1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platform supports and has the facility enabled, it will
125 be requested on all coneections.
127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
130 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
132 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
133 success-DSN messages. Previously the From: header was always the default
134 one for these; the option was ignored.
136 JH/28 Fix the timeout on smtp response to apply to the whole response.
137 Previously it was reset for every read, so a teergrubing peer sending
138 single bytes within the time limit could extend the connection for a
139 long time. Credit to Qualsys Security Advisory Team for the discovery.
141 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
142 delivery address, which leaked information of the results of local
143 forwarding. Change to the original envelope recipient address, per
146 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
147 requested. Previously not bounce was generated and a log entry of
148 error ignored was made.
150 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
152 JH/32 Introduce a general tainting mechanism for values read from the input
153 channel, and values derived from them. Refuse to expand any tainted
154 values, to catch one form of exploit.
156 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
157 was unused and the unexpanded text used for the test. Found and
158 fixed by Ruben Jenster.
160 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
161 an attempt to use a TLS library read routine dereffed a nul pointer,
164 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
165 them in our smtp response.
171 JH/01 Remove code calling the customisable local_scan function, unless a new
172 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
174 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
175 non-signal-safe functions being used.
177 JH/03 Bug 2269: When presented with a received message having a stupidly large
178 number of DKIM-Signature headers, disable DKIM verification to avoid
179 a resource-consumption attack. The limit is set at twenty.
181 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
182 report of oldest_pass in ${authres } in consequence, and separate out
183 some descriptions of reasons for verification fail.
185 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
186 files in the spool were present and unlocked. A queue-runner could spot
187 them, resulting in a duplicate delivery. Fix that by doing the unlock
188 after the unlink. Investigation by Tim Stewart. Take the opportunity to
189 add more error-checking on spoolfile handling while that code is being
192 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
193 No known attacks, no CVE, this is defensive hardening.
195 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
196 a queue-runner could start a delivery while other operations were ongoing.
197 Cutthrough delivery was a common victim, resulting in duplicate delivery.
198 Found and investigated by Tim Stewart. Fix by using the open message data
199 file handle rather than opening another, and not locally closing it (which
200 releases a lock) for that case, while creating the temporary .eml format
201 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
203 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
204 $sender_verify_failure/$recipient_verify_failure to "random".
206 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
209 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
210 Previously this would segfault.
212 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
215 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
216 like zero, since the resolver should be doing this for us, But we need one
217 as a CNAME but no MX presence gets the CNAME returned; we need to check
218 that doesn't point to an MX to declare it "no MX returned" rather than
219 "error, loop". A new main option is added so the older capability of
220 following some limited number of chain links is maintained.
222 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
224 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
225 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
226 now-deprecated earlier definitions used only specified the range up to TLS
227 1.2 (in the older-version library docs).
229 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
231 JH/15 Rework TLS client-side context management. Stop using a global, and
232 explicitly pass a context around. This enables future use of TLS for
233 connections to service-daemons (eg. malware scanning) while a client smtp
234 connection is using TLS; with cutthrough connections this is quite likely.
236 JH/16 Fix ARC verification to do AS checks in reverse order.
238 JH/17 Support a "tls" option on the ${readsocket } expansion item.
240 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
241 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
242 Previously the "utf8" would be re-prepended for every additional message.
244 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
245 Previously thery were accepted, resulting in issues when attempting to
246 forward messages to a non-supporting MTA.
248 PP/02 Let -n work with printing macros too, not just options.
250 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
251 one parent address was copied, and bogus data was used at delivery-logging
252 time. Either a crash (after delivery) or bogus log data could result.
253 Discovery and analysis by Tim Stewart.
255 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
256 Previously if the string ended mid-character, we did not insert the
257 promised '?' replacement.
259 PP/04 Documentation: current string operators work on bytes, not codepoints.
261 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
262 should pack well giving a smaller memory footprint so better caching and
263 therefore performance. Group the declarations where this can't be done so
264 that the byte-sized flag variables are not interspersed among pointer
265 variables, giving a better chance of good packing by the compiler.
267 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
268 non-null, to avoid issues with sites running BATV. Previously reports were
269 sent with an empty envelope sender so looked like bounces.
271 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
272 The ignore_error flag wasn't being returned from the filter subprocess so
273 was not set for later routers. Investigation and fix by Matthias Kurz.
275 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
276 and a msg:complete for the whole, when a message is manually removed using
277 -Mrm. Developement by Matthias Kurz, hacked on by JH.
279 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
280 a "Gnu special" function, asprintf() in the DB utility binary builds; I
281 hope that is portable enough.
283 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
284 requiring a known-CA anchor certificate; make it now rely entirely on the
285 TLSA as an anchor. Checking the name on the leaf cert against the name
286 on the A-record for the host is still done for TA (but not for EE mode).
288 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
289 included in delivery lines for non-proxied connections, when compiled with
290 SUPPORT_SOCKS and running with proxy logging enabled.
292 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
293 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
294 move the existing event to fire before the normal logging of message
295 failure so that custom logging is bracketed by normal logging.
