname()
3625 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3626 runs only) independently of the message's sender address. This meant
3627 that, if the 4xx error was in fact related to the sender, a different
3628 message to the same recipient with a different sender could confuse
3629 things. In particular, this can happen when sending to a greylisting
3630 server, but other circumstances could also provoke similar problems.
3631 I have changed the default so that the retry time for these errors is now
3632 based a combination of the sender and recipient addresses. This change
3633 can be overridden by setting address_retry_include_sender=false in the
3636 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3637 remote server are returned as part of bounce messages. This was not
3638 happening for LMTP over a pipe (the lmtp transport), but now it is the
3639 same for both kinds of LMTP.
3641 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3642 in header lines that were in fact CNAMEs. This is no longer the case.
3644 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3645 and queue runs started by the daemon processed all messages. This has
3646 been fixed so that -R and -S can now usefully be given with -q<time>.
3648 PH/40 Import PCRE release 6.7 (fixes some bugs).
3650 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3652 PH/42 Give an error if -q is specified more than once.
3654 PH/43 Renamed the variables $interface_address and $interface_port as
3655 $received_ip_address and $received_port, to make it clear that these
3656 values apply to message reception, and not to the outgoing interface when
3657 a message is delivered. (The old names remain recognized, of course.)
3659 PH/44 There was no timeout on the connect() call when using a Unix domain
3660 socket in the ${readsocket expansion. There now is.
3662 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3663 be meaningful with "accept".
3665 SC/01 Eximstats V1.43
3666 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3668 SC/02 Eximstats V1.44
3669 Use a glob alias rather than an array ref in the generated
3670 parser. This improves both readability and performance.
3672 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3673 Collect SpamAssassin and rejection statistics.
3674 Don't display local sender or destination tables unless
3675 there is data to show.
3676 Added average volumes into the top table text output.
3678 SC/04 Eximstats V1.46
3679 Collect data on the number of addresses (recipients)
3680 as well as the number of messages.
3682 SC/05 Eximstats V1.47
3683 Added 'Message too big' to the list of mail rejection
3684 reasons (thanks to Marco Gaiarin).
3686 SC/06 Eximstats V1.48
3687 Mainlog lines which have GMT offsets and are too short to
3688 have a flag are now skipped.
3690 SC/07 Eximstats V1.49 (Alain Williams)
3691 Added the -emptyok flag.
3693 SC/08 Eximstats V1.50
3694 Fixes for obtaining the IP address from reject messages.
3696 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3697 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3698 whitespace changes from 4.64-PH/27
3700 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3703 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3704 are found, allow negative numbers in numeric criteria)
3706 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3708 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3709 to match changes made in 4.64-PH/43
3711 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3713 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3714 "domain" is actually an IP address, and force a failure. This locks out
3715 those revolvers/nameservers that support "A-for-A" lookups, in
3716 contravention of the specifications.
3718 PH/48 When a host name was looked up from an IP address, and the subsequent
3719 forward lookup of the name timed out, the host name was left in
3720 $sender_host_name, contrary to the specification.
3722 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3723 restricted to single-key lookups, Exim was not diagnosing an error if
3724 * or *@ was used with a query-style lookup.
3726 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3728 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3729 long ago, but noone interested enough thought of it. Let's just say that
3730 the "1.1" means that there are some new functions that weren't there at
3731 some point in the past.
3733 PH/51 Error processing for expansion failure of helo_data from an smtp
3734 transport during callout processing was broken.
3736 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3737 tested/used via the -bh/-bhc/-bs options.
3739 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3740 bug, fixed in subsequent PCRE releases).
3742 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3743 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3745 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3751 SC/01 Use a glob alias rather than an array ref in eximstats generated
3752 parser. This improves both readability and performance.
3754 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3755 Don't display local sender or destination tables in eximstats unless
3756 there is data to show.
3757 Added average volumes into the eximstats top table text output.
3759 SC/03 Collect data on the number of addresses (recipients) as well
3760 as the number of messages in eximstats.
3762 TF/01 Correct an error in the documentation for the redirect router. Exim
3763 does (usually) call initgroups() when daemonizing.
3765 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3766 with consistent privilege compared to when running as a daemon.
3768 TF/03 Note in the spec that $authenticated_id is not set for local
3769 submissions from trusted users.
3771 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3772 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3774 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3775 by adding some example configuration directives to the default
3776 configuration file. A little bit of work is required to uncomment the
3777 directives and define how usernames and passwords are checked, but
3778 there is now a framework to start from.
3780 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3781 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3782 without this. I don't know how relevant this is to other LDAP libraries.
3784 PH/02 Add the verb name to the "unknown ACL verb" error.
3786 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3788 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3790 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3791 directory not expanded when it should be if an expanded home directory
3792 was set for the address (which is overridden by the transport).
3794 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3797 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3798 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3799 because it is too late at that time, and has no effect.
3801 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3802 security issue with \' (bugzilla #107). I could not use the
3803 PQescapeStringConn() function, because it needs a PGconn value as one of
3806 PH/08 When testing addresses using -bt, indicate those final addresses that
3807 are duplicates that would not cause an additional delivery. At least one
3808 person was confused, thinking that -bt output corresponded to deliveries.
3809 (Suppressing duplicates isn't a good idea as you lose the information
3810 about possibly different redirections that led to the duplicates.)
3812 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3813 systems where poll() doesn't work, in particular OS X.
3815 PH/10 Added more information to debugging output for retry time not reached.
3817 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3818 operations in malware.c.
3820 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3823 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3824 syslog_duplication set false, log lines that would normally be written
3825 both the the main log and to the reject log were not written to syslog at
3828 PH/14 In the default configuration, change the use of "message" in ACL warn
3829 statements to "add_header".
3831 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3832 not followed by a command (e.g. "seen endif").
3834 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3835 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3838 PH/17 Added extra conditions to the default value of delay_warning_condition
3842 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3843 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3844 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3847 The Auto-Submitted: and various List- headers are standardised, whereas I
3848 don't think Precedence: ever was.
3850 PH/18 Refactored debugging code in route_finduser() to show more information,
3851 in particular, the error code if getpwnam() issues one.
3853 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3854 This is apparently needed in addition to the PH/07 change above to avoid
3855 any possible encoding problems.
3857 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3858 but not after initializing Perl.
3860 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3861 output them only if debugging. By default they are written stderr,
3862 apparently, which is not desirable.
3864 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3867 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3870 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3872 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3873 authenticated or an ident call has been made. Suppress the default
3874 values for $authenticated_id and $authenticated_sender (but permit -oMai
3875 and -oMas) when testing with -bh.
3877 PH/24 Re-jigged the order of the tests in the default configuration so that the
3878 tests for valid domains and recipients precede the DNS black list and CSA
3879 tests, on the grounds that those ones are more expensive.
3881 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3882 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3883 This bug exists in every version of Exim that I still have, right back to
3886 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3887 However, an attempt to turn on case-sensitivity in a regex key by
3888 including (?-i) didn't work because the subject string was already
3889 lowercased, and the effects were non-intuitive. It turns out that a
3890 one-line patch can be used to allow (?-i) to work as expected.
3896 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3897 other effects) broke the use of negated acl sub-conditions.
3899 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3902 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3903 "Deny" causes Exim to reject the incoming connection with a 554 error.
3904 Unfortunately, if there is a major crisis, such as a disk failure,
3905 tcp-wrappers gives "deny", whereas what one would like would be some
3906 kind of temporary error. A kludge has been added to help with this.
3907 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3908 554 error is used if errno is still zero or contains ENOENT (which occurs
3909 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3912 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3914 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3915 errors. Otherwise a message that provokes a temporary error (when other
3916 messages do not) can cause a whole host to time out.
3918 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3919 addresses were routed directly to files or pipes from a redirect router.
3920 File deliveries just didn't batch; pipe deliveries might have suffered
3923 PH/06 A failure to get a lock for a hints database would erroneously always say
3924 "Failed to get write lock", even when it was really a read lock.
3926 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3927 of 0600. This has been changed to use the value of the lockfile_mode
3928 option (which defaults to 0600).
3930 PH/08 Applied small patch from the Sieve maintainer.
3932 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3933 folder from quota calculations, a direct delivery into this folder messed
3934 up the contents of the maildirsize file. This was because the regex was
3935 used only to exclude .Trash (or whatever) when the size of the mailbox
3936 was calculated. There was no check that a delivery was happening into an
3937 excluded directory. This bug has been fixed by ignoring all quota
3938 processing for deliveries into excluded directories.
3940 PH/10 Added the maildirfolder_create_regex option to appendfile.
3946 PH/01 The code for finding all the local interface addresses on a FreeBSD
3947 system running IPv6 was broken. This may well have applied to all BSD
3948 systems, as well as to others that have similar system calls. The broken
3949 code found IPv4 interfaces correctly, but gave incorrect values for the
3950 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3951 that it would not match correctly against @[] and not recognize the IPv6
3954 PH/02 The ipliteral router was not recognizing addresses of the form user@
3955 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3957 PH/03 Added disable_ipv6.
3959 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3960 decoded form, because it is most often used to construct To: headers
3961 lines in autoreplies, and the decoded form may well be syntactically
3962 invalid. However, $reply_address has leading white space removed, and all
3963 newlines turned into spaces so that the autoreply transport does not
3966 PH/05 If group was specified without a user on a router, and no group or user
3967 was specified on a transport, the group from the router was ignored.
3969 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3970 for visible compile-time settings that can be used to change these
3971 numbers, for those that want even more. Backwards compatibility with old
3972 spool files has been maintained. However, going back to a previous Exim
3973 release will lost any variables that are in spool files.
3975 PH/07 Two small changes when running in the test harness: increase delay when
3976 passing a TCP/IP connection to a new process, in case the original
3977 process has to generate a bounce, and remove special handling of
3978 127.0.0.2 (sic), which is no longer necessary.
3980 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3981 be the same on different OS.
3983 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3986 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3987 whether --show-vars was specified or not
3989 JJ/02 exipick: Added support for new ACL variable spool format introduced
3992 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3993 syntactically invalid From: or Reply-to: line, and a filter used this to
3994 generate an autoreply, and therefore failed to obtain an address for the
3995 autoreply, Exim could try to deliver to a non-existent relative file
3996 name, causing unrelated and misleading errors. What now happens is that
3997 it logs this as a hard delivery error, but does not attempt to create a
4000 PH/11 The exinext utility has a -C option for testing purposes, but although
4001 the given file was scanned by exinext itself; it wasn't being passed on
4002 when Exim was called.
4004 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4005 an end-of-file indication when reading a command response.
4007 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4008 compiled. In many other places in Exim, IPv6 addresses are always
4009 recognized, so I have changed this. It also means that IPv4 domain
4010 literals of the form [IPV4:n.n.n.n] are now always recognized.
4012 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4013 used if the router is not running as root, for example, when verifying at
4014 ACL time, or when using -bh. The debugging output from this situation was
4015 non-existent - all you got was a failure to exec. I have made two
4018 (a) Failures to set uid/gid, the current directory, or a process leader
4019 in a subprocess such as that created by queryprogram now generate
4020 suitable debugging output when -d is set.
4022 (b) The queryprogram router detects when it is not running as root,
4023 outputs suitable debugging information if -d is set, and then runs
4024 the subprocess without attempting to change uid/gid.
4026 PH/15 Minor change to Makefile for building test_host (undocumented testing
4029 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4030 additional section of a DNS packet that returns MX or SRV records.
4031 Instead, it always explicitly searches for A/AAAA records. This avoids
4032 major problems that occur when a DNS server includes only records of one
4033 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4034 fixed another bug: if SRV records were looked up and the corresponding
4035 address records were *not* found in the additional section, the port
4036 values from the SRV records were lost.
4038 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4039 using the correct key (the original address) when searching the retry
4040 rules in order to find which one to use for generating the retry hint.
4042 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4043 adding the default one. Similarly, if it contains a Reply-To: header, the
4044 errors_reply_to option, if set, is not used.
4046 PH/19 When calculating a retry time, Exim used to measure the "time since
4047 failure" by looking at the "first failed" field in the retry record. Now
4048 it does not use this if it is later than than the arrival time of the
4049 message. Instead it uses the arrival time. This makes for better
4050 behaviour in cases where some deliveries succeed, thus re-setting the
4051 "first failed" field. An example is a quota failure for a huge message
4052 when small messages continue to be delivered. Without this change, the
4053 "time since failure" will always be short, possible causing more frequent
4054 delivery attempts for the huge message than are intended.
4055 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4057 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4058 $1, $2, $3) because the numerical variables can be reset during some
4059 expansion items (e.g. "match"), thereby losing the authentication data.
4061 PH/21 Make -bV show the size of off_t variables so that the test suite can
4062 decide whether to run tests for quotas > 2G.
4064 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4065 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4066 filecount value is greater than 2G or if a quota value is greater than 2G
4067 on a system where the size of off_t is not greater than 4, a panic error
4070 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4071 never match. The debug and -bh output now contains an explicit error
4072 message indicating a malformed IPv4 address or mask.
4074 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4075 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4076 PH/23 above applies.
4078 PH/25 Do not write to syslog when running in the test harness. The only
4079 occasion when this arises is a failure to open the main or panic logs
4080 (for which there is an explicit test).
4082 PH/26 Added the /no_tell option to "control=freeze".
4084 PH/27 If a host name lookup failed very early in a connection, for example, if
4085 the IP address matched host_lookup and the reverse lookup yielded a name
4086 that did not have a forward lookup, an error message of the form "no IP
4087 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4088 could be logged. Now it outputs the IP address instead of "NULL".
4090 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4091 allows the sender and the authenticated sender to be set when
4092 submitting a message from within Exim. Since child_open_exim() is
4093 documented for local_scan(), the new function should be too.
4095 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4096 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4097 results in an empty string is now treated as unset.
4099 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4101 PH/31 Added sender_verify_fail logging option.
4103 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4104 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4105 tidied the source and removed it altogether.
4107 PH/33 When a queue run was abandoned because the load average was too high, a
4108 log line was always written; now it is written only if the queue_run log
4109 selector is set. In addition, the log line for abandonment now contains
4110 information about the queue run such as the pid. This is always present
4111 in "start" and "stop" lines but was omitted from the "abandon" line.
4113 PH/34 Omit spaces between a header name and the colon in the error message that
4114 is given when verify = headers_syntax fails (if there are lots of them,
4115 the message gets confusing).
4117 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4118 names, as there are now some PTR records that contain slashes. This check
4119 is only to protect against broken name servers that fall over on strange
4120 characters, so the fact that it applies to all lookups doesn't matter.
4122 PH/36 Now that the new test suite is complete, we can remove some of the
4123 special code in Exim that was needed for the old test suite. For example,
4124 sorting DNS records because real resolvers return them in an arbitrary
4125 order. The new test suite's fake resolver always returns records in the
4128 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4129 bounces) except when queue_only is set, to avoid logging races between
4130 the different processes.
4132 PH/38 Panic-die if .include specifies a non-absolute path.
4134 PH/39 A tweak to the "H" retry rule from its user.
4136 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4137 a label. They prevented compilation on older perls.
4139 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4140 a warning to be raised on newish perls.
4142 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4143 on queue. Changes to match documented behaviour of showing count of
4144 messages matching specified criteria.
4146 PH/40 Changed the default ident timeout from 30s to 5s.
4148 PH/41 Added support for the use of login_cap features, on those BSD systems
4149 that have them, for controlling the resources used by pipe deliveries.
4151 PH/42 The content-scanning code uses fopen() to create files in which to put
4152 message data. Previously it was not paying any attention to the mode of
4153 the files. Exim runs with umask(0) because the rest of the code creates
4154 files with open(), and sets the required mode explicitly. Thus, these
4155 files were ending up world-writeable. This was not a big issue, because,
4156 being within the spool directory, they were not world-accessible. I have
4157 created a function called modefopen, which takes an additional mode
4158 argument. It sets umask(777), creates the file, chmods it to the required
4159 mode, then resets the umask. All the relevant calls to fopen() in the
4160 content scanning code have been changed to use this function.
4162 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4163 to 24 hours. This avoids potential overflow problems when processing G
4164 and H retry rules. I suspect nobody ever tinkers with this value.
4166 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4168 PH/45 When the plaintext authenticator is running as a client, the server's
4169 challenges are checked to ensure they are valid base64 strings. By
4170 default, the authentication attempt is cancelled if an invalid string is
4171 received. Setting client_ignore_invalid_base64 true ignores these errors.
4172 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4173 they are received. Thus, the responses can be made to depend on the
4174 challenges. If an invalid string is ignored, an empty string is placed in
4177 PH/46 Messages that are created by the autoreply transport now contains a
4178 References: header, in accordance with RFCs 2822 and 3834.
4180 PH/47 Added authenticated_sender_force to the smtp transport.
4182 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4184 PH/49 Installed latest patch from the Sieve maintainer.
4186 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4187 was also set, the appendfile transport was unnecessarily scanning a
4188 directory of message files (e.g. for maildir delivery) to find the count
4189 of files (along with the size), even though it did not need this
4190 information. It now does the scan only if it needs to find either the
4191 size of the count of files.
4193 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4195 PH/52 Two bugs concerned with error handling when the smtp transport is
4198 (i) Exim was not creating retry information for temporary errors given
4199 for individual recipients after the DATA command when the smtp transport
4200 was used in LMTP mode. This meant that they could be retried too
4201 frequently, and not timed out correctly.
4203 (ii) Exim was setting the flag that allows error details to be returned
4204 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4205 recipients that were returned after the DATA command.
4207 PH/53 This is related to PH/52, but is more general: for any failing address,
4208 when detailed error information was permitted to be returned to the
4209 sender, but the error was temporary, then after the final timeout, only
4210 "retry timeout exceeded" was returned. Now it returns the full error as
4211 well as "retry timeout exceeded".
4213 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4214 do this, and (what is worse) MTAs that accept it.
4216 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4217 will now be deprecated.
4219 PH/56 New os.c-cygwin from the Cygwin maintainer.
4221 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4222 formats (previously only available in exim formats via -bpr, -bpru,
4223 and -bpra. Now also available in native and exiqgrep formats)
4225 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4226 with very large, slow to parse queues
4228 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4230 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4232 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4233 responses to authentication challenges, though it was showing the
4234 challenges; (ii) I've removed the CR characters from the debug output for
4237 PH/58 Allow for the insertion of a newline as well as a space when a string
4238 is turned into more than one encoded-word during RFC 2047 encoding. The
4239 Sieve code now uses this.
4241 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4242 data_4xx, lost_connection, tls_required.
4244 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4245 message was being sent as an SMTP response.
4247 PH/61 Add -l and -k options to exicyclog.
4249 PH/62 When verifying, if an address was redirected to one new address, so that
4250 verification continued, and the new address failed or deferred after
4251 having set something in $address_data, the value of $address_data was not
4252 passed back to the ACL. This was different to the case when no
4253 redirection occurred. The value is now passed back in both cases.
4255 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4256 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4257 use login_cap.h, so on its own it isn't the distinguishing feature. The
4258 new name refers directly to the setclassresources() function.
4260 PH/65 Added configuration files for NetBSD3.
4262 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4264 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4265 is preferred over IPv4.
4267 PH/68 The bounce_return_message and bounce_return_body options were not being
4268 honoured for bounces generated during the reception of non-SMTP messages.
4269 In particular, this applied to messages rejected by the ACL. This bug has
4270 been fixed. However, if bounce_return_message is true and bounce_return_
4271 body is false, the headers that are returned for a non-SMTP message
4272 include only those that have been read before the error was detected.
4273 (In the case of an ACL rejection, they have all been read.)
4275 PH/69 The HTML version of the specification is now built in a directory called
4276 spec_html instead of spec.html, because the latter looks like a path with
4277 a MIME-type, and this confuses some software.
4279 PH/70 Catch two compiler warnings in sieve.c.
4281 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4282 function verify_get_ident() calls ip_connect() to connect a socket, but
4283 if the "connect()" function timed out, ip_connect() used to close the
4284 socket. However, verify_get_ident() also closes the socket later, and in
4285 between Exim writes to the log, which may get opened at this point. When
4286 the socket was closed in ip_connect(), the log could get the same file
4287 descriptor number as the socket. This naturally causes chaos. The fix is
4288 not to close the socket in ip_connect(); the socket should be closed by
4289 the function that creates it. There was only one place in the code where
4290 this was missing, in the iplookup router, which I don't think anybody now
4291 uses, but I've fixed it anyway.
4293 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4294 well as to direct DNS lookups. Otherwise the handling of names in host
4295 lists is inconsistent and therefore confusing.
4301 PH/01 Two changes to the default runtime configuration:
4303 (1) Move the checks for relay_from_hosts and authenticated clients from
4304 after to before the (commented out) DNS black list checks.
4306 (2) Add control=submission to the relay_from_hosts and authenticated
4307 clients checks, on the grounds that messages accepted by these
4308 statements are most likely to be submissions.
4310 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4312 (1) Generate an error if the third argument for the ${prvs expansion is
4315 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4318 (3) Reset the variables that are obtained from the first argument of
4319 ${prvscheck and used in the second argument before leaving the code,
4320 because their memory is reclaimed, so using them afterwards may do
4323 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4324 one (it's much easier than Tom thought :-).
4326 (5) Because of (4), we can now allow for the use of $prvscheck_result
4327 inside the third argument.
4329 PH/03 For some reason, the default setting of PATH when running a command from
4330 a pipe transport was just "/usr/bin". I have changed it to
4333 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4334 anything to be listed in the output from -bV.
4336 PH/05 When a filter generated an autoreply, the entire To: header line was
4337 quoted in the delivery log line, like this:
4339 => >A.N.Other <ano@some.domain> <original@ddress> ...
4341 This has been changed so that it extracts the operative address. There
4342 may be more than one such address. If so, they are comma-separated, like
4345 => >ano@some.domain,ona@other.domain <original@ddress> ...
4347 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4348 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4349 not being looked up in the rDNS to get a host name, Exim was showing the
4350 IP address twice in Received: lines, even though the IP addresses were
4351 identical. For example:
4353 Received: from [1.2.3.4] (helo=[1.2.3.4])
4355 However, if the real host name was known, it was omitting the HELO data
4356 if it matched the actual IP address. This has been tidied up so that it
4357 doesn't show the same IP address twice.
4359 PH/07 When both +timestamp and +memory debugging was on, the value given by
4360 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4361 called by the debug printing, thereby overwriting the timestamp buffer.
4362 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4364 PH/08 When the original message was included in an autoreply transport, it
4365 always said "this is a copy of the message, including all the headers",
4366 even if body_only or headers_only was set. It now gives an appropriate
4369 PH/09 Applied a patch from the Sieve maintainer which:
4371 o fixes some comments
4372 o adds the (disabled) notify extension core
4373 o adds some debug output for the result of if/elsif tests
4374 o points to the current vacation draft in the documentation
4375 and documents the missing references header update
4379 o fixes a bug in processing the envelope test (when testing
4380 multiple envelope elements, the last element determined the
4383 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4384 Electronic Mail") by including:
4386 Auto-submitted: auto-generated
4388 in the messages that it generates (bounce messages and others, such as
4389 warnings). In the case of bounce messages for non-SMTP messages, there was
4390 also a typo: it was using "Auto_submitted" (underscore instead of
4391 hyphen). Since every message generated by Exim is necessarily in response
4392 to another message, thes have all been changed to:
4394 Auto-Submitted: auto-replied
4396 in accordance with these statements in the RFC:
4398 The auto-replied keyword:
4400 - SHOULD be used on messages sent in direct response to another
4401 message by an automatic process,
4403 - MUST NOT be used on manually-generated messages,
4405 - MAY be used on Delivery Status Notifications (DSNs) and Message
4406 Disposition Notifications (MDNs),
4408 - MUST NOT be used on messages generated by automatic or periodic
4409 processes, except for messages which are automatic responses to
4412 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4413 to the default Received: header definition.
4415 PH/12 Added log selector acl_warn_skipped (default on).
4417 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4418 variables because (a) they are in the wrong storage pool and (b) even if
4419 they were copied, it wouldn't work properly because of the caching.
4421 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4422 checking when decoding. Apparently there are clients that generate
4423 overlong encoded strings. Why am I not surprised?
4425 PH/15 If the first argument of "${if match_address" was not empty, but did not
4426 contain an "@" character, Exim crashed. Now it writes a panic log message
4427 and treats the condition as false.
4429 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4431 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4432 extension "envelope-auth". The code is finished and in agreement with
4433 other implementations, but there is no documentation so far and in fact,
4434 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4435 not changing the active code.
4437 Print executed "if" and "elsif" statements when debugging is used. This
4438 helps a great deal to understand what a filter does.
4440 Document more things not specified clearly in RFC3028. I had all this
4441 sorted out, when out of a sudden new issues came to my mind. Oops."
4443 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4446 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4447 canonical form (as documented). However, after a host name lookup from
4448 the IP address, check_host() was doing a simple string comparison with
4449 addresses acquired from the DNS when checking that the found name did
4450 have the original IP as one of its addresses. Since any found IPv6
4451 addresses are likely to be in abbreviated form, the comparison could
4452 fail. Luckily, there already exists a function for doing the comparison
4453 by converting both addresses to binary, so now that is used instead of
4454 the text comparison.
4456 PH/20 There was another similar case to PH/19, when a complete host name was
4457 given in a host list; looking up its IP address could give an abbreviated
4458 form, whereas the current host's name might or might not be abbreviated.
4459 The same fix has been applied.
4465 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4466 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4469 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4470 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4472 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4474 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4475 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4476 Also, for "Auto-Submitted", treat anything other than "no" as
4477 non-personal, in accordance with RFC 3834. (Previously it treated
4478 anything starting "auto-" as non-personal.)
4480 TF/01 The control=submission/name=... option had a problem with syntax
4481 errors if the name included a slash character. The /name= option
4482 now slurps the rest of the string, so it can include any characters
4483 but it must come last in the list of options (after /sender_retain
4486 PH/05 Some modifications to the interface to the fake nameserver for the new
4494 TK/01 Added the "success_on_redirect" address verification option. See
4495 NewStuff for rationale and an example.
4497 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4499 PH/02 Patch to exigrep to allow it to work on syslog lines.
4501 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4502 fread() to skip over the body file's header line, because in Cygwin the
4503 header line is locked and is inaccessible.
4505 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4506 co-exist for some time) to make it clear that it is the Exim ID that is
4507 referenced, not the Message-ID: header line.
4509 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4510 string_format() function, because snprintf() does not exist on all
4513 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4514 ${stat: expansion item.
4516 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4517 protocol synchronization error", to keep the pedants happy.
4519 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4520 well as for IRIX systems, when gcc is being used. See the host.c source
4523 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4525 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4528 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4529 transport and to the smtp transport in LMTP mode.
4531 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4533 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4534 run a filter in a subprocess. This could lead to confusion in subsequent
4535 lookups in the parent process. There should also be a search_tidyup() at
4536 the end of the subprocess.
4538 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4539 only if the host matched helo_try_verify_hosts, which caused the
4540 verification to occur when the EHLO/HELO command was issued. The ACL just
4541 tested the remembered result. Now, if a previous verification attempt has
4542 not happened, "verify = helo" does it there and then.
4544 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4546 TK/03 Fix log output including CR from clamd.
4548 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4549 exist provoked a memory error which could cause a segfault.
4551 PH/15 Installed PCRE 6.2
4553 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4555 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4556 of the problem. Specifically, suggested +O2 rather than +O1 for the
4559 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4561 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4562 record, the port was not correctly logged when the outgoing_port log
4563 selector was set (it logged the transort's default port).
4565 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4566 fallback_hosts, and "hosts" in the smtp transport.
4568 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4569 host errors such as "Connection refused".
4571 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4572 authentication with radiusclient 0.4.9:
4574 - Error returned from rc_read_config was caught wrongly
4575 - Username/password not passed on to radius server due to wrong length.
4577 The presumption is that some radiusclient API changes for 4.51/PH/17
4578 were not taken care of correctly. The code is still untested by me (my
4579 Linux distribution still has 0.3.2 of radiusclient), but it was
4580 contributed by a Radius user.
4582 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4583 expanding the "port" option of the smtp transport.
4585 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4586 while reading a MIME header. Thanks to Tom Hughes for a patch.
4588 PH/24 Include config.h inside local_scan.h so that configuration settings are
4591 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4592 that in an ACL for RCPT (for example), you can examine exactly what was
4595 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4596 commands, but it was not correctly comparing the address with the actual
4597 client host address. Thus, it would show the EHLO address in Received:
4598 header lines when this was not necessary.
4600 PH/27 Added the % operator to ${eval:}.
4602 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4603 it should be ignoring failures (because with -C, for example, it has lost
4604 privilege). It wasn't ignoring creation failures other than "already
4607 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4610 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4611 because an input line was too long, either on its own, or by virtue of
4612 too many continuations, the temporary file was not being removed, and the
4613 return code was incorrect.
4615 PH/31 Missing "BOOL" in function definition in filtertest.c.
4617 PH/32 Applied Sieve patches from the maintainer.
4619 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4621 PH/33 Added "verify = not_blind".
4623 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4624 Local/Makefile (with some defaults set). These are used in built scripts
4625 such as exicyclog, but they have never been used in the exim_install
4626 script (though there are many overriding facilities there). I have
4627 arranged that the exim_install script now takes note of these two
4630 PH/35 Installed configuration files for Dragonfly.
4632 PH/36 When a locally submitted message by a trusted user did not contain a
4633 From: header, and the sender address was obtained from -f or from an SMTP
4634 MAIL command, and the trusted user did not use -F to supply a sender
4635 name, $originator_name was incorrectly used when constructing a From:
4636 header. Furthermore, $originator_name was used for submission mode
4637 messages from external hosts without From: headers in a similar way,
4638 which is clearly wrong.
4640 PH/37 Added control=suppress_local_fixups.
4642 PH/38 When log_selector = +received_sender was set, and the addition of the
4643 sender made the log line's construction buffer exactly full, or one byte
4644 less than full, an overflow happened when the terminating "\n" was
4647 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4648 when the result of a list match is failure because a DNS lookup failed.
4650 PH/40 RM_COMMAND is now used in the building process.
4652 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4653 the "build-* directories that it finds.
4655 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4656 address in a domain literal was a prefix of an interface address.
4658 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4659 when verifying a sender address, unless rewrite_headers is false.
4661 PH/44 Wrote a long comment about why errors_to addresses are verified as
4662 recipients, not senders.
4664 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4665 the ratelimit ACL was added.
4667 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4669 PH/46 Added extra information about PostgreSQL errors to the error string.
4671 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4672 avoiding the need to install special zones in a real server. This is
4673 backwards compatible; if it can't find the fake resolver, it drops back.
4674 Thus, both old and new test suites can be run.
4676 TF/02 Added util/ratelimit.pl
4678 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4679 clock is set back in time.
4681 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4682 Candler <B.Candler@pobox.com>.
4684 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4685 OK for addresses that are the result of redirections.
4687 PH/48 A number of further additions for the benefit of the new test suite,
4688 including a fake gethostbyname() that interfaces to the fake DNS resolver
4691 TF/06 The fix for widen_domains has also been applied to qualify_single and
4692 search_parents which are the other dnslookup options that can cause
4695 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4698 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4700 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4701 service" if it encountered an MX record with an empty target hostname.
4702 The message is now "an MX or SRV record indicated no SMTP service".
4704 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4705 if the DNS of the sending site is misconfigured. This is quite a
4706 common situation. This change restores the behaviour of treating a
4707 helo verification defer as a failure.
4709 PH/51 If self=fail was set on a router, the bounce message did not include the
4710 actual error message.
4716 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4718 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4719 command itself ended in error, the underlying message about the transport
4720 filter timeout was being overwritten with the pipe command error. Now the
4721 underlying error message should be appended to the second error message.
4723 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4725 PH/02 Reduce the amount of output that "make" produces by default. Full output
4726 can still be requested.
4728 PH/03 The warning log line about a condition test deferring for a "warn" verb
4729 was being output only once per connection, rather than after each
4730 occurrence (because it was using the same function as for successful
4731 "warn" verbs). This seems wrong, so I have changed it.
4733 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4734 it should not have, which might have caused a crash in the right
4735 circumstances, but probably never did.
4737 PH/04 Installed a modified version of Tony Finch's patch to make submission
4738 mode fix the return path as well as the Sender: header line, and to
4739 add a /name= option so that you can make the user's friendly name appear
4742 TF/03 Added the control = fakedefer ACL modifier.
4744 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4745 Mark Lowes for thorough testing.
4747 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4749 TK/03 Merged latest SRS patch from Miles Wilton.
4751 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4752 with the definition in sysexits.h (which is #included earlier).
4753 Fortunately, Exim does not actually use EX_OK. The code used to try to
4754 preserve the sysexits.h value, by assuming that macro definitions were
4755 scanned for macro replacements. I have been disabused of this notion,
4756 so now the code just undefines EX_OK before #including unistd.h.
4758 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4759 in the smtp transport. When a block could not be written in a single
4760 write() function, the timeout was being re-applied to each part-write.
4761 This seems wrong - if the receiver was accepting one byte at a time it
4762 would take for ever. The timeout is now adjusted when this happens. It
4763 doesn't have to be particularly precise.
4765 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4766 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4768 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4769 without a domain if the check to <postmaster@domain> fails.
4771 SC/01 Eximstats: added -xls and the ability to specify output files
4772 (patch written by Frank Heydlauf).
4774 SC/02 Eximstats: use FileHandles for outputting results.
4776 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4778 SC/04 Eximstats: fixed display of large numbers with -nvr option
4780 SC/05 Eximstats: fixed merging of reports with empty tables.
4782 SC/06 Eximstats: added the -include_original_destination flag
4784 SC/07 Eximstats: removed tabs and trailing whitespace.
4786 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4788 TK/06 MBOX spool code: Add real "From " MBOX separator line
4789 so the .eml file is really in mbox format (even though
4790 most programs do not really care). Patch from Alex Miller.
4792 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4793 The latter is generated from $received_to and is only set if the
4794 message has one envelope recipient. SA can use these headers,
4795 obviously out-of-the-box. Patch from Alex Miller.
4797 PH/08 The ${def test on a variable was returning false if the variable's
4798 value was "0", contrary to what the specification has always said!
4799 The result should be true unless the variable is empty.
4801 PH/09 The syntax error of a character other than { following "${if
4802 def:variable_name" (after optional whitespace) was not being diagnosed.
4803 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4804 accidental colon was present, for example, could give incorrect results.
4806 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4807 result is used (not including appendfile, where other changes are about
4810 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4811 This involved changing a lot of size variables from int to off_t. It
4812 should work with maildirs and everything.
4814 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4815 spamd dying while we are connected to it.
4817 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4820 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4821 function for BDB 4.3.
4823 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4825 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4826 the sources for PCRE have been re-arranged and more files are now
4829 PH/15 The code I had for printing potentially long long variables in PH/11
4830 above was not the best (it lost precision). The length of off_t variables
4831 is now inspected at build time, and an appropriate printing format (%ld
4832 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4833 to be "long long int" or "long int". This is needed for the internal
4834 formatting function string_vformat().
4836 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4837 the configuration file to be ":syslog", then the script "guesses" where
4838 the logs files are, rather than using the compiled in default. In our
4839 case the guess is not the same as the compiled default, so the script
4840 suddenly stopped working when I started to use syslog. The patch checks
4841 to see if log_file_path is "". If so, it attempts to read it from exim
4842 with no configuration file to get the compiled in version, before it
4843 falls back to the previous guessing code."
4845 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4846 implementing BATV in an Exim configuration. See NewStuff for the gory
4849 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4850 Makefile that are specific to HP-UX.
4852 PH/18 If the "use_postmaster" option was set for a recipient callout together
4853 with the "random" option, the postmaster address was used as the MAIL
4854 FROM address for the random test, but not for the subsequent recipient
4855 test. It is now used for both.
4857 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4858 patch removes a few documentation additions to RFC 3028, because the
4859 latest draft now contains them. It adds the new en;ascii-case comparator
4860 and a new error check for 8bit text in MIME parts. Comparator and
4861 require names are now matched exactly. I enabled the subaddress
4862 extension, but it is not well tested yet (read: it works for me)."
4864 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4865 rework some of the code of TK/09 above to avoid the hardwired use of
4866 "%lld" and "long long". Replaced the call to snprintf() with a call to
4869 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4870 records point to non-existent hosts", "retry timeout exceeded", and
4871 "retry time not reached for any host after a long failure period".
4873 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4874 experimental DomainKeys support:
4876 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4877 (2) On an error such as an illegally used "control", the wrong name for
4878 the control was given.
4880 These problems did NOT occur unless DomainKeys support was compiled.
4882 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4884 PH/24 Added ${if match_ip condition.
4886 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4887 databases so that it will be absolutely obvious if a crash occurs in the
4888 DB library. This is a regular occurrence (often caused by mis-matched
4891 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4892 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4893 picked up on a user's system that detects such things. There doesn't seem
4894 to be a gcc warning option for this - only an attribute that has to be
4895 put on the function's prototype. It seems that in Fedora Core 4 they have
4896 set this on a number of new functions. No doubt there will be more in due
4899 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4900 specify a transport. However, if an address that was verified by such a
4901 router was the subject of a callout, Exim crashed because it tried to
4902 read the rcpt_include_affixes from the non-existent transport. Now it
4903 just assumes that the setting of that option is false. This bug was
4904 introduced by 4.51/PH/31.
4906 PH/28 Changed -d+all to exclude +memory, because that information is very
4907 rarely of interest, but it makes the output a lot bigger. People tend to
4908 do -d+all out of habit.
4910 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4911 code in os-type was giving problems when libc.so lives in lib64, like on
4914 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4915 aren't the modern standard, and it seems that some systems' include files
4916 don't always have them. Exim was already checking for some of the newer
4917 ones like T_AAAA, and defining it itself. I've added checks for all the
4918 record types that Exim uses.
4920 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4921 not automatically generating a new one, as it is supposed to. This
4922 prevented TLS from working. If the file did exist, but contained invalid
4923 data, a new version was generated, as expected. It was only the case of a
4924 non-existent file that was broken.
4926 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4927 with a change in libdomainkeys > 0.64.
4929 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4930 from DNS. If the selector record carries the flag, it now has
4931 precedence over the domain-wide flag.
4933 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4935 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4936 the use of a port name that isn't defined in /etc/services) occurred, the
4937 message was deferred as in a normal delivery, and thus remained on the
4938 spool, instead of being failed because of the mua_wrapper setting. This
4939 is now fixed, and I tidied up some of the mua_wrapper messages at the
4942 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4943 the messages in a hash of arrays rather than using individual hashes.
4944 This is a bit cleaner and results in dramatic memory savings, albeit
4945 at a slight CPU cost.
4947 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4948 as requested by Marc Sherman.
4950 SC/10 Eximstats: added histograms for user specified patterns as requested
4953 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4955 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4956 fopen() in the content-scanning modules that did not already have it.
4962 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4963 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4965 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4967 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4969 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4970 to test Sieve filters that use "vacation".
4972 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4973 that changes the way the GnuTLS parameters are stored in the cache file.
4974 The new format can be generated externally. For backward compatibility,
4975 if the data in the cache doesn't make sense, Exim assumes it has read an
4976 old-format file, and it generates new data and writes a new file. This
4977 means that you can't go back to an older release without removing the
4980 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4981 work if there are any delivery delays because "one_time" forces the
4982 parent to be marked "delivered", so its unseen clone is never tried
4983 again. For this reason, Exim now forbids the simultaneous setting of
4986 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4987 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4988 entry if you want to know the details. The fix, however, neglected to
4989 consider the case when local delivery batching is involved. The test for
4990 "previously delivered" was not happening when checking to see if an
4991 address could be batched with a previous (undelivered) one; under
4992 certain circumstances this could lead to multiple deliveries to the same
4995 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4996 in its include files, and this causes problems building Exim.
4998 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4999 header_syntax) but Exim was just ignoring anything given after a slash.
5000 In particular, this caused confusion with an attempt to use "verify =
5001 reverse_host_lookup/defer_ok". An error is now given when options are
5002 supplied for verify items that do not have them. (Maybe reverse_host_
5003 lookup should have a defer_ok option, but that's a different point.)
5005 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5006 defined by RFC 821) to 2048, because there were problems with some AUTH
5007 commands, and RFC 1869 says the size should be increased for extended
5008 SMTP commands that take arguments.
5010 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5013 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5014 "unknown" error; now it says that the functionality isn't in the binary.
5016 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5017 an address' error message when a string expansion fails (syntax or
5018 whatever). Otherwise the password may appear in the log. Following change
5019 PH/42 below, there is no longer a chance of it appearing in a bounce
5022 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5024 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5025 of its addresses was ever tried. (Bugzilla bug #2.)
5027 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5028 the result incorrectly in the debug output. (It correctly added a newline
5029 to what was transported.)
5031 TF/01 Added $received_time.
5033 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5034 commented out examples of how to interface to a virus scanner and to
5035 SpamAssassin. Also added commented examples of av_scanner and
5036 spamd_address settings.
5038 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5039 and controls are allowed in which ACLs. There were a couple of minor
5040 errors. Some of the entries in the conditions table (which is a table of
5041 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5042 negation of where the condition IS allowed.
5044 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5046 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5047 header file does not have a version number, so I've had to invent a new
5048 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5049 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5050 radiusclient), but it was contributed by a Radius user.
5052 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5053 files or directories, for OpenSSL.
5055 PH/19 When an Exim process that is running as root has to create an Exim log
5056 file, it does so in a subprocess that runs as exim:exim so as to get the
5057 ownership right at creation (otherwise, other Exim processes might see
5058 the file with the wrong ownership). There was no test for failure of this
5059 fork() call, which would lead to the process getting stuck as it waited
5060 for a non-existent subprocess. Forks do occasionally fail when resources
5061 run out. I reviewed all the other calls to fork(); they all seem to check
5064 PH/20 When checking for unexpected SMTP input at connect time (before writing
5065 the banner), Exim was not dealing correctly with a non-positive return
5066 from the read() function. If the client had disconnected by this time,
5067 the result was a log entry for a synchronization error with an empty
5068 string after "input=" when read() returned zero. If read() returned -1
5069 (an event I could not check), uninitialized data bytes were printed.
5070 There were reports of junk text (parts of files, etc) appearing after
5073 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5075 PH/22 Added support for macro redefinition, and (re)definition in between
5076 driver and ACL definitions.
5078 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5079 forgetting to use the resulting value; it was using the unexpanded value.
5081 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5082 hadn't been configured. The fix is from Juergen Kreileder, who
5083 understands it better than I do:
5085 "Here's what I see happening with three configured cyrus_sasl
5086 authenticators configured (plain, login, cram-md5):
5088 On startup auth_cyrus_sasl_init() gets called for each of these.
5089 This means three calls to sasl_listmech() without a specified mech_list.
5090 => SASL tests which mechs of all available mechs actually work
5091 => three warnings about OTP not working
5092 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5094 With the patch, sasl_listmech() also gets called three times. But now
5095 SASL's mech_list option is set to the server_mech specified in the the
5096 authenticator. Or in other words, the answer from sasl_listmech()
5097 gets limited to just the mech you're testing for (which is different
5099 => the return list contains just 'plain' or 'login', 'cram-md5' or
5100 nothing depending on the value of ob->server_mech.
5102 I've just tested the patch: Authentication still works fine,
5103 unavailable mechs specified in the exim configuration are still
5104 caught, and the auth.log warnings about OTP are gone."
5106 PH/25 When debugging is enabled, the contents of the command line are added
5107 to the debugging output, even when log_selector=+arguments is not
5110 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5111 answer is "GNU", and only if the return is "GNU/something" is the answer
5114 PH/27 $acl_verify_message is now set immediately after the failure of a
5115 verification in an ACL, and so is available in subsequent modifiers. In
5116 particular, the message can be preserved by coding like this:
5118 warn !verify = sender
5119 set acl_m0 = $acl_verify_message
5121 Previously, $acl_verify_message was set only while expanding "message"
5122 and "log_message" when a very denied access.
5124 PH/28 Modified OS/os.c-Linux with
5126 -#ifndef OS_LOAD_AVERAGE
5127 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5129 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5130 nomenclature these days.)
5132 PH/29 Installed patch from the Sieve maintainer that adds the options
5133 sieve_useraddress and sieve_subaddress to the redirect router.
5135 PH/30 In these circumstances:
5136 . Two addresses routed to the same list of hosts;
5137 . First host does not offer TLS;
5138 . First host accepts first address;
5139 . First host gives temporary error to second address;
5140 . Second host offers TLS and a TLS session is established;
5141 . Second host accepts second address.
5142 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5143 and peerdn, if requested) that were in fact used only for the second
5146 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5147 attention to any local part prefix or suffix that was matched by the
5148 router that accepted the address. It now behaves in the same way as it
5149 does for delivery: the affixes are removed from the local part unless
5150 rcpt_include_affixes is set on the transport.
5152 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5153 timeout during the DATA phase of an incoming message.
5155 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5156 applied a patch sanctioned by the Sieve maintainer.
5158 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5159 the uid or gid is negative. A case of a negative gid caused this to be
5160 noticed. The fix allows for either to be negative.
5162 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5163 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5166 PH/36 The change PH/12 above was broken. Fixed it.
5168 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5169 the grounds that routing the same address twice would always produce the
5170 same answer. This might have been true once, but it is certainly no
5171 longer true now. Routing a child address may depend on the previous
5172 routing that produced that child. Some complicated redirection strategies
5173 went wrong when messages had multiple recipients, and made Exim's
5174 behaviour dependent on the order in which the addresses were given.
5176 I have moved the duplicate checking until after the routing is complete.
5177 Exim scans the addresses that are assigned to local and remote
5178 transports, and removes any duplicates. This means that more work will be
5179 done, as duplicates will always all be routed, but duplicates are
5180 presumably rare, so I don't expect this is of any significance.
5182 For deliveries to pipes, files, and autoreplies, the duplicate checking
5183 still happens during the routing process, since they are not going to be
5186 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5187 It corrects a timeout issue with spamd. This is Ian's comment: "The
5188 background is that sometimes spamd either never reads data from a
5189 connection it has accepted, or it never writes response data. The exiscan
5190 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5191 blindly assumes that writes won't block so it may never time out."
5193 PH/39 Allow G after quota size as well as K and M.
5195 PH/40 The value set for $authenticated_id in an authenticator may not contain
5196 binary zeroes or newlines because the value is written to log lines and
5197 to spool files. There was no check on this. Now the value is run through
5198 the string_printing() function so that such characters are converted to
5199 printable escape sequences.
5201 PH/41 $message_linecount is a new variable that contains the total number of
5202 lines in the message. Compare $body_linecount, which is the count for the
5205 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5206 bounce and delay warning messages, except in certain special cases, which
5209 (a) An SMTP error message from a remote host;
5210 (b) A message specified in a :fail: redirection;
5211 (c) A message specified in a "fail" command in a system filter;
5212 (d) A message specified in a FAIL return from the queryprogram router;
5213 (e) A message specified by the cannot_route_message router option.
5215 In these cases only, Exim does include the error details in bounce and
5216 warning messages. There are also a few cases where bland messages such
5217 as "unrouteable address" or "local delivery error" are given.
5219 PH/43 $value is now also set for the "else" part of a ${run expansion.
5221 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5222 being worked on, but at least Exim now implements the latest version to
5225 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5226 process to complete was treated as a delivery failure, a timeout while
5227 writing the message to the pipe was logged, but erroneously treated as a
5228 successful delivery. Such timeouts include transport filter timeouts. For
5229 consistency with the overall process timeout, these timeouts are now
5230 treated as errors, giving rise to delivery failures by default. However,
5231 there is now a new Boolean option for the pipe transport called
5232 timeout_defer, which, if set TRUE, converts the failures into defers for
5233 both kinds of timeout. A transport filter timeout is now identified in
5236 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5237 systems where "make" and "gmake" are different, calling "gmake" at top
5238 level broke things. I've arranged for the value of $(MAKE) to be passed
5239 from the Makefile to this script so that it can call the same version of
5243 A note about Exim versions 4.44 and 4.50
5244 ----------------------------------------
5246 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5247 changes of various kinds. As a consequence, a big documentation update was
5248 needed. This delayed the release for rather longer than seemed good, especially
5249 in the light of a couple of (minor) security issues. Therefore, the changes
5250 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5251 release. So 4.44 and 4.50 are in effect two different branches that both start
5254 I have left the 4.50 change log unchanged; it contains all the changes since
5255 4.43. The change log for 4.44 is below; many of its items are identical to
5256 those for 4.50. This seems to be the most sensible way to preserve the
5257 historical information.
5263 1. Minor wording change to the doc/README.SIEVE file.
5265 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5266 computation of the current number of files was incorrect.
5268 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5269 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5272 4. Give more explanation in the error message when the command for a transport
5273 filter fails to execute.
5275 5. There are several places where Exim runs a non-Exim command in a
5276 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5277 was being done only for the command run by the queryprogram router. It is
5278 now done for all such subprocesses. The other cases are: ${run, transport
5279 filters, and the commands run by the lmtp and pipe transports.
5281 6. Added CONFIGURE_GROUP build-time option.
5283 7. Some older OS have a limit of 256 on the maximum number of file
5284 descriptors. Exim was using setrlimit() to set 1000 as a large value
5285 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5286 systems. I've change it so that if it can't get 1000, it tries for 256.
5288 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5289 was an oversight, and furthermore, ever since the addition of extra
5290 controls (e.g. 4.43/32), the checks on when to allow different forms of
5291 "control" were broken. There should now be diagnostics for all cases when a
5292 control that does not make sense is encountered.
5294 9. Added the /retain_sender option to "control=submission".
5296 10. $recipients is now available in the predata ACL (oversight).
5298 11. Tidy the search cache before the fork to do a delivery from a message
5299 received from the command line. Otherwise the child will trigger a lookup
5300 failure and thereby defer the delivery if it tries to use (for example) a
5301 cached ldap connection that the parent has called unbind on.
5303 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5304 of $address_data from the recipient verification was clobbered by the
5305 sender verification.
5307 13. The value of address_data from a sender verification is now available in
5308 $sender_address_data in subsequent conditions in the ACL statement.
5310 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5312 15. Added a new option "connect=<time>" to callout options, to set a different
5315 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5316 was its contents. (It was OK if the option was not defined at all.)
5318 17. A "Completed" log line is now written for messages that are removed from
5319 the spool by the -Mrm option.
5321 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5322 information about exactly what failed.
5324 19. Added -dd to debug only the daemon process.
5326 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5327 handles timeouts, both on the server side and network timeouts. Renamed the
5328 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5330 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5331 It is now set to "smtps".
5333 22. $host_address is now set to the target address during the checking of
5334 ignore_target_hosts.
5336 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5337 being passed; this would have caused $sender_host_name to have been used if
5338 matching the list had actually called for a host name (not very likely,
5339 since this list is usually IP addresses). A host name is now passed as
5342 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5343 code that specifies a non-restarting handler (typically sigaction() in
5344 modern systems) in an attempt to fix a rare and obscure crash bug.
5346 25. Narrowed the window for a race in the daemon that could cause it to ignore
5347 SIGCHLD signals. This is not a major problem, because they are used only to
5348 wake it up if nothing else does.
5350 26. A malformed maildirsize file could cause Exim to calculate negative values
5351 for the mailbox size or file count. Odd effects could occur as a result.
5352 The maildirsize information is now recalculated if the size or filecount
5355 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5356 support for a long time. Removed HAVE_SYS_VFS_H.
5358 28. Installed the latest version of exipick from John Jetmore.
5360 29. In an address list, if the pattern was not a regular expression, an empty
5361 subject address (from a bounce message) matched only if the pattern was an
5362 empty string. Non-empty patterns were not even tested. This was the wrong
5363 because it is perfectly reasonable to use an empty address as part of a
5364 database query. An empty address is now tested by patterns that are
5365 lookups. However, all the other forms of pattern expect the subject to
5366 contain a local part and a domain, and therefore, for them, an empty
5367 address still always fails if the pattern is not itself empty.
5369 30. Exim went into a mad DNS loop when attempting to do a callout where the
5370 host was specified on an smtp transport, and looking it up yielded more
5371 than one IP address.
5373 31. Re-factored the code for checking spool and log partition space into a
5374 function that finds that data and another that does the check. The former
5375 is then used to implement four new variables: $spool_space, $log_space,
5376 $spool_inodes, and $log_inodes.
5378 32. The RFC2047 encoding function was originally intended for short strings
5379 such as real names; it was not keeping to the 75-character limit for
5380 encoded words that the RFC imposes. It now respects the limit, and
5381 generates multiple encoded words if necessary. To be on the safe side, I
5382 have increased the buffer size for the ${rfc2047: expansion operator from
5385 33. It is now permitted to omit both strings after an "if" condition; if the
5386 condition is true, the result is "true". As before, when the second string
5387 is omitted, a false condition yields an empty string. This makes it less
5388 cumbersome to write custom ACL and router conditions.
5390 34. Failure to deliver a bounce message always caused it to be frozen, even if
5391 there was an errors_to setting on the router. The errors_to setting is now
5394 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5395 canonical form (fully expanded) before being placed in
5396 $sender_host_address.
5398 36. The table in the code that translates DNS record types into text (T_A to
5399 "A" for instance) was missing entries for NS and CNAME. It is just possible
5400 that this could have caused confusion if both these types were looked up
5401 for the same domain, because the text type is used as part of Exim's
5402 per-process caching. But the chance of anyone hitting this buglet seems
5405 37. The dnsdb lookup has been extended in a number of ways.
5407 (1) There is a new type, "zns", which walks up the domain tree until it
5408 finds some nameserver records. It should be used with care.
5410 (2) There is a new type, "mxh", which is like "mx" except that it returns
5411 just the host names, not the priorities.
5413 (3) It is now possible to give a list of domains (or IP addresses) to be
5414 looked up. The behaviour when one of the lookups defers can be
5415 controlled by a keyword.
5417 (4) It is now possible to specify the separator character for use when
5418 multiple records are returned.
5420 38. The dnslists ACL condition has been extended: it is now possible to supply
5421 a list of IP addresses and/or domains to be looked up in a particular DNS
5424 39. Added log_selector=+queue_time_overall.
5426 40. When running the queue in the test harness, wait just a tad after forking a
5427 delivery process, to get repeatability of debugging output.
5429 41. Include certificate and key file names in error message when GnuTLS fails
5430 to set them up, because the GnuTLS error message doesn't include the name
5431 of the failing file when there is a problem reading it.
5433 42. Allow both -bf and -bF in the same test run.
5435 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5437 44. The "Exiscan patch" is now merged into the mainline Exim source.
5439 45. Sometimes the final signoff response after QUIT could fail to get
5440 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5441 < 0 before doing a fflush(). This bug looks as though it goes back to the
5442 introduction of TLS in release 3.20, but "sometimes" must have been rare
5443 because the tests only now provoked it.
5445 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5446 (this can affect the format of dates).
5448 47. exim_tidydb, when checking for the continued existence of a message for
5449 which it has found a message-specific retry record, was not finding
5450 messages that were in split spool directories. Consequently, it was
5451 deleting retry records that should have stayed in existence.
5453 48. Steve fixed some bugs in eximstats.
5455 49. The SPA authentication driver was not abandoning authentication and moving
5456 on to the next authenticator when an expansion was forced to fail,
5457 contradicting the general specification for all authenticators. Instead it
5458 was generating a temporary error. It now behaves as specified.
5460 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5461 (the order specifies the preference for clients). The order is now AES256,
5462 AES128, 3DES, ARCFOUR128.
5464 51. Small patch to Sieve code - explicitly set From: when generating an
5467 52. Exim crashed if a remote delivery caused a very long error message to be
5468 recorded - for instance if somebody sent an entire SpamAssassin report back
5469 as a large number of 550 error lines. This bug was coincidentally fixed by
5470 increasing the size of one of Exim's internal buffers (big_buffer) that
5471 happened as part of the Exiscan merge. However, to be on the safe side, I
5472 have made the code more robust (and fixed the comments that describe what
5475 53. Now that there can be additional text after "Completed" in log lines (if
5476 the queue_time_overall log selector is set), a one-byte patch to exigrep
5477 was needed to allow it to recognize "Completed" as not the last thing in
5480 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5481 patch that reportedly fixes this has been added. I am not expert enough to
5482 create a test for it. This is what the patch creator wrote:
5484 "I found a little strange behaviour of ldap code when working with
5485 Windows 2003 AD Domain, where users was placed in more than one
5486 Organization Units. When I tried to give exim partial DN, the exit code
5487 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5488 But simultaneously result of request was absolutely normal ldap result,
5489 so I produce this patch..."
5491 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5492 REFERENCE, so I have modified the code to exclude the patch when that macro
5495 55. Some experimental protocols are using DNS PTR records for new purposes. The
5496 keys for these records are domain names, not reversed IP addresses. The
5497 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5498 leaves it alone. Component reversal etc. now happens only for IP addresses.
5501 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5503 57. Double the size of the debug message buffer (to 2048) so that more of very
5504 long debug lines gets shown.
5506 58. The exicyclog utility now does better if the number of log files to keep
5507 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5509 59. Two changes related to the smtp_active_hostname option:
5511 (1) $smtp_active_hostname is now available as a variable.
5512 (2) The default for smtp_banner uses $smtp_active_hostname instead
5513 of $primary_hostname.
5515 60. The host_aton() function is supposed to be passed a string that is known
5516 to be a valid IP address. However, in the case of IPv6 addresses, it was
5517 not checking this. This is a hostage to fortune. Exim now panics and dies
5518 if the condition is not met. A case was found where this could be provoked
5519 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5520 components; fortuitously, this particular loophole had already been fixed
5521 by change 4.50/55 above.
5523 If there are any other similar loopholes, the new check in host_aton()
5524 itself should stop them being exploited. The report I received stated that
5525 data on the command line could provoke the exploit when Exim was running as
5526 exim, but did not say which command line option was involved. All I could
5527 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5528 running as the user.
5531 61. There was a buffer overflow vulnerability in the SPA authentication code
5532 (which came originally from the Samba project). I have added a test to the
5533 spa_base64_to_bits() function which I hope fixes it.
5536 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5537 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5539 63. The daemon start-up calls getloadavg() while still root for those OS that
5540 need the first call to be done as root, but it missed one case: when
5541 deliver_queue_load_max is set with deliver_drop_privilege. This is
5542 necessary for the benefit of the queue runner, because there is no re-exec
5543 when deliver_drop_privilege is set.
5545 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5546 This has been fixed.
5548 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5549 was in use, was not putting the data itself into the right store pool;
5550 consequently, it could be overwritten for a subsequent message in the same
5551 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5554 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5556 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5557 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5558 the function were treating the return as a boolean value, which happened to
5559 work because 0=false and not-0=true, but is not correct code.
5561 68. The host_aton() function was not handling scoped IPv6 addresses (those
5562 with, for example, "%eth0" on the end) correctly.
5564 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5565 negated items (that is, ~something) in unsigned ints. Some compilers
5566 apparently mutter when there is no cast.
5568 70. If an address verification called from an ACL failed, and did not produce a
5569 user-specific message (i.e. there was only a "system" message), nothing was
5570 put in $acl_verify_message. In this situation, it now puts the system
5573 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5574 session; change 4.31/43 added the unwanted input to the log line - except
5575 that it did not do this in the start of session case. It now does.
5577 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5578 This is wrong and can cause the other end to generate a synchronization
5579 error if it is another Exim or anything else that does the synchronization
5580 check. A QUIT command is no longer sent after a timeout.
5582 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5583 during host lookups.
5585 74. The defer_ok option of callout verification was not working if it was used
5586 when verifying addresses in header lines, that is, for this case:
5588 verify = header_sender/callout=defer_ok
5590 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5591 those file descriptors could be used for SMTP connections. If anything
5592 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5593 could be sent to the SMTP client, causing chaos. The daemon now opens
5594 stdin, stdout, and stderr to /dev/null when it puts itself into the
5597 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5598 log by default. The user can override this with suitable Perl magic.
5600 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5601 the log message when discard triggers, was not working for the DATA ACL or
5602 for the non-SMTP ACL.
5604 78. Error message wording change in sieve.c.
5606 79. If smtp_accept_max_per_host was set, the number of connections could be
5607 restricted to fewer than expected, because the daemon was trying to set up
5608 a new connection before checking whether the processes handling previous
5609 connections had finished. The check for completed processes is now done
5610 earlier. On busy systems, this bug wouldn't be noticed because something
5611 else would have woken the daemon, and it would have reaped the completed
5614 80. If a message was submitted locally by a user whose login name contained one
5615 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5616 It caused a spool format error. I have fixed the spool reading code. A
5617 related problem was that the "from" clause in the Received: line became
5618 illegal because of the space(s). It is now covered by ${quote_local_part.
5620 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5623 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5625 83. Patch from Sieve maintainer to fix unterminated string problem in
5626 "vacation" handling.
5628 84. Some minor changes to the Linux configuration files to help with other
5629 OS variants using glibc.
5631 85. One more patch for Sieve to update vacation handling to latest spec.
5634 ----------------------------------------------------
5635 See the note above about the 4.44 and 4.50 releases.
5636 ----------------------------------------------------
5642 1. Change 4.43/35 introduced a bug that caused file counts to be
5643 incorrectly computed when quota_filecount was set in an appendfile
5646 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5647 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5650 3. Give more explanation in the error message when the command for a transport
5651 filter fails to execute.
5653 4. There are several places where Exim runs a non-Exim command in a
5654 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5655 was being done only for the command run by the queryprogram router. It is
5656 now done for all such subprocesses. The other cases are: ${run, transport
5657 filters, and the commands run by the lmtp and pipe transports.
5659 5. Some older OS have a limit of 256 on the maximum number of file
5660 descriptors. Exim was using setrlimit() to set 1000 as a large value
5661 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5662 systems. I've change it so that if it can't get 1000, it tries for 256.
5664 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5665 was an oversight, and furthermore, ever since the addition of extra
5666 controls (e.g. 4.43/32), the checks on when to allow different forms of
5667 "control" were broken. There should now be diagnostics for all cases when a
5668 control that does not make sense is encountered.
5670 7. $recipients is now available in the predata ACL (oversight).
5672 8. Tidy the search cache before the fork to do a delivery from a message
5673 received from the command line. Otherwise the child will trigger a lookup
5674 failure and thereby defer the delivery if it tries to use (for example) a
5675 cached ldap connection that the parent has called unbind on.
5677 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5678 of $address_data from the recipient verification was clobbered by the
5679 sender verification.
5681 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5682 was its contents. (It was OK if the option was not defined at all.)
5684 11. A "Completed" log line is now written for messages that are removed from
5685 the spool by the -Mrm option.
5687 12. $host_address is now set to the target address during the checking of
5688 ignore_target_hosts.
5690 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5691 being passed; this would have caused $sender_host_name to have been used if
5692 matching the list had actually called for a host name (not very likely,
5693 since this list is usually IP addresses). A host name is now passed as
5696 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5697 code that specifies a non-restarting handler (typically sigaction() in
5698 modern systems) in an attempt to fix a rare and obscure crash bug.
5700 15. Narrowed the window for a race in the daemon that could cause it to ignore
5701 SIGCHLD signals. This is not a major problem, because they are used only to
5702 wake it up if nothing else does.
5704 16. A malformed maildirsize file could cause Exim to calculate negative values
5705 for the mailbox size or file count. Odd effects could occur as a result.
5706 The maildirsize information is now recalculated if the size or filecount
5709 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5710 support for a long time. Removed HAVE_SYS_VFS_H.
5712 18. Updated exipick to current release from John Jetmore.
5714 19. Allow an empty sender to be matched against a lookup in an address list.
5715 Previously the only cases considered were a regular expression, or an
5718 20. Exim went into a mad DNS lookup loop when doing a callout where the
5719 host was specified on the transport, if the DNS lookup yielded more than
5722 21. The RFC2047 encoding function was originally intended for short strings
5723 such as real names; it was not keeping to the 75-character limit for
5724 encoded words that the RFC imposes. It now respects the limit, and
5725 generates multiple encoded words if necessary. To be on the safe side, I
5726 have increased the buffer size for the ${rfc2047: expansion operator from
5729 22. Failure to deliver a bounce message always caused it to be frozen, even if
5730 there was an errors_to setting on the router. The errors_to setting is now
5733 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5734 canonical form (fully expanded) before being placed in
5735 $sender_host_address.
5737 24. Updated eximstats to version 1.33
5739 25. Include certificate and key file names in error message when GnuTLS fails
5740 to set them up, because the GnuTLS error message doesn't include the name
5741 of the failing file when there is a problem reading it.
5743 26. Expand error message when OpenSSL has problems setting up cert/key files.
5746 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5747 (this can affect the format of dates).
5749 28. exim_tidydb, when checking for the continued existence of a message for
5750 which it has found a message-specific retry record, was not finding
5751 messages that were in split spool directories. Consequently, it was
5752 deleting retry records that should have stayed in existence.
5754 29. eximstats updated to version 1.35
5755 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5756 1.35 - bugfix such that pie charts by volume are generated correctly
5758 30. The SPA authentication driver was not abandoning authentication and moving
5759 on to the next authenticator when an expansion was forced to fail,
5760 contradicting the general specification for all authenticators. Instead it
5761 was generating a temporary error. It now behaves as specified.
5763 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5764 (the order specifies the preference for clients). The order is now AES256,
5765 AES128, 3DES, ARCFOUR128.
5767 31. Small patch to Sieve code - explicitly set From: when generating an
5770 32. Exim crashed if a remote delivery caused a very long error message to be
5771 recorded - for instance if somebody sent an entire SpamAssassin report back
5772 as a large number of 550 error lines. This bug was coincidentally fixed by
5773 increasing the size of one of Exim's internal buffers (big_buffer) that
5774 happened as part of the Exiscan merge. However, to be on the safe side, I
5775 have made the code more robust (and fixed the comments that describe what
5778 33. Some experimental protocols are using DNS PTR records for new purposes. The
5779 keys for these records are domain names, not reversed IP addresses. The
5780 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5781 leaves it alone. Component reversal etc. now happens only for IP addresses.
5784 34. The host_aton() function is supposed to be passed a string that is known
5785 to be a valid IP address. However, in the case of IPv6 addresses, it was
5786 not checking this. This is a hostage to fortune. Exim now panics and dies
5787 if the condition is not met. A case was found where this could be provoked
5788 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5789 components; fortuitously, this particular loophole had already been fixed
5790 by change 4.50/55 or 4.44/33 above.
5792 If there are any other similar loopholes, the new check in host_aton()
5793 itself should stop them being exploited. The report I received stated that
5794 data on the command line could provoke the exploit when Exim was running as
5795 exim, but did not say which command line option was involved. All I could
5796 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5797 running as the user.
5800 35. There was a buffer overflow vulnerability in the SPA authentication code
5801 (which came originally from the Samba project). I have added a test to the
5802 spa_base64_to_bits() function which I hope fixes it.
5805 36. The daemon start-up calls getloadavg() while still root for those OS that
5806 need the first call to be done as root, but it missed one case: when
5807 deliver_queue_load_max is set with deliver_drop_privilege. This is
5808 necessary for the benefit of the queue runner, because there is no re-exec
5809 when deliver_drop_privilege is set.
5811 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5812 was in use, was not putting the data itself into the right store pool;
5813 consequently, it could be overwritten for a subsequent message in the same
5814 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5817 38. Sometimes the final signoff response after QUIT could fail to get
5818 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5819 < 0 before doing a fflush(). This bug looks as though it goes back to the
5820 introduction of TLS in release 3.20, but "sometimes" must have been rare
5821 because the tests only now provoked it.
5827 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5828 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5829 Now it returns an integer. A number of calls were still expecting a T/F
5830 return. Fortuitously, in all cases, the tests worked in OK situations,
5831 which is the norm. However, things would have gone wrong on any write
5832 failures on the smtp file descriptor. This function is used when sending
5833 messages over SMTP and also when doing verify callouts.
5835 2. When Exim is called to do synchronous delivery of a locally submitted
5836 message (the -odf or -odi options), it no longer closes stderr before doing
5839 3. Implemented the mua_wrapper option.
5841 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5843 5. Implemented the functions header_remove(), header_testname(),
5844 header_add_at_position(), and receive_remove_recipient(), and exported them
5847 6. If an ACL "warn" statement specified the addition of headers, Exim already
5848 inserted X-ACL-Warn: at the start if there was no header name. However, it
5849 was not making this test for the second and subsequent header lines if
5850 there were newlines in the string. This meant that an invalid header could
5851 be inserted if Exim was badly configured.
5853 7. Allow an ACL "warn" statement to add header lines at the start or after all
5854 the Received: headers, as well as at the end.
5856 8. Added the rcpt_4xx retry error code.
5858 9. Added postmaster_mailfrom=xxx to callout verification option.
5860 10. Added mailfrom=xxxx to the callout verification option, for verify=
5863 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5864 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5866 12. Inserted some casts to stop certain compilers warning when using pointer
5867 differences as field lengths or precisions in printf-type calls (mostly
5868 affecting debugging statements).
5870 13. Added optional readline() support for -be (dynamically loaded).
5872 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5873 same clock tick as a message's arrival, so that its received time was the
5874 same as the "first fail" time on the retry record, and that message
5875 remained on the queue past the ultimate address timeout, every queue runner
5876 would try a delivery (because it was past the ultimate address timeout) but
5877 after another failure, the ultimate address timeout, which should have then
5878 bounced the address, did not kick in. This was a "< instead of <=" error;
5879 in most cases the first failure would have been in the next clock tick
5880 after the received time, and all would be well.
5882 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5883 being recognized when the domain list was tested by the match_domain
5884 condition in an expansion string.
5886 16. Added the ${str2b64: operator.
5888 17. Exim was always calling setrlimit() to set a large limit for the number of
5889 processes, without checking whether the existing limit was already
5890 adequate. (It did check for the limit on file descriptors.) Furthermore,
5891 errors from getrlimit() and setrlimit() were being ignored. Now they are
5892 logged to the main and panic logs, but Exim does carry on, to try to do its
5893 job under whatever limits there are.
5895 18. Imported PCRE 5.0.
5897 19. Trivial typo in log message " temporarily refused connection" (the leading
5900 20. If the log selector return_path_on_delivery was set and an address was
5901 redirected to /dev/null, the delivery process crashed because it assumed
5902 that a return path would always be set for a "successful" delivery. In this
5903 case, the whole delivery is bypassed as an optimization, and therefore no
5906 21. Internal re-arrangement: the function for sending a challenge and reading
5907 a response while authentication was assuming a zero-terminated challenge
5908 string. It's now changed to take a pointer and a length, to allow for
5909 binary data in such strings.
5911 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5913 23. Exim was not respecting finduser_retries when seeking the login of the
5914 uid under which it was called; it was always trying 10 times. (The default
5915 setting of finduser_retries is zero.) Also, it was sleeping after the final
5916 failure, which is pointless.
5918 24. Implemented tls_on_connect_ports.
5920 25. Implemented acl_smtp_predata.
5922 26. If the domain in control=submission is set empty, Exim assumes that the
5923 authenticated id is a complete email address when it generates From: or
5924 Sender: header lines.
5926 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5927 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5928 chown and chgrp in /bin and hostname in /usr/bin.
5930 28. Exim was keeping the "process log" file open after each use, just as it
5931 does for the main log. This opens the possibility of it remaining open for
5932 long periods when the USR1 signal hits a daemon. Occasional processlog
5933 errors were reported, that could have been caused by this. Anyway, it seems
5934 much more sensible not to leave this file open at all, so that is what now
5937 29. The long-running daemon process does not normally write to the log once it
5938 has entered its main loop, and it closes the log before doing so. This is
5939 so that log files can straightforwardly be renamed and moved. However,
5940 there are a couple of unusual error situations where the daemon does write
5941 log entries, and I had neglected to close the log afterwards.
5943 30. The text of an SMTP error response that was received during a remote
5944 delivery was being truncated at 512 bytes. This is too short for some of
5945 the long messages that one sometimes sees. I've increased the limit to
5948 31. It is now possible to make retry rules that apply only when a message has a
5949 specific sender, in particular, an empty sender.
5951 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5952 it possible to be selective about when SMTP synchronization is enforced.
5954 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5956 32. Implemented hosts_connection_nolog.
5958 33. Added an ACL for QUIT.
5960 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5963 35. Added mailbox_size and mailbox_filecount to appendfile.
5965 36. Added control = no_multiline_responses to ACLs.
5967 37. There was a bug in the logic of the code that waits for the clock to tick
5968 in the case where the clock went backwards by a substantial amount such
5969 that the microsecond fraction of "now" was more than the microsecond
5970 fraction of "then" (but the whole seconds number was less).
5972 38. Added support for the libradius Radius client library this is found on
5973 FreeBSD (previously only the radiusclient library was supported).
5979 1. When certain lookups returned multiple values in the form name=value, the
5980 quoting of the values was not always being done properly. Specifically:
5981 (a) If the value started with a double quote, but contained no whitespace,
5983 (b) If the value contained whitespace other than a space character (i.e.
5984 tabs or newlines or carriage returns) it was not quoted.
5985 This fix has been applied to the mysql and pgsql lookups by writing a
5986 separate quoting function and calling it from the lookup code. The fix
5987 should probably also be applied to nisplus, ibase and oracle lookups, but
5988 since I cannot test any of those, I have not disturbed their existing code.
5990 2. A hit in the callout cache for a specific address caused a log line with no
5991 reason for rejecting RCPT. Now it says "Previous (cached) callout
5992 verification failure".
5994 3. There was an off-by-one bug in the queryprogram router. An over-long
5995 return line was truncated at 256 instead of 255 characters, thereby
5996 overflowing its buffer with the terminating zero. As well as fixing this, I
5997 have increased the buffer size to 1024 (and made a note to document this).
5999 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6000 when Exim is waiting for an SMTP response from a remote server, Exim
6001 restarts its select() call on the socket, thereby resetting its timeout.
6002 This is not a problem when such interrupts are rare. Somebody set up a cron
6003 job to run exiwhat every 2 minutes, which is less than the normal select()
6004 timeout (5 or 10 minutes). This meant that the select() timeout never
6005 kicked in because it was always reset. I have fixed this by comparing the
6006 time when an interrupt arrives with the time at the start of the first call
6007 to select(). If more time than the timeout has elapsed, the interrupt is
6008 treated as a timeout.
6010 5. Some internal re-factoring in preparation for the addition of Sieve
6011 extensions (by MH). In particular, the "personal" test is moved to a
6012 separate function, and given an option for scanning Cc: and Bcc: (which is
6013 not set for Exim filters).
6015 6. When Exim created an email address using the login of the caller as the
6016 local part (e.g. when creating a From: or Sender: header line), it was not
6017 quoting the local part when it contained special characters such as @.
6019 7. Installed new OpenBSD configuration files.
6021 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6022 try to make them clearer.
6024 9. Callout options, other than the timeout value, were being ignored when
6025 verifying sender addresses in header lines. For example, when using
6027 verify = header_sender/callout=no_cache
6029 the cache was (incorrectly) being used.
6031 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6032 only the Cygwin environment.
6034 11. When return_path_on_delivery was set as a log selector, if different remote
6035 addresses in the same message used different return paths and parallel
6036 remote delivery occurred, the wrong values would sometimes be logged.
6037 (Whenever a remote delivery process finished, the return path value from
6038 the most recently started remote delivery process was logged.)
6040 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6041 lines when AUTH and/or TLS are in use. This is the "received protocol"
6042 field. Exim used to use "asmtp" for authenticated SMTP, without any
6043 indication (in the protocol name) for TLS use. Now it follows the RFC and
6044 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6045 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6046 names appear in log lines as well as in Received: header lines.
6048 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6049 extensions, and comparison tests, and to fix some bugs.
6051 14. Changes to the "personal" filter test:
6053 (1) The test was buggy in that it was just doing the equivalent of
6054 "contains" tests on header lines. For example, if a user's address was
6055 anne@some.where, the "personal" test would incorrectly be true for
6057 To: susanne@some.where
6059 This test is now done by extracting each address from the header in turn,
6060 and checking the entire address. Other tests that are part of "personal"
6061 are now done using regular expressions (for example, to check local parts
6062 of addresses in From: header lines).
6064 (2) The list of non-personal local parts in From: addresses has been
6065 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6066 taken from the Sieve specification recommendations.
6068 (3) If the message contains any header line starting with "List-" it is
6069 treated as non-personal.
6071 (4) The test for "circular" in the Subject: header line has been removed
6072 because it now seems ill-conceived.
6074 15. Minor typos in src/EDITME comments corrected.
6076 16. Installed latest exipick from John Jetmore.
6078 17. If headers_add on a router specified a text string that was too long for
6079 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6080 of string_sprintf() is now avoided.
6082 18. $message_body_size was not set (it was always zero) when running the DATA
6083 ACL and the local_scan() function.
6085 19. For the "mail" command in an Exim filter, no default was being set for
6086 the once_repeat time, causing a random time value to be used if "once" was
6087 specified. (If the value happened to be <= 0, no repeat happened.) The
6088 default is now 0s, meaning "never repeat". The "vacation" command was OK
6089 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6090 (I found it when inspecting the code).
6092 20. There is now an overall timeout for performing a callout verification. It
6093 defaults to 4 times the callout timeout, which applies to individual SMTP
6094 commands during the callout. The overall timeout applies when there is more
6095 than one host that can be tried. The timeout is checked before trying the
6096 next host. This prevents very long delays if there are a large number of
6097 hosts and all are timing out (e.g. when the network connections are timing
6098 out). The value of the overall timeout can be changed by specifying an
6099 additional sub-option for "callout", called "maxwait". For example:
6101 verify = sender/callout=5s,maxwait=20s
6103 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6104 to the end before writing, but this should make it even safer).
6106 22. Exim was forgetting that it had advertised PIPELINING for the second and
6107 subsequent messages on an SMTP connection. It was also not resetting its
6108 memory on STARTTLS and an internal HELO.
6110 23. When Exim logs an SMTP synchronization error within a session, it now
6111 records whether PIPELINING has been advertised or not.
6113 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6114 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6115 rather than long int.
6117 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6119 26. Added the never_mail option to autoreply.
6125 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6126 crash if the getsockname() call failed; this can happen if a connection is
6127 closed very soon after it is established. The problem was simply in the
6128 order in which certain operations were done, causing Exim to try to write
6129 to the SMTP stream before it had set up the file descriptor. The bug has
6130 been fixed by making things happen in the correct order.
6136 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6137 before the connection was closed, thus losing the rejection response.
6139 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6140 some early Solaris releases, but causes trouble in current releases where
6141 socklen_t is defined.
6143 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6146 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6149 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6150 that releasing the top of it at the end releases what was used for sub-
6151 expansions (unless the block got too big). However, discard this block if
6152 the first thing is a variable or header, so that we can use its block when
6153 it is dynamic (useful for very large $message_headers, for example).
6155 6. Lookups now cache *every* query, not just the most recent. A new, separate
6156 store pool is used for this. It can be recovered when all lookup caches are
6157 flushed. Lookups now release memory at the end of their result strings.
6158 This has involved some general refactoring of the lookup sources.
6160 7. Some code has been added to the store_xxx() functions to reduce the amount
6161 of flapping under certain conditions.
6163 8. log_incoming_interface used to affect only the <= reception log lines. Now
6164 it causes the local interface and port to be added to several more SMTP log
6165 lines, for example "SMTP connection from", and rejection lines.
6167 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6169 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6171 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6172 could be overwritten at the end of the current message (or the start of a
6173 new message if it was set in a HELO ACL). The value is now preserved for
6174 the duration of the SMTP connection.
6176 12. If a transport had a headers_rewrite setting, and a matching header line
6177 contained an unqualified address, that address was qualified, even if it
6178 did not match any rewriting rules. The underlying bug was that the values
6179 of the flags that permit the existence of unqualified sender and recipient
6180 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6181 non-local messages, and by -bnq for local messages) were not being
6182 preserved with the message after it was received.
6184 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6185 "next input=" as part of the text comprising the host identity instead of
6186 the correct text. The code was using the same buffer for two different
6187 strings. However, depending on which order the printing function evaluated
6188 its arguments, the bug did not always show up. Under Linux, for example, my
6189 test suite worked just fine.
6191 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6192 doesn't work with some older versions of Perl. It has been changed to "my",
6193 which in any case is probably the better facility to use.
6195 15. A really picky compiler found some instances of statements for creating
6196 error messages that either had too many or two few arguments for the format
6199 16. The size of the buffer for calls to the DNS resolver has been increased
6200 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6201 for addresses that have a lot of PTR records. This alleviates a problem; it
6202 does not fully solve it.
6204 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6205 buffer now truncates the list and logs the incident, which is the same
6206 action as happens when Exim is looking up a host name and its aliases.
6207 Previously in this situation something unpredictable would happen;
6208 sometimes it was "internal error: store_reset failed".
6210 18. If a server dropped the connection unexpectedly when an Exim client was
6211 using GnuTLS and trying to read a response, the client delivery process
6212 crashed while trying to generate an error log message.
6214 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6215 string, for example:
6217 warn message = H1: something\nH2: something
6219 the text was added as a single header line from Exim's point of view
6220 though it ended up OK in the delivered message. However, searching for the
6221 second and subsequent header lines using $h_h2: did not work. This has been
6222 fixed. Similarly, if a system filter added multiple headers in this way,
6223 the routers could not see them.
6225 20. Expanded the error message when iplsearch is called with an invalid key to
6226 suggest using net-iplsearch in a host list.
6228 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6229 ACLs are no longer actually imposed (and a message to that effect is
6232 22. If a "gecos" field in a passwd entry contained escaped characters, in
6233 particular, if it contained a \" sequence, Exim got it wrong when building
6234 a From: or a Sender: header from that name. A second bug also caused
6235 incorrect handling when an unquoted " was present following a character
6236 that needed quoting.
6238 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6239 was not being matched caselessly.
6241 24. Arranged for all hyphens in the exim.8 source to be escaped with
6244 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6245 properly. Recipient callout cache records were still being keyed to include
6246 the sender, even when use_sender was set false. This led to far more
6247 callouts that were necessary. The sender is no longer included in the key
6248 when use_sender is false.
6250 26. Added "control = submission" modifier to ACLs.
6252 27. Added the ${base62d: operator to decode base 62 numbers.
6254 28. dnsdb lookups can now access SRV records.
6256 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6257 the configuration file.
6259 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6260 (-v) mode. This makes the output for a verbose queue run more intelligible.
6262 31. Added a use_postmaster feature to recipient callouts.
6264 32. Added the $body_zerocount variable, containing the number of binary zero
6265 bytes in the message body.
6267 33. The time of last modification of the "new" subdirectory is now used as the
6268 "mailbox time last read" when there is a quota error for a maildir
6271 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6273 35. Added +ignore_unknown as a special item in host lists.
6275 36. Code for decoding IPv6 addresses in host lists is now included, even if
6276 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6277 address was recognized as an IP address, but was then not correctly decoded
6278 into binary, causing unexpected and incorrect effects when compared with
6285 1. Very minor rewording of debugging text in manualroute to say "list of
6286 hosts" instead of "hostlist".
6288 2. If verify=header_syntax was set, and a header line with an unqualified
6289 address (no domain) and a large number of spaces between the end of the
6290 name and the colon was received, the reception process suffered a buffer
6291 overflow, and (when I tested it) crashed. This was caused by some obsolete
6292 code that should have been removed. The fix is to remove it!
6294 3. When running in the test harness, delay a bit after writing a bounce
6295 message to get a bit more predictability in the log output.
6297 4. Added a call to search_tidyup() just before forking a reception process. In
6298 theory, someone could use a lookup in the expansion of smtp_accept_max_
6299 per_host which, without the tidyup, could leave open a database connection.
6301 5. Added the variables $recipient_data and $sender_data which get set from a
6302 lookup success in an ACL "recipients" or "senders" condition, or a router
6303 "senders" option, similar to $domain_data and $local_part_data.
6305 6. Moved the writing of debug_print from before to after the "senders" test
6308 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6309 problems for message scanning, either using a data ACL, or using
6310 local_scan() because the Received: header was not generated till after they
6311 were called (in order to set the time as the time of reception completion).
6312 I have revised the way this works. The header is now generated after the
6313 body is received, but before the ACL or local_scan() are called. After they
6314 are run, the timestamp in the header is updated.
6320 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6321 before starting a queue runner without re-exec. This happened only when
6322 deliver_drop_privilege was set or when the Exim user was set to root. The
6323 effect of the bug was that timeouts during subsequent deliveries caused
6324 crashes instead of being properly handled. The handler is now left at its
6325 default (and expected) setting.
6327 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6328 message, again when deliver_drop_privilege is set or Exim is run as root.
6329 The bug described in (1) was not present in this case, but the tidying up
6330 of the other signals was missing. I have made the two cases consistent.
6332 3. The ignore_target_hosts setting on a manualroute router was being ignored
6333 for hosts that were looked up using the /MX notation.
6335 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6338 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6339 operated on the sender address. After changing the $sender_address to <>
6340 for the sender address verify, Exim was re-instated it as the original
6341 (before rewriting) address, but remembering that it had rewritten it, so it
6342 wasn't rewriting it again. This bug also had the effect of breaking the
6343 sender address verification caching when the sender address was rewritten.
6345 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6346 This has been changed so that if the ip literal address matches
6347 ignore_target_hosts, the router declines.
6349 7. Added expansion conditions match_domain, match_address, and match_local_
6350 part (NOT match_host).
6352 8. The placeholder for the Received: header didn't have a length field set.
6354 9. Added code to Exim itself and to exim_lock to test for a specific race
6355 condition that could lead to file corruption when using MBX delivery. The
6356 issue is with the lockfile that is created in /tmp. If this file is removed
6357 after a process has opened it but before that process has acquired a lock,
6358 there is the potential for a second process to recreate the file and also
6359 acquire a lock. This could lead to two Exim processes writing to the file
6360 at the same time. The added code performs the same test as UW imapd; it
6361 checks after acquiring the lock that its file descriptor still refers to
6362 the same named file.
6364 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6365 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6368 11. Added the smtp_active_hostname option. If used, this will typically be made
6369 to depend on the incoming interface address. Because $interface_address is
6370 not set up until the daemon has forked a reception process, error responses
6371 that can happen earlier (such as "too many connections") no longer contain
6374 12. If an expansion in a condition on a "warn" statement fails because a lookup
6375 defers, the "warn" statement is abandoned, and the next ACL statement is
6376 processed. Previously this caused the whole ACL to be aborted.
6378 13. Added the iplsearch lookup type.
6380 14. Added ident_timeout as a log selector.
6382 15. Added tls_certificate_verified as a log selector.
6384 16. Added a global option tls_require_ciphers (compare the smtp transport
6385 option of the same name). This controls incoming TLS connections.
6387 17. I finally figured out how to make tls_require_ciphers do a similar thing
6388 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6389 before starting the TLS session.
6391 18. Tabs are now shown as \t in -bP output.
6393 19. If the log selector return_path_on_delivery was set, Exim crashed when
6394 bouncing a message because it had too many Received: header lines.
6396 20. If two routers both had headers_remove settings, and the first one included
6397 a superfluous trailing colon, the final name in the first list and the
6398 first name in the second list were incorrectly joined into one item (with a
6399 colon in the middle).
6405 1. Added -C and -D options to the exinext utility, mainly to make it easier
6406 to include in the automated testing, but these could be helpful when
6407 multiple configurations are in use.
6409 2. The exinext utility was not formatting the output nicely when there was
6410 an alternate port involved in the retry record key, nor when there was a
6411 message id as well (for retries that were specific to a specific message
6412 and a specific host). It was also confused by IPv6 addresses, because of
6413 the additional colons they contain. I have fixed the IPv4 problem, and
6414 patched it up to do a reasonable job for IPv6.
6416 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6417 delivery, the log line now contains "pipelined" if PIPELINING was used.
6419 4. An SMTP transport process used to panic and die if the bind() call to set
6420 an explicit outgoing interface failed. This has been changed; it is now
6421 treated in the same way as a connect() failure.
6423 5. A reference to $sender_host_name in the part of a conditional expansion
6424 that was being skipped was still causing a DNS lookup. This no longer
6427 6. The def: expansion condition was not recognizing references to header lines
6428 that used bh_ and bheader_.
6430 7. Added the _cache feature to named lists.
6432 8. The code for checking quota_filecount in the appendfile transport was
6433 allowing one more file than it should have been.
6435 9. For compatibility with Sendmail, the command line option
6443 and sets the incoming protocol and host name (for trusted callers). The
6444 host name and its colon can be omitted when only the protocol is to be set.
6445 Note the Exim already has two private options, -pd and -ps, that refer to
6446 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6447 "s", but I don't think that's a major issue.
6449 10. A number of refactoring changes to the code, none of which should affect
6452 (a) The number of logging options was getting close to filling up the
6453 32-bit word that was used as a bit map. I have split them into two classes:
6454 those that are passed in the argument to log_write(), and those that are
6455 only ever tested independently outside of that function. These are now in
6456 separate 32-bit words, so there is plenty of room for expansion again.
6457 There is no change in the user interface or the logging behaviour.
6459 (b) When building, for example, log lines, the code previously used a
6460 macro that called string_cat() twice, in order to add two strings. This is
6461 not really sufficiently general. Furthermore, there was one instance where
6462 it was actually wrong because one of the argument was used twice, and in
6463 one call a function was used. (As it happened, calling the function twice
6464 did not affect the overall behaviour.) The macro has been replaced by a
6465 function that can join an arbitrary number of extra strings onto a growing
6468 (c) The code for expansion conditions now uses a table and a binary chop
6469 instead of a serial search (which was left over from when there were very
6470 few conditions). Also, it now recognizes conditions like "pam" even when
6471 the relevant support is not compiled in: a suitably worded error message is
6472 given if an attempt is made to use such a condition.
6474 11. Added ${time_interval:xxxxx}.
6476 12. A bug was causing one of the ddress fields not to be passed back correctly
6477 from remote delivery subprocesses. The field in question was not being
6478 subsequently used, so this caused to problems in practice.
6480 13. Added new log selectors queue_time and deliver_time.
6482 14. Might have fixed a bug in maildirsizefile handling that threw up
6483 "unexpected character" debug warnings, and recalculated the data
6484 unnecessarily. In any case, I expanded the warning message to give more
6487 15. Added the message "Restricted characters in address" to the statements in
6488 the default ACL that block characters like @ and % in local parts.
6490 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6491 Three changes have been made:
6493 (a) There was a serious bug; a negative response to MAIL caused the whole
6494 recipient domain to be cached as invalid, thereby blocking all messages
6495 to all local parts at the same domain, from all senders. This bug has
6496 been fixed. The domain is no longer cached after a negative response to
6497 MAIL if the sender used is not empty.
6499 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6502 (c) A new callout option, "use_sender" has been added for people who want
6503 the modified behaviour.
6509 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6512 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6513 indeed breaks things for older releases.
6515 3. Added additional logging to the case where there is a problem reading data
6516 from a filter that is running in a subprocess using a pipe, in order to
6517 try to track down a specific problem.
6519 4. Testing facility fudge: when running in the test harness and attempting
6520 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6521 getting "No route to host". Convert this to a timeout.
6523 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6526 6. Some OS don't have socklen_t but use size_t instead. This affects the
6527 fifth argument of getsockopt() amongst other things. This is now
6528 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6529 can be set for individual OS. I have set it for SunOS5, OSF1, and
6530 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6531 some earlier ones do not.
6533 7. Change 4.30/15 was not doing the test caselessly.
6535 8. The standard form for an IPv6 address literal was being rejected by address
6536 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6537 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6538 this, as well as the form without the "IPv6" on the front (but only when
6539 address literals are enabled, of course).
6541 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6543 10. Exim crashed if a message with an empty sender address specified by -f
6544 encountered a router with an errors_to setting. This could be provoked only
6545 by a command such as
6549 where an empty string was supplied; "<>" did not hit this bug.
6551 11. Installed PCRE release 4.5.
6553 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6554 remained set. It is now erased.
6556 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6557 times from message ids (which are base 36 rather than the normal 62).
6559 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6560 were being counted as actual protocol errors, and logged if the log
6561 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6562 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6563 and DATA following a set of rejected RCPTs do not count as protocol errors.
6564 In other words, Exim assumes they were pipelined, though this may not
6565 actually be the case. Of course, in all cases the client gets an
6566 appropriate error code.
6568 15. If a lookup fails in an ACL condition, a message about the failure may
6569 be available; it is used if testing the ACL cannot continue, because most
6570 such messages specify what the cause of the deferral is. However, some
6571 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6572 that caused an old message to be retained and used if a later statement
6573 caused a defer, replacing the real cause of the deferral.
6575 16. If an IP address had so many PTR records that the DNS lookup buffer
6576 was not large enough to hold them, Exim could crash while trying to process
6577 the truncated data. It now detects and logs this case.
6579 17. Further to 4.21/58, another change has been made: if (and only if) the
6580 first line of a message (the first header line) ends with CRLF, a bare LF
6581 in a subsequent header line has a space inserted after it, so as not to
6582 terminate the header.
6584 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6585 unnecessarily, used atoi() instead of strtol(), and didn't check the
6586 termination when getting file sizes from file names by regex.
6588 19. Completely re-implemented the support for maildirsize files, in the light
6589 of a number of problems with the previous contributed implementation
6590 (4.30/29). In particular:
6592 . If the quota is zero, the maildirsize file is maintained, but no quota is
6595 . If the maildir directory does not exist, it is created before any attempt
6596 to write a maildirsize file.
6598 . The quota value in the file is just a cache; if the quota is changed in
6599 the transport, the new value overrides.
6601 . A regular expression is available for excluding directories from the
6604 20. The autoreply transport checks the characters in options that define the
6605 message's headers; it allows continued headers, but it was checking with
6606 isspace() after an embedded newline instead of explicitly looking for a
6609 21. If all the "regular" hosts to which an address was routed had passed their
6610 expiry times, and had not reached their retry times, the address was
6611 bounced, even if fallback hosts were defined. Now Exim should go on to try
6614 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6615 equivalent code in the SMTP transport. Some hosts send humungous responses
6616 to HELO/EHLO, more than 1024 it seems.
6618 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6619 gives compiler warnings in some environments. I've now done it "properly",
6622 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6623 (because of problems with the built-in one) was declared to return uschar *
6624 instead of char *, causing compiler failure.
6626 25. Fixed a file descriptor leak when processing alias/forward files.
6628 26. Fixed a minor format string issue in dbfn.c.
6630 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6632 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6633 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6636 29. When a sender address is verified, it is cached, to save repeating the test
6637 when there is more than one recipient in a message. However, when the
6638 verification involves a callout, it is possible for different callout
6639 options to be set for different recipients. It is too complicated to keep
6640 track of this in the cache, so now Exim always runs a verification when a
6641 callout is required, relying on the callout cache for the optimization.
6642 The overhead is duplication of the address routing, but this should not be
6645 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6646 to be verified with callout=postmaster, and the main callout worked but the
6647 postmaster check failed, the verification correctly failed. However, if a
6648 subsequent RCPT command asked for sender verification *without* the
6649 postmaster check, incorrect caching caused this verification also to fail,
6652 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6653 it was not caching the DNS options (qualify_single, search_parents) that
6654 were used when the lookup failed. A subsequent lookup with different
6655 options therefore always gave the same answer, though there were cases
6656 where it should not have. (Example: a "domains = !$mx_any" option on a
6657 dnslookup router: the "domains" option is always processed without any
6658 widening, but the router might have qualify_single set.) Now Exim uses the
6659 cached value only when the same options are set.
6661 32. Added John Jetmore's "exipick" utility to the distribution.
6663 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6664 than a timeout (e.g. a certificate is required, and is not provided), an
6665 Exim server now closes the connection immediately. Previously it waited for
6666 the client to close - but if the client is SSL, it seems that they each
6667 wait for each other, leading to a delay before one of them times out.
6669 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6670 maintained 0.8.x compatibility because I don't think many are using it, and
6671 it is clearly obsolete.
6673 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6676 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6677 client certificate was expired. A simple patch fixes this, though I don't
6678 understand the full logic of why the verify callback is called multiple
6681 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6682 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6683 which causes problems with some clients (such as the Certicom SSL Plus
6684 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6685 disables the coutermeasure allowing Eudora to connect."
6687 38. Exim was not checking that a write() to a log file succeeded. This could
6688 lead to Bad Things if a log got too big, in particular if it hit a file
6689 size limit. Exim now panics and dies if it cannot write to a log file, just
6690 as it does if it cannot open a log file.
6692 39. Modified OS/Makefile-Linux so that it now contains
6694 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6696 The two -D definitions ensure that Exim is compiled with large file
6697 support, which makes it possible to handle log files that are bigger than
6700 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6701 instance) a domain was checked against a named list that involved a lookup,
6702 causing $domain_data to be set, then another domain was checked against the
6703 same list, then the first domain was re-checked, the value of $domain_data
6704 after the final check could be wrong. In particular, if the second check
6705 failed, it could be set empty. This bug probably also applied to
6708 41. The strip_trailing_dot option was not being applied to the address given
6709 with the -f command-line option.
6711 42. The code for reading a message's header from the spool was incrementing
6712 $received_count, but never initializing it. This meant that the value was
6713 incorrect (doubled) while delivering a message in the same process in which
6714 it was received. In the most common configuration of Exim, this never
6715 happens - a fresh exec is done - but it can happen when
6716 deliver_drop_privilege is set.
6718 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6719 it now includes up to 150 characters of the unexpected data in the log
6722 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6723 and building data strings. The size of both of these buffers was 10 000
6724 bytes - far larger than anybody would *ever* want, thought I. Needless to
6725 say, somebody hit the limit. I have increased the maximum line length to
6726 20 000 and the maximum data length of concatenated lines to 100 000. I have
6727 also fixed two bugs, because there was no checking on these buffers. Tsk,
6728 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6729 buffer is too small.
6731 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6732 lsearch lookups. Now it does.
6734 46. When parsing a route_list item in a manualroute router, a fixed-length
6735 buffer was used for the list of hosts. I made this 1024 bytes long,
6736 thinking that nobody would ever have a list of hosts that long. Wrong.
6737 Somebody had a whole pile of complicated expansion conditions, and the
6738 string was silently truncated, leading to an expansion error. It turns out
6739 that it is easier to change to an unlimited length (owing to other changes
6740 that have happened since this code was originally written) than to build
6741 structure for giving a limitation error. The length of the item that
6742 expands into the list of hosts is now unlimited.
6744 47. The lsearch lookup could not handle data where the length of text line was
6745 more than 4095 characters. Such lines were truncated, leading to shortened
6746 data being returned. It should now handle lines of any length.
6748 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6749 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6752 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6753 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6754 no longer happens in comment lines. A list of replacements is now placed
6755 at the head of all of the source files, except those whose only change is
6756 to replace PERL_COMMAND in the very first #! line.
6758 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6759 messages on the queue, with a bottom-up merge sort, using code contributed
6760 by Michael Haardt. This should make operations like -bp somewhat faster on
6761 large queues. It won't affect queue runners, except when queue_run_in_order
6764 51. Installed eximstats 1.31 in the distribution.
6766 52. Added support for SRV lookups to the dnslookup router.
6768 53. If an ACL referred to $message_body or $message_body_end, the value was not
6769 reset for any messages that followed in the same SMTP session.
6771 54. The store-handling optimization for building very long strings was not
6772 differentiating between the different store pools. I don't think this
6773 actually made any difference in practice, but I've tidied it.
6775 55. While running the routers to verify a sender address, $sender_address
6776 was still set to the sender address. This is wrong, because when routing to
6777 send a bounce to the sender, it would be empty. Therefore, I have changed
6778 it so that, while verifying a sender address, $sender_address is set to <>.
6779 (There is no change to what happens when verifying a recipient address.)
6781 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6782 target A or AAAA records (if not already returned) without resetting the
6783 qualify_single or search_parents options of the DNS resolver. These are
6784 inappropriate in this case because the targets of MX and SRV records must
6785 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6786 target that, when qualified, matched something in the local domain. These
6787 two options are now turned off when doing these lookups.
6789 57. It seems that at least some releases of Reiserfs (which does not have the
6790 concept of a fixed number of inodes) returns zero and not -1 for the
6791 number of available inodes. This interacted badly with check_spool_inodes,
6792 which assumed that -1 was the "no such thing" setting. What I have done is
6793 to check that the total number of inodes is greater than zero before doing
6794 the test of how many are available.
6796 58. When a "warn" ACL statement has a log_message modifier, the message is
6797 remembered, and not repeated. This is to avoid a lot of repetition when a
6798 message has many recipients that cause the same warning to be written.
6799 However, Exim was preserving the list of already written lines for an
6800 entire SMTP session, which doesn't seem right. The memory is now reset if a
6801 new message is started.
6803 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6804 debugging output unless log_rewrite was also set.
6806 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6807 of (int)(handle) when we know that handle contains (void *)(-1).
6809 61. The Exim daemon panic-logs an error return when it closes the incoming
6810 connection. However "connection reset by peer" seems to be common, and
6811 isn't really an error worthy of noting specially, so that particular error
6814 62. When Exim is trying to find all the local interfaces, it used to panic and
6815 die if the ioctl to get the interface flags failed. However, it seems that
6816 on at least one OS (Solaris 9) it is possible to have an interface that is
6817 included in the list of interfaces, but for which you get a failure error
6818 for this call. This happens when the interface is not "plumbed" into a
6819 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6820 failure of the "get flags" call assumes that the interface is down.
6822 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6823 makes life easier for people who are doing arithmetic on fields extracted
6824 from dates, where you often get leading zeros that should not be
6825 interpreted as octal.
6827 64. Added qualify_domain to the redirect router, to override the global
6830 65. If a pathologically long header line contained very many addresses (the
6831 report of this problem mentioned 10 000) and each of them was rewritten,
6832 Exim could use up a very large amount of memory. (It kept on making new
6833 copies of the header line as it rewrote, and never released the old ones.)
6834 At the expense of a bit more processing, the header rewriting function has
6835 been changed so that it no longer eats memory in this way.
6837 66. The generation of the Received: header has been moved from the time that a
6838 message starts to be received, to the time that it finishes. The timestamp
6839 in the Received: header should now be very close to that of the <= log
6840 line. There are two side-effects of this change:
6842 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6843 logged header lines no longer include the local Received: line, because
6844 it has not yet been created. The same applies to a copy of the message
6845 that is returned to a non-SMTP sender when a message is rejected.
6847 (b) When a filter file is tested using -bf, no additional Received: header
6848 is added to the test message. After some thought, I decided that this
6851 This change does not affect the value of $received_for. It is still set
6852 after address rewriting, but before local_scan() is called.
6854 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6856 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6857 gave an unhelpful panic error message, and a defer error. I have managed to
6858 change this behaviour so that it now rejects any supplied certificate,
6859 which seems right, as the list of acceptable certificates is empty.
6861 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6862 gave an unhelpful defer error. I have not managed to make this reject any
6863 supplied certificates, but the error message it gives is "no certificate
6864 supplied", which is not helpful.
6866 70. exigrep's output now also includes lines that are not associated with any
6867 message, but which match the given pattern. Implemented by a patch from
6868 Martin Sluka, which also tidied up the Perl a bit.
6870 71. Recipient callout verification, like sender verification, was using <> in
6871 the MAIL FROM command. This isn't really the right thing, since the actual
6872 sender may affect whether the remote host accepts the recipient or not. I
6873 have changed it to use the actual sender in the callout; this means that
6874 the cache record is now keyed on a recipient/sender pair, not just the
6875 recipient address. There doesn't seem to be a real danger of callout loops,
6876 since a callout by the remote host to check the sender would use <>.
6877 [SEE ABOVE: changed after hitting problems.]
6879 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6880 temporary errors. However, in the case of such a code being given after
6881 the end of a data transmission (i.e. after ".") Exim was failing to write
6882 a retry record for the message. (Yes, there was some broken host that was
6883 actually sending 8xx at this point.)
6885 73. An unknown lookup type in a host list could cause Exim to panic-die when
6886 the list was checked. (An example that provoked this was putting <; in the
6887 middle of a list instead of at the start.) If this happened during a DATA
6888 ACL check, a -D file could be left lying around. This kind of configuration
6889 error no longer causes Exim to die; instead it causes a defer error. The
6890 incident is still logged to the main and panic logs.
6892 74. Buglet left over from Exim 3 conversion. The message "too many messages
6893 in one connection" was written to the rejectlog but not the mainlog, except
6894 when address rewriting (yes!) was being logged.
6896 75. Added write_rejectlog option.
6898 76. When a system filter was run not as root (that is, when system_filter_user
6899 was set), the values of the $n variables were not being returned to the
6900 main process; thus, they were not subsequently available in the $sn
6903 77. Added +return_path_on_delivery log selector.
6905 78. A connection timeout was being treated differently from recipients deferred
6906 when testing hosts_max_try with a message that was older than the host's
6907 retry timeout. (The host should not be counted, thus allowing all hosts to
6908 be tried at least once before bouncing.) This may have been the cause of an
6909 occasionally reported bug whereby a message would remain on the queue
6910 longer than the retry timeout, but would be bounced if a delivery was
6911 forced. I say "may" because I never totally pinned down the problem;
6912 setting up timeout/retry tests is difficult. See also the next item.
6914 79. The ultimate address timeout was not being applied to errors that involved
6915 a combination of host plus message (for example, a timeout on a MAIL
6916 command). When an address resolved to a number of possible hosts, and they
6917 were not all tried for each delivery (e.g. because of hosts_max_try), a
6918 message could remain on the queue longer than the retry timeout.
6920 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6923 81. Fixed an obscure SMTP outgoing bug which required at least the following
6924 conditions: (a) there was another message waiting for the same server;
6925 (b) the server returned 5xx to all RCPT commands in the first message so
6926 that the message was not completed; (c) the server dropped the connection
6927 or gave a negative response to the RSET that Exim sends to abort the
6928 transaction. The observed case was a dropped connection after DATA that had
6929 been sent in pipelining mode. That is, the server had advertised PIPELINING
6930 but was not implementing it correctly. The effect of the bug was incorrect
6931 behaviour, such as trying another host, and this could lead to a crash.
6937 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6938 and daemon.c were passed as pointers to ints; they should have been
6939 pointers to socklen_t variables (which are typically unsigned ints).
6941 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6944 3. Fixed a really odd bug that affected only the testing scheme; patching a
6945 certain fixed string in the binary changed the value of another string that
6946 happened to be identical to the end of the original first string.
6948 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6949 name", it returns that address as the IP address. On some operating
6950 systems (e.g. Solaris), it also passes back the IP address string as the
6951 "host name". However, on others (e.g. Linux), it passes back an empty
6952 string. Exim wasn't checking for this, and was changing the host name to an
6953 empty string, assuming it had been canonicalized.
6955 5. Although rare, it is permitted to have more than one PTR record for a given
6956 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6957 all the names associated with an address, because they do in Solaris.
6958 However, it seems that they do not in Linux for data that comes from the
6959 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6960 I found this out when I moved to a new Linux workstation and tried to run
6961 the Exim test suite.
6963 To get round this problem I have changed the code so that it now does its
6964 own call to the DNS to look up PTR records when searching for a host name.
6965 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6966 addresses that are only in /etc/hosts are still found.
6968 This behaviour is, however, controlled by an option called host_lookup_
6969 order, which defaults to "bydns:byaddr". If people want to use the other
6970 order, or indeed, just use one or the other means of lookup, they can
6971 specify it in this variable.
6973 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6974 some operating systems, this comes back from gethostbyaddr() as an empty
6975 string, and this is what Exim used to test for. However, it seems that in
6976 other systems, "." is yielded. Exim now tests for this case too.
6978 7. The values of check_spool_space and check_log_space are now held internally
6979 as a number of kilobytes instead of an absolute number of bytes. If a
6980 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6981 kilobyte. This means that much larger values can be stored.
6983 8. Exim monitor: an attempt to get the action menu when not actually pointing
6984 at a message produces an empty menu entitled "No message selected". This
6985 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6986 no entries in it ("Shell widget menu has zero width and/or height"). So I
6987 have added a single, blank menu entry in this case.
6989 9. Added ${quote_local_part.
6991 10. MIME decoding is now applied to the contents of Subject: header lines when
6994 11. Now that a reference to $sender_host_address automatically causes a reverse
6995 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6996 host lookup before query-style lookups in lists that might use this
6997 variable. This has therefore been abolished, and the "net-" prefix is no
6998 longer necessary for query-style lookups.
7000 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7001 have been a typo for LFLAGS, so it has been changed.
7003 13. The install script calls Exim with "-C /dev/null" in order to find the
7004 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7005 to be output. However, since Exim outputs its version number before the
7006 error, it didn't break the script. It just looked ugly. I fixed this by
7007 always allowing "-C /dev/null" if the caller is root.
7009 14. Ignore overlarge ACL variable number when reading spool file - insurance
7010 against a later release with more variables having written the file.
7012 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7013 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7014 this, as well as the form without the "IPv6" on the front.
7016 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7017 OS/Makefile-Darwin file.
7019 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7020 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7022 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7023 variable. However, some OS use ints for inodes. I've added cast to long int
7024 to get rid of the compiler warning.
7026 19. I had forgotten to lock out "/../" in configuration file names when
7027 ALT_CONFIG_PREFIX was set.
7029 20. Routers used for verification do not need to specify transports. However,
7030 if such a router generated a host list, and callout was configured, Exim
7031 crashed, because it could not find a port number from the (non-existent)
7032 transport. It now assumes port 25 in this circumstance.
7034 21. Added the -t option to exigrep.
7036 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7037 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7038 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7039 them is set, LOOKUP_LSEARCH is forced.
7041 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7043 24. Added sender and host information to the "rejected by local_scan()" log
7044 line; previously there was no indication of these.
7046 25. Added .include_if_exists.
7048 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7049 Linux. It turns out that not all file systems support this. Apparently some
7050 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7051 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7052 not supported on the file descriptor, is now ignored when Exim is trying to
7053 sync a directory. This applies only to Linux.
7055 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7057 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7058 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7061 29. Added support for maildirsize files from supplied patch (modified a bit).
7063 30. The use of :fail: followed by an empty string could lead Exim to respond to
7064 sender verification failures with (e.g.):
7066 550 Verification failed for <xxx>
7067 550 Sender verify failed
7069 where the first response line was missing the '-' that indicates it is not
7070 the final line of the response.
7072 31. The loop for finding the name of the user that called Exim had a hardwired
7073 limit of 10; it now uses the value of finduser_retries, which is used for
7074 all other user lookups.
7076 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7079 33. Exim was neglecting to zero errno before one call of strtol() when
7080 expanding a string and expecting an integer value. On some systems this
7081 resulted in spurious "integer overflow" errors. Also, it was casting the
7082 result into an int without checking.
7084 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7085 did not work. The code looks as if it has *never* worked, though it appears
7086 to have been documented since at least release 1.62. I have made it work.
7088 35. The "timeout_DNS" error in retry rules, also documented since at least
7089 1.62, also never worked. As it isn't clear exactly what this means, and
7090 clearly it isn't a major issue, I have abolished the feature by treating it
7091 as "timeout", and writing a warning to the main and panic logs.
7093 36. The display of retry rules for -brt wasn't always showing the error code
7096 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7097 timeout_connect_A, timeout_connect_MX.
7099 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7100 to the empty sender.
7102 39. The daemon was not analysing the content of -oX till after it had closed
7103 stderr and disconnected from the controlling terminal. This meant that any
7104 syntax errors were only noted on the panic log, and the return code from
7105 the command was 0. By re-arranging the code a little, I've made the
7106 decoding happen first, so such errors now appear on stderr, and the return
7107 code is 1. However, the actual setting up of the sockets still happens in
7108 the disconnected process, so errors there are still only recorded on the
7111 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7112 connections (as happens on some IP stacks) was logged at start up time as
7113 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7114 it from "IPv6 and IPv4", which means that two separate sockets are being
7117 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7118 says whether AF_INET or AF_INET6 was passed as an argument.
7120 42. Exiwhat output was messed up when time zones were included in log
7123 43. Exiwhat now gives more information about the daemon's listening ports,
7124 and whether -tls-on-connect was used.
7126 44. The "port" option of the smtp transport is now expanded.
7128 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7129 silently ignored. Now an error message is written to the main and panic
7132 46. There's a new ACL modifier called "logwrite" which writes to a log file
7133 as soon as it is encountered.
7135 47. Added $local_user_uid and $local_user_gid at routing time.
7137 48. Exim crashed when trying to verify a sender address that was being
7140 49. Exim was recognizing only a space character after ".include". It now also
7141 recognizes a tab character.
7143 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7144 extracting the relevant information from the specification. The man page no
7145 longer contains scrambled data for the -d option, and I've added a section
7146 at the front about calling Exim under different names.
7148 51. Added "extra_headers" argument to the "mail" command in filter files.
7150 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7153 53. Installed eximstats 1.29.
7155 54. Added transport_filter_timeout as a generic transport option.
7157 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7158 Cc: header lines. This was required by RFC 822, but it not required by RFC
7161 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7162 incoming messages that did not have them. Now it does so only if the
7163 message originates locally, that is, if there is no associated remote host
7164 address. When Resent- header lines are present, this applies to the Resent-
7165 lines rather than the non-Resent- lines.
7167 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7168 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7170 58. Messages for configuration errors now include the name of the main
7171 configuration file - useful now that there may be more than one file in a
7172 list (.included file names were always shown).
7174 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7175 for those rare installations that do not start the daemon as root or run it
7176 setuid root. I've cut out the call to initgroups() if the daemon is not
7179 60. The Exim user and group can now be bound into the binary as text strings
7180 that are looked up at the start of Exim's processing.
7182 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7184 62. Added $mailstore_basename variable.
7186 63. Installed patch to sieve.c from Michael Haardt.
7188 64. When Exim failed to open the panic log after failing to open the main log,
7189 the original message it was trying to log was written to stderr and debug
7190 output, but if they were not available (the usual case in production), it
7191 was lost. Now it is written to syslog before the two lines that record the
7192 failures to open the logs.
7194 65. Users' Exim filters run in subprocesses under the user's uid. It is
7195 possible for a "deliver" command or an alias in a "personal" command to
7196 provoke an address rewrite. If logging of address rewriting is configured,
7197 this fails because the process is not running as root or exim. There may be
7198 a better way of dealing with this, but for the moment (because 4.30 needs
7199 to be released), I have disabled address rewrite logging when running a
7200 filter in a non-root, non-exim process.
7206 1. The buildconfig auxiliary program wasn't quoting the value set for
7207 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7208 not defined. This bug was masked in 4.22 by the effect that was fixed in
7211 2. Some messages that were rejected after a message id was allocated were
7212 shown as "incomplete" by exigrep. It no longer does this for messages that
7213 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7215 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7216 have allow_domain_literals set, the ID did not get logged in the <= line.
7217 Domain literals are now always recognized in Message-ID: header lines.
7219 4. The first argument for a ${extract expansion item is the key name or field
7220 number. Leading and trailing spaces in this item were not being ignored,
7221 causing some misleading effects.
7223 5. When deliver_drop_privilege was set, single queue runner processes started
7224 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7225 same command in the process it spins off) were not dropping privilege.
7227 6. When the daemon running as "exim" started a queue runner, it always
7228 re-executed Exim in the spun-off process. This is a waste of effort when
7229 deliver_drop_privilege is set. The new process now just calls the
7230 queue-runner function directly.
7236 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7239 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7240 ignored. Though the use of -oP was forcing the writing of a pid file, it
7241 was always written to the default place.
7243 3. If the message "no IP address found for host xxxx" is generated during
7244 incoming verification, it is now followed by identification of the incoming
7245 connection (so you can more easily find what provoked it).
7247 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7249 5. Added some features to "harden" Exim a bit more against certain attacks:
7251 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7252 be put in Local/Makefile. This is like the never_users runtime option,
7253 but it cannot be overridden. The default setting is "root".
7255 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7256 prefix string with which any file named in a -C command line option
7259 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7260 is retained for -C and -D only if the caller of Exim is root. Without
7261 it, the exim user may also use -C and -D and retain privilege.
7263 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7264 command line option is disabled.
7266 6. Macro names set by the -D option must start with an upper case letter, just
7267 like macro names defined in the configuration file.
7269 7. Added "dereference=" facility to LDAP.
7271 8. Two instances of the typo "uknown" in the source files are fixed.
7273 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7274 the Configure-Makefile script screwed up while processing it.
7276 10. Incorporated PCRE 4.4.
7278 11. The SMTP synchronization check was not operating right at the start of an
7279 SMTP session. For example, it could not catch a HELO sent before the client
7280 waited for the greeting. There is now a check for outstanding input at the
7281 point when the greeting is written. Because of the duplex, asynchronous
7282 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7283 way, but not yet received, when the check is performed.
7285 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7286 on TCP/IP sockets, because this apparently causes some broken clients to
7289 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7290 unchanged) from the Cygwin maintainer.
7292 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7293 was supported instead of testing for mbx. Effectively a typo.
7295 15. The spa authenticator server code was not checking that the input it
7296 received was valid base64.
7298 16. The debug output line for the "set" modifier in ACLs was not showing the
7299 name of the variable that was being set.
7301 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7303 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7304 lookup on its own. Something else was needed to trigger the lookup. For
7305 example, a match in host_lookup or the need for a host name in a host list.
7306 Now, if $sender_host_name is referenced and the host name has not yet been
7307 looked up, a lookup is performed. If the lookup fails, the variable remains
7308 empty, and $host_lookup_failed is set to "1".
7310 19. Added "eqi" as a case-independent comparison operator.
7312 20. The saslauthd authentication condition could segfault if neither service
7313 nor realm was specified.
7315 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7316 error message that was logged was misleading, and incoming SMTP
7317 connections were dropped. The message is now more accurate, and temporary
7318 errors are given to SMTP connections.
7320 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7321 (except RSET) with a 421 error, until QUIT is received. However, it was
7322 failing to send a response to QUIT.
7324 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7325 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7326 ACL is now run after the helo_try_verify_hosts code.
7328 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7329 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7330 case-independent, so other case variants are also recognized). Apparently
7331 some systems use these upper case variants.
7333 25. If more than two messages were waiting for the same host, and a transport
7334 filter was specified for the transport, Exim sent two messages over the
7335 same TCP/IP connection, and then failed with "socket operation on non-
7336 socket" when it tried to send the third.
7338 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7340 27. The extern definition of crypt16() in expand.c was not being excluded when
7341 the OS had its own crypt16() function.
7343 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7344 as a preferred synonym for return_size_limit, both as an option and as an
7347 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7349 30. Changed the default configuration ACL to relax the local part checking rule
7350 for addresses that are not in any local domains. For these addresses,
7351 slashes and pipe symbols are allowed within local parts, but the sequence
7352 /../ is explicitly forbidden.
7354 31. SPA server authentication was not clearing the challenge buffer before
7357 32. log_message in a "warn" ACL statement was writing to the reject log as
7358 well as to the main log, which contradicts the documentation and doesn't
7359 seem right (because no rejection is happening). So I have stopped it.
7361 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7362 However, I am unable to do any testing of this.
7364 34. Fixed an infelicity in the appendfile transport. When checking directories
7365 for a mailbox, to see if any needed to be created, it was accidentally
7366 using path names with one or more superfluous leading slashes; tracing
7367 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7369 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7370 contents are added to the log line that is written for every discarded
7371 recipient. (Previously a log_message setting was ignored.)
7373 36. The ${quote: operator now quotes the string if it is empty.
7375 37. The install script runs exim in order to find its version number. If for
7376 some reason other than non-existence or emptiness, which it checks, it
7377 could not run './exim', it was installing it with an empty version number,
7378 i.e. as "exim-". This error state is now caught, and the installation is
7381 38. An argument was missing from the function that creates an error message
7382 when Exim fails to connect to the socket for saslauthd authentication.
7383 This could cause Exim to crash, or give a corrupted message.
7385 39. Added isip, isip4, and isip6 to ${if conditions.
7387 40. The ACL variables $acl_xx are now saved with the message, and can be
7388 accessed later in routers, transports, and filters.
7390 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7391 strings in the file are not string-expanded.
7393 42. If a MAIL command specified a SIZE value that was too large to fit into an
7394 int variable, the check against message_size_limit failed. Such values are
7395 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7396 day this will have to be increased, but I don't think I want to be around
7397 when emails are that large.
7404 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7405 iconv() is not standard in FreeBSD.
7407 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7408 IPv6 enabled. The observed symptom was a segmentation fault on return from
7409 the function os_common_find_running_interfaces() in src/os.c.
7411 3. In the check_special_case() function in daemon.c I had used "errno" as an
7412 argument name, which causes warnings on some systems. This was basically a
7413 typo, since it was named "eno" in the comments!
7415 4. The code that waits for the clock to tick (at a resolution of some fraction
7416 of a second) so as to ensure message-id uniqueness was always waiting for
7417 at least one whole tick, when it could have waited for less. [This is
7418 almost certainly not relevant at current processor speeds, where it is
7419 unlikely to ever wait at all. But we try to future-proof.]
7421 5. The function that sleeps for a time interval that includes fractions of a
7422 second contained a race. It did not block SIGALRM between setting the
7423 timer, and suspending (a couple of lines later). If the interval was short
7424 and the sigsuspend() was delayed until after it had expired, the suspension
7425 never ended. On busy systems this could lead to processes getting stuck for
7428 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7429 process, before it forks any delivery processes. The open lookup caching
7430 mechanism meant that the open file or database connection was passed into
7431 the delivery process. The problem was that delivery processes always tidy
7432 up cached lookup data. This could cause a problem for the next delivery
7433 process started by the queue runner, because the external queue runner
7434 process does not know about the closure. So the next delivery process
7435 still has data in the lookup cache. In the case of a file lookup, there was
7436 no problem because closing a file descriptor in a subprocess doesn't affect
7437 the parent. However, if the lookup was caching a connection to a database,
7438 the connection was closed, and the second delivery process was likely to
7439 see errors such as "PGSQL: query failed: server closed the connection
7440 unexpectedly". The problem has been fixed by closing all cached lookups
7441 in a queue runner before running a delivery process.
7443 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7444 seem to have the "const" qualifier which it has on other OS. I've
7447 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7448 *of the same type* (client or server) with the same public name that an
7449 error should be diagnosed.
7451 9. When Exim looked up a host name for an IP address, but failed to find the
7452 original IP address when looking up the host name (a safety check), it
7453 output the message "<ip address> does not match any IP for NULL", which was
7454 confusing, to say the least. The bug was that the host name should have
7455 appeared instead of "NULL".
7457 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7458 user that is built into the binary, and the -C or -D options is used, root
7459 privilege is dropped before the configuration file is read. In addition,
7460 logging is switched to stderr instead of the normal log files. If the
7461 configuration then re-defines the Exim user, the unprivileged environment
7462 is probably not what is expected, so Exim logs a panic warning message (but
7465 However, if deliver_drop_privilege is set, the unprivileged state may well
7466 be exactly what is intended, so the warning has been cut out in that case,
7467 and Exim is allowed to try to write to its normal log files.
7473 1. smtp_return_error_details was not giving details for temporary sender
7474 or receiver verification errors.
7476 2. Diagnose a configuration error if two authenticators have the same public
7479 3. Exim used not to create the message log file for a message until the first
7480 delivery attempt. This could be confusing when incoming messages were held
7481 for policy or load reasons. The message log file is now created at the time
7482 the message is received, and an initial "Received" line is written to it.
7484 4. The automatically generated man page for command line options had a minor
7485 bug that caused no ill effects; however, a more serious problem was that
7486 the procedure for building the man page automatically didn't always
7487 operate. Consequently, release 4.20 contains an out-of-date version. This
7488 shouldn't happen again.
7490 5. When building Exim with embedded Perl support, the script that builds the
7491 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7492 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7494 6. The freeze_tell option was not being used for messages that were frozen on
7495 arrival, either by an ACL or by local_scan().
7497 7. Added the smtp_incomplete_transaction log selector.
7499 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7500 was accepting AUTH without a new EHLO.
7502 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7503 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7506 10. Logging of TCP/IP connections (when configured) now happens in the main
7507 daemon process instead of the child process, so that the TCP/IP connection
7508 count is more accurate (but it can never be perfect).
7510 11. The use of "drop" in a nested ACL was not being handled correctly in the
7511 outer ACL. Now, if condition failure induced by the nested "drop" causes
7512 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7513 or "require"), the connection is dropped.
7515 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7516 that yield "discard" can now be used with an "accept" or a "discard" verb,
7517 but an error is generated for any others (because I can't see a useful way
7518 to define what should happen).
7520 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7521 are now processed in the same way as lines in the Exim configuration file.
7522 In particular, continuation lines are supported.
7524 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7526 15. Added -ti meaning -t -i.
7528 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7529 domains, and warn by logging if others are found.
7531 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7532 returned from GIFCONF when Exim is trying to find the list of interfaces on
7533 a host. The code in os.c has been modified to copy each ifreq to an aligned
7534 structure in all cases.
7536 Also, in some cases, the returned ifreq's were being copied to a 'struct
7537 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7538 means the last couple of bytes of an IPv6 address could be chopped if the
7539 ifreq contained only a normal sockaddr (14 bytes storage).
7541 18. Named domain lists were not supported in the hosts_treat_as_local option.
7542 An entry such as +xxxx was not recognized, and was treated as a literal
7545 19. Ensure that header lines added by a DATA ACL are included in the reject log
7546 if the ACL subsequently rejects the message.
7548 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7549 MD5 (which is deprecated).
7551 21. When testing a filter file using -bf, Exim was writing a message when it
7552 took the sender from a "From " line in the message, but it was not doing so
7553 when it took $return_path from a Return-Path: header line. It now does.
7555 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7556 with a valid header line field name (a series of printing characters
7557 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7559 23. Changed "disc" in the source to "disk" to conform to the documentation and
7560 the book and for uniformity.
7562 24. Ignore Sendmail's -Ooption=value command line item.
7564 25. When execve() failed while trying to run a command in a pipe transport,
7565 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7566 could be confused with a return value of 69 from the command itself. This
7567 has been changed to 127, the value the shell returns if it is asked to run
7568 a non-existent command. The wording for the related log line suggests a
7569 non-existent command as the problem.
7571 26. If received_header_text expands to an empty string, do not add a Received:
7572 header line to the message. (Well, it adds a token one on the spool, but
7573 marks it "old" so that it doesn't get used or transmitted.)
7575 27. Installed eximstats 1.28 (addition of -nt option).
7577 28. There was no check for failure on the call to getsockname() in the daemon
7578 code. This can fail if there is a shortage of resources on the system, with
7579 ENOMEM, for example. A temporary error is now given on failure.
7581 29. Contrary to the C standard, it seems that in some environments, the
7582 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7583 program. Exim now does this explicitly; it affects the formatting of
7584 timestamps using strftime().
7586 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7587 complaints. I've now initialized all the variables, to avoid this.
7589 32. Header lines added by a system filter were not being "seen" during
7590 transport-time rewrites.
7592 33. The info_callback() function passed to OpenSSL is set up with type void
7593 (*)(SSL *, int, int), as described somewhere. However, when calling the
7594 function (actually a macro) that sets it up, the type void(*)() is
7595 expected. I've put in a cast to prevent warnings from picky compilers.
7597 34. If a DNS black list lookup found a CNAME record, but there were no A
7598 records associated with the domain it pointed at, Exim crashed.
7600 35. If a DNS black list lookup returned more than one A record, Exim ignored
7601 all but the first. It now scans all returned addresses if a particular IP
7602 value is being sought. In this situation, the contents of the
7603 $dnslist_value variable are a list of all the addresses, separated by a
7606 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7607 to accept a host name and all its aliases if the forward lookup for any of
7608 them yielded the IP address of the incoming connection. Now it accepts only
7609 those names whose forward lookup yields the correct IP address. Any other
7610 names are discarded. This closes a loophole whereby a rogue DNS
7611 administrator could create reverse DNS records to break through a
7612 wildcarded host restriction in an ACL.
7614 37. If a user filter or a system filter that ran in a subprocess used any of
7615 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7616 the wrong values were passed to the pipe command ($thisaddress had the
7617 value of $0, $0 had the value of $1, etc). This bug was introduced by
7618 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7620 38. Improved the line breaking for long SMTP error messages from ACLs.
7621 Previously, if there was no break point between 40 and 75 characters, Exim
7622 left the rest of the message alone. Two changes have been made: (a) I've
7623 reduced the minimum length to 35 characters; (b) if it can't find a break
7624 point between 35 and 75 characters, it looks ahead and uses the first one
7625 that it finds. This may give the occasional overlong line, but at least the
7626 remaining text gets split now.
7628 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7629 file descriptors might be low, and that setting 1000 would always raise it.
7630 It turns out that in some environments, the limit is already over 1000 and
7631 that lowering it causes trouble. So now Exim takes care not to decrease it.
7633 40. When delivering a message, the value of $return_path is set to $sender_
7634 address at the start of routing (routers may change the value). By an
7635 oversight, this default was not being set up when an address was tested by
7636 -bt or -bv, which affected the outcome if any router or filter referred to
7639 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7640 log only when "message" or "log_message" are set. However, if one of the
7641 conditions was an address verification, or a call to a nested ACL, the
7642 messages generated by the underlying test were being passed through. This
7643 no longer happens. The underlying message is available in $acl_verify_
7644 message for both "message" and "log_message" expansions, so it can be
7645 passed through if needed.
7647 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7648 new expansion $bh_ to give the encoded byte string without charset
7649 translation. Translation happens only if iconv() is available; HAVE_ICONV
7650 indicates this at build time. HEADERS_CHARSET gives the charset to
7651 translate to; headers_charset can change it in the configuration, and
7652 "headers charset" can change it in an individual filter file.
7654 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7655 that creates RFC 2047 encoded "words" labels them as that charset instead
7656 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7657 expansion operator; (ii) when Exim creates a From: line for a local
7658 message; (iii) when a header line is rewritten to include a "phrase" part.
7660 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7661 buggy, causing it to skip the first lines of messages whose message ID
7662 ended in 'D'. This would not have bitten before Exim release 4.14, because
7663 message IDs were unlikely to end in 'D' before then. The effect was to have
7664 incorrect size information for certain domains.
7666 45. #include "config.h" was missing at the start of the crypt16.c module. This
7667 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7670 46. If there was a timeout during a "random" callout check, Exim treated it as
7671 a failure of the random address, and carried on sending RSET and the real
7672 address. If the delay was just some slowness somewhere, the response to the
7673 original RCPT would be taken as a response to RSET and so on, causing
7674 mayhem of various kinds.
7676 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7677 when I implemented it. It didn't allow for the fact that some option values
7678 may legitimately be negative (e.g. size_addition), and it didn't even do
7679 the right test for positive values.
7681 48. Domain names in DNS records are case-independent. Exim always looks them up
7682 in lower case. Some resolvers return domain names in exactly the case they
7683 appear in the zone file, that is, they may contain uppercase letters. Not
7684 all resolvers do this - some return always lower case. Exim was treating a
7685 change of case by a resolver as a change of domain, similar to a widening
7686 of a domain abbreviation. This triggered its re-routing code and so it was
7687 trying to route what was effectively the same domain again. This normally
7688 caused routing to fail (because the router wouldn't handle the domain
7689 twice). Now Exim checks for this case specially, and just changes the
7690 casing of the domain that it ultimately uses when it transmits the message
7693 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7696 50. If a filter generated a file delivery with a non-absolute name (possible if
7697 no home directory exists for the router), the forbid_file option was not
7700 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7701 the existing equality matching.
7703 52. Exim was using ints instead of ino_t variables in some places where it was
7704 dealing with inode numbers.
7706 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7707 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7708 called TMPDIR, and if it finds it is different, it changes its value.
7710 54. The smtp_printf() function is now made available to local_scan() so
7711 additional output lines can be written before returning. There is also an
7712 smtp_fflush() function to enable the detection of a dropped connection.
7713 The variables smtp_input and smtp_batched_input are exported to
7716 55. Changed the default runtime configuration: the message "Unknown user"
7717 has been removed from the ACL, and instead placed on the localuser router,
7718 using the cannot_route_message feature. This means that any verification
7719 failures that generate their own messages won't get overridden. Similarly,
7720 the "Unrouteable address" message that was in the ACL for unverifiable
7721 relay addresses has also been removed.
7723 56. Added hosts_avoid_esmtp to the smtp transport.
7725 57. The exicyclog script was not checking for the esoteric option
7726 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7727 will work only if exicyclog is run under the appropriate euid.
7729 58. Following a discussion on the list, the rules by which Exim recognises line
7730 endings on incoming messages have been changed. The -dropcr and drop_cr
7731 options are now no-ops, retained only for backwards compatibility. The
7732 following line terminators are recognized: LF CRLF CR. However, special
7733 processing applies to CR:
7735 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7736 nor a local message in the state where . is a terminator.
7738 (ii) If a bare CR is encountered in a header line, an extra space is added
7739 after the line terminator so as not to end the header. The reasoning
7740 behind this is that bare CRs in header lines are most likely either
7741 to be mistakes, or people trying to play silly games.
7743 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7744 was being incorrectly given as 18 bytes larger than it should have been.
7745 This is a VOB (very old bug).
7747 60. This may never have affected anything current, but just in case it has:
7748 When the local host is found other than at the start of a list of hosts,
7749 the local host, those with the same MX, and any that follow, are discarded.
7750 When the list in question was part of a longer list of hosts, the following
7751 hosts (not currently being processed) were also being discarded. This no
7752 longer happens. I'm not sure if this situation could ever has previously
7755 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7758 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7759 header. This is something that is recommended in a new Internet Draft, and
7760 is something that is documented as being done by Sendmail. There are two
7761 possible values. For messages generated by the autoreply transport, Exim
7764 Auto-Submitted: auto-replied
7766 whereas for all other generated messages (e.g. bounces) it adds
7768 Auto-Submitted: auto-generated
7770 63. The "personal" condition in filters now includes a test for the
7771 Auto-Submitted: header. If it contains the string "auto-" the message it
7772 not considered personal.
7774 64. Added rcpt_include_affixes as a generic transport option.
7776 65. Added queue_only_override (default true).
7778 66. Added the syslog_duplication option.
7780 67. If what should have been the first header line of a message consisted of
7781 a space followed by a colon, Exim was mis-interpreting it as a header line.
7782 It isn't of course - it is syntactically invalid and should therefore be
7783 treated as the start of the message body. The misbehaviour could have
7784 caused a number of strange effects, including loss of data in subsequent
7785 header lines, and spool format errors.
7787 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7788 client host had authenticated. This control can now be exercised by an ACL
7789 for more flexibility.
7791 69. By default, callouts do not happen when testing with -bh. There is now a
7792 variant, -bhc, which does actually run the callout code, including
7793 consulting and updating the callout cache.
7795 70. Added support for saslauthd authentication, courtesy of Alexander
7798 71. If statvfs() failed on the spool or log directories while checking their
7799 size for availability, Exim confusingly gave the error "space shortage".
7800 Furthermore, in debugging mode it crashed with a floating point exception.
7801 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7802 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7803 serious problem, Exim now writes to the main and panic logs when this
7804 happens, with details of the failure. It then refuses to accept the
7805 incoming message, giving the message "spool directory problem" or "log
7806 directory problem" with a 421 code for SMTP messages.
7808 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7809 0, 1, and 2 exist, because some OS complain for execs without them (see
7810 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7811 descriptors. However, the code omitted to check that the open succeeded,
7812 causing mysterious errors if for some reason the permissions on /dev/null
7813 got screwed. Now Exim writes a message to the main and panic logs, and
7814 bombs out if it can't open /dev/null.
7816 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7817 interact so that it is all more flexible. It is supposed to remain
7818 backwards compatible. Also added extra_local_interfaces.
7820 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7821 to bomb out with an assertion failure - to the client this appears as a
7822 connection drop. This problem occurs in the part of the code that was taken
7823 from the Samba project. Fortunately, the assertion is in a very simple
7824 function, so I have fixed this by reproducing the function inline in the
7825 one place where it is called, and arranging for authentication to fail
7826 instead of killing the process with assert().
7828 75. The SPA client code was not working when the server requested OEM rather
7829 than Unicode encoding.
7831 76. Added code to make require_files with a specific uid setting more usable in
7832 the case where statting the file as root fails - usually a non-root-mounted
7833 NFS file system. When this happens and the failure is EACCES, Exim now
7834 forks a subprocess and does the per-uid checking as the relevant uid.
7836 77. Added process_log_path.
7838 78. If log_file_path was not explicitly set, a setting of check_log_space or
7839 check_log_inodes was ignored.
7841 79. If a space check for the spool or log partitions fails, the incident is now
7842 logged. Of course, in the latter case the data may get lost...
7844 80. Added the %p formatting code to string_format() so that it can be used to
7845 print addresses in debug_print(). Adjusted all the address printing in the
7846 debugging in store.c to use %p rather than %d.
7848 81. There was a concern that a line of code in smtp_in.c could overflow a
7849 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7850 initially expressed, the concern was not well-founded, because trailing
7851 spaces are removed early. However, if the trailing spaces were followed by
7852 a NULL, they did not get removed, so the overflow was possible. Two fixes
7855 (a) I re-wrote the offending code in a cleaner fashion.
7856 (b) If an incoming SMTP command contains a NULL character, it is rejected
7859 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7860 runs initgroups(), so that if the Exim user is in any additional groups,
7861 they will be used during message reception.
7867 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob