1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Fix "spam" ACL condition. Previously, tainted values for the "name"
59 argument resulted in a trap. There is no reason to disallow such; this
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
74 JH/01 Avoid costly startup code when not strictly needed. This reduces time
75 for some exim process initialisations. It does mean that the logging
76 of TLS configuration problems is only done for the daemon startup.
78 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
80 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
81 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
84 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
85 DKIM signing is being done. Previously a transport_filter would always
86 disable CHUNKING, falling back to traditional DATA.
88 JH/05 Regard command-line receipients as tainted.
90 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
92 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
93 PAM library frees one of the arguments given to it, despite the
94 documentation. Therefore a plain malloc must be used.
96 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
97 on-stack buffers were used, resulting in a taint trap when DSN information
98 copied from a received message was written into the buffer.
100 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
101 the ordering of its ARC headers. This caused a crash.
103 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
104 a new record was being constructed with information from the peer, a trap
107 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
108 installation would get error messages from DMARC verify, when it hit the
109 nonexistent file indicated by the default. Distros wanting DMARC enabled
110 should both provide the file and set the option.
111 Also enforce no DMARC verification for command-line sourced messages.
113 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
114 could, depending on the platform, hang at the STARTTLS response.
116 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
117 message on a connection. Previously if one message had ARC headers and
118 the following one did not, a crash could result when adding an
119 Authentication-Results: header.
121 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
122 Exim main code and Exim-related utities. The introduction of taint
123 tracking also did many adjustments to string handling. Since then, eximon
124 frequently terminated with an assert failure.
126 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
127 check for 452 responses. This slightly helps the inefficieny of doing
128 a large alias-expansion into a recipient-limited target. The max_rcpt
129 transport option still applies (and at the current default, will override
130 the new feature). The check is done for either cause of synch, and forces
131 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
132 connection. The new facility is not tunable at this time.
134 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
135 library live data was being used, so the results became garbage. Make
136 copies while it is still usable.
138 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
139 on delivery deferred (==) and failed (**) lines (if a delivery was
140 attemtped). Previously it was only on completion (=>) lines.
142 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
143 for the expansion of the server_scram_iter and server_scram_salt options.
145 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
146 are now specifically given a NO_DATA response without hitting the system
147 resolver. The library goes on to do the now-standard TXT lookup.
148 Use of dnsdb lookups is not affected.
150 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
151 only retrieve the errormessage once. Previously two calls to dlerror()
152 were used, and the second one (for mainlog/paniclog) retrieved null
155 JH/20 Taint checking: disallow use of tainted data for
156 - the appendfile transport file and directory options
157 - the pipe transport command
158 - the autoreply transport file, log and once options
159 - file names used by the redirect router (including filter files)
161 - paths used by single-key lookups
162 Previously this was permitted.
164 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
165 adjusted the size of a major service buffer; this failed because the
166 buffer was in use at the time. Change to a compile-time increase in the
167 buffer size, when this authenticator is compiled into exim.
169 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
170 previous fast-mode was untenable in the face of glibs using mmap to
171 support larger malloc requests.
173 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
174 New values supported, if defined on system where compiled:
175 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
176 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
178 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
179 running a limited number of proceses in parallel, a benefit is gained. The
180 amount varies with the platform hardware and load. The use of the option
181 queue_run_in_order means we cannot do this, as ordering becomes
184 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
185 had introduced a string-copy (for ensuring NUL-termination) which was not
186 appropriate for that case, which can include embedded NUL bytes in the
187 block of data. Investigation showed the copy to actually be needless, the
188 data being length-specified.
190 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
191 done during a receiving connection, and both used TLS, global info was
192 used rather than per-connection info for tracking the state of data
193 queued for transmission. This could result in a connection hang.
195 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
196 Previously, when delivering serveral messages down a single connection
197 only the first would provide a SIZE. This was due to the size information
198 not being properly tracked.
200 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
201 TAI (at 37 seconds currently), pretend to be in UTC for time-related
202 expansion and logging. Previously, spurious values such as a future
203 minute could be seen.
205 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
206 it could crash from a null-deref. This could also affect the
207 ${addresses: } operator and ${readsock } item.
209 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
210 message following a mime one, the variable was not reset.
212 JH/30 When an pipelined-connect fails at the first response, assume incorrect
213 cached capability (perhaps the peer reneged?) and immediately retry in
216 JH/31 Fix spurious detection of timeout while writing to transport filter.
218 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
219 an attempt to copy the string was made before checking it.
221 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
222 taint of the lookup key was maintained; we now regard the presence in the
223 filesystem as sufficient validation.
225 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
226 argument is supplied.
228 JH/35 The dsearch lookup now requires that the directory is an absolute path.
229 Previously this was not checked, and nonempty relative paths made an
230 access under Exim's current working directory.
232 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
233 Previously no event was raised.
235 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
236 parameter supplied by the sender MAIL FROM command. Previously it was
237 ignored, and only the check_spool_space option value for the required
240 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
241 the size of the signing public-key. Previously it was instead giving
242 the size of the signature hash.
244 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
245 the default. See the (new) dkim_verify_min_keysizes option.
247 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
248 and an ACL use a lookup for checking either the local_part or domain,
249 stale data could be accessed. Ensure that variable references are
250 dropped between messages.
252 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
253 by the client was not checked as pointing within response data before
254 being used. A malicious client could thus cause an out-of-bounds read and
255 possibly gain authentication. Fix by adding the check.
257 JH/42 Internationalisation: change the default for downconversion in the smtp
258 transport to be "if needed". Previously it was "as previously set" for
259 the message, which usually meant "if needed" for message-submission but
260 "no" for everything else. However, MTAs have been seen using SMTPUTF8
261 even when the envelope addresses did not need it, resulting in forwarding
262 failures to non-supporting MTAs. A downconvert in such cases will be
263 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
264 transport. The change does mean that addresses needing conversion will
265 be converted when previously a delivery failure would occur.
267 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
268 response was received it would be used unchecked in a fail-DSN, violating
269 standards on line-length limits. Truncate if needed.
271 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
272 doesn't work. (Additionally add a new main config option to configure the
279 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
280 side, to decode a TLS 1.3 packet capture.
282 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
283 Previously the default library behaviour applied, sending two, each in
286 JH/03 Debug output for ACL now gives the config file name and line number for
289 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
291 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
293 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
294 buffer overrun for (non-chunking) other transports.
296 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
297 TLS1.3, means that a server rejecting a client certificate is not visible
298 to the client until the first read of encrypted data (typically the
299 response to EHLO). Add detection for that case and treat it as a failed
300 TLS connection attempt, so that the normal retry-in-clear can work (if
301 suitably configured).
303 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
304 and/or domain. Found and fixed by Jason Betts.
306 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
307 configuration). If a CNAME target was not a wellformed name pattern, a
310 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
311 the OS reports them interleaved with other addresses.
313 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
314 used both for input and for a verify callout, both encrypted, SMTP
315 responses being sent by the server could be lost. This resulted in
316 dropped connections and sometimes bounces generated by a peer sending
319 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
320 string. Previously it was possible to cause undefined behaviour in a
321 library routine (usually a crash). Found by "zerons".
323 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
326 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
327 API was removed, so update to use the newer ones.
329 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
330 any timeout set, is taking a long time. Previously we would hang on to a
331 rotated logfile "forever" if the input was arriving with long gaps
332 (a previous attempt to fix addressed lack, for a long time, of initial
335 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
336 shared (NFS) environment. The length of the tempfile name is now
337 4 + 16 ("hdr.$message_exim_id") which might break on file
338 systems which restrict the file name length to lower values.
341 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
342 shared (NFS) environment.
344 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
345 did for all versions <4.90). Notably -M, -m, --invert, -I may be
348 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
349 on some platforms for bit 31.
351 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
352 to changes apparently associated with TLS1.3 handling some of the APIs
353 previously used were either nonfunctional or inappropriate. Strings
354 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
355 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
356 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
357 This affects log line X= elements, the $tls_{in,out}_cipher variables,
358 and the use of specific cipher names in the encrypted= ACL condition.
360 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
362 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
363 verification result was not updated unless hosts_require_ocsp applied.
365 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
366 queue_list_requires_admin set to false, non-admin users were denied the
369 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
370 directory-of-certs mode. Previously they were advertised despite the
373 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
374 A single TCP connection by a client will now hold a TLS connection open
375 for multiple message deliveries, by default. Previously the default was to
378 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
379 default. If built with the facility, DANE will be used. The facility
380 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
382 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
383 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
384 must be defined and you must still, unless you define DISABLE_TLS, manage
385 the the include-dir and library-file requirements that go with that
386 choice. Non-TLS builds are still supported.
388 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
391 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
392 default. If the platform supports and has the facility enabled, it will
393 be requested on all coneections.
395 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
396 controlled by the build-time option SUPPORT_PIPE_CONNECT.
398 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
400 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
401 success-DSN messages. Previously the From: header was always the default
402 one for these; the option was ignored.
404 JH/28 Fix the timeout on smtp response to apply to the whole response.
405 Previously it was reset for every read, so a teergrubing peer sending
406 single bytes within the time limit could extend the connection for a
407 long time. Credit to Qualsys Security Advisory Team for the discovery.
409 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
410 delivery address, which leaked information of the results of local
411 forwarding. Change to the original envelope recipient address, per
414 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
415 requested. Previously not bounce was generated and a log entry of
416 error ignored was made.
418 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
420 JH/32 Introduce a general tainting mechanism for values read from the input
421 channel, and values derived from them. Refuse to expand any tainted
422 values, to catch one form of exploit.
424 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
425 was unused and the unexpanded text used for the test. Found and
426 fixed by Ruben Jenster.
428 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
429 an attempt to use a TLS library read routine dereffed a nul pointer,
432 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
433 them in our smtp response.
435 JH/36 Have the general router option retry_use_local_part default to true when
436 any of the restrictive preconditions are set (to anything). Previously it
437 was only for check_local user. The change removes one item of manual
438 configuration which is required for proper retries when a remote router
439 handles a subset of addresses for a domain.
441 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
442 link count into consideration.
444 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
445 caused the extension of big_buffer, the following lines were ignored.
447 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
448 accordance with RFC 2308. Previously there was no expiry, so a longlived
449 receive process (eg. due to ACL delays) versus a short SOA value could
452 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
454 JH/39 Promote DMARC support to mainline.
456 JH/40 Bug 2452: Add a References: header to DSNs.
458 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
459 parameters. The relevant library call is documented as "Deprecated: This
460 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
461 3.6.0, DH parameters are negotiated following RFC7919."
463 HS/06 Change the default of dnssec_request_domains to "*"
465 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
466 carried on and emitted a BDAT command, even when PIPELINING was not
469 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
470 buffer was used for the filename, resulting in a trap when tainted
471 arguments (eg. $domain) were used.
473 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
474 recommended to avoid a possible server-load attack. The feature can be
475 re-enabled via the openssl_options main cofiguration option.
477 JH/45 local_scan API: documented the current smtp_printf() call. This changed
478 for version 4.90 - adding a "more data" boolean to the arguments.
479 Bumped the ABI version number also, this having been missed previously;
480 release versions 4.90 to 4.92.3 inclusive were effectively broken in
481 respect of usage of smtp_printf() by either local_scan code or libraries
482 accessed via the ${dlfunc } expansion item. Both will need coding
483 adjustment for any calls to smtp_printf() to match the new function
484 signature; a FALSE value for the new argument is always safe.
486 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
487 the file-offset (which the Linux syscall does, and exim expects); this
488 resulted in an indefinite loop.
490 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
491 to do ARC verification. The Authentication-Results: header line added
492 by the configuration then had no ARC item.
498 JH/01 Remove code calling the customisable local_scan function, unless a new
499 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
501 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
502 non-signal-safe functions being used.
504 JH/03 Bug 2269: When presented with a received message having a stupidly large
505 number of DKIM-Signature headers, disable DKIM verification to avoid
506 a resource-consumption attack. The limit is set at twenty.
508 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
509 report of oldest_pass in ${authres } in consequence, and separate out
510 some descriptions of reasons for verification fail.
512 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
513 files in the spool were present and unlocked. A queue-runner could spot
514 them, resulting in a duplicate delivery. Fix that by doing the unlock
515 after the unlink. Investigation by Tim Stewart. Take the opportunity to
516 add more error-checking on spoolfile handling while that code is being
519 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
520 No known attacks, no CVE, this is defensive hardening.
522 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
523 a queue-runner could start a delivery while other operations were ongoing.
524 Cutthrough delivery was a common victim, resulting in duplicate delivery.
525 Found and investigated by Tim Stewart. Fix by using the open message data
526 file handle rather than opening another, and not locally closing it (which
527 releases a lock) for that case, while creating the temporary .eml format
528 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
530 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
531 $sender_verify_failure/$recipient_verify_failure to "random".
533 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
536 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
537 Previously this would segfault.
539 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
542 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
543 like zero, since the resolver should be doing this for us, But we need one
544 as a CNAME but no MX presence gets the CNAME returned; we need to check
545 that doesn't point to an MX to declare it "no MX returned" rather than
546 "error, loop". A new main option is added so the older capability of
547 following some limited number of chain links is maintained.
549 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
551 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
552 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
553 now-deprecated earlier definitions used only specified the range up to TLS
554 1.2 (in the older-version library docs).
556 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
558 JH/15 Rework TLS client-side context management. Stop using a global, and
559 explicitly pass a context around. This enables future use of TLS for
560 connections to service-daemons (eg. malware scanning) while a client smtp
561 connection is using TLS; with cutthrough connections this is quite likely.
563 JH/16 Fix ARC verification to do AS checks in reverse order.
565 JH/17 Support a "tls" option on the ${readsocket } expansion item.
567 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
568 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
569 Previously the "utf8" would be re-prepended for every additional message.
571 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
572 Previously thery were accepted, resulting in issues when attempting to
573 forward messages to a non-supporting MTA.
575 PP/02 Let -n work with printing macros too, not just options.
577 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
578 one parent address was copied, and bogus data was used at delivery-logging
579 time. Either a crash (after delivery) or bogus log data could result.
580 Discovery and analysis by Tim Stewart.
582 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
583 Previously if the string ended mid-character, we did not insert the
584 promised '?' replacement.
586 PP/04 Documentation: current string operators work on bytes, not codepoints.
588 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
589 should pack well giving a smaller memory footprint so better caching and
590 therefore performance. Group the declarations where this can't be done so
591 that the byte-sized flag variables are not interspersed among pointer
592 variables, giving a better chance of good packing by the compiler.
594 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
595 non-null, to avoid issues with sites running BATV. Previously reports were
596 sent with an empty envelope sender so looked like bounces.
598 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
599 The ignore_error flag wasn't being returned from the filter subprocess so
600 was not set for later routers. Investigation and fix by Matthias Kurz.
602 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
603 and a msg:complete for the whole, when a message is manually removed using
604 -Mrm. Developement by Matthias Kurz, hacked on by JH.
606 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
607 a "Gnu special" function, asprintf() in the DB utility binary builds; I
608 hope that is portable enough.
610 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
611 requiring a known-CA anchor certificate; make it now rely entirely on the
612 TLSA as an anchor. Checking the name on the leaf cert against the name
613 on the A-record for the host is still done for TA (but not for EE mode).
615 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
616 included in delivery lines for non-proxied connections, when compiled with
617 SUPPORT_SOCKS and running with proxy logging enabled.
619 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
620 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
621 move the existing event to fire before the normal logging of message
622 failure so that custom logging is bracketed by normal logging.
624 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
625 msg:fail:internal event. Developement by Matthias Kurz.
627 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
628 far too small for todays use of crypto signatures stored there. Go all
629 the way to the max DNS message size of 64kB, even though this might be
630 overmuch for IOT constrained device use.
632 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
633 copy a string over itself. The library routine is documented as not
634 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
636 JH/32 For main options check_spool_space and check_inode_space, where the
637 platform supports 64b integers, support more than the previous 2^31 kB
638 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
639 the previous G, M, k.
641 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
642 $authenticated_fail_id variable on authentication failure. Previously
645 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
646 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
647 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
648 GNUTLS_SEC_PARAM_MEDIUM.
650 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
651 side. Previously we would continue as if no SNI had been received.
653 JH/36 Harden the handling of string-lists. When a list consisted of a sole
654 "<" character, which should be a list-separator specification, we walked
655 off past the nul-terimation.
657 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
658 causes) even when the retry time is not yet met. Previously they were
659 not, meaning that when (say) an account was over-quota and temp-rejecting,
660 and multiple senders' messages were queued, only one sender would get
661 notified on each configured delay_warning cycle.
663 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
665 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
666 cipher-suites, an error can be left on the stack even for a succeeding
667 accept; this results in impossible error messages when a later operation
670 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
671 return error codes indicating retry. Under TLS1.3 this becomes required.
673 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
674 it only wrote the new authenticators, resulting in a lack of tracking of
675 peer changes of ESMTP extensions until the next cache flush.
677 JH/41 Fix the loop reading a message header line to check for integer overflow,
678 and more-often against header_maxsize. Previously a crafted message could
679 induce a crash of the recive process; now the message is cleanly rejected.
681 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
682 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
688 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
689 When redis_servers is set to a list of > 1 element, and the Redis servers
690 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
691 case of MOVED into a DEFER case instead, thus moving the query onto the
692 next server in the list. For a cluster of N elements, all N servers must
693 be defined in redis_servers.
695 GF/02 Catch and remove uninitialized value warning in exiqsumm
696 Check for existence of @ARGV before looking at $ARGV[0]
698 JH/01 Replace the store_release() internal interface with store_newblock(),
699 which internalises the check required to safely use the old one, plus
700 the allocate and data copy operations duplicated in both (!) of the
701 extant use locations.
703 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
704 modifier. This matches the restriction on the commandline.
706 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
707 Previously only the last row was returned.
709 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
710 we assumed that tags in the header were well-formed, and parsed the
711 element content after inspecting only the first char of the tag.
712 Assumptions at that stage could crash the receive process on malformed
715 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
716 While running the DKIM ACL we operate on the Permanent memory pool so that
717 variables created with "set" persist to the DATA ACL. Also (at any time)
718 DNS lookups that fail create cache records using the Permanent pool. But
719 expansions release any allocations made on the current pool - so a dnsdb
720 lookup expansion done in the DKIM ACL releases the memory used for the
721 DNS negative-cache, and bad things result. Solution is to switch to the
722 Main pool for expansions.
723 While we're in that code, add checks on the DNS cache during store_reset,
724 active in the testsuite.
725 Problem spotted, and debugging aided, by Wolfgang Breyha.
727 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
728 When none of the hosts presented to a transport match an already-open
729 connection, close it and proceed with the list. Previously we would
730 queue the message. Spotted by Lena with Yahoo, probably involving
733 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
734 Previously a spurious "250 OK id=" response was appended to the proper
737 JH/08 The "support for" informational output now, which built with Content
738 Scanning support, has a line for the malware scanner interfaces compiled
739 in. Interface can be individually included or not at build time.
741 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
742 by the template makefile "src/EDITME". The "STREAM" support for an older
743 ClamAV interface method is removed.
745 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
746 rows affected is given instead).
748 JH/11 The runtime Berkeley DB library version is now additionally output by
749 "exim -d -bV". Previously only the compile-time version was shown.
751 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
752 SMTP connection. Previously, when one had more receipients than the
753 first, an abortive onward connection was made. Move to full support for
754 multiple onward connections in sequence, handling cutthrough connection
755 for all multi-message initiating connections.
757 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
758 routers. Previously, a multi-recipient message would fail to match the
759 onward-connection opened for the first recipient, and cause its closure.
761 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
762 a timeout on read on a GnuTLS initiating connection, resulting in the
763 initiating connection being dropped. This mattered most when the callout
764 was marked defer_ok. Fix to keep the two timeout-detection methods
767 JH/15 Relax results from ACL control request to enable cutthrough, in
768 unsupported situations, from error to silently (except under debug)
769 ignoring. This covers use with PRDR, frozen messages, queue-only and
772 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
774 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
775 metadata, resulting in a crash in free().
777 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
778 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
779 Broken also in d185889f4, with init system revamp.
781 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
782 Previously we abruptly closed the connection after reading a malware-
783 found indication; now we go on to read the "scan ok" response line,
786 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
787 ACL. Previously, a crash would result.
789 JH/19 Speed up macro lookups during configuration file read, by skipping non-
790 macro text after a replacement (previously it was only once per line) and
791 by skipping builtin macros when searching for an uppercase lead character.
793 JH/20 DANE support moved from Experimental to mainline. The Makefile control
794 for the build is renamed.
796 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
797 was allocated for every new TLS startup, meaning one per message. Fix
798 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
800 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
801 reported the original. Fix to report (as far as possible) the ACL
802 result replacing the original.
804 JH/23 Fix memory leak during multi-message connections using STARTTLS under
805 OpenSSL. Certificate information is loaded for every new TLS startup,
806 and the resources needed to be freed.
808 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
810 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
813 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
814 DATA response info to the (existing) per-recipient response info for
815 the "C=" log element. It can have useful tracking info from the
816 destination system. Patch from Simon Arlott.
818 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
819 length value. Previously this would segfault.
821 HS/02 Support Avast multiline protoocol, this allows passing flags to
822 newer versions of the scanner.
824 JH/28 Ensure that variables possibly set during message acceptance are marked
825 dead before release of memory in the daemon loop. This stops complaints
826 about them when the debug_store option is enabled. Discovered specifically
827 for sender_rate_period, but applies to a whole set of variables.
828 Do the same for the queue-runner and queue-list loops, for variables set
829 from spool message files. Do the same for the SMTP per-message loop, for
830 certain variables indirectly set in ACL operations.
832 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
833 as a multi-recipient message from a mailinglist manager). The coding had
834 an arbitrary cutoff number of characters while checking for more input;
835 enforced by writing a NUL into the buffer. This corrupted long / fast
836 input. The problem was exposed more widely when more pipelineing of SMTP
837 responses was introduced, and one Exim system was feeding another.
838 The symptom is log complaints of SMTP syntax error (NUL chars) on the
839 receiving system, and refused recipients seen by the sending system
840 (propating to people being dropped from mailing lists).
841 Discovered and pinpointed by David Carter.
843 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
844 replaced by the ${authresults } expansion.
846 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
848 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
849 allows proper process termination in container environments.
851 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
852 Previously the "final dot" had a newline after it; ensure it is CR,LF.
854 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
855 and "err_perm", deprecated since 4.83 when the RFC-defined words
856 "temperror" and "permerror" were introduced.
858 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
859 transport-filters or DKIM-signing. The restriction was lost in the
860 consolidation of verify-callout and delivery SMTP handling.
861 Extend the restriction to also cover ARC-signing.
863 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
864 in defer=pass mode supply a 450 to the initiator. Previously the message
867 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
868 tls_require_ciphers is used as before.
870 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
871 "pass_unscanned". Only tmpfails from the scanner are written to
872 the paniclog, as they may require admin intervention (permission
873 denied, license issues). Other scanner errors (like decompression
874 bombs) do not cause a paniclog entry.
876 JH/36 Fix reinitialisation of DKIM logging variable between messages.
877 Previously it was possible to log spurious information in receive log
880 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
881 triggered odd behaviour from Outlook Express clients.
883 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
886 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
887 since the IETF WG has not yet settled on that versus the original
888 "bare" representation.
890 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
891 Previously the millisecond value corrupted the output.
892 Fix also for syslog_pid=no and log_selector +pid, for which the pid
893 corrupted the output.
899 JH/01 Rework error string handling in TLS interface so that the caller in
900 more cases is responsible for logging. This permits library-sourced
901 string to be attached to addresses during delivery, and collapses
902 pairs of long lines into single ones.
904 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
905 during configuration. Wildcards are allowed and expanded.
907 JH/02 Rework error string handling in DKIM to pass more info back to callers.
908 This permits better logging.
910 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
911 do not close it down and have the child transport start it up again on
912 the passed-on TCP connection. Instead, proxy the child (and any
913 subsequent ones) for TLS via a unix-domain socket channel. Logging is
914 affected: the continued delivery log lines do not have any DNSSEC, TLS
915 Certificate or OCSP information. TLS cipher information is still logged.
917 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
918 identical IP addresses on different listening ports. Will also affect
921 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
922 add noisy ifdef guards to special-case this sillyness.
923 Patch from Bernd Kuhls.
925 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
926 than 255 are no longer allowed.
928 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
929 Disable the session-cache too, which might reduce our load. Since we
930 currrectly use a new context for every connection, both as server and
931 client, there is no benefit for these.
932 GnuTLS appears to not support tickets server-side by default (we don't
933 call gnutls_session_ticket_enable_server()) but client side is enabled
934 by default on recent versions (3.1.3 +) unless the PFS priority string
937 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
938 <https://reproducible-builds.org/specs/source-date-epoch/>.
940 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
941 the check for any unsuccessful recipients did not notice the limit, and
942 erroneously found still-pending ones.
944 JH/08 Pipeline CHUNKING command and data together, on kernels that support
945 MSG_MORE. Only in-clear (not on TLS connections).
947 JH/09 Avoid using a temporary file during transport using dkim. Unless a
948 transport-filter is involved we can buffer the headers in memory for
949 creating the signature, and read the spool data file once for the
950 signature and again for transmission.
952 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
953 4.77 as the kernel support then wasn't solid, having issues in 64bit
954 mode. Now, it's been long enough. Add support for FreeBSD also.
956 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
957 case where the routing stage had gathered several addresses to send to
958 a host before calling the transport for the first, we previously failed
959 to close down TLS in the old transport process before passing the TCP
960 connection to the new process. The new one sent a STARTTLS command
961 which naturally failed, giving a failed delivery and bloating the retry
962 database. Investigation and fix prototype from Wolfgang Breyha.
964 JH/12 Fix check on SMTP command input synchronisation. Previously there were
965 false-negatives in the check that the sender had not preempted a response
966 or prompt from Exim (running as a server), due to that code's lack of
967 awareness of the SMTP input buffering.
969 PP/04 Add commandline_checks_require_admin option.
970 Exim drops privileges sanely, various checks such as -be aren't a
971 security problem, as long as you trust local users with access to their
972 own account. When invoked by services which pass untrusted data to
973 Exim, this might be an issue. Set this option in main configuration
974 AND make fixes to the calling application, such as using `--` to stop
977 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
978 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
979 responses to those, into a single TLS record each way (this usually means
980 a single packet). As a side issue, smtp_enforce_sync now works on TLS
983 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
984 affects you only if you're dancing at the edge of the param size limits.
985 If you are, and this message makes sense to you, then: raise the
986 configured limit or use OpenSSL 1.1. Nothing we can do for older
989 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
990 cmdline element to get the documented default one. Previously it was
993 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
994 are unsuable for non-message-specific reasons - eg. connection timeout,
995 banner-time rejection.
997 JH/16 Fix logging of delivery remote port, when specified by router, under
1000 PP/06 Repair manualroute's ability to take options in any order, even if one
1001 is the name of a transport.
1004 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1006 JH/17 Change the list-building routines interface to use the expanding-string
1007 triplet model, for better allocation and copying behaviour.
1009 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1010 Previously it was constructed the first time a possibly-matching string
1011 was met in the configuration file input during startup; now it is done
1014 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1015 compatible one, to avoid the (poorly documented) possibility of a config
1016 file in the working directory redirecting the DB files, possibly correpting
1017 some existing file. CVE-2017-10140 assigned for BDB.
1019 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1020 cache-hot. Previously, although the result was properly cached, the
1021 initial verify call returned a defer.
1023 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1024 the main verify for receipient in uncached-mode.
1026 JH/22 Retire historical build files to an "unsupported" subdir. These are
1027 defined as "ones for which we have no current evidence of testing".
1029 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1030 if present. Previously it was ignored.
1032 JH/24 Start using specified-initialisers in C structure init coding. This is
1033 a C99 feature (it's 2017, so now considered safe).
1035 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1036 if was a fixed-sized field and bitmask ops via macros; it is now more
1039 PP/07 GitHub PR 56: Apply MariaDB build fix.
1040 Patch provided by Jaroslav Škarvada.
1042 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1043 during Coverity cleanups [4.87 JH/47]
1044 Diagnosis and fix provided by Michael Fischer v. Mollard.
1046 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1047 the right size to place the terminating semicolon on its own folded
1048 line, the header hash was calculated to an incorrect value thanks to
1049 the (relaxed) space the fold became.
1051 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1052 and confused the parent.
1054 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1055 which could crash as a result. This could lead to undeliverable messages.
1057 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1060 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1061 matters on fast-turnover and PID-randomising systems, which were getting
1062 out-of-order delivery.
1064 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1065 a possibly-overlapping copy. The symptom was that "Remote host closed
1066 connection in response to HELO" was logged instead of the actual 4xx
1069 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1070 Previously only that bufferd was discarded, resulting in SYMTP command
1073 JH/32 DKIM: when a message has multiple signatures matching an identity given
1074 in dkim_verify_signers, run the dkim acl once for each. Previously only
1075 one run was done. Bug 2189.