297 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
298 msg:fail:internal event. Developement by Matthias Kurz.
300 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
301 far too small for todays use of crypto signatures stored there. Go all
302 the way to the max DNS message size of 64kB, even though this might be
303 overmuch for IOT constrained device use.
305 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
306 copy a string over itself. The library routine is documented as not
307 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
309 JH/32 For main options check_spool_space and check_inode_space, where the
310 platform supports 64b integers, support more than the previous 2^31 kB
311 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
312 the previous G, M, k.
314 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
315 $authenticated_fail_id variable on authentication failure. Previously
318 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
319 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
320 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
321 GNUTLS_SEC_PARAM_MEDIUM.
323 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
324 side. Previously we would continue as if no SNI had been received.
326 JH/36 Harden the handling of string-lists. When a list consisted of a sole
327 "<" character, which should be a list-separator specification, we walked
328 off past the nul-terimation.
330 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
331 causes) even when the retry time is not yet met. Previously they were
332 not, meaning that when (say) an account was over-quota and temp-rejecting,
333 and multiple senders' messages were queued, only one sender would get
334 notified on each configured delay_warning cycle.
336 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
338 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
339 cipher-suites, an error can be left on the stack even for a succeeding
340 accept; this results in impossible error messages when a later operation
343 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
344 return error codes indicating retry. Under TLS1.3 this becomes required.
346 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
347 it only wrote the new authenticators, resulting in a lack of tracking of
348 peer changes of ESMTP extensions until the next cache flush.
350 JH/41 Fix the loop reading a message header line to check for integer overflow,
351 and more-often against header_maxsize. Previously a crafted message could
352 induce a crash of the recive process; now the message is cleanly rejected.
354 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
355 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
361 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
362 When redis_servers is set to a list of > 1 element, and the Redis servers
363 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
364 case of MOVED into a DEFER case instead, thus moving the query onto the
365 next server in the list. For a cluster of N elements, all N servers must
366 be defined in redis_servers.
368 GF/02 Catch and remove uninitialized value warning in exiqsumm
369 Check for existence of @ARGV before looking at $ARGV[0]
371 JH/01 Replace the store_release() internal interface with store_newblock(),
372 which internalises the check required to safely use the old one, plus
373 the allocate and data copy operations duplicated in both (!) of the
374 extant use locations.
376 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
377 modifier. This matches the restriction on the commandline.
379 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
380 Previously only the last row was returned.
382 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
383 we assumed that tags in the header were well-formed, and parsed the
384 element content after inspecting only the first char of the tag.
385 Assumptions at that stage could crash the receive process on malformed
388 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
389 While running the DKIM ACL we operate on the Permanent memory pool so that
390 variables created with "set" persist to the DATA ACL. Also (at any time)
391 DNS lookups that fail create cache records using the Permanent pool. But
392 expansions release any allocations made on the current pool - so a dnsdb
393 lookup expansion done in the DKIM ACL releases the memory used for the
394 DNS negative-cache, and bad things result. Solution is to switch to the
395 Main pool for expansions.
396 While we're in that code, add checks on the DNS cache during store_reset,
397 active in the testsuite.
398 Problem spotted, and debugging aided, by Wolfgang Breyha.
400 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
401 When none of the hosts presented to a transport match an already-open
402 connection, close it and proceed with the list. Previously we would
403 queue the message. Spotted by Lena with Yahoo, probably involving
406 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
407 Previously a spurious "250 OK id=" response was appended to the proper
410 JH/08 The "support for" informational output now, which built with Content
411 Scanning support, has a line for the malware scanner interfaces compiled
412 in. Interface can be individually included or not at build time.
414 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
415 by the template makefile "src/EDITME". The "STREAM" support for an older
416 ClamAV interface method is removed.
418 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
419 rows affected is given instead).
421 JH/11 The runtime Berkeley DB library version is now additionally output by
422 "exim -d -bV". Previously only the compile-time version was shown.
424 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
425 SMTP connection. Previously, when one had more receipients than the
426 first, an abortive onward connection was made. Move to full support for
427 multiple onward connections in sequence, handling cutthrough connection
428 for all multi-message initiating connections.
430 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
431 routers. Previously, a multi-recipient message would fail to match the
432 onward-connection opened for the first recipient, and cause its closure.
434 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
435 a timeout on read on a GnuTLS initiating connection, resulting in the
436 initiating connection being dropped. This mattered most when the callout
437 was marked defer_ok. Fix to keep the two timeout-detection methods
440 JH/15 Relax results from ACL control request to enable cutthrough, in
441 unsupported situations, from error to silently (except under debug)
442 ignoring. This covers use with PRDR, frozen messages, queue-only and
445 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
447 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
448 metadata, resulting in a crash in free().
450 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
451 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
452 Broken also in d185889f4, with init system revamp.
454 