1077 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1078 "panic the current process" to "deliberately defer". The panic log is
1079 still written with the problem list name; the mail and reject logs now
1080 get a temp-reject line for the message that was being handled, saying
1081 something like "domains check lookup or other defer". The SMTP 451
1082 message is still "Temporary local problem".
1084 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1085 A crafted sequence of BDAT commands could result in in-use memory beeing
1086 freed. CVE-2017-16943.
1088 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1089 from SMTP input. Previously it was always done; now only done for DATA
1090 and not BDAT commands. CVE-2017-16944.
1092 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1093 to the message (such as an overlong header line). Previously this was
1094 not done and we did not exit BDAT mode. Followon from the previous item
1095 though a different problem.
1101 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1102 than -2003 did; needs libidn2 in addition to libidn.
1104 JH/02 The path option on a pipe transport is now expanded before use.
1106 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1107 Patch provided by "Björn", documentation fix added too.
1109 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1110 missing a wire-to-host endian conversion.
1112 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1113 close after a BDAT command line could be taken as a following command,
1114 giving a synch failure. Fix by only checking for synch immediately
1115 before acknowledging the chunk.
1117 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1118 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1119 macro. Patches provided by Josh Soref.
1121 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1122 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1123 by IANA as a service extension. However, John Klensin suggests that we
1126 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1127 the dkim code may be unix-mode line endings rather than smtp wire-format
1128 CRLF, so prepend a CR to any bare LF.
1130 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1131 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1133 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1134 the first were themselves being wrongly included in the feed into dkim
1135 processing; with most chunk sizes in use this resulted in an incorrect
1136 body hash calculated value.
1138 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1139 DKIM signature block, for verification. Although advised against by
1140 standards it is specifically not ruled illegal.
1142 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1144 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1145 missing a body hash (the bh= tag).
1147 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1148 It seems that HAProxy sends the Proxy Protocol information in clear and
1149 only then does a TLS startup, so do the same.
1151 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1152 TCP connections (such as for Spamd) unless the daemon successfully set
1153 Fast Open mode on its listening sockets. This fixes breakage seen on
1154 too-old kernels or those not configured for Fast Open, at the cost of
1155 requiring both directions being enabled for TFO, and TFO never being used
1156 by non-daemon-related Exim processes.
1158 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1159 endings, at least on the first header line. Try to canonify any that get
1160 past that check, despite the cost.
1162 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1163 now limited to an arbitrary five deep, while parsing addresses with the
1164 strip_excess_angle_brackets option enabled.
1166 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1167 instead leave the unprompted TLS handshake in socket buffer for the
1168 TLS library to consume.
1170 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1172 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1174 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1175 region in one operation, for speed, and this leaves assigned pointers
1176 dangling. Add checks run only under the testsuite which checks all
1177 variables at a store-reset and panics on a dangling pointer; add code
1178 explicitly nulling out all the variables discovered. Fixes one known
1179 bug: a transport crash, where a dangling pointer for $sending_ip_address
1180 originally assigned in a verify callout, is re-used.
1182 PP/06 Drop '.' from @INC in various Perl scripts.
1184 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1186 PP/08 Reduce a number of compilation warnings under clang; building with
1187 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1188 should be warning-free.
1190 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1192 HS/01 Fix portability problems introduced by PP/08 for platforms where
1193 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1195 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1196 chunk. This allows us to accept broken chunked messages. We need a more
1197 general solution here.
1199 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1200 already-broken messages in the queue.
1202 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1204 JH/19 Fix reference counting bug in routing-generated-address tracking.
1210 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1211 supports it and a size is available (ie. the sending peer gave us one).
1213 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1214 years of being deprecated). The replacements are the ACLs
1215 acl_smtp_mime and acl_not_smtp_mime.
1217 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1218 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1219 or even an in-clear connection were permitted. Now, if the host lookup
1220 was dnssec and dane was requested then the host is only used if the
1221 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1222 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1223 if one fails this test.
1224 This means that a poorly-configured remote DNS will make it incommunicado;
1225 but it protects against a DNS-interception attack on it.
1227 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1228 non-noisy when a race steals the message being considered.
1230 JH/05 If main configuration option tls_certificate is unset, generate a
1231 self-signed certificate for inbound TLS connections.
1233 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1234 in rewrites and routers.
1236 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1237 and logged a warning sing 4.83; now they are a configuration file error.
1239 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1240 (lacking @domain). Apply the same qualification processing as RCPT.
1242 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1244 JH/10 Support ${sha256:} applied to a string (as well as the previous
1247 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1248 a cutthrough deliver is pending, as we always want to make a connection.
1249 This also avoids re-routing the message when later placing the cutthrough
1250 connection after a verify cache hit.
1251 Do not update it with the verify result either.
1253 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1254 when routing results in more than one destination address.
1256 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1257 signing (which inhibits the cutthrough capability). Previously only
1258 the presence of an option was tested; now an expansion evaluating as
1259 empty is permissible (obviously it should depend only on data available
1260 when the cutthrough connection is made).
1262 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1263 the relevant preceding SMTP command did not note the pipelining mode.
1265 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1266 Previously they were not counted.
1268 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1269 as one having no matching records. Previously we deferred the message
1270 that needed the lookup.
1272 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1273 distinguished as "(=".
1275 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1276 for missing MX records. Previously it only worked for missing A records.
1278 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1280 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1281 after the data-go-ahead and data-ack. Patch from Jason Betts.
1283 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1284 even for a "none" policy. Patch from Tony Meyer.
1286 JH/22 Fix continued use of a connection for further deliveries. If a port was
1287 specified by a router, it must also match for the delivery to be
1290 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1291 When one of the recipients of a message was unsuitable for the connection
1292 (has no matching addresses), we lost track of needing to mark it
1293 deferred. As a result mail would be lost.
1295 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1297 JH/25 Decoding ACL controls is now done using a binary search; the source code
1298 takes up less space and should be simpler to maintain. Merge the ACL
1299 condition decode tables also, with similar effect.
1301 JH/26 Fix problem with one_time used on a redirect router which returned the
1302 parent address unchanged. A retry would see the parent address marked as
1303 delivered, so not attempt the (identical) child. As a result mail would
1306 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1307 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1308 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1311 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1312 Main config options check_{log,spool}_{inodes,space} are now
1313 100 inodes, 10MB unless set otherwise in the configuration.
1315 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1316 Previously it only applied to the main-section connection policy
1319 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1321 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1322 by me. Added RFC7919 DH primes as an alternative.
1324 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1325 are not in the system include path.
1327 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1328 GnuTLS, when a session startup failed (eg because the client disconnected)
1329 Exim did stdio operations after fclose. This was exposed by a recent
1330 change which nulled out the file handle after the fclose.
1332 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1333 signed directly by the cert-signing cert, rather than an intermediate
1334 OCSP-signing cert. This is the model used by LetsEncrypt.
1336 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1338 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1339 an incoming connection.
1341 HS/02 Bug 1802: Do not half-close the connection after sending a request
1344 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1345 fallback to "prime256v1".
1347 JH/34 SECURITY: Use proper copy of DATA command in error message.
1348 Could leak key material. Remotely exploitable. CVE-2016-9963.
1354 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1355 and 3.4.4 - once the server is enabled to respond to an OCSP request
1356 it does even when not requested, resulting in a stapling non-aware
1357 client dropping the TLS connection.
1359 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1360 support variable-length bit vectors. No functional change.
1362 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1363 The I= interface field on outgoing lines is now after the H= remote
1364 host field, same as incoming lines. There is a separate
1365 outgoing_interface log selector which allows you to disable the
1368 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1369 If not running log_selector +smtp_connection the mainlog would be held
1370 open indefinitely after a "too many connections" event, including to a
1371 deleted file after a log rotate. Leave the per net connection logging
1372 leaving it open for efficiency as that will be quickly detected by the
1373 check on the next write.
1375 HS/01 Bug 1671: Fix post transport crash.
1376 Processing the wait-<transport> messages could crash the delivery
1377 process if the message IDs didn't exist for some reason. When
1378 using 'split_spool_directory=yes' the construction of the spool
1379 file name failed already, exposing the same netto behaviour.
1381 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1382 mime_regex ACL conditions.
1384 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1385 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1386 to HELO, local diagnostic string.
1388 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1389 log line to debug. Even when configured with a tls authenticator many
1390 client connections are expected to not authenticate in this way, so
1391 an authenticate fail is not an error.
1393 HS/02 Add the Exim version string to the process info. This way exiwhat
1394 gives some more detail about the running daemon.
1396 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1397 matter for fast-change records such as DNSBLs.
1399 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1400 retry record, even if constant. There may be multiple transports with
1401 different interface settings and the retry behaviour needs to be kept
1404 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1406 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1408 JH/10 Bug 840: fix log_defer_output option of pipe transport
1410 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1411 hosts_randomize. This matters a lot when combined with mua_wrapper.
1413 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1414 ${quote_pgsql:<string>} operator.
1416 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1418 JH/14 Change header file name for freeradius-client. Relevant if compiling
1419 with Radius support; from the Gentoo tree and checked under Fedora.
1421 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1423 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1424 option transport_filter, meaning no filtering.
1426 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1428 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1429 defaults to "*" (all hosts). The variable is now available when not built
1430 with TLS, default unset, mainly to enable keeping the testsuite sane.
1431 If a server certificate is not supplied (via tls_certificate) an error is
1432 logged, and clients will find TLS connections fail on startup. Presumably
1433 they will retry in-clear.
1434 Packagers of Exim are strongly encouraged to create a server certificate
1435 at installation time.
1437 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1438 with the $config_file variable.
1440 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1441 in transport context, after the attempt, and per-recipient. The latter type
1442 is per host attempted. The event data is the error message, and the errno
1443 information encodes the lookup type (A vs. MX) used for the (first) host,
1444 and the trailing two digits of the smtp 4xx response.
1446 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1447 to write to mainlog (or rejectlog, paniclog) in the window between file
1448 creation and permissions/ownership being changed. Particularly affects
1449 installations where exicyclog is run as root, rather than exim user;
1450 result is that the running daemon panics and dies.
1452 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1454 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1455 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1456 "pri" and "weight". Note that the previous implicit priority given by the
1457 list order is no longer honoured.
1459 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1460 for DKIM processing.
1462 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1463 by defining SUPPORT_SOCKS.
1465 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1466 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1467 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1468 variables are renamed to proxy_{local,external}_{address,port}.
1470 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1471 for a build by defining SUPPORT_I18N
1473 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1474 of the query string, and make ${quote_redis:} do that quoting.
1476 JH/29 Move Events support from Experimental to mainline, enabled by default
1477 and removable for a build by defining DISABLE_EVENT.
1479 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1481 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1482 cached by the daemon.
1484 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1485 by defining LOOKUP_REDIS. The libhiredis library is required.
1487 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1488 keys are given for lookup.
1490 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1491 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1492 only supported when built with TLS support. The PolarSSL SHA routines
1493 are still used when the TLS library is too old for convenient support.
1495 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1496 openssl_options), for security. OpenSSL forces this from version 1.1.0
1497 server-side so match that on older versions.
1499 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1500 allocation for $value could be released as the expansion processing
1501 concluded, but leaving the global pointer active for it.
1503 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1504 and to use the domains and local_parts ACL conditions.
1506 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1507 incorrectly not doubled on cutthrough transmission, hence seen as a
1508 body-termination at the receiving system - resulting in truncated mails.
1509 Commonly the sender saw a TCP-level error, and retransmitted the message
1510 via the normal store-and-forward channel. This could result in duplicates
1511 received - but deduplicating mailstores were liable to retain only the
1512 initial truncated version.
1514 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1516 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1518 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1519 we're in there, support oversigning also; bug 1309.
1521 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1523 HS/04 Add support for keep_environment and add_environment options.
1525 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1526 either intentional arithmetic overflow during PRNG, or testing config-
1529 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1530 delivery resulted in actual delivery. Cancel cutthrough before DATA
1533 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1534 rejects a recipient: pass the reject to the originator.
1536 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1537 Many were false-positives and ignorable, but it's worth fixing the
1540 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1541 for the new environment-manipulation done at startup. Move the routines
1542 from being local to tls.c to being global via the os.c file.
1544 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1545 an extract embedded as result-arg for a map, the first arg for extract
1546 is unavailable so we cannot tell if this is a numbered or keyed
1547 extraction. Accept either.
1553 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1556 JH/02 The smtp transport option "multi_domain" is now expanded.
1558 JH/03 The smtp transport now requests PRDR by default, if the server offers
1561 JH/04 Certificate name checking on server certificates, when exim is a client,
1562 is now done by default. The transport option tls_verify_cert_hostnames
1563 can be used to disable this per-host. The build option
1564 EXPERIMENTAL_CERTNAMES is withdrawn.
1566 JH/05 The value of the tls_verify_certificates smtp transport and main options
1567 default to the word "system" to access the system default CA bundle.
1568 For GnuTLS, only version 3.0.20 or later.
1570 JH/06 Verification of the server certificate for a TLS connection is now tried
1571 (but not required) by default. The verification status is now logged by
1572 default, for both outbound TLS and client-certificate supplying inbound
1575 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1578 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1579 Status Notification (bounce) messages are now MIME format per RFC 3464.
1580 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1581 under the control of the dsn_advertise_hosts option, and routers may
1582 have a dsn_lasthop option.
1584 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1585 default, modifiable by a malware= option. The list separator for
1586 the options can now be changed in the usual way. Bug 68.
1588 JH/10 The smtp_receive_timeout main option is now expanded before use.
1590 JH/11 The incoming_interface log option now also enables logging of the
1591 local interface on delivery outgoing connections.
1593 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1594 if the interface and destination host and port all match.
1596 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1599 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1600 Patch from Andrew Lewis.
1602 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1603 now supports optional time-restrictions, weighting, and priority
1604 modifiers per server. Patch originally by <rommer@active.by>.
1606 JH/16 The spamd_address main option now supports a mixed list of local
1607 and remote servers. Remote servers can be IPv6 addresses, and
1608 specify a port-range.
1610 JH/17 Bug 68: The spamd_address main option now supports an optional
1611 timeout value per server.
1613 JH/18 Bug 1581: Router and transport options headers_add/remove can
1614 now have the list separator specified.
1616 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1619 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1622 JH/21 Support for the A6 type of dns record is withdrawn.
1624 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1625 rather than the verbs used.
1627 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1628 from 255 to 1024 chars.
1630 JH/24 Verification callouts now attempt to use TLS by default.
1632 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1633 are generic router options now. The defaults didn't change.
1635 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1636 Original patch from Alexander Shikoff, worked over by JH.
1638 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1639 files if ZCAT_COMMAND is not executable.
1641 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1643 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1645 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1646 Normally benign, it bites when the pair was led to by a CNAME;
1647 modern usage is to not canonicalize the domain to a CNAME target
1648 (and we were inconsistent anyway for A-only vs AAAA+A).
1650 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1652 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1653 when evaluating $sender_host_dnssec.
1655 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1656 $sender_helo_dnssec variable.
1658 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1660 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1662 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1664 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1665 documented as working, but never had. Support all but $spam_report.
1667 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1668 added for tls authenticator.
1670 HS/03 Add perl_taintmode main config option
1676 TL/01 When running the test suite, the README says that variables such as
1677 no_msglog_check are global and can be placed anywhere in a specific
1678 test's script, however it was observed that placement needed to be near
1679 the beginning for it to behave that way. Changed the runtest perl
1680 script to read through the entire script once to detect and set these
1681 variables, reset to the beginning of the script, and then run through
1682 the script parsing/test process like normal.
1684 TL/02 The BSD's have an arc4random API. One of the functions to induce
1685 adding randomness was arc4random_stir(), but it has been removed in
1686 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1687 function when detected.
1689 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1690 cause callback expansion.
1692 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1693 syntax errors in an expansion can be treated as a string instead of
1694 logging or causing an error, due to the internal use of bool_lax
1695 instead of bool when processing it.
1697 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1698 server certificates when making smtp deliveries.
1700 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1702 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1704 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1706 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1707 Merged patch from Sebastian Wiedenroth.
1709 JH/05 Fix results-pipe from transport process. Several recipients, combined
1710 with certificate use, exposed issues where response data items split
1711 over buffer boundaries were not parsed properly. This eventually
1712 resulted in duplicates being sent. This issue only became common enough
1713 to notice due to the introduction of connection certificate information,
1714 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1716 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1717 size buffer was used, resulting in syntax errors when an expansion
1720 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1721 version 3.3.6 or later.
1723 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1724 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1725 both become "event_action", the variables become $event_name, $event_data
1726 and $event_defer_errno. There is a new variable $verify_mode, usable in
1727 routers, transports and related events. The tls:cert event is now also
1728 raised for inbound connections, if the main configuration event_action
1731 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1732 early OCSP support, but no stapling (appears to be less than 1.0.0).
1734 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1735 server certificate names available under the smtp transport option
1736 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1739 JH/10 Time-related extraction expansions from certificates now use the main
1740 option "timezone" setting for output formatting, and are consistent
1741 between OpenSSL and GnuTLS compilations. Bug 1541.
1743 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1744 encoded parameter in the incoming message. Bug 1558.
1746 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1747 include certificate info, eximon was claiming there were spoolfile
1750 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1752 JH/14 Log delivery-related information more consistently, using the sequence
1753 "H=<name> [<ip>]" wherever possible.
1755 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1756 are problematic for Debian distribution, omit them from the release
1759 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1761 JH/16 Fix string representation of time values on 64bit time_t architectures.
1764 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1765 output list separator was used.
1770 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1771 checkers that were complaining about end of non-void function with no
1774 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1775 This was a regression introduced in 4.83 by another bugfix.
1777 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1779 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1780 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1786 TF/01 Correctly close the server side of TLS when forking for delivery.
1788 When a message was received over SMTP with TLS, Exim failed to clear up
1789 the incoming connection properly after forking off the child process to
1790 deliver the message. In some situations the subsequent outgoing
1791 delivery connection happened to have the same fd number as the incoming
1792 connection previously had. Exim would try to use TLS and fail, logging
1793 a "Bad file descriptor" error.
1795 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1796 utilities have not been installed.
1798 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1799 temporary space as the ACL may create new global variables.
1801 TL/01 LDAP support uses per connection or global context settings, depending
1802 upon the detected version of the libraries at build time.
1804 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1805 to extract and use the src ip:port in logging and expansions as if it
1806 were a direct connection from the outside internet. PPv2 support was
1807 updated based on HAProxy spec change in May 2014.
1809 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1811 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1812 Properly escape header and check for NULL return.
1814 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1817 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1819 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1820 characters in header names, implemented as a verify condition.
1821 Contributed by Michael Fischer v. Mollard.
1823 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1824 results permerror and temperror. Previous values are deprecated but
1825 still accepted. In a future release, err_perm and err_temp will be
1826 completely removed, which will be a backward incompatibility if the
1827 ACL tests for either of these two old results. Patch contributed by
1828 user bes-internal on the mailing list.
1830 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1832 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1833 selectors, in both main and reject logs.
1835 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1838 JH/07 Add malware type "sock" for talking to simple daemon.
1840 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1842 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1843 routers/transports under cutthrough routing.
1845 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1846 numbers. Touch up "bool" conditional to keep the same definition.
1848 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1850 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1852 JH/12 Expand items in router/transport headers_add or headers_remove lists
1853 individually rather than the list as a whole. Bug 1452.
1855 Required for reasonable handling of multiple headers_ options when
1856 they may be empty; requires that headers_remove items with embedded
1857 colons must have them doubled (or the list-separator changed).
1859 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1860 view the policy declared in the DMARC record. Currently, $dmarc_status
1861 is a combined value of both the record presence and the result of the
1864 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1866 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1867 dnslookup router and the smtp transport (applying to the forward
1870 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1871 of ldap servers used for a specific lookup. Patch provided by Heiko
1874 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1875 New variable $lookup_dnssec_authenticated for observability.
1877 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1878 Patch submitted by Lars Timman.
1880 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1882 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1883 Requires trusted mode and valid format message id, aborts otherwise.
1884 Patch contributed by Heiko Schlichting.
1886 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1887 certextract with support for various fields. Bug 1358.
1889 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1890 is requested by default, modifiable by smtp transport option
1893 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1894 operate on certificate variables to give certificate fingerprints
1895 Also new ${sha256:cert_variable}.
1897 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1899 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1900 Christian Aistleitner.
1902 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1904 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1905 file. Patch from Wolfgang Breyha.
1907 JH/25 Expand the coverage of the delivery $host and $host_address to
1908 client authenticators run in verify callout. Bug 1476.
1910 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1911 align with daemon_smtp_ports. Bug 72.
1913 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1914 support and error reporting did not work properly.
1916 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1917 and is readable. Patch from Andrew Colin Kissa.
1919 TL/14 Enhance documentation of ${run expansion and how it parses the
1920 commandline after expansion, particularly in the case when an
1921 unquoted variable expansion results in an empty value.
1923 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1925 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1926 by renaming away from it. Addresses GH issue 15, reported by
1929 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1931 TL/15 SECURITY: prevent double expansion in math comparison functions
1932 (can expand unsanitized data). Not remotely exploitable.
1939 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1941 PP/02 Make -n do something, by making it not do something.
1942 When combined with -bP, the name of an option is not output.
1944 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1947 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1948 $sender_host_name and config options to manage this, and basic check
1951 PP/05 DSCP support for outbound connections and control modifier for inbound.
1953 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1954 (Only plugin which currently uses this is kerberos4, which nobody should
1955 be using, but we should make it available and other future plugins might
1956 conceivably use it, even though it would break NAT; stuff *should* be
1957 using channel bindings instead).
1959 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1960 name; added for Sendmail compatibility; requires admin caller.
1961 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1962 just ignore it); requires trusted caller.
1963 Also parse but ignore: -Ac -Am -X<logfile>
1966 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1968 TL/02 Add +smtp_confirmation as a default logging option.
1970 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1971 Patch by Magnus Holmgren from 2007-02-20.
1973 TL/04 Bugzilla 1281 - Spec typo.
1974 Bugzilla 1283 - Spec typo.
1975 Bugzilla 1290 - Spec grammar fixes.
1977 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1979 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1981 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1982 mailing list by by Dmitry Isaikin.
1984 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1986 JH/02 Support "G" suffix to numbers in ${if comparisons.
1988 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1990 NM/01 Bugzilla 1197 - Spec typo
1991 Bugzilla 1196 - Spec examples corrections
1993 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1995 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1996 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1999 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2000 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2002 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2003 "acl {{name}{arg}...}", and optional args on acl condition
2006 JH/05 Permit multiple router/transport headers_add/remove lines.
2008 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2010 JH/07 Avoid using a waiting database for a single-message-only transport.
2011 Performance patch from Paul Fisher. Bugzilla 1262.
2013 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2016 JH/09 Add $headers_added variable, with content from use of ACL modifier
2017 add_header (but not yet added to the message). Bugzilla 199.
2019 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2020 Pulled from Bugzilla 817 by Wolfgang Breyha.
2022 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2024 (nb: this is the same fix as in Exim 4.80.1)
2026 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2029 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2032 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2034 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2035 advertise SMTP AUTH mechanism to us, instead of a generic
2036 protocol violation error. Also, make Exim more robust to bad
2037 data from the Dovecot auth socket.
2039 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2041 When a queue runner is handling a message, Exim first routes the
2042 recipient addresses, during which it prunes them based on the retry
2043 hints database. After that it attempts to deliver the message to
2044 any remaining recipients. It then updates the hints database using
2047 So if a recipient address works intermittently, it can get repeatedly
2048 deferred at routing time. The retry hints record remains fresh so the
2049 address never reaches the final cutoff time.
2051 This is a fairly common occurrence when a user is bumping up against
2052 their storage quota. Exim had some logic in its local delivery code
2053 to deal with this. However it did not apply to per-recipient defers
2054 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2056 This change adds a proper retry rule check during routing so that the
2057 final cutoff time is checked against the message's age. We only do
2058 this check if there is an address retry record and there is not a
2059 domain retry record; this implies that previous attempts to handle
2060 the address had the retry_use_local_parts option turned on. We use
2061 this as an approximation for the destination being like a local
2062 delivery, as in LMTP.
2064 I suspect this new check makes the old local delivery cutoff check
2065 redundant, but I have not verified this so I left the code in place.
2067 TF/02 Correct gecos expansion when From: is a prefix of the username.
2069 Test 0254 submits a message to Exim with the header
2073 When I ran the test suite under the user fanf2, Exim expanded
2074 the header to contain my full name, whereas it should have added
2075 a Resent-Sender: header. It erroneously treats any prefix of the
2076 username as equal to the username.
2078 This change corrects that bug.
2080 GF/01 DCC debug and logging tidyup
2081 Error conditions log to paniclog rather than rejectlog.
2082 Debug lines prefixed by "DCC: " to remove any ambiguity.
2084 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2086 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2087 Bug spotted by Jeremy Harris; was flawed since initial commit.
2088 Would have resulted in OCSP responses post-SNI triggering an Exim
2089 NULL dereference and crash.
2091 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2093 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2094 Bug detection, analysis and fix by Samuel Thibault.
2095 Bugzilla 1331, Debian bug #698092.
2097 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2099 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2100 Server implementation by Todd Lyons, client by JH.
2101 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2102 config variable "prdr_enable" controls whether the server
2103 advertises the facility. If the client requests PRDR a new
2104 acl_data_smtp_prdr ACL is called once for each recipient, after
2105 the body content is received and before the acl_smtp_data ACL.
2106 The client is controlled by both of: a hosts_try_prdr option
2107 on the smtp transport, and the server advertisement.
2108 Default client logging of deliveries and rejections involving
2109 PRDR are flagged with the string "PRDR".
2111 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2112 fclose(). Diagnosis by Todd Lyons.
2114 PP/17 Update configure.default to handle IPv6 localhost better.
2115 Patch by Alain Williams (plus minor tweaks).
2118 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2119 This is now consistent with GnuTLS, and is now documented: the
2120 previous undocumented portable approach to treating the option as
2121 unset was to force an expansion failure. That still works, and
2122 an empty string is now equivalent.
2124 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2125 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2126 not performing validation itself.
2128 PP/20 Added force_command boolean option to pipe transport.
2129 Patch from Nick Koston, of cPanel Inc.
2131 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2134 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2136 PP/21 Fix eximon continuous updating with timestamped log-files.
2137 Broken in a format-string cleanup in 4.80, missed when I repaired the
2138 other false fix of the same issue.
2139 Report and fix from Heiko Schlichting.
2142 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2143 Report from Prashanth Katuri.
2145 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2146 It's SecureTransport, so affects any MacOS clients which use the
2147 system-integrated TLS libraries, including email clients.
2149 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2150 using a MIME ACL for non-SMTP local injection.
2151 Report and assistance in diagnosis by Warren Baker.
2153 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2155 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2157 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2158 last id that failed so it may be referenced in subsequent ACL's.
2160 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2163 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2164 ldap library initialization, allowing self-signed CA's to be
2165 used. Also properly sets require_cert option later in code by
2166 using NULL (global ldap config) instead of ldap handle (per
2167 session). Bug diagnosis and testing by alxgomz.
2169 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2170 the src/util/ subdirectory.
2172 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2173 renamed to Transport Post Delivery Action by Jeremy Harris, as
2176 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2177 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2178 redis_servers = needs to be configured which will be used by the redis
2179 lookup. Patch from Warren Baker, of The Packet Hub.
2181 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2183 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2184 hostname or reverse DNS when processing a host list. Used suggestions
2185 from multiple comments on this bug.
2187 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2189 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2190 Missed a few lines, added it to make the runtest require no keyboard
2193 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2194 contains upper case chars. Make router use caseful_local_part.
2196 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2197 support when GnuTLS has been built with p11-kit.
2203 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2205 This, or similar/improved, will also be change PP/11 of 4.82.
2211 PP/01 Handle short writes when writing local log-files.
2212 In practice, only affects FreeBSD (8 onwards).
2213 Bugzilla 1053, with thanks to Dmitry Isaikin.
2215 NM/01 Bugzilla 949 - Documentation tweak
2217 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2220 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2222 PP/02 Implemented gsasl authenticator.
2224 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2226 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2227 `pkg-config foo` for cflags/libs.
2229 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2230 with rest of GSASL and with heimdal_gssapi.
2232 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2233 `pkg-config foo` for cflags/libs for the TLS implementation.
2235 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2236 properties get this fed in as external SSF. A number of robustness
2237 and debugging improvements to the cyrus_sasl authenticator.
2239 PP/08 cyrus_sasl server now expands the server_realm option.
2241 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2242 Patch by Jeremy Harris.
2244 PP/10 Added dbmjz lookup type.
2246 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2248 PP/12 MAIL args handles TAB as well as SP, for better interop with
2249 non-compliant senders.
2250 Analysis and variant patch by Todd Lyons.
2252 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2253 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2254 Patch from Dirk Mueller <dmueller@suse.com>
2256 PP/13 tls_peerdn now print-escaped for spool files.
2257 Observed some $tls_peerdn in wild which contained \n, which resulted
2258 in spool file corruption.
2260 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2261 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2262 or write after TLS renegotiation, which otherwise led to messages
2265 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2266 as a tracking header (ie: a signed header comes before the signature).
2267 Patch from Wolfgang Breyha.
2269 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2270 comma-sep list; embedded commas doubled.
2272 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2274 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2276 Report and patch from Dmitry Banschikov.
2278 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2279 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2280 failures. We appear to now support TLS1.1+ with Exim.
2282 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2283 lets Exim select keys and certificates based upon TLS SNI from client.
2284 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2285 before an outbound SMTP session. New log_selector, +tls_sni.
2287 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2288 NULL dereference. Report and patch from Alun Jones.
2290 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2291 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2292 Not seeing resolver debug output on NetBSD, but suspect this is a
2293 resolver implementation change.
2295 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2296 Left warnings. Added "eximon gdb" invocation mode.
2298 PP/21 Defaulting "accept_8bitmime" to true, not false.
2300 PP/22 Added -bw for inetd wait mode support.
2302 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2303 locate the relevant includes and libraries. Made this the default.
2305 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2306 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2308 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2309 This may cause build issues on older platforms.
2311 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2312 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2313 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2314 Added SNI support via GnuTLS too.
2315 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2317 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2319 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2320 Applied second patch from Janne, implementing suggestion to default
2321 multiple-strings-in-record handling to match SPF spec.
2323 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2325 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2326 read-only, out of scope).
2327 Patch from Wolfgang Breyha, report from Stuart Northfield.
2329 PP/29 Fix three issues highlighted by clang analyser static analysis.
2330 Only crash-plausible issue would require the Cambridge-specific
2331 iplookup router and a misconfiguration.
2332 Report from Marcin Mirosław.
2334 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2336 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2337 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2338 As part of this, removing so much warning spew let me fix some minor
2339 real issues in debug logging.
2341 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2342 assignment on my part. Fixed.
2344 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2345 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2346 Janne Snabb (who went above and beyond: thank you).
2348 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2349 string otherwise requires a connection and a bunch more work and it's
2350 relatively easy to get wrong. Should also expose TLS library linkage
2353 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2354 64-bit ${eval} (JH/03).
2356 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2357 GNU libc to support some of the 64-bit stuff, should not lead to
2358 conflicts. Defined before os.h is pulled in, so if a given platform
2359 needs to override this, it can.
2361 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2362 protection layer was required, which is not implemented.
2363 Bugzilla 1254, patch from Wolfgang Breyha.
2365 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2366 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2367 tls_dhparam take prime identifiers. Also unbreak combination of
2368 OpenSSL+DH_params+TLSSNI.
2370 PP/39 Disable SSLv2 by default in OpenSSL support.
2376 PP/01 Solaris build fix for Oracle's LDAP libraries.
2377 Bugzilla 1109, patch from Stephen Usher.
2379 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2381 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2384 TF/02 Fix a couple more cases where we did not log the error message
2385 when unlink() failed. See also change 4.74-TF/03.
2387 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2388 lock up or crash if it happened to be inside a call to libc when it
2389 got a SIGUSR1 from exiwhat.
2391 The SIGUSR1 handler appends the current process status to the process
2392 log which is later printed by exiwhat. It used to use the general
2393 purpose logging code to do this, but several functions it calls are
2394 not safe for signals.
2396 The new output code in the SIGUSR1 handler is specific to the process
2397 log, and simple enough that it's easy to inspect for signal safety.
2398 Removing some special cases also simplifies the general logging code.
2399 Removing the spurious timestamps from the process log simplifies
2402 TF/04 Improved ratelimit ACL condition.
2404 The /noupdate option has been deprecated in favour of /readonly which
2405 has clearer semantics. The /leaky, /strict, and /readonly update modes
2406 are mutually exclusive. The update mode is no longer included in the
2407 database key; it just determines when the database is updated. (This
2408 means that when you upgrade Exim will forget old rate measurements.)
2410 Exim now checks that the per_* options are used with an update mode that
2411 makes sense for the current ACL. For example, when Exim is processing a
2412 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2413 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2414 must specify per_mail/readonly. If you omit the update mode it defaults to
2415 /leaky where that makes sense (as before) or /readonly where required.
2417 The /noupdate option is now undocumented but still supported for
2418 backwards compatibility. It is equivalent to /readonly except that in
2419 ACLs where /readonly is required you may specify /leaky/noupdate or
2420 /strict/noupdate which are treated the same as /readonly.
2422 A useful new feature is the /count= option. This is a generalization
2423 of the per_byte option, so that you can measure the throughput of other
2424 aggregate values. For example, the per_byte option is now equivalent
2425 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2427 The per_rcpt option has been generalized using the /count= mechanism
2428 (though it's more complicated than the per_byte equivalence). When it is
2429 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2430 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2431 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2432 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2433 Note that using per_rcpt with a non-readonly update mode in more than
2434 one ACL will cause the recipients to be double-counted. (The per_mail
2435 and per_byte options don't have this problem.)
2437 The handling of very low rates has changed slightly. If the computed rate
2438 is less than the event's count (usually one) then this event is the first
2439 after a long gap. In this case the rate is set to the same as this event's
2440 count, so that the first message of a spam run is counted properly.
2442 The major new feature is a mechanism for counting the rate of unique
2443 events. The new per_addr option counts the number of different
2444 recipients that someone has sent messages to in the last time period. It
2445 behaves like per_rcpt if all the recipient addresses are different, but
2446 duplicate recipient addresses do not increase the measured rate. Like
2447 the /count= option this is a general mechanism, so the per_addr option
2448 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2449 example, measure the rate that a client uses different sender addresses
2450 with the options per_mail/unique=$sender_address. There are further
2451 details in the main documentation.
2453 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2455 TF/06 Removed a few PCRE remnants.
2457 TF/07 Automatically extract Exim's version number from tags in the git
2458 repository when doing development or release builds.
2460 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2461 Bugzilla 879. Patch from Paul Fisher.
2463 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2464 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2467 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2469 PP/05 Variable $av_failed, true if the AV scanner deferred.
2470 Bugzilla 1078. Patch from John Horne.
2472 PP/06 Stop make process more reliably on build failure.
2473 Bugzilla 1087. Patch from Heiko Schlittermann.
2475 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2476 Bugzilla 1089. Patch from Heiko Schlittermann.
2478 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2479 Bugzilla 1131. Patch from Holger Weiß.
2481 PP/09 Handle IPv6 addresses with SPF.
2482 Bugzilla 860. Patch from Wolfgang Breyha.
2484 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2486 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2489 PP/11 match_* no longer expand right-hand-side by default.
2490 New compile-time build option, EXPAND_LISTMATCH_RHS.
2491 New expansion conditions, "inlist", "inlisti".
2493 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2495 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2497 PP/14 fix log_write() format string regression from TF/03.
2498 Bugzilla 1152. Patch from Dmitry Isaikin.
2504 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2506 PP/02 Harmonised TLS library version reporting; only show if debugging.
2507 Layout now matches that introduced for other libraries in 4.74 PP/03.
2509 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2511 PP/04 New "dns_use_edns0" global option.
2513 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2516 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2517 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2519 TK/01 Updated PolarSSL code to 0.14.2.
2520 Bugzilla 1097. Patch from Andreas Metzler.
2522 PP/07 Catch divide-by-zero in ${eval:...}.
2523 Fixes bugzilla 1102.
2525 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2528 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2529 format-string attack -- SECURITY: remote arbitrary code execution.
2531 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2532 time unintentionally subject to list matching rules, letting the header
2533 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2534 arbitrary string expansion). This allowed for information disclosure.
2536 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2537 INT_MIN/-1 -- value coerced to INT_MAX.
2543 NM/01 Workaround for PCRE version dependency in version reporting
2546 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2547 This fixes portability to compilers other than gcc, notably
2548 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2550 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2551 makefiles for portability to HP-UX and POSIX correctness.
2553 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2554 Also via indented variable definition in the Makefile.
2555 (Debugging by Oliver Heesakkers).
2557 PP/02 Restore caching of spamd results with expanded spamd_address.
2558 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2560 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2561 Improves build reliability. Fix from: Frank Elsner
2563 NM/02 Fix wide character breakage in the rfc2047 coding
2564 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2566 NM/03 Allow underscore in dnslist lookups
2567 Fixes bug 1026. Patch from Graeme Fowler
2569 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2570 Code patches from Adam Ciarcinski of NetBSD.
2572 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2575 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2576 is logged, to avoid truncation. Patch from John Horne.
2578 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2579 Patch from Jakob Hirsch.
2581 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2582 SQL string expansion failure details.
2583 Patch from Andrey Oktyabrski.
2585 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2586 Patch from Simon Arlott.
2588 PP/09 New lookups functionality failed to compile on old gcc which rejects
2589 extern declarations in function scope.
2590 Patch from Oliver Fleischmann
2592 PP/10 Use sig_atomic_t for flags set from signal handlers.
2593 Check getgroups() return and improve debugging.
2594 Fixed developed for diagnosis in bug 927 (which turned out to be
2597 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2598 Patch from Mark Zealey.
2600 PP/12 Bugzilla 1056: Improved spamd server selection.
2601 Patch from Mark Zealey.
2603 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2604 Based on patch from Heiko Schlittermann.
2606 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2607 Patch from Uwe Doering, sign-off by Michael Haardt.
2609 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2610 variable declaration deep within a block. Bug and patch from
2613 PP/15 lookups-Makefile IRIX compatibility coercion.
2615 PP/16 Make DISABLE_DKIM build knob functional.
2617 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2618 Patch by Simon Arlott
2620 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2621 variable argument macros. Our copy now differs from upstream.
2627 TF/01 Failure to get a lock on a hints database can have serious
2628 consequences so log it to the panic log.
2630 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2631 controlled using the smtp_confirmation log selector.
2633 TF/03 Include the error message when we fail to unlink a spool file.
2635 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2636 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2637 for maintaining out-of-tree patches for some time.
2639 PP/01 Bugzilla 139: Documentation and portability issues.
2640 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2641 Handle per-OS dynamic-module compilation flags.
2643 PP/02 Let /dev/null have normal permissions.
2644 The 4.73 fixes were a little too stringent and complained about the
2645 permissions on /dev/null. Exempt it from some checks.
2646 Reported by Andreas M. Kirchwitz.
2648 PP/03 Report version information for many libraries, including
2649 Exim version information for dynamically loaded libraries. Created
2650 version.h, now support a version extension string for distributors
2651 who patch heavily. Dynamic module ABI change.
2653 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2654 privilege escalation vulnerability whereby the Exim run-time user
2655 can cause root to append content of the attacker's choosing to
2658 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2661 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2662 If dropping privileges for untrusted macros, we disabled normal logging
2663 on the basis that it would fail; for the Exim run-time user, this is not
2664 the case, and it resulted in successful deliveries going unlogged.
2665 Fixed. Reported by Andreas Metzler.
2671 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2672 only prepend for the Resent-* case. Fixes regression introduced in
2673 Exim 4.70 by NM/22 for Bugzilla 607.
2675 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2676 increasing numbers of administrators be bitten by this.
2678 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2680 PP/03 Bugzilla 994: added openssl_options main configuration option.
2682 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2684 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2686 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2688 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2689 without a peer certificate, leading to a segfault because of an
2690 assumption that peers always have certificates. Be a little more
2691 paranoid. Problem reported by Martin Tscholak.
2693 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2694 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2695 NB: ClamAV planning to remove STREAM in "middle of 2010".
2696 CL also introduces -bmalware, various -d+acl logging additions and
2697 more caution in buffer sizes.
2699 PP/09 Implemented reverse_ip expansion operator.
2701 PP/10 Bugzilla 937: provide a "debug" ACL control.
2703 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2705 PP/12 Bugzilla 973: Implement --version.
2707 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2709 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2711 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2713 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2714 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2715 ignore trailing whitespace.
2717 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2719 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2722 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2723 Notification from Dr Andrew Aitchison.
2725 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2726 ExtendedDetectionInfo response format.
2727 Notification from John Horne.
2729 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2732 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2733 XSL and documented dependency on system catalogs, with examples of how
2736 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2739 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2740 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2741 configuration file which is writeable by the Exim user or group.
2743 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2744 of configuration files to cover files specified with the -C option if
2745 they are going to be used with root privileges, not just the default
2748 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2749 option (effectively making it always true).
2751 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2752 files to be used while preserving root privileges.
2754 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2755 that rogue child processes cannot use them.
2757 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2758 run-time user, instead of root.
2760 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2761 Exim run-time user without dropping privileges.
2763 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2764 result string, instead of calling string_vformat() twice with the same
2767 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2768 for other users. Others should always drop root privileges if they use
2769 -C on the command line, even for a whitelisted configure file.
2771 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2773 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2779 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2780 $data_path, and $header_path variables; fixed documentation bugs and
2783 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2784 exipick to access non-standard spools, including the "frozen" queue
2787 NM/01 Bugzilla 965: Support mysql stored procedures.
2788 Patch from Alain Williams
2790 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2792 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2793 Patch from Andreas Metzler
2795 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2796 Patch from Kirill Miazine
2798 NM/05 Bugzilla 671: Added umask to procmail example.
2800 JJ/03 installed exipick 20100323.0, fixing doc bug
2802 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2803 directory. Notification and patch from Dan Rosenberg.
2805 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2807 TK/02 Improve log output when DKIM signing operation fails.
2809 MH/01 Treat the transport option dkim_domain as a colon separated
2810 list, not as a single string, and sign the message with each element,
2811 omitting multiple occurences of the same signer.
2813 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2814 Bugzilla 985, 986. Patch by Simon Arlott
2816 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2817 Patch by Simon Arlott
2819 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2820 MBX locking. Notification from Dan Rosenberg.
2826 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2828 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2830 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2832 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2834 NM/04 Bugzilla 915: Flag broken perl installation during build.
2840 TK/01 Added patch by Johannes Berg that expands the main option
2841 "spamd_address" if it starts with a dollar sign.
2843 TK/02 Write list of recipients to X-Envelope-Sender header when building
2844 the mbox-format spool file for content scanning (suggested by Jakob
2847 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2848 (http://www.dcc-servers.net/) support via dccifd. Activated by
2849 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2851 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2852 by Mark Daniel Reidel <mr@df.eu>.
2854 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2855 When building exim an external PCRE library is now needed -
2856 PCRE is a system library on the majority of modern systems.
2857 See entry on PCRE_LIBS in EDITME file.
2859 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2860 conversation. Added nologin parameter to request.
2861 Patch contributed by Kirill Miazine.
2863 TF/01 Do not log submission mode rewrites if they do not change the address.
2865 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2867 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2868 log files in place. Contributed by Roberto Lima.
2870 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2872 TF/03 Bugzilla 615: When checking the local_parts router precondition
2873 after a local_part_suffix or local_part_prefix option, Exim now
2874 does not use the address's named list lookup cache, since this
2875 contains cached lookups for the whole local part.
2877 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2878 Robert Millan. Documentation is in experimental-spec.txt.
2880 TF/04 Bugzilla 668: Fix parallel build (make -j).
2882 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2884 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2885 Patch provided by Jan Srzednicki.
2887 TF/05 Leading white space used to be stripped from $spam_report which
2888 wrecked the formatting. Now it is preserved.
2890 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2891 that they are available at delivery time.
2893 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2895 TF/08 TLS error reporting now respects the incoming_interface and
2896 incoming_port log selectors.
2898 TF/09 Produce a more useful error message if an SMTP transport's hosts
2899 setting expands to an empty string.
2901 NM/06 Bugzilla 744: EXPN did not work under TLS.
2902 Patch provided by Phil Pennock.
2904 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2905 Patch provided by Richard Godbee.
2907 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2908 acl_smtp_notquit, added index entry.
2910 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2911 Patch provided by Eugene Bujak.
2913 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2914 accept(). Patch provided by Maxim Dounin.
2916 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2917 Patch provided by Phil Pennock.
2919 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2921 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2922 Patch provided by Brad "anomie" Jorsch.
2924 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2925 Patch provided by Dean Brooks.
2927 TK/05 Add native DKIM support (does not depend on external libraries).
2929 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2930 Patch provided by Graeme Fowler.
2932 NM/16 Bugzilla 851: Documentation example syntax fix.
2934 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2936 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2939 NM/19 Bugzilla 745: TLS version reporting.
2940 Patch provided by Phil Pennock.
2942 NM/20 Bugzilla 167: bool: condition support.
2943 Patch provided by Phil Pennock.
2945 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2946 clients. Patch provided by Phil Pennock.
2948 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2949 Patch provided by Brad "anomie" Jorsch.
2951 NM/23 Bugzilla 687: Fix misparses in eximstats.
2952 Patch provided by Heiko Schlittermann.
2954 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2955 Patch provided by Heiko Schlittermann.
2957 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2958 plus update to original patch.
2960 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2962 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2963 Patch provided by David Brownlee.
2965 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2967 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2969 NM/30 Bugzilla 888: TLS documentation bugfixes.
2971 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2973 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2974 Unlike the original bugzilla I have changed all shell scripts in src tree.
2976 NM/33 Bugzilla 898: Transport filter timeout fix.
2977 Patch by Todd Rinaldo.
2979 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2980 Patch by Serge Demonchaux.
2982 NM/35 Bugzilla 39: Base64 decode bug fixes.
2983 Patch by Jakob Hirsch.
2985 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2987 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2989 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2991 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2997 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2998 ALT-N's libdkim that I have put here:
2999 http://duncanthrax.net/exim-experimental/
3001 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3002 were called 'true' and it seems that C99 defines that as a reserved
3003 keyword to be used with 'bool' variable types. That means you could
3004 not include C99-style headers which use bools without triggering
3005 build errors in sieve.c.
3007 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3008 as mailq or other aliases. Changed the --help handling significantly
3009 to do whats expected. exim_usage() emits usage/help information.
3011 SC/01 Added the -bylocaldomain option to eximstats.
3013 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3015 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3017 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3023 PH/01 Another patch from the Sieve maintainer.
3025 PH/02 When an IPv6 address is converted to a string for single-key lookup
3026 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3027 dots are used instead of colons so that keys in lsearch files need not
3028 contain colons. This was done some time before quoting was made available
3029 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3030 (notated using the quote facility) so as to distinguish them from IPv4
3031 keys. This meant that lookups for IP addresses in host lists did not work
3032 for iplsearch lookups.
3034 This has been fixed by arranging for IPv6 addresses to be expressed with
3035 colons if the lookup type is iplsearch. This is not incompatible, because
3036 previously such lookups could never work.
3038 The situation is now rather anomalous, since one *can* have colons in
3039 ordinary lsearch keys. However, making the change in all cases is
3040 incompatible and would probably break a number of configurations.
3042 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3045 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3046 conversion specification without a maximum field width, thereby enabling
3047 a rogue spamd server to cause a buffer overflow. While nobody in their
3048 right mind would setup Exim to query an untrusted spamd server, an
3049 attacker that gains access to a server running spamd could potentially
3050 exploit this vulnerability to run arbitrary code as the Exim user.
3052 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3053 $primary_hostname instead of what libspf2 thinks the hosts name is.
3055 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3056 a directory entry by the name of the lookup key. Previously, if a
3057 symlink pointed to a non-existing file or a file in a directory that
3058 Exim lacked permissions to read, a lookup for a key matching that
3059 symlink would fail. Now it is enough that a matching directory entry
3060 exists, symlink or not. (Bugzilla 503.)
3062 PH/03 The body_linecount and body_zerocount variables are now exported in the
3065 PH/04 Added the $dnslist_matched variable.
3067 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3068 This means they are set thereafter only if the connection becomes
3071 PH/06 Added the client_condition to authenticators so that some can be skipped
3072 by clients under certain conditions.
3074 PH/07 The error message for a badly-placed control=no_multiline_responses left
3075 "_responses" off the end of the name.
3077 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3079 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3080 (without spaces) instead of just copying the configuration text.
3082 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3084 PH/11 Added $max_received_linelength.
3086 PH/12 Added +ignore_defer and +include_defer to host lists.
3088 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3089 way in which PCRE > 7.0 is built.
3091 PH/14 Implemented queue_only_load_latch.
3093 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3094 MAIL command. The effect was to mangle the value on 64-bit systems.
3096 PH/16 Another patch from the Sieve maintainer.
3098 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3100 PH/18 If a system quota error occurred while trying to create the file for
3101 a maildir delivery, the message "Mailbox is full" was not appended to the
3102 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3103 only to a quota excession during the actual writing of the file.
3105 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3106 characters?) which causes problems in log lines. The DN values are now
3107 passed through string_printing() before being added to log lines.
3109 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3110 and InterBase are left for another time.)
3112 PH/21 Added message_body_newlines option.
3114 PH/22 Guard against possible overflow in moan_check_errorcopy().
3116 PH/23 POSIX allows open() to be a macro; guard against that.
3118 PH/24 If the recipient of an error message contained an @ in the local part
3119 (suitably quoted, of course), incorrect values were put in $domain and
3120 $local_part during the evaluation of errors_copy.
3126 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3127 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3130 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3131 issue a MAIL command.
3133 PH/02 In an ACL statement such as
3135 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3137 if a client was not listed at all, or was listed with a value other than
3138 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3139 the condition was not true (as it should be), so access was not denied.
3140 The bug was that the ! inversion was incorrectly passed on to the second
3141 item. This has been fixed.
3143 PH/03 Added additional dnslists conditions == and =& which are different from
3144 = and & when the dns lookup returns more than one IP address.
3146 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3147 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3149 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3150 FSYNC, which compiles an option called disable_fsync that allows for
3151 bypassing fsync(). The documentation is heavily laced with warnings.
3153 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3155 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3156 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3157 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3158 including adding "make clean"; (3) Added -fPIC when compiling the test
3159 dynamically loaded module, to get rid of a warning.
3161 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3162 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3163 = 0s. The bug is otherwise harmless.
3165 PH/07 There was a bug in the dovecot authenticator such that the value of
3166 $auth1 could be overwritten, and so not correctly preserved, after a
3167 successful authentication. This usually meant that the value preserved by
3168 the server_setid option was incorrect.
3170 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3172 PH/09 Installed PCRE release 7.0.
3174 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3175 run for batched SMTP input. It is now run at the start of every message
3176 in the batch. While fixing this I discovered that the process information
3177 (output by running exiwhat) was not always getting set for -bs and -bS
3178 input. This is fixed, and it now also says "batched" for BSMTP.
3180 PH/11 Added control=no_pipelining.
3182 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3183 patch, slightly modified), and move the expansion of helo_data till after
3184 the connection is made in the smtp transport (so it can use these
3187 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3189 PH/14 Added log_selector = +pid.
3191 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3193 PH/16 Add ${if forany and ${if forall.
3195 PH/17 Added dsn_from option to vary the From: line in DSNs.
3197 PH/18 Flush SMTP output before performing a callout, unless control =
3198 no_callout_flush is set.
3200 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3201 was true (the default) a successful delivery failed to delete the retry
3202 item, thus causing premature timeout of the address. The bug is now
3205 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3207 PH/21 Long custom messages for fakedefer and fakereject are now split up
3208 into multiline responses in the same way that messages for "deny" and
3209 other ACL rejections are.
3211 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3212 with slight modification.
3214 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3215 draft, changing the syntax and factoring some duplicate code".
3217 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3218 for deliveries of the second and subsequent messages over the same SMTP
3221 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3222 ${reduce, with only minor "tidies".
3224 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3226 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3227 expansion side effects.
3229 PH/27 When a message times out after an over-quota error from an Exim-imposed
3230 quota, the bounce message says "mailbox is full". This message was not
3231 being given when it was a system quota that was exceeded. It now should
3234 MH/03 Made $recipients available in local_scan(). local_scan() already has
3235 better access to the recipient list through recipients_list[], but
3236 $recipients can be useful in postmaster-provided expansion strings.
3238 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3239 in the case of a MAIL command with additional options following the
3240 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3241 were accidentally chopped off.
3243 PH/29 SMTP synchronization checks are implemented when a command is read -
3244 there is a check that no more input is waiting when there shouldn't be
3245 any. However, for some commands, a delay in an ACL can mean that it is
3246 some time before the response is written. In this time, more input might
3247 arrive, invalidly. So now there are extra checks after an ACL has run for
3248 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3249 pipelining has not been advertised.
3251 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3253 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3254 returned in the SMTP response when smtp_return_error_details was false.
3255 This has been fixed.
3257 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3258 instead of the C I/O that was originally supplied, because problems were
3259 reported on Solaris.
3261 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3262 Exim which did not show up earlier: it was assuming that a call to
3263 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3264 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3265 was a macro that became an assignment, so it seemed to work. This has
3266 changed to a proper function call with a void return, hence the compile
3267 error. Exim's code has been fixed.
3269 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3272 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3274 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3276 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3277 because it turns out to be too big or there is a timeout) there is no
3278 "Completed" line in the log. When some messages of this type were
3279 selected by exigrep, they were listed as "not completed". Others were
3280 picked up by some special patterns. I have improved the selection
3281 criteria to be more general.
3283 PH/38 The host_find_failed option in the manualroute router can now be set
3284 to "ignore", to completely ignore a host whose IP address cannot be
3285 found. If all hosts are ignored, the behaviour is controlled by the new
3286 host_all_ignored option.
3288 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3289 homing or because of multiple MX records with /mx) generated more than
3290 one IP address, and the following item turned out to be the local host,
3291 all the secondary addresses of the first item were incorrectly removed
3292 from the list, along with the local host and any following hosts (which
3293 is what is supposed to happen).
3295 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3296 whoever called Exim into the -H file. In the case of the daemon it was
3297 behaving confusingly. When first started, it used values for whoever
3298 started the daemon, but after a SIGHUP it used the Exim user (because it
3299 calls itself on a restart). I have changed the code so that it now always
3302 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3303 message are rejected with the same error (e.g. no authentication or bad
3304 sender address), and a DATA command is nevertheless sent (as can happen
3305 with PIPELINING or a stupid MUA), the error message that was given to the
3306 RCPT commands is included in the rejection of the DATA command. This is
3307 intended to be helpful for MUAs that show only the final error to their
3310 PH/42 Another patch from the Sieve maintainer.
3312 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3313 Eximstats - Fixed some broken HTML links and added missing column headers
3315 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3316 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3318 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3320 PH/43 Yet another patch from the Sieve maintainer.
3322 PH/44 I found a way to check for a TCP/IP connection going away before sending
3323 the response to the final '.' that terminates a message, but only in the
3324 case where the client has not sent further data following the '.'
3325 (unfortunately, this is allowed). However, in many cases there won't be
3326 any further data because there won't be any more messages to send. A call
3327 to select() can be used: if it shows that the input is "ready", there is
3328 either input waiting, or the socket has been closed. An attempt to read
3329 the next input character can distinguish the two cases. Previously, Exim
3330 would have sent an OK response which the client would never have see.
3331 This could lead to message repetition. This fix should cure that, at
3332 least in a lot of common cases.
3334 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3335 advertised in response to EHLO.
3341 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3342 fixed by 4.65/MH/01 (is this a record?) are fixed:
3344 (i) An empty string was always treated as zero by the numeric comparison
3345 operators. This behaviour has been restored.
3347 (ii) It is documented that the numeric comparison operators always treat
3348 their arguments as decimal numbers. This was broken in that numbers
3349 starting with 0 were being interpreted as octal.
3351 While fixing these problems I realized that there was another issue that
3352 hadn't been noticed. Values of message_size_limit (both the global option
3353 and the transport option) were treated as octal if they started with 0.
3354 The documentation was vague. These values are now always treated as
3355 decimal, and I will make that clear in the documentation.
3361 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3362 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3365 MH/01 Don't check that the operands of numeric comparison operators are
3366 integers when their expansion is in "skipping" mode (fixes bug
3367 introduced by 4.64-PH/07).
3369 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3370 child addresses, Exim now panics and dies. Previously, because the count
3371 is held in a short int, deliveries were likely to be lost. As such a
3372 large number of recipients for a single message is ridiculous
3373 (performance will be very, very poor), I have chosen to impose a limit
3374 rather than extend the field.
3380 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3381 leftover -K file (the existence of which was triggered by #402).
3382 While we were at it, introduced process PID as part of the -K
3383 filename. This should rule out race conditions when creating
3386 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3387 processing considerably. Previous code took too long for large mails,
3388 triggering a timeout which in turn triggers #401.
3390 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3391 in the DK code in transports.c. sendfile() is not really portable,
3392 hence the _LINUX specificness.
3394 TF/01 In the add_headers option to the mail command in an Exim filter,
3395 there was a bug that Exim would claim a syntax error in any
3396 header after the first one which had an odd number of characters
3399 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3400 callout verification, Exim cached a "reject" for the entire domain. This
3401 is correct for most verifications, but it is not correct for a recipient
3402 verification with use_sender or use_postmaster set, because in that case
3403 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3404 case of MAIL FROM:<> rejection from other early rejections (e.g.
3405 rejection of HELO). When verifying a recipient using a non-null MAIL
3406 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3407 Whatever the result of the callout, the value of the domain cache is
3408 left unchanged (for any other kind of callout, getting as far as trying
3409 RCPT means that the domain itself is ok).
3411 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3414 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3415 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3416 session, the daemon ignores ECONNECT errors and logs others; it now
3417 ignores EPIPE as well.
3419 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3420 (quoted-printable decoding).
3422 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3423 later the small subsequent patch to fix an introduced bug.
3425 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3427 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3429 PH/08 An error is now given if message_size_limit is specified negative.
3431 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3432 to be given (somewhat) arbitrary names.
3434 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3437 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3438 miscellaneous code fixes
3440 PH/10 Added the log_reject_target ACL modifier to specify where to log
3443 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3444 hostname. This is wrong, because it relates to the incoming message (and
3445 probably the interface on which it is arriving) and not to the outgoing
3446 callout (which could be using a different interface). This has been
3447 changed to use the value of the helo_data option from the smtp transport
3448 instead - this is what is used when a message is actually being sent. If
3449 there is no remote transport (possible with a router that sets up host
3450 addresses), $smtp_active_hostname is used.
3452 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3453 tweaks were necessary in order to get it to work (see also 21 below):
3454 (a) The code assumed that strncpy() returns a negative number on buffer
3455 overflow, which isn't the case. Replaced with Exim's string_format()
3457 (b) There were several signed/unsigned issues. I just did the minimum
3458 hacking in of casts. There is scope for a larger refactoring.
3459 (c) The code used strcasecmp() which is not a standard C function.
3460 Replaced with Exim's strcmpic() function.
3461 (d) The code set only $1; it now sets $auth1 as well.
3462 (e) A simple test gave the error "authentication client didn't specify
3463 service in request". It would seem that Dovecot has changed its
3464 interface. Fortunately there's a specification; I followed it and
3465 changed what the client sends and it appears to be working now.
3467 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3470 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3471 address A is aliased to B and C, where B exists and C does not. Without
3472 -v the output is "A verified" because verification stops after a
3473 successful redirection if more than one address is generated. However,
3474 with -v the child addresses are also verified. Exim was outputting "A
3475 failed to verify" and then showing the successful verification for C,
3476 with its parentage. It now outputs "B failed to verify", showing B's
3477 parentage before showing the successful verification of C.
3479 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3480 look up a TXT record in a specific list after matching in a combined
3483 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3484 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3485 they consult the DNS. I had assumed they would set it the way they
3486 wanted; and indeed my experiments on Linux seem to show that in some
3487 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3488 To be on the safe side, however, I have now made the interface to
3489 host_find_byname() similar to host_find_bydns(), with an argument
3490 containing the DNS resolver options. The host_find_byname() function now
3491 sets these options at its start, just as host_find_bydns() does. The smtp
3492 transport options dns_qualify_single and dns_search_parents are passed to
3493 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3494 of host_find_byname() use the default settings of RES_DEFNAMES
3495 (qualify_single) but not RES_DNSRCH (search_parents).
3497 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3498 spool_read_header() do less string testing, by means of a preliminary
3499 switch on the second character of optional "-foo" lines. (This is
3500 overdue, caused by the large number of possibilities that now exist.
3501 Originally there were few.) While I was there, I also converted the
3502 str(n)cmp tests so they don't re-test the leading "-" and the first
3503 character, in the hope this might squeeze out yet more improvement.
3505 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3506 flag allowing group syntax was set by the header_syntax check but not
3507 turned off, possible causing trouble later; (2) The flag was not being
3508 set at all for the header_verify test, causing "group"-style headers to
3509 be rejected. I have now set it in this case, and also caused header_
3510 verify to ignore an empty address taken from a group. While doing this, I
3511 came across some other cases where the code for allowing group syntax
3512 while scanning a header line wasn't quite right (mostly, not resetting
3513 the flag correctly in the right place). These bugs could have caused
3514 trouble for malformed header lines. I hope it is now all correct.
3516 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3517 with the "reply" argument non-NULL. The code, however (which originally
3518 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3519 but it didn't always do it. This confused somebody who was copying the
3520 code for some other use. I have removed all the tests.
3522 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3523 feature that was used to support insecure browsers during the U.S. crypto
3524 embargo. It requires special client support, and Exim is probably the
3525 only MTA that supported it -- and would never use it because real RSA is
3526 always available. This code has been removed, because it had the bad
3527 effect of slowing Exim down by computing (never used) parameters for the
3528 RSA_EXPORT functionality.
3530 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3531 authenticator to fail if there's a tab character in the incoming data
3532 (there should never be unless someone is messing about, as it's supposed
3533 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3534 if the connection is using TLS or if the remote IP is the same as the
3535 local IP, and the "valid-client-cert option" if a client certificate has
3538 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3539 authenticators. This can be used for authorization after authentication
3540 succeeds. (In the case of plaintext, it servers for both authentication
3543 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3544 if any retry times were supplied.
3546 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3547 connection, where there is no client IP address to check. In this
3548 situation, the verify now always succeeds.
3550 PH/25 Applied John Jetmore's -Mset patch.
3552 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3554 PH/27 In a string expansion for a processed (not raw) header when multiple
3555 headers of the same name were present, leading whitespace was being
3556 removed from all of them, but trailing whitespace was being removed only
3557 from the last one. Now trailing whitespace is removed from each header
3558 before concatenation. Completely empty headers in a concatenation (as
3559 before) are ignored.
3561 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3562 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3564 PH/29 [Removed. This was a change that I later backed out, and forgot to
3565 correct the ChangeLog entry (that I had efficiently created) before
3566 committing the later change.]
3568 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3569 address errors (4xx response to RCPT) over the same connection as other
3570 messages routed to the same hosts. Such deliveries are always "forced",
3571 so retry times are not inspected. This resulted in far too many retries
3572 for the affected addresses. The effect occurred only when there were more
3573 hosts than the hosts_max_try setting in the smtp transport when it had
3574 the 4xx errors. Those hosts that it had tried were not added to the list
3575 of hosts for which the message was waiting, so if all were tried, there
3576 was no problem. Two fixes have been applied:
3578 (i) If there are any address or message errors in an SMTP delivery, none
3579 of the hosts (tried or untried) are now added to the list of hosts
3580 for which the message is waiting, so the message should not be a
3581 candidate for sending over the same connection that was used for a
3582 successful delivery of some other message. This seems entirely
3583 reasonable: after all the message is NOT "waiting for some host".
3584 This is so "obvious" that I'm not sure why it wasn't done
3585 previously. Hope I haven't missed anything, but it can't do any
3586 harm, as the worst effect is to miss an optimization.
3588 (ii) If, despite (i), such a delivery is accidentally attempted, the
3589 routing retry time is respected, so at least it doesn't keep
3590 hammering the server.
3592 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3593 in ${readsocket because some servers need this prod.
3595 PH/32 Added some extra debug output when updating a wait-xxx database.
3597 PH/33 The hint "could be header name not terminated by colon", which has been
3598 given for certain expansion errors for a long time, was not being given
3599 for the ${if def:h_colon_omitted{... case.
3601 PH/34 The spec says: "With one important exception, whenever a domain list is
3602 being scanned, $domain contains the subject domain." There was at least
3603 one case where this was not true.
3605 PH/35 The error "getsockname() failed: connection reset by peer" was being
3606 written to the panic log as well as the main log, but it isn't really
3607 panic-worthy as it just means the connection died rather early on. I have
3608 removed the panic log writing for the ECONNRESET error when getsockname()
3611 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3612 runs only) independently of the message's sender address. This meant
3613 that, if the 4xx error was in fact related to the sender, a different
3614 message to the same recipient with a different sender could confuse
3615 things. In particular, this can happen when sending to a greylisting
3616 server, but other circumstances could also provoke similar problems.
3617 I have changed the default so that the retry time for these errors is now
3618 based a combination of the sender and recipient addresses. This change
3619 can be overridden by setting address_retry_include_sender=false in the
3622 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3623 remote server are returned as part of bounce messages. This was not
3624 happening for LMTP over a pipe (the lmtp transport), but now it is the
3625 same for both kinds of LMTP.
3627 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3628 in header lines that were in fact CNAMEs. This is no longer the case.
3630 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3631 and queue runs started by the daemon processed all messages. This has
3632 been fixed so that -R and -S can now usefully be given with -q<time>.
3634 PH/40 Import PCRE release 6.7 (fixes some bugs).
3636 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3638 PH/42 Give an error if -q is specified more than once.
3640 PH/43 Renamed the variables $interface_address and $interface_port as
3641 $received_ip_address and $received_port, to make it clear that these
3642 values apply to message reception, and not to the outgoing interface when
3643 a message is delivered. (The old names remain recognized, of course.)
3645 PH/44 There was no timeout on the connect() call when using a Unix domain
3646 socket in the ${readsocket expansion. There now is.
3648 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3649 be meaningful with "accept".
3651 SC/01 Eximstats V1.43
3652 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3654 SC/02 Eximstats V1.44
3655 Use a glob alias rather than an array ref in the generated
3656 parser. This improves both readability and performance.
3658 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3659 Collect SpamAssassin and rejection statistics.
3660 Don't display local sender or destination tables unless
3661 there is data to show.
3662 Added average volumes into the top table text output.
3664 SC/04 Eximstats V1.46
3665 Collect data on the number of addresses (recipients)
3666 as well as the number of messages.
3668 SC/05 Eximstats V1.47
3669 Added 'Message too big' to the list of mail rejection
3670 reasons (thanks to Marco Gaiarin).
3672 SC/06 Eximstats V1.48
3673 Mainlog lines which have GMT offsets and are too short to
3674 have a flag are now skipped.
3676 SC/07 Eximstats V1.49 (Alain Williams)
3677 Added the -emptyok flag.
3679 SC/08 Eximstats V1.50
3680 Fixes for obtaining the IP address from reject messages.
3682 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3683 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3684 whitespace changes from 4.64-PH/27
3686 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3689 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3690 are found, allow negative numbers in numeric criteria)
3692 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3694 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3695 to match changes made in 4.64-PH/43
3697 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3699 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3700 "domain" is actually an IP address, and force a failure. This locks out
3701 those revolvers/nameservers that support "A-for-A" lookups, in
3702 contravention of the specifications.
3704 PH/48 When a host name was looked up from an IP address, and the subsequent
3705 forward lookup of the name timed out, the host name was left in
3706 $sender_host_name, contrary to the specification.
3708 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3709 restricted to single-key lookups, Exim was not diagnosing an error if
3710 * or *@ was used with a query-style lookup.
3712 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3714 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3715 long ago, but noone interested enough thought of it. Let's just say that
3716 the "1.1" means that there are some new functions that weren't there at
3717 some point in the past.
3719 PH/51 Error processing for expansion failure of helo_data from an smtp
3720 transport during callout processing was broken.
3722 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3723 tested/used via the -bh/-bhc/-bs options.
3725 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3726 bug, fixed in subsequent PCRE releases).
3728 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3729 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3731 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3737 SC/01 Use a glob alias rather than an array ref in eximstats generated
3738 parser. This improves both readability and performance.
3740 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3741 Don't display local sender or destination tables in eximstats unless
3742 there is data to show.
3743 Added average volumes into the eximstats top table text output.
3745 SC/03 Collect data on the number of addresses (recipients) as well
3746 as the number of messages in eximstats.
3748 TF/01 Correct an error in the documentation for the redirect router. Exim
3749 does (usually) call initgroups() when daemonizing.
3751 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3752 with consistent privilege compared to when running as a daemon.
3754 TF/03 Note in the spec that $authenticated_id is not set for local
3755 submissions from trusted users.
3757 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3758 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3760 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3761 by adding some example configuration directives to the default
3762 configuration file. A little bit of work is required to uncomment the
3763 directives and define how usernames and passwords are checked, but
3764 there is now a framework to start from.
3766 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3767 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3768 without this. I don't know how relevant this is to other LDAP libraries.
3770 PH/02 Add the verb name to the "unknown ACL verb" error.
3772 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3774 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3776 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3777 directory not expanded when it should be if an expanded home directory
3778 was set for the address (which is overridden by the transport).
3780 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3783 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3784 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3785 because it is too late at that time, and has no effect.
3787 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3788 security issue with \' (bugzilla #107). I could not use the
3789 PQescapeStringConn() function, because it needs a PGconn value as one of
3792 PH/08 When testing addresses using -bt, indicate those final addresses that
3793 are duplicates that would not cause an additional delivery. At least one
3794 person was confused, thinking that -bt output corresponded to deliveries.
3795 (Suppressing duplicates isn't a good idea as you lose the information
3796 about possibly different redirections that led to the duplicates.)
3798 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3799 systems where poll() doesn't work, in particular OS X.
3801 PH/10 Added more information to debugging output for retry time not reached.
3803 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3804 operations in malware.c.
3806 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3809 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3810 syslog_duplication set false, log lines that would normally be written
3811 both the the main log and to the reject log were not written to syslog at
3814 PH/14 In the default configuration, change the use of "message" in ACL warn
3815 statements to "add_header".
3817 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3818 not followed by a command (e.g. "seen endif").
3820 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3821 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3824 PH/17 Added extra conditions to the default value of delay_warning_condition
3828 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3829 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3830 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3833 The Auto-Submitted: and various List- headers are standardised, whereas I
3834 don't think Precedence: ever was.
3836 PH/18 Refactored debugging code in route_finduser() to show more information,
3837 in particular, the error code if getpwnam() issues one.
3839 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3840 This is apparently needed in addition to the PH/07 change above to avoid
3841 any possible encoding problems.
3843 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3844 but not after initializing Perl.
3846 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3847 output them only if debugging. By default they are written stderr,
3848 apparently, which is not desirable.
3850 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3853 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3856 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3858 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3859 authenticated or an ident call has been made. Suppress the default
3860 values for $authenticated_id and $authenticated_sender (but permit -oMai
3861 and -oMas) when testing with -bh.
3863 PH/24 Re-jigged the order of the tests in the default configuration so that the
3864 tests for valid domains and recipients precede the DNS black list and CSA
3865 tests, on the grounds that those ones are more expensive.
3867 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3868 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3869 This bug exists in every version of Exim that I still have, right back to
3872 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3873 However, an attempt to turn on case-sensitivity in a regex key by
3874 including (?-i) didn't work because the subject string was already
3875 lowercased, and the effects were non-intuitive. It turns out that a
3876 one-line patch can be used to allow (?-i) to work as expected.
3882 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3883 other effects) broke the use of negated acl sub-conditions.
3885 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3888 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3889 "Deny" causes Exim to reject the incoming connection with a 554 error.
3890 Unfortunately, if there is a major crisis, such as a disk failure,
3891 tcp-wrappers gives "deny", whereas what one would like would be some
3892 kind of temporary error. A kludge has been added to help with this.
3893 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3894 554 error is used if errno is still zero or contains ENOENT (which occurs
3895 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3898 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3900 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3901 errors. Otherwise a message that provokes a temporary error (when other
3902 messages do not) can cause a whole host to time out.
3904 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3905 addresses were routed directly to files or pipes from a redirect router.
3906 File deliveries just didn't batch; pipe deliveries might have suffered
3909 PH/06 A failure to get a lock for a hints database would erroneously always say
3910 "Failed to get write lock", even when it was really a read lock.
3912 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3913 of 0600. This has been changed to use the value of the lockfile_mode
3914 option (which defaults to 0600).
3916 PH/08 Applied small patch from the Sieve maintainer.
3918 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3919 folder from quota calculations, a direct delivery into this folder messed
3920 up the contents of the maildirsize file. This was because the regex was
3921 used only to exclude .Trash (or whatever) when the size of the mailbox
3922 was calculated. There was no check that a delivery was happening into an
3923 excluded directory. This bug has been fixed by ignoring all quota
3924 processing for deliveries into excluded directories.
3926 PH/10 Added the maildirfolder_create_regex option to appendfile.
3932 PH/01 The code for finding all the local interface addresses on a FreeBSD
3933 system running IPv6 was broken. This may well have applied to all BSD
3934 systems, as well as to others that have similar system calls. The broken
3935 code found IPv4 interfaces correctly, but gave incorrect values for the
3936 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3937 that it would not match correctly against @[] and not recognize the IPv6
3940 PH/02 The ipliteral router was not recognizing addresses of the form user@
3941 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3943 PH/03 Added disable_ipv6.
3945 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3946 decoded form, because it is most often used to construct To: headers
3947 lines in autoreplies, and the decoded form may well be syntactically
3948 invalid. However, $reply_address has leading white space removed, and all
3949 newlines turned into spaces so that the autoreply transport does not
3952 PH/05 If group was specified without a user on a router, and no group or user
3953 was specified on a transport, the group from the router was ignored.
3955 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3956 for visible compile-time settings that can be used to change these
3957 numbers, for those that want even more. Backwards compatibility with old
3958 spool files has been maintained. However, going back to a previous Exim
3959 release will lost any variables that are in spool files.
3961 PH/07 Two small changes when running in the test harness: increase delay when
3962 passing a TCP/IP connection to a new process, in case the original
3963 process has to generate a bounce, and remove special handling of
3964 127.0.0.2 (sic), which is no longer necessary.
3966 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3967 be the same on different OS.
3969 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3972 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3973 whether --show-vars was specified or not
3975 JJ/02 exipick: Added support for new ACL variable spool format introduced
3978 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3979 syntactically invalid From: or Reply-to: line, and a filter used this to
3980 generate an autoreply, and therefore failed to obtain an address for the
3981 autoreply, Exim could try to deliver to a non-existent relative file
3982 name, causing unrelated and misleading errors. What now happens is that
3983 it logs this as a hard delivery error, but does not attempt to create a
3986 PH/11 The exinext utility has a -C option for testing purposes, but although
3987 the given file was scanned by exinext itself; it wasn't being passed on
3988 when Exim was called.
3990 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3991 an end-of-file indication when reading a command response.
3993 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3994 compiled. In many other places in Exim, IPv6 addresses are always
3995 recognized, so I have changed this. It also means that IPv4 domain
3996 literals of the form [IPV4:n.n.n.n] are now always recognized.
3998 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3999 used if the router is not running as root, for example, when verifying at
4000 ACL time, or when using -bh. The debugging output from this situation was
4001 non-existent - all you got was a failure to exec. I have made two
4004 (a) Failures to set uid/gid, the current directory, or a process leader
4005 in a subprocess such as that created by queryprogram now generate
4006 suitable debugging output when -d is set.
4008 (b) The queryprogram router detects when it is not running as root,
4009 outputs suitable debugging information if -d is set, and then runs
4010 the subprocess without attempting to change uid/gid.
4012 PH/15 Minor change to Makefile for building test_host (undocumented testing
4015 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4016 additional section of a DNS packet that returns MX or SRV records.
4017 Instead, it always explicitly searches for A/AAAA records. This avoids
4018 major problems that occur when a DNS server includes only records of one
4019 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4020 fixed another bug: if SRV records were looked up and the corresponding
4021 address records were *not* found in the additional section, the port
4022 values from the SRV records were lost.
4024 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4025 using the correct key (the original address) when searching the retry
4026 rules in order to find which one to use for generating the retry hint.
4028 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4029 adding the default one. Similarly, if it contains a Reply-To: header, the
4030 errors_reply_to option, if set, is not used.
4032 PH/19 When calculating a retry time, Exim used to measure the "time since
4033 failure" by looking at the "first failed" field in the retry record. Now
4034 it does not use this if it is later than than the arrival time of the
4035 message. Instead it uses the arrival time. This makes for better
4036 behaviour in cases where some deliveries succeed, thus re-setting the
4037 "first failed" field. An example is a quota failure for a huge message
4038 when small messages continue to be delivered. Without this change, the
4039 "time since failure" will always be short, possible causing more frequent
4040 delivery attempts for the huge message than are intended.
4041 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4043 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4044 $1, $2, $3) because the numerical variables can be reset during some
4045 expansion items (e.g. "match"), thereby losing the authentication data.
4047 PH/21 Make -bV show the size of off_t variables so that the test suite can
4048 decide whether to run tests for quotas > 2G.
4050 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4051 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4052 filecount value is greater than 2G or if a quota value is greater than 2G
4053 on a system where the size of off_t is not greater than 4, a panic error
4056 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4057 never match. The debug and -bh output now contains an explicit error
4058 message indicating a malformed IPv4 address or mask.
4060 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4061 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4062 PH/23 above applies.
4064 PH/25 Do not write to syslog when running in the test harness. The only
4065 occasion when this arises is a failure to open the main or panic logs
4066 (for which there is an explicit test).
4068 PH/26 Added the /no_tell option to "control=freeze".
4070 PH/27 If a host name lookup failed very early in a connection, for example, if
4071 the IP address matched host_lookup and the reverse lookup yielded a name
4072 that did not have a forward lookup, an error message of the form "no IP
4073 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4074 could be logged. Now it outputs the IP address instead of "NULL".
4076 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4077 allows the sender and the authenticated sender to be set when
4078 submitting a message from within Exim. Since child_open_exim() is
4079 documented for local_scan(), the new function should be too.
4081 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4082 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4083 results in an empty string is now treated as unset.
4085 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4087 PH/31 Added sender_verify_fail logging option.
4089 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4090 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4091 tidied the source and removed it altogether.
4093 PH/33 When a queue run was abandoned because the load average was too high, a
4094 log line was always written; now it is written only if the queue_run log
4095 selector is set. In addition, the log line for abandonment now contains
4096 information about the queue run such as the pid. This is always present
4097 in "start" and "stop" lines but was omitted from the "abandon" line.
4099 PH/34 Omit spaces between a header name and the colon in the error message that
4100 is given when verify = headers_syntax fails (if there are lots of them,
4101 the message gets confusing).
4103 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4104 names, as there are now some PTR records that contain slashes. This check
4105 is only to protect against broken name servers that fall over on strange
4106 characters, so the fact that it applies to all lookups doesn't matter.
4108 PH/36 Now that the new test suite is complete, we can remove some of the
4109 special code in Exim that was needed for the old test suite. For example,
4110 sorting DNS records because real resolvers return them in an arbitrary
4111 order. The new test suite's fake resolver always returns records in the
4114 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4115 bounces) except when queue_only is set, to avoid logging races between
4116 the different processes.
4118 PH/38 Panic-die if .include specifies a non-absolute path.
4120 PH/39 A tweak to the "H" retry rule from its user.
4122 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4123 a label. They prevented compilation on older perls.
4125 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4126 a warning to be raised on newish perls.
4128 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4129 on queue. Changes to match documented behaviour of showing count of
4130 messages matching specified criteria.
4132 PH/40 Changed the default ident timeout from 30s to 5s.
4134 PH/41 Added support for the use of login_cap features, on those BSD systems
4135 that have them, for controlling the resources used by pipe deliveries.
4137 PH/42 The content-scanning code uses fopen() to create files in which to put
4138 message data. Previously it was not paying any attention to the mode of
4139 the files. Exim runs with umask(0) because the rest of the code creates
4140 files with open(), and sets the required mode explicitly. Thus, these
4141 files were ending up world-writeable. This was not a big issue, because,
4142 being within the spool directory, they were not world-accessible. I have
4143 created a function called modefopen, which takes an additional mode
4144 argument. It sets umask(777), creates the file, chmods it to the required
4145 mode, then resets the umask. All the relevant calls to fopen() in the
4146 content scanning code have been changed to use this function.
4148 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4149 to 24 hours. This avoids potential overflow problems when processing G
4150 and H retry rules. I suspect nobody ever tinkers with this value.
4152 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4154 PH/45 When the plaintext authenticator is running as a client, the server's
4155 challenges are checked to ensure they are valid base64 strings. By
4156 default, the authentication attempt is cancelled if an invalid string is
4157 received. Setting client_ignore_invalid_base64 true ignores these errors.
4158 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4159 they are received. Thus, the responses can be made to depend on the
4160 challenges. If an invalid string is ignored, an empty string is placed in
4163 PH/46 Messages that are created by the autoreply transport now contains a
4164 References: header, in accordance with RFCs 2822 and 3834.
4166 PH/47 Added authenticated_sender_force to the smtp transport.
4168 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4170 PH/49 Installed latest patch from the Sieve maintainer.
4172 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4173 was also set, the appendfile transport was unnecessarily scanning a
4174 directory of message files (e.g. for maildir delivery) to find the count
4175 of files (along with the size), even though it did not need this
4176 information. It now does the scan only if it needs to find either the
4177 size of the count of files.
4179 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4181 PH/52 Two bugs concerned with error handling when the smtp transport is
4184 (i) Exim was not creating retry information for temporary errors given
4185 for individual recipients after the DATA command when the smtp transport
4186 was used in LMTP mode. This meant that they could be retried too
4187 frequently, and not timed out correctly.
4189 (ii) Exim was setting the flag that allows error details to be returned
4190 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4191 recipients that were returned after the DATA command.
4193 PH/53 This is related to PH/52, but is more general: for any failing address,
4194 when detailed error information was permitted to be returned to the
4195 sender, but the error was temporary, then after the final timeout, only
4196 "retry timeout exceeded" was returned. Now it returns the full error as
4197 well as "retry timeout exceeded".
4199 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4200 do this, and (what is worse) MTAs that accept it.
4202 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4203 will now be deprecated.
4205 PH/56 New os.c-cygwin from the Cygwin maintainer.
4207 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4208 formats (previously only available in exim formats via -bpr, -bpru,
4209 and -bpra. Now also available in native and exiqgrep formats)
4211 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4212 with very large, slow to parse queues
4214 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4216 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4218 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4219 responses to authentication challenges, though it was showing the
4220 challenges; (ii) I've removed the CR characters from the debug output for
4223 PH/58 Allow for the insertion of a newline as well as a space when a string
4224 is turned into more than one encoded-word during RFC 2047 encoding. The
4225 Sieve code now uses this.
4227 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4228 data_4xx, lost_connection, tls_required.
4230 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4231 message was being sent as an SMTP response.
4233 PH/61 Add -l and -k options to exicyclog.
4235 PH/62 When verifying, if an address was redirected to one new address, so that
4236 verification continued, and the new address failed or deferred after
4237 having set something in $address_data, the value of $address_data was not
4238 passed back to the ACL. This was different to the case when no
4239 redirection occurred. The value is now passed back in both cases.
4241 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4242 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4243 use login_cap.h, so on its own it isn't the distinguishing feature. The
4244 new name refers directly to the setclassresources() function.
4246 PH/65 Added configuration files for NetBSD3.
4248 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4250 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4251 is preferred over IPv4.
4253 PH/68 The bounce_return_message and bounce_return_body options were not being
4254 honoured for bounces generated during the reception of non-SMTP messages.
4255 In particular, this applied to messages rejected by the ACL. This bug has
4256 been fixed. However, if bounce_return_message is true and bounce_return_
4257 body is false, the headers that are returned for a non-SMTP message
4258 include only those that have been read before the error was detected.
4259 (In the case of an ACL rejection, they have all been read.)
4261 PH/69 The HTML version of the specification is now built in a directory called
4262 spec_html instead of spec.html, because the latter looks like a path with
4263 a MIME-type, and this confuses some software.
4265 PH/70 Catch two compiler warnings in sieve.c.
4267 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4268 function verify_get_ident() calls ip_connect() to connect a socket, but
4269 if the "connect()" function timed out, ip_connect() used to close the
4270 socket. However, verify_get_ident() also closes the socket later, and in
4271 between Exim writes to the log, which may get opened at this point. When
4272 the socket was closed in ip_connect(), the log could get the same file
4273 descriptor number as the socket. This naturally causes chaos. The fix is
4274 not to close the socket in ip_connect(); the socket should be closed by
4275 the function that creates it. There was only one place in the code where
4276 this was missing, in the iplookup router, which I don't think anybody now
4277 uses, but I've fixed it anyway.
4279 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4280 well as to direct DNS lookups. Otherwise the handling of names in host
4281 lists is inconsistent and therefore confusing.
4287 PH/01 Two changes to the default runtime configuration:
4289 (1) Move the checks for relay_from_hosts and authenticated clients from
4290 after to before the (commented out) DNS black list checks.
4292 (2) Add control=submission to the relay_from_hosts and authenticated
4293 clients checks, on the grounds that messages accepted by these
4294 statements are most likely to be submissions.
4296 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4298 (1) Generate an error if the third argument for the ${prvs expansion is
4301 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4304 (3) Reset the variables that are obtained from the first argument of
4305 ${prvscheck and used in the second argument before leaving the code,
4306 because their memory is reclaimed, so using them afterwards may do
4309 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4310 one (it's much easier than Tom thought :-).
4312 (5) Because of (4), we can now allow for the use of $prvscheck_result
4313 inside the third argument.
4315 PH/03 For some reason, the default setting of PATH when running a command from
4316 a pipe transport was just "/usr/bin". I have changed it to
4319 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4320 anything to be listed in the output from -bV.
4322 PH/05 When a filter generated an autoreply, the entire To: header line was
4323 quoted in the delivery log line, like this:
4325 => >A.N.Other <ano@some.domain> <original@ddress> ...
4327 This has been changed so that it extracts the operative address. There
4328 may be more than one such address. If so, they are comma-separated, like
4331 => >ano@some.domain,ona@other.domain <original@ddress> ...
4333 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4334 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4335 not being looked up in the rDNS to get a host name, Exim was showing the
4336 IP address twice in Received: lines, even though the IP addresses were
4337 identical. For example:
4339 Received: from [1.2.3.4] (helo=[1.2.3.4])
4341 However, if the real host name was known, it was omitting the HELO data
4342 if it matched the actual IP address. This has been tidied up so that it
4343 doesn't show the same IP address twice.
4345 PH/07 When both +timestamp and +memory debugging was on, the value given by
4346 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4347 called by the debug printing, thereby overwriting the timestamp buffer.
4348 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4350 PH/08 When the original message was included in an autoreply transport, it
4351 always said "this is a copy of the message, including all the headers",
4352 even if body_only or headers_only was set. It now gives an appropriate
4355 PH/09 Applied a patch from the Sieve maintainer which:
4357 o fixes some comments
4358 o adds the (disabled) notify extension core
4359 o adds some debug output for the result of if/elsif tests
4360 o points to the current vacation draft in the documentation
4361 and documents the missing references header update
4365 o fixes a bug in processing the envelope test (when testing
4366 multiple envelope elements, the last element determined the
4369 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4370 Electronic Mail") by including:
4372 Auto-submitted: auto-generated
4374 in the messages that it generates (bounce messages and others, such as
4375 warnings). In the case of bounce messages for non-SMTP messages, there was
4376 also a typo: it was using "Auto_submitted" (underscore instead of
4377 hyphen). Since every message generated by Exim is necessarily in response
4378 to another message, thes have all been changed to:
4380 Auto-Submitted: auto-replied
4382 in accordance with these statements in the RFC:
4384 The auto-replied keyword:
4386 - SHOULD be used on messages sent in direct response to another
4387 message by an automatic process,
4389 - MUST NOT be used on manually-generated messages,
4391 - MAY be used on Delivery Status Notifications (DSNs) and Message
4392 Disposition Notifications (MDNs),
4394 - MUST NOT be used on messages generated by automatic or periodic
4395 processes, except for messages which are automatic responses to
4398 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4399 to the default Received: header definition.
4401 PH/12 Added log selector acl_warn_skipped (default on).
4403 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4404 variables because (a) they are in the wrong storage pool and (b) even if
4405 they were copied, it wouldn't work properly because of the caching.
4407 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4408 checking when decoding. Apparently there are clients that generate
4409 overlong encoded strings. Why am I not surprised?
4411 PH/15 If the first argument of "${if match_address" was not empty, but did not
4412 contain an "@" character, Exim crashed. Now it writes a panic log message
4413 and treats the condition as false.
4415 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4417 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4418 extension "envelope-auth". The code is finished and in agreement with
4419 other implementations, but there is no documentation so far and in fact,
4420 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4421 not changing the active code.
4423 Print executed "if" and "elsif" statements when debugging is used. This
4424 helps a great deal to understand what a filter does.
4426 Document more things not specified clearly in RFC3028. I had all this
4427 sorted out, when out of a sudden new issues came to my mind. Oops."
4429 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4432 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4433 canonical form (as documented). However, after a host name lookup from
4434 the IP address, check_host() was doing a simple string comparison with
4435 addresses acquired from the DNS when checking that the found name did
4436 have the original IP as one of its addresses. Since any found IPv6
4437 addresses are likely to be in abbreviated form, the comparison could
4438 fail. Luckily, there already exists a function for doing the comparison
4439 by converting both addresses to binary, so now that is used instead of
4440 the text comparison.
4442 PH/20 There was another similar case to PH/19, when a complete host name was
4443 given in a host list; looking up its IP address could give an abbreviated
4444 form, whereas the current host's name might or might not be abbreviated.
4445 The same fix has been applied.
4451 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4452 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4455 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4456 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4458 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4460 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4461 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4462 Also, for "Auto-Submitted", treat anything other than "no" as
4463 non-personal, in accordance with RFC 3834. (Previously it treated
4464 anything starting "auto-" as non-personal.)
4466 TF/01 The control=submission/name=... option had a problem with syntax
4467 errors if the name included a slash character. The /name= option
4468 now slurps the rest of the string, so it can include any characters
4469 but it must come last in the list of options (after /sender_retain
4472 PH/05 Some modifications to the interface to the fake nameserver for the new
4480 TK/01 Added the "success_on_redirect" address verification option. See
4481 NewStuff for rationale and an example.
4483 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4485 PH/02 Patch to exigrep to allow it to work on syslog lines.
4487 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4488 fread() to skip over the body file's header line, because in Cygwin the
4489 header line is locked and is inaccessible.
4491 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4492 co-exist for some time) to make it clear that it is the Exim ID that is
4493 referenced, not the Message-ID: header line.
4495 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4496 string_format() function, because snprintf() does not exist on all
4499 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4500 ${stat: expansion item.
4502 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4503 protocol synchronization error", to keep the pedants happy.
4505 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4506 well as for IRIX systems, when gcc is being used. See the host.c source
4509 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4511 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4514 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4515 transport and to the smtp transport in LMTP mode.
4517 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4519 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4520 run a filter in a subprocess. This could lead to confusion in subsequent
4521 lookups in the parent process. There should also be a search_tidyup() at
4522 the end of the subprocess.
4524 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4525 only if the host matched helo_try_verify_hosts, which caused the
4526 verification to occur when the EHLO/HELO command was issued. The ACL just
4527 tested the remembered result. Now, if a previous verification attempt has
4528 not happened, "verify = helo" does it there and then.
4530 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4532 TK/03 Fix log output including CR from clamd.
4534 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4535 exist provoked a memory error which could cause a segfault.
4537 PH/15 Installed PCRE 6.2
4539 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4541 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4542 of the problem. Specifically, suggested +O2 rather than +O1 for the
4545 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4547 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4548 record, the port was not correctly logged when the outgoing_port log
4549 selector was set (it logged the transort's default port).
4551 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4552 fallback_hosts, and "hosts" in the smtp transport.
4554 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4555 host errors such as "Connection refused".
4557 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4558 authentication with radiusclient 0.4.9:
4560 - Error returned from rc_read_config was caught wrongly
4561 - Username/password not passed on to radius server due to wrong length.
4563 The presumption is that some radiusclient API changes for 4.51/PH/17
4564 were not taken care of correctly. The code is still untested by me (my
4565 Linux distribution still has 0.3.2 of radiusclient), but it was
4566 contributed by a Radius user.
4568 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4569 expanding the "port" option of the smtp transport.
4571 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4572 while reading a MIME header. Thanks to Tom Hughes for a patch.
4574 PH/24 Include config.h inside local_scan.h so that configuration settings are
4577 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4578 that in an ACL for RCPT (for example), you can examine exactly what was
4581 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4582 commands, but it was not correctly comparing the address with the actual
4583 client host address. Thus, it would show the EHLO address in Received:
4584 header lines when this was not necessary.
4586 PH/27 Added the % operator to ${eval:}.
4588 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4589 it should be ignoring failures (because with -C, for example, it has lost
4590 privilege). It wasn't ignoring creation failures other than "already
4593 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4596 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4597 because an input line was too long, either on its own, or by virtue of
4598 too many continuations, the temporary file was not being removed, and the
4599 return code was incorrect.
4601 PH/31 Missing "BOOL" in function definition in filtertest.c.
4603 PH/32 Applied Sieve patches from the maintainer.
4605 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4607 PH/33 Added "verify = not_blind".
4609 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4610 Local/Makefile (with some defaults set). These are used in built scripts
4611 such as exicyclog, but they have never been used in the exim_install
4612 script (though there are many overriding facilities there). I have
4613 arranged that the exim_install script now takes note of these two
4616 PH/35 Installed configuration files for Dragonfly.
4618 PH/36 When a locally submitted message by a trusted user did not contain a
4619 From: header, and the sender address was obtained from -f or from an SMTP
4620 MAIL command, and the trusted user did not use -F to supply a sender
4621 name, $originator_name was incorrectly used when constructing a From:
4622 header. Furthermore, $originator_name was used for submission mode
4623 messages from external hosts without From: headers in a similar way,
4624 which is clearly wrong.
4626 PH/37 Added control=suppress_local_fixups.
4628 PH/38 When log_selector = +received_sender was set, and the addition of the
4629 sender made the log line's construction buffer exactly full, or one byte
4630 less than full, an overflow happened when the terminating "\n" was
4633 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4634 when the result of a list match is failure because a DNS lookup failed.
4636 PH/40 RM_COMMAND is now used in the building process.
4638 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4639 the "build-* directories that it finds.
4641 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4642 address in a domain literal was a prefix of an interface address.
4644 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4645 when verifying a sender address, unless rewrite_headers is false.
4647 PH/44 Wrote a long comment about why errors_to addresses are verified as
4648 recipients, not senders.
4650 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4651 the ratelimit ACL was added.
4653 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4655 PH/46 Added extra information about PostgreSQL errors to the error string.
4657 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4658 avoiding the need to install special zones in a real server. This is
4659 backwards compatible; if it can't find the fake resolver, it drops back.
4660 Thus, both old and new test suites can be run.
4662 TF/02 Added util/ratelimit.pl
4664 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4665 clock is set back in time.
4667 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4668 Candler <B.Candler@pobox.com>.
4670 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4671 OK for addresses that are the result of redirections.
4673 PH/48 A number of further additions for the benefit of the new test suite,
4674 including a fake gethostbyname() that interfaces to the fake DNS resolver
4677 TF/06 The fix for widen_domains has also been applied to qualify_single and
4678 search_parents which are the other dnslookup options that can cause
4681 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4684 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4686 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4687 service" if it encountered an MX record with an empty target hostname.
4688 The message is now "an MX or SRV record indicated no SMTP service".
4690 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4691 if the DNS of the sending site is misconfigured. This is quite a
4692 common situation. This change restores the behaviour of treating a
4693 helo verification defer as a failure.
4695 PH/51 If self=fail was set on a router, the bounce message did not include the
4696 actual error message.
4702 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4704 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4705 command itself ended in error, the underlying message about the transport
4706 filter timeout was being overwritten with the pipe command error. Now the
4707 underlying error message should be appended to the second error message.
4709 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4711 PH/02 Reduce the amount of output that "make" produces by default. Full output
4712 can still be requested.
4714 PH/03 The warning log line about a condition test deferring for a "warn" verb
4715 was being output only once per connection, rather than after each
4716 occurrence (because it was using the same function as for successful
4717 "warn" verbs). This seems wrong, so I have changed it.
4719 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4720 it should not have, which might have caused a crash in the right
4721 circumstances, but probably never did.
4723 PH/04 Installed a modified version of Tony Finch's patch to make submission
4724 mode fix the return path as well as the Sender: header line, and to
4725 add a /name= option so that you can make the user's friendly name appear
4728 TF/03 Added the control = fakedefer ACL modifier.
4730 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4731 Mark Lowes for thorough testing.
4733 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4735 TK/03 Merged latest SRS patch from Miles Wilton.
4737 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4738 with the definition in sysexits.h (which is #included earlier).
4739 Fortunately, Exim does not actually use EX_OK. The code used to try to
4740 preserve the sysexits.h value, by assuming that macro definitions were
4741 scanned for macro replacements. I have been disabused of this notion,
4742 so now the code just undefines EX_OK before #including unistd.h.
4744 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4745 in the smtp transport. When a block could not be written in a single
4746 write() function, the timeout was being re-applied to each part-write.
4747 This seems wrong - if the receiver was accepting one byte at a time it
4748 would take for ever. The timeout is now adjusted when this happens. It
4749 doesn't have to be particularly precise.
4751 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4752 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4754 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4755 without a domain if the check to <postmaster@domain> fails.
4757 SC/01 Eximstats: added -xls and the ability to specify output files
4758 (patch written by Frank Heydlauf).
4760 SC/02 Eximstats: use FileHandles for outputting results.
4762 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4764 SC/04 Eximstats: fixed display of large numbers with -nvr option
4766 SC/05 Eximstats: fixed merging of reports with empty tables.
4768 SC/06 Eximstats: added the -include_original_destination flag
4770 SC/07 Eximstats: removed tabs and trailing whitespace.
4772 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4774 TK/06 MBOX spool code: Add real "From " MBOX separator line
4775 so the .eml file is really in mbox format (even though
4776 most programs do not really care). Patch from Alex Miller.
4778 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4779 The latter is generated from $received_to and is only set if the
4780 message has one envelope recipient. SA can use these headers,
4781 obviously out-of-the-box. Patch from Alex Miller.
4783 PH/08 The ${def test on a variable was returning false if the variable's
4784 value was "0", contrary to what the specification has always said!
4785 The result should be true unless the variable is empty.
4787 PH/09 The syntax error of a character other than { following "${if
4788 def:variable_name" (after optional whitespace) was not being diagnosed.
4789 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4790 accidental colon was present, for example, could give incorrect results.
4792 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4793 result is used (not including appendfile, where other changes are about
4796 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4797 This involved changing a lot of size variables from int to off_t. It
4798 should work with maildirs and everything.
4800 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4801 spamd dying while we are connected to it.
4803 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4806 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4807 function for BDB 4.3.
4809 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4811 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4812 the sources for PCRE have been re-arranged and more files are now
4815 PH/15 The code I had for printing potentially long long variables in PH/11
4816 above was not the best (it lost precision). The length of off_t variables
4817 is now inspected at build time, and an appropriate printing format (%ld
4818 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4819 to be "long long int" or "long int". This is needed for the internal
4820 formatting function string_vformat().
4822 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4823 the configuration file to be ":syslog", then the script "guesses" where
4824 the logs files are, rather than using the compiled in default. In our
4825 case the guess is not the same as the compiled default, so the script
4826 suddenly stopped working when I started to use syslog. The patch checks
4827 to see if log_file_path is "". If so, it attempts to read it from exim
4828 with no configuration file to get the compiled in version, before it
4829 falls back to the previous guessing code."
4831 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4832 implementing BATV in an Exim configuration. See NewStuff for the gory
4835 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4836 Makefile that are specific to HP-UX.
4838 PH/18 If the "use_postmaster" option was set for a recipient callout together
4839 with the "random" option, the postmaster address was used as the MAIL
4840 FROM address for the random test, but not for the subsequent recipient
4841 test. It is now used for both.
4843 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4844 patch removes a few documentation additions to RFC 3028, because the
4845 latest draft now contains them. It adds the new en;ascii-case comparator
4846 and a new error check for 8bit text in MIME parts. Comparator and
4847 require names are now matched exactly. I enabled the subaddress
4848 extension, but it is not well tested yet (read: it works for me)."
4850 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4851 rework some of the code of TK/09 above to avoid the hardwired use of
4852 "%lld" and "long long". Replaced the call to snprintf() with a call to
4855 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4856 records point to non-existent hosts", "retry timeout exceeded", and
4857 "retry time not reached for any host after a long failure period".
4859 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4860 experimental DomainKeys support:
4862 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4863 (2) On an error such as an illegally used "control", the wrong name for
4864 the control was given.
4866 These problems did NOT occur unless DomainKeys support was compiled.
4868 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4870 PH/24 Added ${if match_ip condition.
4872 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4873 databases so that it will be absolutely obvious if a crash occurs in the
4874 DB library. This is a regular occurrence (often caused by mis-matched
4877 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4878 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4879 picked up on a user's system that detects such things. There doesn't seem
4880 to be a gcc warning option for this - only an attribute that has to be
4881 put on the function's prototype. It seems that in Fedora Core 4 they have
4882 set this on a number of new functions. No doubt there will be more in due
4885 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4886 specify a transport. However, if an address that was verified by such a
4887 router was the subject of a callout, Exim crashed because it tried to
4888 read the rcpt_include_affixes from the non-existent transport. Now it
4889 just assumes that the setting of that option is false. This bug was
4890 introduced by 4.51/PH/31.
4892 PH/28 Changed -d+all to exclude +memory, because that information is very
4893 rarely of interest, but it makes the output a lot bigger. People tend to
4894 do -d+all out of habit.
4896 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4897 code in os-type was giving problems when libc.so lives in lib64, like on
4900 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4901 aren't the modern standard, and it seems that some systems' include files
4902 don't always have them. Exim was already checking for some of the newer
4903 ones like T_AAAA, and defining it itself. I've added checks for all the
4904 record types that Exim uses.
4906 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4907 not automatically generating a new one, as it is supposed to. This
4908 prevented TLS from working. If the file did exist, but contained invalid
4909 data, a new version was generated, as expected. It was only the case of a
4910 non-existent file that was broken.
4912 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4913 with a change in libdomainkeys > 0.64.
4915 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4916 from DNS. If the selector record carries the flag, it now has
4917 precedence over the domain-wide flag.
4919 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4921 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4922 the use of a port name that isn't defined in /etc/services) occurred, the
4923 message was deferred as in a normal delivery, and thus remained on the
4924 spool, instead of being failed because of the mua_wrapper setting. This
4925 is now fixed, and I tidied up some of the mua_wrapper messages at the
4928 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4929 the messages in a hash of arrays rather than using individual hashes.
4930 This is a bit cleaner and results in dramatic memory savings, albeit
4931 at a slight CPU cost.
4933 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4934 as requested by Marc Sherman.
4936 SC/10 Eximstats: added histograms for user specified patterns as requested
4939 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4941 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4942 fopen() in the content-scanning modules that did not already have it.
4948 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4949 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4951 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4953 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4955 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4956 to test Sieve filters that use "vacation".
4958 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4959 that changes the way the GnuTLS parameters are stored in the cache file.
4960 The new format can be generated externally. For backward compatibility,
4961 if the data in the cache doesn't make sense, Exim assumes it has read an
4962 old-format file, and it generates new data and writes a new file. This
4963 means that you can't go back to an older release without removing the
4966 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4967 work if there are any delivery delays because "one_time" forces the
4968 parent to be marked "delivered", so its unseen clone is never tried
4969 again. For this reason, Exim now forbids the simultaneous setting of
4972 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4973 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4974 entry if you want to know the details. The fix, however, neglected to
4975 consider the case when local delivery batching is involved. The test for
4976 "previously delivered" was not happening when checking to see if an
4977 address could be batched with a previous (undelivered) one; under
4978 certain circumstances this could lead to multiple deliveries to the same
4981 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4982 in its include files, and this causes problems building Exim.
4984 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4985 header_syntax) but Exim was just ignoring anything given after a slash.
4986 In particular, this caused confusion with an attempt to use "verify =
4987 reverse_host_lookup/defer_ok". An error is now given when options are
4988 supplied for verify items that do not have them. (Maybe reverse_host_
4989 lookup should have a defer_ok option, but that's a different point.)
4991 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4992 defined by RFC 821) to 2048, because there were problems with some AUTH
4993 commands, and RFC 1869 says the size should be increased for extended
4994 SMTP commands that take arguments.
4996 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4999 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5000 "unknown" error; now it says that the functionality isn't in the binary.
5002 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5003 an address' error message when a string expansion fails (syntax or
5004 whatever). Otherwise the password may appear in the log. Following change
5005 PH/42 below, there is no longer a chance of it appearing in a bounce
5008 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5010 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5011 of its addresses was ever tried. (Bugzilla bug #2.)
5013 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5014 the result incorrectly in the debug output. (It correctly added a newline
5015 to what was transported.)
5017 TF/01 Added $received_time.
5019 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5020 commented out examples of how to interface to a virus scanner and to
5021 SpamAssassin. Also added commented examples of av_scanner and
5022 spamd_address settings.
5024 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5025 and controls are allowed in which ACLs. There were a couple of minor
5026 errors. Some of the entries in the conditions table (which is a table of
5027 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5028 negation of where the condition IS allowed.
5030 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5032 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5033 header file does not have a version number, so I've had to invent a new
5034 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5035 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5036 radiusclient), but it was contributed by a Radius user.
5038 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5039 files or directories, for OpenSSL.
5041 PH/19 When an Exim process that is running as root has to create an Exim log
5042 file, it does so in a subprocess that runs as exim:exim so as to get the
5043 ownership right at creation (otherwise, other Exim processes might see
5044 the file with the wrong ownership). There was no test for failure of this
5045 fork() call, which would lead to the process getting stuck as it waited
5046 for a non-existent subprocess. Forks do occasionally fail when resources
5047 run out. I reviewed all the other calls to fork(); they all seem to check
5050 PH/20 When checking for unexpected SMTP input at connect time (before writing
5051 the banner), Exim was not dealing correctly with a non-positive return
5052 from the read() function. If the client had disconnected by this time,
5053 the result was a log entry for a synchronization error with an empty
5054 string after "input=" when read() returned zero. If read() returned -1
5055 (an event I could not check), uninitialized data bytes were printed.
5056 There were reports of junk text (parts of files, etc) appearing after
5059 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5061 PH/22 Added support for macro redefinition, and (re)definition in between
5062 driver and ACL definitions.
5064 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5065 forgetting to use the resulting value; it was using the unexpanded value.
5067 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5068 hadn't been configured. The fix is from Juergen Kreileder, who
5069 understands it better than I do:
5071 "Here's what I see happening with three configured cyrus_sasl
5072 authenticators configured (plain, login, cram-md5):
5074 On startup auth_cyrus_sasl_init() gets called for each of these.
5075 This means three calls to sasl_listmech() without a specified mech_list.
5076 => SASL tests which mechs of all available mechs actually work
5077 => three warnings about OTP not working
5078 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5080 With the patch, sasl_listmech() also gets called three times. But now
5081 SASL's mech_list option is set to the server_mech specified in the the
5082 authenticator. Or in other words, the answer from sasl_listmech()
5083 gets limited to just the mech you're testing for (which is different
5085 => the return list contains just 'plain' or 'login', 'cram-md5' or
5086 nothing depending on the value of ob->server_mech.
5088 I've just tested the patch: Authentication still works fine,
5089 unavailable mechs specified in the exim configuration are still
5090 caught, and the auth.log warnings about OTP are gone."
5092 PH/25 When debugging is enabled, the contents of the command line are added
5093 to the debugging output, even when log_selector=+arguments is not
5096 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5097 answer is "GNU", and only if the return is "GNU/something" is the answer
5100 PH/27 $acl_verify_message is now set immediately after the failure of a
5101 verification in an ACL, and so is available in subsequent modifiers. In
5102 particular, the message can be preserved by coding like this:
5104 warn !verify = sender
5105 set acl_m0 = $acl_verify_message
5107 Previously, $acl_verify_message was set only while expanding "message"
5108 and "log_message" when a very denied access.
5110 PH/28 Modified OS/os.c-Linux with
5112 -#ifndef OS_LOAD_AVERAGE
5113 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5115 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5116 nomenclature these days.)
5118 PH/29 Installed patch from the Sieve maintainer that adds the options
5119 sieve_useraddress and sieve_subaddress to the redirect router.
5121 PH/30 In these circumstances:
5122 . Two addresses routed to the same list of hosts;
5123 . First host does not offer TLS;
5124 . First host accepts first address;
5125 . First host gives temporary error to second address;
5126 . Second host offers TLS and a TLS session is established;
5127 . Second host accepts second address.
5128 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5129 and peerdn, if requested) that were in fact used only for the second
5132 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5133 attention to any local part prefix or suffix that was matched by the
5134 router that accepted the address. It now behaves in the same way as it
5135 does for delivery: the affixes are removed from the local part unless
5136 rcpt_include_affixes is set on the transport.
5138 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5139 timeout during the DATA phase of an incoming message.
5141 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5142 applied a patch sanctioned by the Sieve maintainer.
5144 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5145 the uid or gid is negative. A case of a negative gid caused this to be
5146 noticed. The fix allows for either to be negative.
5148 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5149 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5152 PH/36 The change PH/12 above was broken. Fixed it.
5154 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5155 the grounds that routing the same address twice would always produce the
5156 same answer. This might have been true once, but it is certainly no
5157 longer true now. Routing a child address may depend on the previous
5158 routing that produced that child. Some complicated redirection strategies
5159 went wrong when messages had multiple recipients, and made Exim's
5160 behaviour dependent on the order in which the addresses were given.
5162 I have moved the duplicate checking until after the routing is complete.
5163 Exim scans the addresses that are assigned to local and remote
5164 transports, and removes any duplicates. This means that more work will be
5165 done, as duplicates will always all be routed, but duplicates are
5166 presumably rare, so I don't expect this is of any significance.
5168 For deliveries to pipes, files, and autoreplies, the duplicate checking
5169 still happens during the routing process, since they are not going to be
5172 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5173 It corrects a timeout issue with spamd. This is Ian's comment: "The
5174 background is that sometimes spamd either never reads data from a
5175 connection it has accepted, or it never writes response data. The exiscan
5176 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5177 blindly assumes that writes won't block so it may never time out."
5179 PH/39 Allow G after quota size as well as K and M.
5181 PH/40 The value set for $authenticated_id in an authenticator may not contain
5182 binary zeroes or newlines because the value is written to log lines and
5183 to spool files. There was no check on this. Now the value is run through
5184 the string_printing() function so that such characters are converted to
5185 printable escape sequences.
5187 PH/41 $message_linecount is a new variable that contains the total number of
5188 lines in the message. Compare $body_linecount, which is the count for the
5191 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5192 bounce and delay warning messages, except in certain special cases, which
5195 (a) An SMTP error message from a remote host;
5196 (b) A message specified in a :fail: redirection;
5197 (c) A message specified in a "fail" command in a system filter;
5198 (d) A message specified in a FAIL return from the queryprogram router;
5199 (e) A message specified by the cannot_route_message router option.
5201 In these cases only, Exim does include the error details in bounce and
5202 warning messages. There are also a few cases where bland messages such
5203 as "unrouteable address" or "local delivery error" are given.
5205 PH/43 $value is now also set for the "else" part of a ${run expansion.
5207 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5208 being worked on, but at least Exim now implements the latest version to
5211 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5212 process to complete was treated as a delivery failure, a timeout while
5213 writing the message to the pipe was logged, but erroneously treated as a
5214 successful delivery. Such timeouts include transport filter timeouts. For
5215 consistency with the overall process timeout, these timeouts are now
5216 treated as errors, giving rise to delivery failures by default. However,
5217 there is now a new Boolean option for the pipe transport called
5218 timeout_defer, which, if set TRUE, converts the failures into defers for
5219 both kinds of timeout. A transport filter timeout is now identified in
5222 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5223 systems where "make" and "gmake" are different, calling "gmake" at top
5224 level broke things. I've arranged for the value of $(MAKE) to be passed
5225 from the Makefile to this script so that it can call the same version of
5229 A note about Exim versions 4.44 and 4.50
5230 ----------------------------------------
5232 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5233 changes of various kinds. As a consequence, a big documentation update was
5234 needed. This delayed the release for rather longer than seemed good, especially
5235 in the light of a couple of (minor) security issues. Therefore, the changes
5236 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5237 release. So 4.44 and 4.50 are in effect two different branches that both start
5240 I have left the 4.50 change log unchanged; it contains all the changes since
5241 4.43. The change log for 4.44 is below; many of its items are identical to
5242 those for 4.50. This seems to be the most sensible way to preserve the
5243 historical information.
5249 1. Minor wording change to the doc/README.SIEVE file.
5251 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5252 computation of the current number of files was incorrect.
5254 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5255 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5258 4. Give more explanation in the error message when the command for a transport
5259 filter fails to execute.
5261 5. There are several places where Exim runs a non-Exim command in a
5262 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5263 was being done only for the command run by the queryprogram router. It is
5264 now done for all such subprocesses. The other cases are: ${run, transport
5265 filters, and the commands run by the lmtp and pipe transports.
5267 6. Added CONFIGURE_GROUP build-time option.
5269 7. Some older OS have a limit of 256 on the maximum number of file
5270 descriptors. Exim was using setrlimit() to set 1000 as a large value
5271 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5272 systems. I've change it so that if it can't get 1000, it tries for 256.
5274 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5275 was an oversight, and furthermore, ever since the addition of extra
5276 controls (e.g. 4.43/32), the checks on when to allow different forms of
5277 "control" were broken. There should now be diagnostics for all cases when a
5278 control that does not make sense is encountered.
5280 9. Added the /retain_sender option to "control=submission".
5282 10. $recipients is now available in the predata ACL (oversight).
5284 11. Tidy the search cache before the fork to do a delivery from a message
5285 received from the command line. Otherwise the child will trigger a lookup
5286 failure and thereby defer the delivery if it tries to use (for example) a
5287 cached ldap connection that the parent has called unbind on.
5289 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5290 of $address_data from the recipient verification was clobbered by the
5291 sender verification.
5293 13. The value of address_data from a sender verification is now available in
5294 $sender_address_data in subsequent conditions in the ACL statement.
5296 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5298 15. Added a new option "connect=<time>" to callout options, to set a different
5301 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5302 was its contents. (It was OK if the option was not defined at all.)
5304 17. A "Completed" log line is now written for messages that are removed from
5305 the spool by the -Mrm option.
5307 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5308 information about exactly what failed.
5310 19. Added -dd to debug only the daemon process.
5312 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5313 handles timeouts, both on the server side and network timeouts. Renamed the
5314 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5316 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5317 It is now set to "smtps".
5319 22. $host_address is now set to the target address during the checking of
5320 ignore_target_hosts.
5322 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5323 being passed; this would have caused $sender_host_name to have been used if
5324 matching the list had actually called for a host name (not very likely,
5325 since this list is usually IP addresses). A host name is now passed as
5328 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5329 code that specifies a non-restarting handler (typically sigaction() in
5330 modern systems) in an attempt to fix a rare and obscure crash bug.
5332 25. Narrowed the window for a race in the daemon that could cause it to ignore
5333 SIGCHLD signals. This is not a major problem, because they are used only to
5334 wake it up if nothing else does.
5336 26. A malformed maildirsize file could cause Exim to calculate negative values
5337 for the mailbox size or file count. Odd effects could occur as a result.
5338 The maildirsize information is now recalculated if the size or filecount
5341 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5342 support for a long time. Removed HAVE_SYS_VFS_H.
5344 28. Installed the latest version of exipick from John Jetmore.
5346 29. In an address list, if the pattern was not a regular expression, an empty
5347 subject address (from a bounce message) matched only if the pattern was an
5348 empty string. Non-empty patterns were not even tested. This was the wrong
5349 because it is perfectly reasonable to use an empty address as part of a
5350 database query. An empty address is now tested by patterns that are
5351 lookups. However, all the other forms of pattern expect the subject to
5352 contain a local part and a domain, and therefore, for them, an empty
5353 address still always fails if the pattern is not itself empty.
5355 30. Exim went into a mad DNS loop when attempting to do a callout where the
5356 host was specified on an smtp transport, and looking it up yielded more
5357 than one IP address.
5359 31. Re-factored the code for checking spool and log partition space into a
5360 function that finds that data and another that does the check. The former
5361 is then used to implement four new variables: $spool_space, $log_space,
5362 $spool_inodes, and $log_inodes.
5364 32. The RFC2047 encoding function was originally intended for short strings
5365 such as real names; it was not keeping to the 75-character limit for
5366 encoded words that the RFC imposes. It now respects the limit, and
5367 generates multiple encoded words if necessary. To be on the safe side, I
5368 have increased the buffer size for the ${rfc2047: expansion operator from
5371 33. It is now permitted to omit both strings after an "if" condition; if the
5372 condition is true, the result is "true". As before, when the second string
5373 is omitted, a false condition yields an empty string. This makes it less
5374 cumbersome to write custom ACL and router conditions.
5376 34. Failure to deliver a bounce message always caused it to be frozen, even if
5377 there was an errors_to setting on the router. The errors_to setting is now
5380 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5381 canonical form (fully expanded) before being placed in
5382 $sender_host_address.
5384 36. The table in the code that translates DNS record types into text (T_A to
5385 "A" for instance) was missing entries for NS and CNAME. It is just possible
5386 that this could have caused confusion if both these types were looked up
5387 for the same domain, because the text type is used as part of Exim's
5388 per-process caching. But the chance of anyone hitting this buglet seems
5391 37. The dnsdb lookup has been extended in a number of ways.
5393 (1) There is a new type, "zns", which walks up the domain tree until it
5394 finds some nameserver records. It should be used with care.
5396 (2) There is a new type, "mxh", which is like "mx" except that it returns
5397 just the host names, not the priorities.
5399 (3) It is now possible to give a list of domains (or IP addresses) to be
5400 looked up. The behaviour when one of the lookups defers can be
5401 controlled by a keyword.
5403 (4) It is now possible to specify the separator character for use when
5404 multiple records are returned.
5406 38. The dnslists ACL condition has been extended: it is now possible to supply
5407 a list of IP addresses and/or domains to be looked up in a particular DNS
5410 39. Added log_selector=+queue_time_overall.
5412 40. When running the queue in the test harness, wait just a tad after forking a
5413 delivery process, to get repeatability of debugging output.
5415 41. Include certificate and key file names in error message when GnuTLS fails
5416 to set them up, because the GnuTLS error message doesn't include the name
5417 of the failing file when there is a problem reading it.
5419 42. Allow both -bf and -bF in the same test run.
5421 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5423 44. The "Exiscan patch" is now merged into the mainline Exim source.
5425 45. Sometimes the final signoff response after QUIT could fail to get
5426 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5427 < 0 before doing a fflush(). This bug looks as though it goes back to the
5428 introduction of TLS in release 3.20, but "sometimes" must have been rare
5429 because the tests only now provoked it.
5431 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5432 (this can affect the format of dates).
5434 47. exim_tidydb, when checking for the continued existence of a message for
5435 which it has found a message-specific retry record, was not finding
5436 messages that were in split spool directories. Consequently, it was
5437 deleting retry records that should have stayed in existence.
5439 48. Steve fixed some bugs in eximstats.
5441 49. The SPA authentication driver was not abandoning authentication and moving
5442 on to the next authenticator when an expansion was forced to fail,
5443 contradicting the general specification for all authenticators. Instead it
5444 was generating a temporary error. It now behaves as specified.
5446 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5447 (the order specifies the preference for clients). The order is now AES256,
5448 AES128, 3DES, ARCFOUR128.
5450 51. Small patch to Sieve code - explicitly set From: when generating an
5453 52. Exim crashed if a remote delivery caused a very long error message to be
5454 recorded - for instance if somebody sent an entire SpamAssassin report back
5455 as a large number of 550 error lines. This bug was coincidentally fixed by
5456 increasing the size of one of Exim's internal buffers (big_buffer) that
5457 happened as part of the Exiscan merge. However, to be on the safe side, I
5458 have made the code more robust (and fixed the comments that describe what
5461 53. Now that there can be additional text after "Completed" in log lines (if
5462 the queue_time_overall log selector is set), a one-byte patch to exigrep
5463 was needed to allow it to recognize "Completed" as not the last thing in
5466 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5467 patch that reportedly fixes this has been added. I am not expert enough to
5468 create a test for it. This is what the patch creator wrote:
5470 "I found a little strange behaviour of ldap code when working with
5471 Windows 2003 AD Domain, where users was placed in more than one
5472 Organization Units. When I tried to give exim partial DN, the exit code
5473 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5474 But simultaneously result of request was absolutely normal ldap result,
5475 so I produce this patch..."
5477 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5478 REFERENCE, so I have modified the code to exclude the patch when that macro
5481 55. Some experimental protocols are using DNS PTR records for new purposes. The
5482 keys for these records are domain names, not reversed IP addresses. The
5483 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5484 leaves it alone. Component reversal etc. now happens only for IP addresses.
5487 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5489 57. Double the size of the debug message buffer (to 2048) so that more of very
5490 long debug lines gets shown.
5492 58. The exicyclog utility now does better if the number of log files to keep
5493 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5495 59. Two changes related to the smtp_active_hostname option:
5497 (1) $smtp_active_hostname is now available as a variable.
5498 (2) The default for smtp_banner uses $smtp_active_hostname instead
5499 of $primary_hostname.
5501 60. The host_aton() function is supposed to be passed a string that is known
5502 to be a valid IP address. However, in the case of IPv6 addresses, it was
5503 not checking this. This is a hostage to fortune. Exim now panics and dies
5504 if the condition is not met. A case was found where this could be provoked
5505 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5506 components; fortuitously, this particular loophole had already been fixed
5507 by change 4.50/55 above.
5509 If there are any other similar loopholes, the new check in host_aton()
5510 itself should stop them being exploited. The report I received stated that
5511 data on the command line could provoke the exploit when Exim was running as
5512 exim, but did not say which command line option was involved. All I could
5513 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5514 running as the user.
5517 61. There was a buffer overflow vulnerability in the SPA authentication code
5518 (which came originally from the Samba project). I have added a test to the
5519 spa_base64_to_bits() function which I hope fixes it.
5522 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5523 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5525 63. The daemon start-up calls getloadavg() while still root for those OS that
5526 need the first call to be done as root, but it missed one case: when
5527 deliver_queue_load_max is set with deliver_drop_privilege. This is
5528 necessary for the benefit of the queue runner, because there is no re-exec
5529 when deliver_drop_privilege is set.
5531 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5532 This has been fixed.
5534 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5535 was in use, was not putting the data itself into the right store pool;
5536 consequently, it could be overwritten for a subsequent message in the same
5537 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5540 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5542 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5543 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5544 the function were treating the return as a boolean value, which happened to
5545 work because 0=false and not-0=true, but is not correct code.
5547 68. The host_aton() function was not handling scoped IPv6 addresses (those
5548 with, for example, "%eth0" on the end) correctly.
5550 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5551 negated items (that is, ~something) in unsigned ints. Some compilers
5552 apparently mutter when there is no cast.
5554 70. If an address verification called from an ACL failed, and did not produce a
5555 user-specific message (i.e. there was only a "system" message), nothing was
5556 put in $acl_verify_message. In this situation, it now puts the system
5559 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5560 session; change 4.31/43 added the unwanted input to the log line - except
5561 that it did not do this in the start of session case. It now does.
5563 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5564 This is wrong and can cause the other end to generate a synchronization
5565 error if it is another Exim or anything else that does the synchronization
5566 check. A QUIT command is no longer sent after a timeout.
5568 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5569 during host lookups.
5571 74. The defer_ok option of callout verification was not working if it was used
5572 when verifying addresses in header lines, that is, for this case:
5574 verify = header_sender/callout=defer_ok
5576 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5577 those file descriptors could be used for SMTP connections. If anything
5578 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5579 could be sent to the SMTP client, causing chaos. The daemon now opens
5580 stdin, stdout, and stderr to /dev/null when it puts itself into the
5583 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5584 log by default. The user can override this with suitable Perl magic.
5586 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5587 the log message when discard triggers, was not working for the DATA ACL or
5588 for the non-SMTP ACL.
5590 78. Error message wording change in sieve.c.
5592 79. If smtp_accept_max_per_host was set, the number of connections could be
5593 restricted to fewer than expected, because the daemon was trying to set up
5594 a new connection before checking whether the processes handling previous
5595 connections had finished. The check for completed processes is now done
5596 earlier. On busy systems, this bug wouldn't be noticed because something
5597 else would have woken the daemon, and it would have reaped the completed
5600 80. If a message was submitted locally by a user whose login name contained one
5601 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5602 It caused a spool format error. I have fixed the spool reading code. A
5603 related problem was that the "from" clause in the Received: line became
5604 illegal because of the space(s). It is now covered by ${quote_local_part.
5606 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5609 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5611 83. Patch from Sieve maintainer to fix unterminated string problem in
5612 "vacation" handling.
5614 84. Some minor changes to the Linux configuration files to help with other
5615 OS variants using glibc.
5617 85. One more patch for Sieve to update vacation handling to latest spec.
5620 ----------------------------------------------------
5621 See the note above about the 4.44 and 4.50 releases.
5622 ----------------------------------------------------
5628 1. Change 4.43/35 introduced a bug that caused file counts to be
5629 incorrectly computed when quota_filecount was set in an appendfile
5632 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5633 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5636 3. Give more explanation in the error message when the command for a transport
5637 filter fails to execute.
5639 4. There are several places where Exim runs a non-Exim command in a
5640 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5641 was being done only for the command run by the queryprogram router. It is
5642 now done for all such subprocesses. The other cases are: ${run, transport
5643 filters, and the commands run by the lmtp and pipe transports.
5645 5. Some older OS have a limit of 256 on the maximum number of file
5646 descriptors. Exim was using setrlimit() to set 1000 as a large value
5647 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5648 systems. I've change it so that if it can't get 1000, it tries for 256.
5650 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5651 was an oversight, and furthermore, ever since the addition of extra
5652 controls (e.g. 4.43/32), the checks on when to allow different forms of
5653 "control" were broken. There should now be diagnostics for all cases when a
5654 control that does not make sense is encountered.
5656 7. $recipients is now available in the predata ACL (oversight).
5658 8. Tidy the search cache before the fork to do a delivery from a message
5659 received from the command line. Otherwise the child will trigger a lookup
5660 failure and thereby defer the delivery if it tries to use (for example) a
5661 cached ldap connection that the parent has called unbind on.
5663 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5664 of $address_data from the recipient verification was clobbered by the
5665 sender verification.
5667 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5668 was its contents. (It was OK if the option was not defined at all.)
5670 11. A "Completed" log line is now written for messages that are removed from
5671 the spool by the -Mrm option.
5673 12. $host_address is now set to the target address during the checking of
5674 ignore_target_hosts.
5676 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5677 being passed; this would have caused $sender_host_name to have been used if
5678 matching the list had actually called for a host name (not very likely,
5679 since this list is usually IP addresses). A host name is now passed as
5682 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5683 code that specifies a non-restarting handler (typically sigaction() in
5684 modern systems) in an attempt to fix a rare and obscure crash bug.
5686 15. Narrowed the window for a race in the daemon that could cause it to ignore
5687 SIGCHLD signals. This is not a major problem, because they are used only to
5688 wake it up if nothing else does.
5690 16. A malformed maildirsize file could cause Exim to calculate negative values
5691 for the mailbox size or file count. Odd effects could occur as a result.
5692 The maildirsize information is now recalculated if the size or filecount
5695 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5696 support for a long time. Removed HAVE_SYS_VFS_H.
5698 18. Updated exipick to current release from John Jetmore.
5700 19. Allow an empty sender to be matched against a lookup in an address list.
5701 Previously the only cases considered were a regular expression, or an
5704 20. Exim went into a mad DNS lookup loop when doing a callout where the
5705 host was specified on the transport, if the DNS lookup yielded more than
5708 21. The RFC2047 encoding function was originally intended for short strings
5709 such as real names; it was not keeping to the 75-character limit for
5710 encoded words that the RFC imposes. It now respects the limit, and
5711 generates multiple encoded words if necessary. To be on the safe side, I
5712 have increased the buffer size for the ${rfc2047: expansion operator from
5715 22. Failure to deliver a bounce message always caused it to be frozen, even if
5716 there was an errors_to setting on the router. The errors_to setting is now
5719 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5720 canonical form (fully expanded) before being placed in
5721 $sender_host_address.
5723 24. Updated eximstats to version 1.33
5725 25. Include certificate and key file names in error message when GnuTLS fails
5726 to set them up, because the GnuTLS error message doesn't include the name
5727 of the failing file when there is a problem reading it.
5729 26. Expand error message when OpenSSL has problems setting up cert/key files.
5732 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5733 (this can affect the format of dates).
5735 28. exim_tidydb, when checking for the continued existence of a message for
5736 which it has found a message-specific retry record, was not finding
5737 messages that were in split spool directories. Consequently, it was
5738 deleting retry records that should have stayed in existence.
5740 29. eximstats updated to version 1.35
5741 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5742 1.35 - bugfix such that pie charts by volume are generated correctly
5744 30. The SPA authentication driver was not abandoning authentication and moving
5745 on to the next authenticator when an expansion was forced to fail,
5746 contradicting the general specification for all authenticators. Instead it
5747 was generating a temporary error. It now behaves as specified.
5749 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5750 (the order specifies the preference for clients). The order is now AES256,
5751 AES128, 3DES, ARCFOUR128.
5753 31. Small patch to Sieve code - explicitly set From: when generating an
5756 32. Exim crashed if a remote delivery caused a very long error message to be
5757 recorded - for instance if somebody sent an entire SpamAssassin report back
5758 as a large number of 550 error lines. This bug was coincidentally fixed by
5759 increasing the size of one of Exim's internal buffers (big_buffer) that
5760 happened as part of the Exiscan merge. However, to be on the safe side, I
5761 have made the code more robust (and fixed the comments that describe what
5764 33. Some experimental protocols are using DNS PTR records for new purposes. The
5765 keys for these records are domain names, not reversed IP addresses. The
5766 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5767 leaves it alone. Component reversal etc. now happens only for IP addresses.
5770 34. The host_aton() function is supposed to be passed a string that is known
5771 to be a valid IP address. However, in the case of IPv6 addresses, it was
5772 not checking this. This is a hostage to fortune. Exim now panics and dies
5773 if the condition is not met. A case was found where this could be provoked
5774 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5775 components; fortuitously, this particular loophole had already been fixed
5776 by change 4.50/55 or 4.44/33 above.
5778 If there are any other similar loopholes, the new check in host_aton()
5779 itself should stop them being exploited. The report I received stated that
5780 data on the command line could provoke the exploit when Exim was running as
5781 exim, but did not say which command line option was involved. All I could
5782 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5783 running as the user.
5786 35. There was a buffer overflow vulnerability in the SPA authentication code
5787 (which came originally from the Samba project). I have added a test to the
5788 spa_base64_to_bits() function which I hope fixes it.
5791 36. The daemon start-up calls getloadavg() while still root for those OS that
5792 need the first call to be done as root, but it missed one case: when
5793 deliver_queue_load_max is set with deliver_drop_privilege. This is
5794 necessary for the benefit of the queue runner, because there is no re-exec
5795 when deliver_drop_privilege is set.
5797 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5798 was in use, was not putting the data itself into the right store pool;
5799 consequently, it could be overwritten for a subsequent message in the same
5800 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5803 38. Sometimes the final signoff response after QUIT could fail to get
5804 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5805 < 0 before doing a fflush(). This bug looks as though it goes back to the
5806 introduction of TLS in release 3.20, but "sometimes" must have been rare
5807 because the tests only now provoked it.
5813 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5814 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5815 Now it returns an integer. A number of calls were still expecting a T/F
5816 return. Fortuitously, in all cases, the tests worked in OK situations,
5817 which is the norm. However, things would have gone wrong on any write
5818 failures on the smtp file descriptor. This function is used when sending
5819 messages over SMTP and also when doing verify callouts.
5821 2. When Exim is called to do synchronous delivery of a locally submitted
5822 message (the -odf or -odi options), it no longer closes stderr before doing
5825 3. Implemented the mua_wrapper option.
5827 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5829 5. Implemented the functions header_remove(), header_testname(),
5830 header_add_at_position(), and receive_remove_recipient(), and exported them
5833 6. If an ACL "warn" statement specified the addition of headers, Exim already
5834 inserted X-ACL-Warn: at the start if there was no header name. However, it
5835 was not making this test for the second and subsequent header lines if
5836 there were newlines in the string. This meant that an invalid header could
5837 be inserted if Exim was badly configured.
5839 7. Allow an ACL "warn" statement to add header lines at the start or after all
5840 the Received: headers, as well as at the end.
5842 8. Added the rcpt_4xx retry error code.
5844 9. Added postmaster_mailfrom=xxx to callout verification option.
5846 10. Added mailfrom=xxxx to the callout verification option, for verify=
5849 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5850 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5852 12. Inserted some casts to stop certain compilers warning when using pointer
5853 differences as field lengths or precisions in printf-type calls (mostly
5854 affecting debugging statements).
5856 13. Added optional readline() support for -be (dynamically loaded).
5858 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5859 same clock tick as a message's arrival, so that its received time was the
5860 same as the "first fail" time on the retry record, and that message
5861 remained on the queue past the ultimate address timeout, every queue runner
5862 would try a delivery (because it was past the ultimate address timeout) but
5863 after another failure, the ultimate address timeout, which should have then
5864 bounced the address, did not kick in. This was a "< instead of <=" error;
5865 in most cases the first failure would have been in the next clock tick
5866 after the received time, and all would be well.
5868 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5869 being recognized when the domain list was tested by the match_domain
5870 condition in an expansion string.
5872 16. Added the ${str2b64: operator.
5874 17. Exim was always calling setrlimit() to set a large limit for the number of
5875 processes, without checking whether the existing limit was already
5876 adequate. (It did check for the limit on file descriptors.) Furthermore,
5877 errors from getrlimit() and setrlimit() were being ignored. Now they are
5878 logged to the main and panic logs, but Exim does carry on, to try to do its
5879 job under whatever limits there are.
5881 18. Imported PCRE 5.0.
5883 19. Trivial typo in log message " temporarily refused connection" (the leading
5886 20. If the log selector return_path_on_delivery was set and an address was
5887 redirected to /dev/null, the delivery process crashed because it assumed
5888 that a return path would always be set for a "successful" delivery. In this
5889 case, the whole delivery is bypassed as an optimization, and therefore no
5892 21. Internal re-arrangement: the function for sending a challenge and reading
5893 a response while authentication was assuming a zero-terminated challenge
5894 string. It's now changed to take a pointer and a length, to allow for
5895 binary data in such strings.
5897 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5899 23. Exim was not respecting finduser_retries when seeking the login of the
5900 uid under which it was called; it was always trying 10 times. (The default
5901 setting of finduser_retries is zero.) Also, it was sleeping after the final
5902 failure, which is pointless.
5904 24. Implemented tls_on_connect_ports.
5906 25. Implemented acl_smtp_predata.
5908 26. If the domain in control=submission is set empty, Exim assumes that the
5909 authenticated id is a complete email address when it generates From: or
5910 Sender: header lines.
5912 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5913 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5914 chown and chgrp in /bin and hostname in /usr/bin.
5916 28. Exim was keeping the "process log" file open after each use, just as it
5917 does for the main log. This opens the possibility of it remaining open for
5918 long periods when the USR1 signal hits a daemon. Occasional processlog
5919 errors were reported, that could have been caused by this. Anyway, it seems
5920 much more sensible not to leave this file open at all, so that is what now
5923 29. The long-running daemon process does not normally write to the log once it
5924 has entered its main loop, and it closes the log before doing so. This is
5925 so that log files can straightforwardly be renamed and moved. However,
5926 there are a couple of unusual error situations where the daemon does write
5927 log entries, and I had neglected to close the log afterwards.
5929 30. The text of an SMTP error response that was received during a remote
5930 delivery was being truncated at 512 bytes. This is too short for some of
5931 the long messages that one sometimes sees. I've increased the limit to
5934 31. It is now possible to make retry rules that apply only when a message has a
5935 specific sender, in particular, an empty sender.
5937 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5938 it possible to be selective about when SMTP synchronization is enforced.
5940 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5942 32. Implemented hosts_connection_nolog.
5944 33. Added an ACL for QUIT.
5946 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5949 35. Added mailbox_size and mailbox_filecount to appendfile.
5951 36. Added control = no_multiline_responses to ACLs.
5953 37. There was a bug in the logic of the code that waits for the clock to tick
5954 in the case where the clock went backwards by a substantial amount such
5955 that the microsecond fraction of "now" was more than the microsecond
5956 fraction of "then" (but the whole seconds number was less).
5958 38. Added support for the libradius Radius client library this is found on
5959 FreeBSD (previously only the radiusclient library was supported).
5965 1. When certain lookups returned multiple values in the form name=value, the
5966 quoting of the values was not always being done properly. Specifically:
5967 (a) If the value started with a double quote, but contained no whitespace,
5969 (b) If the value contained whitespace other than a space character (i.e.
5970 tabs or newlines or carriage returns) it was not quoted.
5971 This fix has been applied to the mysql and pgsql lookups by writing a
5972 separate quoting function and calling it from the lookup code. The fix
5973 should probably also be applied to nisplus, ibase and oracle lookups, but
5974 since I cannot test any of those, I have not disturbed their existing code.
5976 2. A hit in the callout cache for a specific address caused a log line with no
5977 reason for rejecting RCPT. Now it says "Previous (cached) callout
5978 verification failure".
5980 3. There was an off-by-one bug in the queryprogram router. An over-long
5981 return line was truncated at 256 instead of 255 characters, thereby
5982 overflowing its buffer with the terminating zero. As well as fixing this, I
5983 have increased the buffer size to 1024 (and made a note to document this).
5985 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5986 when Exim is waiting for an SMTP response from a remote server, Exim
5987 restarts its select() call on the socket, thereby resetting its timeout.
5988 This is not a problem when such interrupts are rare. Somebody set up a cron
5989 job to run exiwhat every 2 minutes, which is less than the normal select()
5990 timeout (5 or 10 minutes). This meant that the select() timeout never
5991 kicked in because it was always reset. I have fixed this by comparing the
5992 time when an interrupt arrives with the time at the start of the first call
5993 to select(). If more time than the timeout has elapsed, the interrupt is
5994 treated as a timeout.
5996 5. Some internal re-factoring in preparation for the addition of Sieve
5997 extensions (by MH). In particular, the "personal" test is moved to a
5998 separate function, and given an option for scanning Cc: and Bcc: (which is
5999 not set for Exim filters).
6001 6. When Exim created an email address using the login of the caller as the
6002 local part (e.g. when creating a From: or Sender: header line), it was not
6003 quoting the local part when it contained special characters such as @.
6005 7. Installed new OpenBSD configuration files.
6007 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6008 try to make them clearer.
6010 9. Callout options, other than the timeout value, were being ignored when
6011 verifying sender addresses in header lines. For example, when using
6013 verify = header_sender/callout=no_cache
6015 the cache was (incorrectly) being used.
6017 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6018 only the Cygwin environment.
6020 11. When return_path_on_delivery was set as a log selector, if different remote
6021 addresses in the same message used different return paths and parallel
6022 remote delivery occurred, the wrong values would sometimes be logged.
6023 (Whenever a remote delivery process finished, the return path value from
6024 the most recently started remote delivery process was logged.)
6026 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6027 lines when AUTH and/or TLS are in use. This is the "received protocol"
6028 field. Exim used to use "asmtp" for authenticated SMTP, without any
6029 indication (in the protocol name) for TLS use. Now it follows the RFC and
6030 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6031 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6032 names appear in log lines as well as in Received: header lines.
6034 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6035 extensions, and comparison tests, and to fix some bugs.
6037 14. Changes to the "personal" filter test:
6039 (1) The test was buggy in that it was just doing the equivalent of
6040 "contains" tests on header lines. For example, if a user's address was
6041 anne@some.where, the "personal" test would incorrectly be true for
6043 To: susanne@some.where
6045 This test is now done by extracting each address from the header in turn,
6046 and checking the entire address. Other tests that are part of "personal"
6047 are now done using regular expressions (for example, to check local parts
6048 of addresses in From: header lines).
6050 (2) The list of non-personal local parts in From: addresses has been
6051 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6052 taken from the Sieve specification recommendations.
6054 (3) If the message contains any header line starting with "List-" it is
6055 treated as non-personal.
6057 (4) The test for "circular" in the Subject: header line has been removed
6058 because it now seems ill-conceived.
6060 15. Minor typos in src/EDITME comments corrected.
6062 16. Installed latest exipick from John Jetmore.
6064 17. If headers_add on a router specified a text string that was too long for
6065 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6066 of string_sprintf() is now avoided.
6068 18. $message_body_size was not set (it was always zero) when running the DATA
6069 ACL and the local_scan() function.
6071 19. For the "mail" command in an Exim filter, no default was being set for
6072 the once_repeat time, causing a random time value to be used if "once" was
6073 specified. (If the value happened to be <= 0, no repeat happened.) The
6074 default is now 0s, meaning "never repeat". The "vacation" command was OK
6075 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6076 (I found it when inspecting the code).
6078 20. There is now an overall timeout for performing a callout verification. It
6079 defaults to 4 times the callout timeout, which applies to individual SMTP
6080 commands during the callout. The overall timeout applies when there is more
6081 than one host that can be tried. The timeout is checked before trying the
6082 next host. This prevents very long delays if there are a large number of
6083 hosts and all are timing out (e.g. when the network connections are timing
6084 out). The value of the overall timeout can be changed by specifying an
6085 additional sub-option for "callout", called "maxwait". For example:
6087 verify = sender/callout=5s,maxwait=20s
6089 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6090 to the end before writing, but this should make it even safer).
6092 22. Exim was forgetting that it had advertised PIPELINING for the second and
6093 subsequent messages on an SMTP connection. It was also not resetting its
6094 memory on STARTTLS and an internal HELO.
6096 23. When Exim logs an SMTP synchronization error within a session, it now
6097 records whether PIPELINING has been advertised or not.
6099 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6100 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6101 rather than long int.
6103 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6105 26. Added the never_mail option to autoreply.
6111 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6112 crash if the getsockname() call failed; this can happen if a connection is
6113 closed very soon after it is established. The problem was simply in the
6114 order in which certain operations were done, causing Exim to try to write
6115 to the SMTP stream before it had set up the file descriptor. The bug has
6116 been fixed by making things happen in the correct order.
6122 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6123 before the connection was closed, thus losing the rejection response.
6125 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6126 some early Solaris releases, but causes trouble in current releases where
6127 socklen_t is defined.
6129 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6132 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6135 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6136 that releasing the top of it at the end releases what was used for sub-
6137 expansions (unless the block got too big). However, discard this block if
6138 the first thing is a variable or header, so that we can use its block when
6139 it is dynamic (useful for very large $message_headers, for example).
6141 6. Lookups now cache *every* query, not just the most recent. A new, separate
6142 store pool is used for this. It can be recovered when all lookup caches are
6143 flushed. Lookups now release memory at the end of their result strings.
6144 This has involved some general refactoring of the lookup sources.
6146 7. Some code has been added to the store_xxx() functions to reduce the amount
6147 of flapping under certain conditions.
6149 8. log_incoming_interface used to affect only the <= reception log lines. Now
6150 it causes the local interface and port to be added to several more SMTP log
6151 lines, for example "SMTP connection from", and rejection lines.
6153 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6155 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6157 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6158 could be overwritten at the end of the current message (or the start of a
6159 new message if it was set in a HELO ACL). The value is now preserved for
6160 the duration of the SMTP connection.
6162 12. If a transport had a headers_rewrite setting, and a matching header line
6163 contained an unqualified address, that address was qualified, even if it
6164 did not match any rewriting rules. The underlying bug was that the values
6165 of the flags that permit the existence of unqualified sender and recipient
6166 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6167 non-local messages, and by -bnq for local messages) were not being
6168 preserved with the message after it was received.
6170 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6171 "next input=" as part of the text comprising the host identity instead of
6172 the correct text. The code was using the same buffer for two different
6173 strings. However, depending on which order the printing function evaluated
6174 its arguments, the bug did not always show up. Under Linux, for example, my
6175 test suite worked just fine.
6177 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6178 doesn't work with some older versions of Perl. It has been changed to "my",
6179 which in any case is probably the better facility to use.
6181 15. A really picky compiler found some instances of statements for creating
6182 error messages that either had too many or two few arguments for the format
6185 16. The size of the buffer for calls to the DNS resolver has been increased
6186 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6187 for addresses that have a lot of PTR records. This alleviates a problem; it
6188 does not fully solve it.
6190 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6191 buffer now truncates the list and logs the incident, which is the same
6192 action as happens when Exim is looking up a host name and its aliases.
6193 Previously in this situation something unpredictable would happen;
6194 sometimes it was "internal error: store_reset failed".
6196 18. If a server dropped the connection unexpectedly when an Exim client was
6197 using GnuTLS and trying to read a response, the client delivery process
6198 crashed while trying to generate an error log message.
6200 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6201 string, for example:
6203 warn message = H1: something\nH2: something
6205 the text was added as a single header line from Exim's point of view
6206 though it ended up OK in the delivered message. However, searching for the
6207 second and subsequent header lines using $h_h2: did not work. This has been
6208 fixed. Similarly, if a system filter added multiple headers in this way,
6209 the routers could not see them.
6211 20. Expanded the error message when iplsearch is called with an invalid key to
6212 suggest using net-iplsearch in a host list.
6214 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6215 ACLs are no longer actually imposed (and a message to that effect is
6218 22. If a "gecos" field in a passwd entry contained escaped characters, in
6219 particular, if it contained a \" sequence, Exim got it wrong when building
6220 a From: or a Sender: header from that name. A second bug also caused
6221 incorrect handling when an unquoted " was present following a character
6222 that needed quoting.
6224 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6225 was not being matched caselessly.
6227 24. Arranged for all hyphens in the exim.8 source to be escaped with
6230 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6231 properly. Recipient callout cache records were still being keyed to include
6232 the sender, even when use_sender was set false. This led to far more
6233 callouts that were necessary. The sender is no longer included in the key
6234 when use_sender is false.
6236 26. Added "control = submission" modifier to ACLs.
6238 27. Added the ${base62d: operator to decode base 62 numbers.
6240 28. dnsdb lookups can now access SRV records.
6242 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6243 the configuration file.
6245 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6246 (-v) mode. This makes the output for a verbose queue run more intelligible.
6248 31. Added a use_postmaster feature to recipient callouts.
6250 32. Added the $body_zerocount variable, containing the number of binary zero
6251 bytes in the message body.
6253 33. The time of last modification of the "new" subdirectory is now used as the
6254 "mailbox time last read" when there is a quota error for a maildir
6257 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6259 35. Added +ignore_unknown as a special item in host lists.
6261 36. Code for decoding IPv6 addresses in host lists is now included, even if
6262 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6263 address was recognized as an IP address, but was then not correctly decoded
6264 into binary, causing unexpected and incorrect effects when compared with
6271 1. Very minor rewording of debugging text in manualroute to say "list of
6272 hosts" instead of "hostlist".
6274 2. If verify=header_syntax was set, and a header line with an unqualified
6275 address (no domain) and a large number of spaces between the end of the
6276 name and the colon was received, the reception process suffered a buffer
6277 overflow, and (when I tested it) crashed. This was caused by some obsolete
6278 code that should have been removed. The fix is to remove it!
6280 3. When running in the test harness, delay a bit after writing a bounce
6281 message to get a bit more predictability in the log output.
6283 4. Added a call to search_tidyup() just before forking a reception process. In
6284 theory, someone could use a lookup in the expansion of smtp_accept_max_
6285 per_host which, without the tidyup, could leave open a database connection.
6287 5. Added the variables $recipient_data and $sender_data which get set from a
6288 lookup success in an ACL "recipients" or "senders" condition, or a router
6289 "senders" option, similar to $domain_data and $local_part_data.
6291 6. Moved the writing of debug_print from before to after the "senders" test
6294 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6295 problems for message scanning, either using a data ACL, or using
6296 local_scan() because the Received: header was not generated till after they
6297 were called (in order to set the time as the time of reception completion).
6298 I have revised the way this works. The header is now generated after the
6299 body is received, but before the ACL or local_scan() are called. After they
6300 are run, the timestamp in the header is updated.
6306 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6307 before starting a queue runner without re-exec. This happened only when
6308 deliver_drop_privilege was set or when the Exim user was set to root. The
6309 effect of the bug was that timeouts during subsequent deliveries caused
6310 crashes instead of being properly handled. The handler is now left at its
6311 default (and expected) setting.
6313 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6314 message, again when deliver_drop_privilege is set or Exim is run as root.
6315 The bug described in (1) was not present in this case, but the tidying up
6316 of the other signals was missing. I have made the two cases consistent.
6318 3. The ignore_target_hosts setting on a manualroute router was being ignored
6319 for hosts that were looked up using the /MX notation.
6321 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6324 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6325 operated on the sender address. After changing the $sender_address to <>
6326 for the sender address verify, Exim was re-instated it as the original
6327 (before rewriting) address, but remembering that it had rewritten it, so it
6328 wasn't rewriting it again. This bug also had the effect of breaking the
6329 sender address verification caching when the sender address was rewritten.
6331 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6332 This has been changed so that if the ip literal address matches
6333 ignore_target_hosts, the router declines.
6335 7. Added expansion conditions match_domain, match_address, and match_local_
6336 part (NOT match_host).
6338 8. The placeholder for the Received: header didn't have a length field set.
6340 9. Added code to Exim itself and to exim_lock to test for a specific race
6341 condition that could lead to file corruption when using MBX delivery. The
6342 issue is with the lockfile that is created in /tmp. If this file is removed
6343 after a process has opened it but before that process has acquired a lock,
6344 there is the potential for a second process to recreate the file and also
6345 acquire a lock. This could lead to two Exim processes writing to the file
6346 at the same time. The added code performs the same test as UW imapd; it
6347 checks after acquiring the lock that its file descriptor still refers to
6348 the same named file.
6350 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6351 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6354 11. Added the smtp_active_hostname option. If used, this will typically be made
6355 to depend on the incoming interface address. Because $interface_address is
6356 not set up until the daemon has forked a reception process, error responses
6357 that can happen earlier (such as "too many connections") no longer contain
6360 12. If an expansion in a condition on a "warn" statement fails because a lookup
6361 defers, the "warn" statement is abandoned, and the next ACL statement is
6362 processed. Previously this caused the whole ACL to be aborted.
6364 13. Added the iplsearch lookup type.
6366 14. Added ident_timeout as a log selector.
6368 15. Added tls_certificate_verified as a log selector.
6370 16. Added a global option tls_require_ciphers (compare the smtp transport
6371 option of the same name). This controls incoming TLS connections.
6373 17. I finally figured out how to make tls_require_ciphers do a similar thing
6374 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6375 before starting the TLS session.
6377 18. Tabs are now shown as \t in -bP output.
6379 19. If the log selector return_path_on_delivery was set, Exim crashed when
6380 bouncing a message because it had too many Received: header lines.
6382 20. If two routers both had headers_remove settings, and the first one included
6383 a superfluous trailing colon, the final name in the first list and the
6384 first name in the second list were incorrectly joined into one item (with a
6385 colon in the middle).
6391 1. Added -C and -D options to the exinext utility, mainly to make it easier
6392 to include in the automated testing, but these could be helpful when
6393 multiple configurations are in use.
6395 2. The exinext utility was not formatting the output nicely when there was
6396 an alternate port involved in the retry record key, nor when there was a
6397 message id as well (for retries that were specific to a specific message
6398 and a specific host). It was also confused by IPv6 addresses, because of
6399 the additional colons they contain. I have fixed the IPv4 problem, and
6400 patched it up to do a reasonable job for IPv6.
6402 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6403 delivery, the log line now contains "pipelined" if PIPELINING was used.
6405 4. An SMTP transport process used to panic and die if the bind() call to set
6406 an explicit outgoing interface failed. This has been changed; it is now
6407 treated in the same way as a connect() failure.
6409 5. A reference to $sender_host_name in the part of a conditional expansion
6410 that was being skipped was still causing a DNS lookup. This no longer
6413 6. The def: expansion condition was not recognizing references to header lines
6414 that used bh_ and bheader_.
6416 7. Added the _cache feature to named lists.
6418 8. The code for checking quota_filecount in the appendfile transport was
6419 allowing one more file than it should have been.
6421 9. For compatibility with Sendmail, the command line option
6429 and sets the incoming protocol and host name (for trusted callers). The
6430 host name and its colon can be omitted when only the protocol is to be set.
6431 Note the Exim already has two private options, -pd and -ps, that refer to
6432 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6433 "s", but I don't think that's a major issue.
6435 10. A number of refactoring changes to the code, none of which should affect
6438 (a) The number of logging options was getting close to filling up the
6439 32-bit word that was used as a bit map. I have split them into two classes:
6440 those that are passed in the argument to log_write(), and those that are
6441 only ever tested independently outside of that function. These are now in
6442 separate 32-bit words, so there is plenty of room for expansion again.
6443 There is no change in the user interface or the logging behaviour.
6445 (b) When building, for example, log lines, the code previously used a
6446 macro that called string_cat() twice, in order to add two strings. This is
6447 not really sufficiently general. Furthermore, there was one instance where
6448 it was actually wrong because one of the argument was used twice, and in
6449 one call a function was used. (As it happened, calling the function twice
6450 did not affect the overall behaviour.) The macro has been replaced by a
6451 function that can join an arbitrary number of extra strings onto a growing
6454 (c) The code for expansion conditions now uses a table and a binary chop
6455 instead of a serial search (which was left over from when there were very
6456 few conditions). Also, it now recognizes conditions like "pam" even when
6457 the relevant support is not compiled in: a suitably worded error message is
6458 given if an attempt is made to use such a condition.
6460 11. Added ${time_interval:xxxxx}.
6462 12. A bug was causing one of the ddress fields not to be passed back correctly
6463 from remote delivery subprocesses. The field in question was not being
6464 subsequently used, so this caused to problems in practice.
6466 13. Added new log selectors queue_time and deliver_time.
6468 14. Might have fixed a bug in maildirsizefile handling that threw up
6469 "unexpected character" debug warnings, and recalculated the data
6470 unnecessarily. In any case, I expanded the warning message to give more
6473 15. Added the message "Restricted characters in address" to the statements in
6474 the default ACL that block characters like @ and % in local parts.
6476 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6477 Three changes have been made:
6479 (a) There was a serious bug; a negative response to MAIL caused the whole
6480 recipient domain to be cached as invalid, thereby blocking all messages
6481 to all local parts at the same domain, from all senders. This bug has
6482 been fixed. The domain is no longer cached after a negative response to
6483 MAIL if the sender used is not empty.
6485 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6488 (c) A new callout option, "use_sender" has been added for people who want
6489 the modified behaviour.
6495 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6498 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6499 indeed breaks things for older releases.
6501 3. Added additional logging to the case where there is a problem reading data
6502 from a filter that is running in a subprocess using a pipe, in order to
6503 try to track down a specific problem.
6505 4. Testing facility fudge: when running in the test harness and attempting
6506 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6507 getting "No route to host". Convert this to a timeout.
6509 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6512 6. Some OS don't have socklen_t but use size_t instead. This affects the
6513 fifth argument of getsockopt() amongst other things. This is now
6514 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6515 can be set for individual OS. I have set it for SunOS5, OSF1, and
6516 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6517 some earlier ones do not.
6519 7. Change 4.30/15 was not doing the test caselessly.
6521 8. The standard form for an IPv6 address literal was being rejected by address
6522 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6523 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6524 this, as well as the form without the "IPv6" on the front (but only when
6525 address literals are enabled, of course).
6527 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6529 10. Exim crashed if a message with an empty sender address specified by -f
6530 encountered a router with an errors_to setting. This could be provoked only
6531 by a command such as
6535 where an empty string was supplied; "<>" did not hit this bug.
6537 11. Installed PCRE release 4.5.
6539 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6540 remained set. It is now erased.
6542 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6543 times from message ids (which are base 36 rather than the normal 62).
6545 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6546 were being counted as actual protocol errors, and logged if the log
6547 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6548 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6549 and DATA following a set of rejected RCPTs do not count as protocol errors.
6550 In other words, Exim assumes they were pipelined, though this may not
6551 actually be the case. Of course, in all cases the client gets an
6552 appropriate error code.
6554 15. If a lookup fails in an ACL condition, a message about the failure may
6555 be available; it is used if testing the ACL cannot continue, because most
6556 such messages specify what the cause of the deferral is. However, some
6557 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6558 that caused an old message to be retained and used if a later statement
6559 caused a defer, replacing the real cause of the deferral.
6561 16. If an IP address had so many PTR records that the DNS lookup buffer
6562 was not large enough to hold them, Exim could crash while trying to process
6563 the truncated data. It now detects and logs this case.
6565 17. Further to 4.21/58, another change has been made: if (and only if) the
6566 first line of a message (the first header line) ends with CRLF, a bare LF
6567 in a subsequent header line has a space inserted after it, so as not to
6568 terminate the header.
6570 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6571 unnecessarily, used atoi() instead of strtol(), and didn't check the
6572 termination when getting file sizes from file names by regex.
6574 19. Completely re-implemented the support for maildirsize files, in the light
6575 of a number of problems with the previous contributed implementation
6576 (4.30/29). In particular:
6578 . If the quota is zero, the maildirsize file is maintained, but no quota is
6581 . If the maildir directory does not exist, it is created before any attempt
6582 to write a maildirsize file.
6584 . The quota value in the file is just a cache; if the quota is changed in
6585 the transport, the new value overrides.
6587 . A regular expression is available for excluding directories from the
6590 20. The autoreply transport checks the characters in options that define the
6591 message's headers; it allows continued headers, but it was checking with
6592 isspace() after an embedded newline instead of explicitly looking for a
6595 21. If all the "regular" hosts to which an address was routed had passed their
6596 expiry times, and had not reached their retry times, the address was
6597 bounced, even if fallback hosts were defined. Now Exim should go on to try
6600 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6601 equivalent code in the SMTP transport. Some hosts send humungous responses
6602 to HELO/EHLO, more than 1024 it seems.
6604 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6605 gives compiler warnings in some environments. I've now done it "properly",
6608 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6609 (because of problems with the built-in one) was declared to return uschar *
6610 instead of char *, causing compiler failure.
6612 25. Fixed a file descriptor leak when processing alias/forward files.
6614 26. Fixed a minor format string issue in dbfn.c.
6616 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6618 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6619 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6622 29. When a sender address is verified, it is cached, to save repeating the test
6623 when there is more than one recipient in a message. However, when the
6624 verification involves a callout, it is possible for different callout
6625 options to be set for different recipients. It is too complicated to keep
6626 track of this in the cache, so now Exim always runs a verification when a
6627 callout is required, relying on the callout cache for the optimization.
6628 The overhead is duplication of the address routing, but this should not be
6631 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6632 to be verified with callout=postmaster, and the main callout worked but the
6633 postmaster check failed, the verification correctly failed. However, if a
6634 subsequent RCPT command asked for sender verification *without* the
6635 postmaster check, incorrect caching caused this verification also to fail,
6638 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6639 it was not caching the DNS options (qualify_single, search_parents) that
6640 were used when the lookup failed. A subsequent lookup with different
6641 options therefore always gave the same answer, though there were cases
6642 where it should not have. (Example: a "domains = !$mx_any" option on a
6643 dnslookup router: the "domains" option is always processed without any
6644 widening, but the router might have qualify_single set.) Now Exim uses the
6645 cached value only when the same options are set.
6647 32. Added John Jetmore's "exipick" utility to the distribution.
6649 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6650 than a timeout (e.g. a certificate is required, and is not provided), an
6651 Exim server now closes the connection immediately. Previously it waited for
6652 the client to close - but if the client is SSL, it seems that they each
6653 wait for each other, leading to a delay before one of them times out.
6655 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6656 maintained 0.8.x compatibility because I don't think many are using it, and
6657 it is clearly obsolete.
6659 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6662 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6663 client certificate was expired. A simple patch fixes this, though I don't
6664 understand the full logic of why the verify callback is called multiple
6667 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6668 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6669 which causes problems with some clients (such as the Certicom SSL Plus
6670 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6671 disables the coutermeasure allowing Eudora to connect."
6673 38. Exim was not checking that a write() to a log file succeeded. This could
6674 lead to Bad Things if a log got too big, in particular if it hit a file
6675 size limit. Exim now panics and dies if it cannot write to a log file, just
6676 as it does if it cannot open a log file.
6678 39. Modified OS/Makefile-Linux so that it now contains
6680 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6682 The two -D definitions ensure that Exim is compiled with large file
6683 support, which makes it possible to handle log files that are bigger than
6686 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6687 instance) a domain was checked against a named list that involved a lookup,
6688 causing $domain_data to be set, then another domain was checked against the
6689 same list, then the first domain was re-checked, the value of $domain_data
6690 after the final check could be wrong. In particular, if the second check
6691 failed, it could be set empty. This bug probably also applied to
6694 41. The strip_trailing_dot option was not being applied to the address given
6695 with the -f command-line option.
6697 42. The code for reading a message's header from the spool was incrementing
6698 $received_count, but never initializing it. This meant that the value was
6699 incorrect (doubled) while delivering a message in the same process in which
6700 it was received. In the most common configuration of Exim, this never
6701 happens - a fresh exec is done - but it can happen when
6702 deliver_drop_privilege is set.
6704 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6705 it now includes up to 150 characters of the unexpected data in the log
6708 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6709 and building data strings. The size of both of these buffers was 10 000
6710 bytes - far larger than anybody would *ever* want, thought I. Needless to
6711 say, somebody hit the limit. I have increased the maximum line length to
6712 20 000 and the maximum data length of concatenated lines to 100 000. I have
6713 also fixed two bugs, because there was no checking on these buffers. Tsk,
6714 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6715 buffer is too small.
6717 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6718 lsearch lookups. Now it does.
6720 46. When parsing a route_list item in a manualroute router, a fixed-length
6721 buffer was used for the list of hosts. I made this 1024 bytes long,
6722 thinking that nobody would ever have a list of hosts that long. Wrong.
6723 Somebody had a whole pile of complicated expansion conditions, and the
6724 string was silently truncated, leading to an expansion error. It turns out
6725 that it is easier to change to an unlimited length (owing to other changes
6726 that have happened since this code was originally written) than to build
6727 structure for giving a limitation error. The length of the item that
6728 expands into the list of hosts is now unlimited.
6730 47. The lsearch lookup could not handle data where the length of text line was
6731 more than 4095 characters. Such lines were truncated, leading to shortened
6732 data being returned. It should now handle lines of any length.
6734 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6735 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6738 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6739 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6740 no longer happens in comment lines. A list of replacements is now placed
6741 at the head of all of the source files, except those whose only change is
6742 to replace PERL_COMMAND in the very first #! line.
6744 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6745 messages on the queue, with a bottom-up merge sort, using code contributed
6746 by Michael Haardt. This should make operations like -bp somewhat faster on
6747 large queues. It won't affect queue runners, except when queue_run_in_order
6750 51. Installed eximstats 1.31 in the distribution.
6752 52. Added support for SRV lookups to the dnslookup router.
6754 53. If an ACL referred to $message_body or $message_body_end, the value was not
6755 reset for any messages that followed in the same SMTP session.
6757 54. The store-handling optimization for building very long strings was not
6758 differentiating between the different store pools. I don't think this
6759 actually made any difference in practice, but I've tidied it.
6761 55. While running the routers to verify a sender address, $sender_address
6762 was still set to the sender address. This is wrong, because when routing to
6763 send a bounce to the sender, it would be empty. Therefore, I have changed
6764 it so that, while verifying a sender address, $sender_address is set to <>.
6765 (There is no change to what happens when verifying a recipient address.)
6767 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6768 target A or AAAA records (if not already returned) without resetting the
6769 qualify_single or search_parents options of the DNS resolver. These are
6770 inappropriate in this case because the targets of MX and SRV records must
6771 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6772 target that, when qualified, matched something in the local domain. These
6773 two options are now turned off when doing these lookups.
6775 57. It seems that at least some releases of Reiserfs (which does not have the
6776 concept of a fixed number of inodes) returns zero and not -1 for the
6777 number of available inodes. This interacted badly with check_spool_inodes,
6778 which assumed that -1 was the "no such thing" setting. What I have done is
6779 to check that the total number of inodes is greater than zero before doing
6780 the test of how many are available.
6782 58. When a "warn" ACL statement has a log_message modifier, the message is
6783 remembered, and not repeated. This is to avoid a lot of repetition when a
6784 message has many recipients that cause the same warning to be written.
6785 However, Exim was preserving the list of already written lines for an
6786 entire SMTP session, which doesn't seem right. The memory is now reset if a
6787 new message is started.
6789 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6790 debugging output unless log_rewrite was also set.
6792 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6793 of (int)(handle) when we know that handle contains (void *)(-1).
6795 61. The Exim daemon panic-logs an error return when it closes the incoming
6796 connection. However "connection reset by peer" seems to be common, and
6797 isn't really an error worthy of noting specially, so that particular error
6800 62. When Exim is trying to find all the local interfaces, it used to panic and
6801 die if the ioctl to get the interface flags failed. However, it seems that
6802 on at least one OS (Solaris 9) it is possible to have an interface that is
6803 included in the list of interfaces, but for which you get a failure error
6804 for this call. This happens when the interface is not "plumbed" into a
6805 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6806 failure of the "get flags" call assumes that the interface is down.
6808 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6809 makes life easier for people who are doing arithmetic on fields extracted
6810 from dates, where you often get leading zeros that should not be
6811 interpreted as octal.
6813 64. Added qualify_domain to the redirect router, to override the global
6816 65. If a pathologically long header line contained very many addresses (the
6817 report of this problem mentioned 10 000) and each of them was rewritten,
6818 Exim could use up a very large amount of memory. (It kept on making new
6819 copies of the header line as it rewrote, and never released the old ones.)
6820 At the expense of a bit more processing, the header rewriting function has
6821 been changed so that it no longer eats memory in this way.
6823 66. The generation of the Received: header has been moved from the time that a
6824 message starts to be received, to the time that it finishes. The timestamp
6825 in the Received: header should now be very close to that of the <= log
6826 line. There are two side-effects of this change:
6828 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6829 logged header lines no longer include the local Received: line, because
6830 it has not yet been created. The same applies to a copy of the message
6831 that is returned to a non-SMTP sender when a message is rejected.
6833 (b) When a filter file is tested using -bf, no additional Received: header
6834 is added to the test message. After some thought, I decided that this
6837 This change does not affect the value of $received_for. It is still set
6838 after address rewriting, but before local_scan() is called.
6840 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6842 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6843 gave an unhelpful panic error message, and a defer error. I have managed to
6844 change this behaviour so that it now rejects any supplied certificate,
6845 which seems right, as the list of acceptable certificates is empty.
6847 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6848 gave an unhelpful defer error. I have not managed to make this reject any
6849 supplied certificates, but the error message it gives is "no certificate
6850 supplied", which is not helpful.
6852 70. exigrep's output now also includes lines that are not associated with any
6853 message, but which match the given pattern. Implemented by a patch from
6854 Martin Sluka, which also tidied up the Perl a bit.
6856 71. Recipient callout verification, like sender verification, was using <> in
6857 the MAIL FROM command. This isn't really the right thing, since the actual
6858 sender may affect whether the remote host accepts the recipient or not. I
6859 have changed it to use the actual sender in the callout; this means that
6860 the cache record is now keyed on a recipient/sender pair, not just the
6861 recipient address. There doesn't seem to be a real danger of callout loops,
6862 since a callout by the remote host to check the sender would use <>.
6863 [SEE ABOVE: changed after hitting problems.]
6865 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6866 temporary errors. However, in the case of such a code being given after
6867 the end of a data transmission (i.e. after ".") Exim was failing to write
6868 a retry record for the message. (Yes, there was some broken host that was
6869 actually sending 8xx at this point.)
6871 73. An unknown lookup type in a host list could cause Exim to panic-die when
6872 the list was checked. (An example that provoked this was putting <; in the
6873 middle of a list instead of at the start.) If this happened during a DATA
6874 ACL check, a -D file could be left lying around. This kind of configuration
6875 error no longer causes Exim to die; instead it causes a defer error. The
6876 incident is still logged to the main and panic logs.
6878 74. Buglet left over from Exim 3 conversion. The message "too many messages
6879 in one connection" was written to the rejectlog but not the mainlog, except
6880 when address rewriting (yes!) was being logged.
6882 75. Added write_rejectlog option.
6884 76. When a system filter was run not as root (that is, when system_filter_user
6885 was set), the values of the $n variables were not being returned to the
6886 main process; thus, they were not subsequently available in the $sn
6889 77. Added +return_path_on_delivery log selector.
6891 78. A connection timeout was being treated differently from recipients deferred
6892 when testing hosts_max_try with a message that was older than the host's
6893 retry timeout. (The host should not be counted, thus allowing all hosts to
6894 be tried at least once before bouncing.) This may have been the cause of an
6895 occasionally reported bug whereby a message would remain on the queue
6896 longer than the retry timeout, but would be bounced if a delivery was
6897 forced. I say "may" because I never totally pinned down the problem;
6898 setting up timeout/retry tests is difficult. See also the next item.
6900 79. The ultimate address timeout was not being applied to errors that involved
6901 a combination of host plus message (for example, a timeout on a MAIL
6902 command). When an address resolved to a number of possible hosts, and they
6903 were not all tried for each delivery (e.g. because of hosts_max_try), a
6904 message could remain on the queue longer than the retry timeout.
6906 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6909 81. Fixed an obscure SMTP outgoing bug which required at least the following
6910 conditions: (a) there was another message waiting for the same server;
6911 (b) the server returned 5xx to all RCPT commands in the first message so
6912 that the message was not completed; (c) the server dropped the connection
6913 or gave a negative response to the RSET that Exim sends to abort the
6914 transaction. The observed case was a dropped connection after DATA that had
6915 been sent in pipelining mode. That is, the server had advertised PIPELINING
6916 but was not implementing it correctly. The effect of the bug was incorrect
6917 behaviour, such as trying another host, and this could lead to a crash.
6923 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6924 and daemon.c were passed as pointers to ints; they should have been
6925 pointers to socklen_t variables (which are typically unsigned ints).
6927 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6930 3. Fixed a really odd bug that affected only the testing scheme; patching a
6931 certain fixed string in the binary changed the value of another string that
6932 happened to be identical to the end of the original first string.
6934 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6935 name", it returns that address as the IP address. On some operating
6936 systems (e.g. Solaris), it also passes back the IP address string as the
6937 "host name". However, on others (e.g. Linux), it passes back an empty
6938 string. Exim wasn't checking for this, and was changing the host name to an
6939 empty string, assuming it had been canonicalized.
6941 5. Although rare, it is permitted to have more than one PTR record for a given
6942 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6943 all the names associated with an address, because they do in Solaris.
6944 However, it seems that they do not in Linux for data that comes from the
6945 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6946 I found this out when I moved to a new Linux workstation and tried to run
6947 the Exim test suite.
6949 To get round this problem I have changed the code so that it now does its
6950 own call to the DNS to look up PTR records when searching for a host name.
6951 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6952 addresses that are only in /etc/hosts are still found.
6954 This behaviour is, however, controlled by an option called host_lookup_
6955 order, which defaults to "bydns:byaddr". If people want to use the other
6956 order, or indeed, just use one or the other means of lookup, they can
6957 specify it in this variable.
6959 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6960 some operating systems, this comes back from gethostbyaddr() as an empty
6961 string, and this is what Exim used to test for. However, it seems that in
6962 other systems, "." is yielded. Exim now tests for this case too.
6964 7. The values of check_spool_space and check_log_space are now held internally
6965 as a number of kilobytes instead of an absolute number of bytes. If a
6966 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6967 kilobyte. This means that much larger values can be stored.
6969 8. Exim monitor: an attempt to get the action menu when not actually pointing
6970 at a message produces an empty menu entitled "No message selected". This
6971 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6972 no entries in it ("Shell widget menu has zero width and/or height"). So I
6973 have added a single, blank menu entry in this case.
6975 9. Added ${quote_local_part.
6977 10. MIME decoding is now applied to the contents of Subject: header lines when
6980 11. Now that a reference to $sender_host_address automatically causes a reverse
6981 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6982 host lookup before query-style lookups in lists that might use this
6983 variable. This has therefore been abolished, and the "net-" prefix is no
6984 longer necessary for query-style lookups.
6986 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6987 have been a typo for LFLAGS, so it has been changed.
6989 13. The install script calls Exim with "-C /dev/null" in order to find the
6990 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6991 to be output. However, since Exim outputs its version number before the
6992 error, it didn't break the script. It just looked ugly. I fixed this by
6993 always allowing "-C /dev/null" if the caller is root.
6995 14. Ignore overlarge ACL variable number when reading spool file - insurance
6996 against a later release with more variables having written the file.
6998 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6999 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7000 this, as well as the form without the "IPv6" on the front.
7002 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7003 OS/Makefile-Darwin file.
7005 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7006 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7008 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7009 variable. However, some OS use ints for inodes. I've added cast to long int
7010 to get rid of the compiler warning.
7012 19. I had forgotten to lock out "/../" in configuration file names when
7013 ALT_CONFIG_PREFIX was set.
7015 20. Routers used for verification do not need to specify transports. However,
7016 if such a router generated a host list, and callout was configured, Exim
7017 crashed, because it could not find a port number from the (non-existent)
7018 transport. It now assumes port 25 in this circumstance.
7020 21. Added the -t option to exigrep.
7022 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7023 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7024 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7025 them is set, LOOKUP_LSEARCH is forced.
7027 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7029 24. Added sender and host information to the "rejected by local_scan()" log
7030 line; previously there was no indication of these.
7032 25. Added .include_if_exists.
7034 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7035 Linux. It turns out that not all file systems support this. Apparently some
7036 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7037 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7038 not supported on the file descriptor, is now ignored when Exim is trying to
7039 sync a directory. This applies only to Linux.
7041 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7043 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7044 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7047 29. Added support for maildirsize files from supplied patch (modified a bit).
7049 30. The use of :fail: followed by an empty string could lead Exim to respond to
7050 sender verification failures with (e.g.):
7052 550 Verification failed for <xxx>
7053 550 Sender verify failed
7055 where the first response line was missing the '-' that indicates it is not
7056 the final line of the response.
7058 31. The loop for finding the name of the user that called Exim had a hardwired
7059 limit of 10; it now uses the value of finduser_retries, which is used for
7060 all other user lookups.
7062 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7065 33. Exim was neglecting to zero errno before one call of strtol() when
7066 expanding a string and expecting an integer value. On some systems this
7067 resulted in spurious "integer overflow" errors. Also, it was casting the
7068 result into an int without checking.
7070 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7071 did not work. The code looks as if it has *never* worked, though it appears
7072 to have been documented since at least release 1.62. I have made it work.
7074 35. The "timeout_DNS" error in retry rules, also documented since at least
7075 1.62, also never worked. As it isn't clear exactly what this means, and
7076 clearly it isn't a major issue, I have abolished the feature by treating it
7077 as "timeout", and writing a warning to the main and panic logs.
7079 36. The display of retry rules for -brt wasn't always showing the error code
7082 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7083 timeout_connect_A, timeout_connect_MX.
7085 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7086 to the empty sender.
7088 39. The daemon was not analysing the content of -oX till after it had closed
7089 stderr and disconnected from the controlling terminal. This meant that any
7090 syntax errors were only noted on the panic log, and the return code from
7091 the command was 0. By re-arranging the code a little, I've made the
7092 decoding happen first, so such errors now appear on stderr, and the return
7093 code is 1. However, the actual setting up of the sockets still happens in
7094 the disconnected process, so errors there are still only recorded on the
7097 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7098 connections (as happens on some IP stacks) was logged at start up time as
7099 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7100 it from "IPv6 and IPv4", which means that two separate sockets are being
7103 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7104 says whether AF_INET or AF_INET6 was passed as an argument.
7106 42. Exiwhat output was messed up when time zones were included in log
7109 43. Exiwhat now gives more information about the daemon's listening ports,
7110 and whether -tls-on-connect was used.
7112 44. The "port" option of the smtp transport is now expanded.
7114 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7115 silently ignored. Now an error message is written to the main and panic
7118 46. There's a new ACL modifier called "logwrite" which writes to a log file
7119 as soon as it is encountered.
7121 47. Added $local_user_uid and $local_user_gid at routing time.
7123 48. Exim crashed when trying to verify a sender address that was being
7126 49. Exim was recognizing only a space character after ".include". It now also
7127 recognizes a tab character.
7129 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7130 extracting the relevant information from the specification. The man page no
7131 longer contains scrambled data for the -d option, and I've added a section
7132 at the front about calling Exim under different names.
7134 51. Added "extra_headers" argument to the "mail" command in filter files.
7136 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7139 53. Installed eximstats 1.29.
7141 54. Added transport_filter_timeout as a generic transport option.
7143 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7144 Cc: header lines. This was required by RFC 822, but it not required by RFC
7147 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7148 incoming messages that did not have them. Now it does so only if the
7149 message originates locally, that is, if there is no associated remote host
7150 address. When Resent- header lines are present, this applies to the Resent-
7151 lines rather than the non-Resent- lines.
7153 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7154 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7156 58. Messages for configuration errors now include the name of the main
7157 configuration file - useful now that there may be more than one file in a
7158 list (.included file names were always shown).
7160 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7161 for those rare installations that do not start the daemon as root or run it
7162 setuid root. I've cut out the call to initgroups() if the daemon is not
7165 60. The Exim user and group can now be bound into the binary as text strings
7166 that are looked up at the start of Exim's processing.
7168 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7170 62. Added $mailstore_basename variable.
7172 63. Installed patch to sieve.c from Michael Haardt.
7174 64. When Exim failed to open the panic log after failing to open the main log,
7175 the original message it was trying to log was written to stderr and debug
7176 output, but if they were not available (the usual case in production), it
7177 was lost. Now it is written to syslog before the two lines that record the
7178 failures to open the logs.
7180 65. Users' Exim filters run in subprocesses under the user's uid. It is
7181 possible for a "deliver" command or an alias in a "personal" command to
7182 provoke an address rewrite. If logging of address rewriting is configured,
7183 this fails because the process is not running as root or exim. There may be
7184 a better way of dealing with this, but for the moment (because 4.30 needs
7185 to be released), I have disabled address rewrite logging when running a
7186 filter in a non-root, non-exim process.
7192 1. The buildconfig auxiliary program wasn't quoting the value set for
7193 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7194 not defined. This bug was masked in 4.22 by the effect that was fixed in
7197 2. Some messages that were rejected after a message id was allocated were
7198 shown as "incomplete" by exigrep. It no longer does this for messages that
7199 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7201 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7202 have allow_domain_literals set, the ID did not get logged in the <= line.
7203 Domain literals are now always recognized in Message-ID: header lines.
7205 4. The first argument for a ${extract expansion item is the key name or field
7206 number. Leading and trailing spaces in this item were not being ignored,
7207 causing some misleading effects.
7209 5. When deliver_drop_privilege was set, single queue runner processes started
7210 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7211 same command in the process it spins off) were not dropping privilege.
7213 6. When the daemon running as "exim" started a queue runner, it always
7214 re-executed Exim in the spun-off process. This is a waste of effort when
7215 deliver_drop_privilege is set. The new process now just calls the
7216 queue-runner function directly.
7222 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7225 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7226 ignored. Though the use of -oP was forcing the writing of a pid file, it
7227 was always written to the default place.
7229 3. If the message "no IP address found for host xxxx" is generated during
7230 incoming verification, it is now followed by identification of the incoming
7231 connection (so you can more easily find what provoked it).
7233 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7235 5. Added some features to "harden" Exim a bit more against certain attacks:
7237 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7238 be put in Local/Makefile. This is like the never_users runtime option,
7239 but it cannot be overridden. The default setting is "root".
7241 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7242 prefix string with which any file named in a -C command line option
7245 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7246 is retained for -C and -D only if the caller of Exim is root. Without
7247 it, the exim user may also use -C and -D and retain privilege.
7249 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7250 command line option is disabled.
7252 6. Macro names set by the -D option must start with an upper case letter, just
7253 like macro names defined in the configuration file.
7255 7. Added "dereference=" facility to LDAP.
7257 8. Two instances of the typo "uknown" in the source files are fixed.
7259 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7260 the Configure-Makefile script screwed up while processing it.
7262 10. Incorporated PCRE 4.4.
7264 11. The SMTP synchronization check was not operating right at the start of an
7265 SMTP session. For example, it could not catch a HELO sent before the client
7266 waited for the greeting. There is now a check for outstanding input at the
7267 point when the greeting is written. Because of the duplex, asynchronous
7268 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7269 way, but not yet received, when the check is performed.
7271 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7272 on TCP/IP sockets, because this apparently causes some broken clients to
7275 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7276 unchanged) from the Cygwin maintainer.
7278 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7279 was supported instead of testing for mbx. Effectively a typo.
7281 15. The spa authenticator server code was not checking that the input it
7282 received was valid base64.
7284 16. The debug output line for the "set" modifier in ACLs was not showing the
7285 name of the variable that was being set.
7287 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7289 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7290 lookup on its own. Something else was needed to trigger the lookup. For
7291 example, a match in host_lookup or the need for a host name in a host list.
7292 Now, if $sender_host_name is referenced and the host name has not yet been
7293 looked up, a lookup is performed. If the lookup fails, the variable remains
7294 empty, and $host_lookup_failed is set to "1".
7296 19. Added "eqi" as a case-independent comparison operator.
7298 20. The saslauthd authentication condition could segfault if neither service
7299 nor realm was specified.
7301 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7302 error message that was logged was misleading, and incoming SMTP
7303 connections were dropped. The message is now more accurate, and temporary
7304 errors are given to SMTP connections.
7306 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7307 (except RSET) with a 421 error, until QUIT is received. However, it was
7308 failing to send a response to QUIT.
7310 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7311 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7312 ACL is now run after the helo_try_verify_hosts code.
7314 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7315 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7316 case-independent, so other case variants are also recognized). Apparently
7317 some systems use these upper case variants.
7319 25. If more than two messages were waiting for the same host, and a transport
7320 filter was specified for the transport, Exim sent two messages over the
7321 same TCP/IP connection, and then failed with "socket operation on non-
7322 socket" when it tried to send the third.
7324 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7326 27. The extern definition of crypt16() in expand.c was not being excluded when
7327 the OS had its own crypt16() function.
7329 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7330 as a preferred synonym for return_size_limit, both as an option and as an
7333 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7335 30. Changed the default configuration ACL to relax the local part checking rule
7336 for addresses that are not in any local domains. For these addresses,
7337 slashes and pipe symbols are allowed within local parts, but the sequence
7338 /../ is explicitly forbidden.
7340 31. SPA server authentication was not clearing the challenge buffer before
7343 32. log_message in a "warn" ACL statement was writing to the reject log as
7344 well as to the main log, which contradicts the documentation and doesn't
7345 seem right (because no rejection is happening). So I have stopped it.
7347 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7348 However, I am unable to do any testing of this.
7350 34. Fixed an infelicity in the appendfile transport. When checking directories
7351 for a mailbox, to see if any needed to be created, it was accidentally
7352 using path names with one or more superfluous leading slashes; tracing
7353 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7355 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7356 contents are added to the log line that is written for every discarded
7357 recipient. (Previously a log_message setting was ignored.)
7359 36. The ${quote: operator now quotes the string if it is empty.
7361 37. The install script runs exim in order to find its version number. If for
7362 some reason other than non-existence or emptiness, which it checks, it
7363 could not run './exim', it was installing it with an empty version number,
7364 i.e. as "exim-". This error state is now caught, and the installation is
7367 38. An argument was missing from the function that creates an error message
7368 when Exim fails to connect to the socket for saslauthd authentication.
7369 This could cause Exim to crash, or give a corrupted message.
7371 39. Added isip, isip4, and isip6 to ${if conditions.
7373 40. The ACL variables $acl_xx are now saved with the message, and can be
7374 accessed later in routers, transports, and filters.
7376 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7377 strings in the file are not string-expanded.
7379 42. If a MAIL command specified a SIZE value that was too large to fit into an
7380 int variable, the check against message_size_limit failed. Such values are
7381 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7382 day this will have to be increased, but I don't think I want to be around
7383 when emails are that large.
7390 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7391 iconv() is not standard in FreeBSD.
7393 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7394 IPv6 enabled. The observed symptom was a segmentation fault on return from
7395 the function os_common_find_running_interfaces() in src/os.c.
7397 3. In the check_special_case() function in daemon.c I had used "errno" as an
7398 argument name, which causes warnings on some systems. This was basically a
7399 typo, since it was named "eno" in the comments!
7401 4. The code that waits for the clock to tick (at a resolution of some fraction
7402 of a second) so as to ensure message-id uniqueness was always waiting for
7403 at least one whole tick, when it could have waited for less. [This is
7404 almost certainly not relevant at current processor speeds, where it is
7405 unlikely to ever wait at all. But we try to future-proof.]
7407 5. The function that sleeps for a time interval that includes fractions of a
7408 second contained a race. It did not block SIGALRM between setting the
7409 timer, and suspending (a couple of lines later). If the interval was short
7410 and the sigsuspend() was delayed until after it had expired, the suspension
7411 never ended. On busy systems this could lead to processes getting stuck for
7414 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7415 process, before it forks any delivery processes. The open lookup caching
7416 mechanism meant that the open file or database connection was passed into
7417 the delivery process. The problem was that delivery processes always tidy
7418 up cached lookup data. This could cause a problem for the next delivery
7419 process started by the queue runner, because the external queue runner
7420 process does not know about the closure. So the next delivery process
7421 still has data in the lookup cache. In the case of a file lookup, there was
7422 no problem because closing a file descriptor in a subprocess doesn't affect
7423 the parent. However, if the lookup was caching a connection to a database,
7424 the connection was closed, and the second delivery process was likely to
7425 see errors such as "PGSQL: query failed: server closed the connection
7426 unexpectedly". The problem has been fixed by closing all cached lookups
7427 in a queue runner before running a delivery process.
7429 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7430 seem to have the "const" qualifier which it has on other OS. I've
7433 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7434 *of the same type* (client or server) with the same public name that an
7435 error should be diagnosed.
7437 9. When Exim looked up a host name for an IP address, but failed to find the
7438 original IP address when looking up the host name (a safety check), it
7439 output the message "<ip address> does not match any IP for NULL", which was
7440 confusing, to say the least. The bug was that the host name should have
7441 appeared instead of "NULL".
7443 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7444 user that is built into the binary, and the -C or -D options is used, root
7445 privilege is dropped before the configuration file is read. In addition,
7446 logging is switched to stderr instead of the normal log files. If the
7447 configuration then re-defines the Exim user, the unprivileged environment
7448 is probably not what is expected, so Exim logs a panic warning message (but
7451 However, if deliver_drop_privilege is set, the unprivileged state may well
7452 be exactly what is intended, so the warning has been cut out in that case,
7453 and Exim is allowed to try to write to its normal log files.
7459 1. smtp_return_error_details was not giving details for temporary sender
7460 or receiver verification errors.
7462 2. Diagnose a configuration error if two authenticators have the same public
7465 3. Exim used not to create the message log file for a message until the first
7466 delivery attempt. This could be confusing when incoming messages were held
7467 for policy or load reasons. The message log file is now created at the time
7468 the message is received, and an initial "Received" line is written to it.
7470 4. The automatically generated man page for command line options had a minor
7471 bug that caused no ill effects; however, a more serious problem was that
7472 the procedure for building the man page automatically didn't always
7473 operate. Consequently, release 4.20 contains an out-of-date version. This
7474 shouldn't happen again.
7476 5. When building Exim with embedded Perl support, the script that builds the
7477 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7478 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7480 6. The freeze_tell option was not being used for messages that were frozen on
7481 arrival, either by an ACL or by local_scan().
7483 7. Added the smtp_incomplete_transaction log selector.
7485 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7486 was accepting AUTH without a new EHLO.
7488 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7489 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7492 10. Logging of TCP/IP connections (when configured) now happens in the main
7493 daemon process instead of the child process, so that the TCP/IP connection
7494 count is more accurate (but it can never be perfect).
7496 11. The use of "drop" in a nested ACL was not being handled correctly in the
7497 outer ACL. Now, if condition failure induced by the nested "drop" causes
7498 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7499 or "require"), the connection is dropped.
7501 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7502 that yield "discard" can now be used with an "accept" or a "discard" verb,
7503 but an error is generated for any others (because I can't see a useful way
7504 to define what should happen).
7506 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7507 are now processed in the same way as lines in the Exim configuration file.
7508 In particular, continuation lines are supported.
7510 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7512 15. Added -ti meaning -t -i.
7514 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7515 domains, and warn by logging if others are found.
7517 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7518 returned from GIFCONF when Exim is trying to find the list of interfaces on
7519 a host. The code in os.c has been modified to copy each ifreq to an aligned
7520 structure in all cases.
7522 Also, in some cases, the returned ifreq's were being copied to a 'struct
7523 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7524 means the last couple of bytes of an IPv6 address could be chopped if the
7525 ifreq contained only a normal sockaddr (14 bytes storage).
7527 18. Named domain lists were not supported in the hosts_treat_as_local option.
7528 An entry such as +xxxx was not recognized, and was treated as a literal
7531 19. Ensure that header lines added by a DATA ACL are included in the reject log
7532 if the ACL subsequently rejects the message.
7534 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7535 MD5 (which is deprecated).
7537 21. When testing a filter file using -bf, Exim was writing a message when it
7538 took the sender from a "From " line in the message, but it was not doing so
7539 when it took $return_path from a Return-Path: header line. It now does.
7541 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7542 with a valid header line field name (a series of printing characters
7543 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7545 23. Changed "disc" in the source to "disk" to conform to the documentation and
7546 the book and for uniformity.
7548 24. Ignore Sendmail's -Ooption=value command line item.
7550 25. When execve() failed while trying to run a command in a pipe transport,
7551 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7552 could be confused with a return value of 69 from the command itself. This
7553 has been changed to 127, the value the shell returns if it is asked to run
7554 a non-existent command. The wording for the related log line suggests a
7555 non-existent command as the problem.
7557 26. If received_header_text expands to an empty string, do not add a Received:
7558 header line to the message. (Well, it adds a token one on the spool, but
7559 marks it "old" so that it doesn't get used or transmitted.)
7561 27. Installed eximstats 1.28 (addition of -nt option).
7563 28. There was no check for failure on the call to getsockname() in the daemon
7564 code. This can fail if there is a shortage of resources on the system, with
7565 ENOMEM, for example. A temporary error is now given on failure.
7567 29. Contrary to the C standard, it seems that in some environments, the
7568 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7569 program. Exim now does this explicitly; it affects the formatting of
7570 timestamps using strftime().
7572 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7573 complaints. I've now initialized all the variables, to avoid this.
7575 32. Header lines added by a system filter were not being "seen" during
7576 transport-time rewrites.
7578 33. The info_callback() function passed to OpenSSL is set up with type void
7579 (*)(SSL *, int, int), as described somewhere. However, when calling the
7580 function (actually a macro) that sets it up, the type void(*)() is
7581 expected. I've put in a cast to prevent warnings from picky compilers.
7583 34. If a DNS black list lookup found a CNAME record, but there were no A
7584 records associated with the domain it pointed at, Exim crashed.
7586 35. If a DNS black list lookup returned more than one A record, Exim ignored
7587 all but the first. It now scans all returned addresses if a particular IP
7588 value is being sought. In this situation, the contents of the
7589 $dnslist_value variable are a list of all the addresses, separated by a
7592 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7593 to accept a host name and all its aliases if the forward lookup for any of
7594 them yielded the IP address of the incoming connection. Now it accepts only
7595 those names whose forward lookup yields the correct IP address. Any other
7596 names are discarded. This closes a loophole whereby a rogue DNS
7597 administrator could create reverse DNS records to break through a
7598 wildcarded host restriction in an ACL.
7600 37. If a user filter or a system filter that ran in a subprocess used any of
7601 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7602 the wrong values were passed to the pipe command ($thisaddress had the
7603 value of $0, $0 had the value of $1, etc). This bug was introduced by
7604 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7606 38. Improved the line breaking for long SMTP error messages from ACLs.
7607 Previously, if there was no break point between 40 and 75 characters, Exim
7608 left the rest of the message alone. Two changes have been made: (a) I've
7609 reduced the minimum length to 35 characters; (b) if it can't find a break
7610 point between 35 and 75 characters, it looks ahead and uses the first one
7611 that it finds. This may give the occasional overlong line, but at least the
7612 remaining text gets split now.
7614 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7615 file descriptors might be low, and that setting 1000 would always raise it.
7616 It turns out that in some environments, the limit is already over 1000 and
7617 that lowering it causes trouble. So now Exim takes care not to decrease it.
7619 40. When delivering a message, the value of $return_path is set to $sender_
7620 address at the start of routing (routers may change the value). By an
7621 oversight, this default was not being set up when an address was tested by
7622 -bt or -bv, which affected the outcome if any router or filter referred to
7625 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7626 log only when "message" or "log_message" are set. However, if one of the
7627 conditions was an address verification, or a call to a nested ACL, the
7628 messages generated by the underlying test were being passed through. This
7629 no longer happens. The underlying message is available in $acl_verify_
7630 message for both "message" and "log_message" expansions, so it can be
7631 passed through if needed.
7633 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7634 new expansion $bh_ to give the encoded byte string without charset
7635 translation. Translation happens only if iconv() is available; HAVE_ICONV
7636 indicates this at build time. HEADERS_CHARSET gives the charset to
7637 translate to; headers_charset can change it in the configuration, and
7638 "headers charset" can change it in an individual filter file.
7640 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7641 that creates RFC 2047 encoded "words" labels them as that charset instead
7642 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7643 expansion operator; (ii) when Exim creates a From: line for a local
7644 message; (iii) when a header line is rewritten to include a "phrase" part.
7646 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7647 buggy, causing it to skip the first lines of messages whose message ID
7648 ended in 'D'. This would not have bitten before Exim release 4.14, because
7649 message IDs were unlikely to end in 'D' before then. The effect was to have
7650 incorrect size information for certain domains.
7652 45. #include "config.h" was missing at the start of the crypt16.c module. This
7653 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7656 46. If there was a timeout during a "random" callout check, Exim treated it as
7657 a failure of the random address, and carried on sending RSET and the real
7658 address. If the delay was just some slowness somewhere, the response to the
7659 original RCPT would be taken as a response to RSET and so on, causing
7660 mayhem of various kinds.
7662 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7663 when I implemented it. It didn't allow for the fact that some option values
7664 may legitimately be negative (e.g. size_addition), and it didn't even do
7665 the right test for positive values.
7667 48. Domain names in DNS records are case-independent. Exim always looks them up
7668 in lower case. Some resolvers return domain names in exactly the case they
7669 appear in the zone file, that is, they may contain uppercase letters. Not
7670 all resolvers do this - some return always lower case. Exim was treating a
7671 change of case by a resolver as a change of domain, similar to a widening
7672 of a domain abbreviation. This triggered its re-routing code and so it was
7673 trying to route what was effectively the same domain again. This normally
7674 caused routing to fail (because the router wouldn't handle the domain
7675 twice). Now Exim checks for this case specially, and just changes the
7676 casing of the domain that it ultimately uses when it transmits the message
7679 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7682 50. If a filter generated a file delivery with a non-absolute name (possible if
7683 no home directory exists for the router), the forbid_file option was not
7686 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7687 the existing equality matching.
7689 52. Exim was using ints instead of ino_t variables in some places where it was
7690 dealing with inode numbers.
7692 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7693 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7694 called TMPDIR, and if it finds it is different, it changes its value.
7696 54. The smtp_printf() function is now made available to local_scan() so
7697 additional output lines can be written before returning. There is also an
7698 smtp_fflush() function to enable the detection of a dropped connection.
7699 The variables smtp_input and smtp_batched_input are exported to
7702 55. Changed the default runtime configuration: the message "Unknown user"
7703 has been removed from the ACL, and instead placed on the localuser router,
7704 using the cannot_route_message feature. This means that any verification
7705 failures that generate their own messages won't get overridden. Similarly,
7706 the "Unrouteable address" message that was in the ACL for unverifiable
7707 relay addresses has also been removed.
7709 56. Added hosts_avoid_esmtp to the smtp transport.
7711 57. The exicyclog script was not checking for the esoteric option
7712 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7713 will work only if exicyclog is run under the appropriate euid.
7715 58. Following a discussion on the list, the rules by which Exim recognises line
7716 endings on incoming messages have been changed. The -dropcr and drop_cr
7717 options are now no-ops, retained only for backwards compatibility. The
7718 following line terminators are recognized: LF CRLF CR. However, special
7719 processing applies to CR:
7721 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7722 nor a local message in the state where . is a terminator.
7724 (ii) If a bare CR is encountered in a header line, an extra space is added
7725 after the line terminator so as not to end the header. The reasoning
7726 behind this is that bare CRs in header lines are most likely either
7727 to be mistakes, or people trying to play silly games.
7729 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7730 was being incorrectly given as 18 bytes larger than it should have been.
7731 This is a VOB (very old bug).
7733 60. This may never have affected anything current, but just in case it has:
7734 When the local host is found other than at the start of a list of hosts,
7735 the local host, those with the same MX, and any that follow, are discarded.
7736 When the list in question was part of a longer list of hosts, the following
7737 hosts (not currently being processed) were also being discarded. This no
7738 longer happens. I'm not sure if this situation could ever has previously
7741 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7744 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7745 header. This is something that is recommended in a new Internet Draft, and
7746 is something that is documented as being done by Sendmail. There are two
7747 possible values. For messages generated by the autoreply transport, Exim
7750 Auto-Submitted: auto-replied
7752 whereas for all other generated messages (e.g. bounces) it adds
7754 Auto-Submitted: auto-generated
7756 63. The "personal" condition in filters now includes a test for the
7757 Auto-Submitted: header. If it contains the string "auto-" the message it
7758 not considered personal.
7760 64. Added rcpt_include_affixes as a generic transport option.
7762 65. Added queue_only_override (default true).
7764 66. Added the syslog_duplication option.
7766 67. If what should have been the first header line of a message consisted of
7767 a space followed by a colon, Exim was mis-interpreting it as a header line.
7768 It isn't of course - it is syntactically invalid and should therefore be
7769 treated as the start of the message body. The misbehaviour could have
7770 caused a number of strange effects, including loss of data in subsequent
7771 header lines, and spool format errors.
7773 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7774 client host had authenticated. This control can now be exercised by an ACL
7775 for more flexibility.
7777 69. By default, callouts do not happen when testing with -bh. There is now a
7778 variant, -bhc, which does actually run the callout code, including
7779 consulting and updating the callout cache.
7781 70. Added support for saslauthd authentication, courtesy of Alexander
7784 71. If statvfs() failed on the spool or log directories while checking their
7785 size for availability, Exim confusingly gave the error "space shortage".
7786 Furthermore, in debugging mode it crashed with a floating point exception.
7787 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7788 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7789 serious problem, Exim now writes to the main and panic logs when this
7790 happens, with details of the failure. It then refuses to accept the
7791 incoming message, giving the message "spool directory problem" or "log
7792 directory problem" with a 421 code for SMTP messages.
7794 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7795 0, 1, and 2 exist, because some OS complain for execs without them (see
7796 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7797 descriptors. However, the code omitted to check that the open succeeded,
7798 causing mysterious errors if for some reason the permissions on /dev/null
7799 got screwed. Now Exim writes a message to the main and panic logs, and
7800 bombs out if it can't open /dev/null.
7802 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7803 interact so that it is all more flexible. It is supposed to remain
7804 backwards compatible. Also added extra_local_interfaces.
7806 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7807 to bomb out with an assertion failure - to the client this appears as a
7808 connection drop. This problem occurs in the part of the code that was taken
7809 from the Samba project. Fortunately, the assertion is in a very simple
7810 function, so I have fixed this by reproducing the function inline in the
7811 one place where it is called, and arranging for authentication to fail
7812 instead of killing the process with assert().
7814 75. The SPA client code was not working when the server requested OEM rather
7815 than Unicode encoding.
7817 76. Added code to make require_files with a specific uid setting more usable in
7818 the case where statting the file as root fails - usually a non-root-mounted
7819 NFS file system. When this happens and the failure is EACCES, Exim now
7820 forks a subprocess and does the per-uid checking as the relevant uid.
7822 77. Added process_log_path.
7824 78. If log_file_path was not explicitly set, a setting of check_log_space or
7825 check_log_inodes was ignored.
7827 79. If a space check for the spool or log partitions fails, the incident is now
7828 logged. Of course, in the latter case the data may get lost...
7830 80. Added the %p formatting code to string_format() so that it can be used to
7831 print addresses in debug_print(). Adjusted all the address printing in the
7832 debugging in store.c to use %p rather than %d.
7834 81. There was a concern that a line of code in smtp_in.c could overflow a
7835 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7836 initially expressed, the concern was not well-founded, because trailing
7837 spaces are removed early. However, if the trailing spaces were followed by
7838 a NULL, they did not get removed, so the overflow was possible. Two fixes
7841 (a) I re-wrote the offending code in a cleaner fashion.
7842 (b) If an incoming SMTP command contains a NULL character, it is rejected
7845 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7846 runs initgroups(), so that if the Exim user is in any additional groups,
7847 they will be used during message reception.
7853 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob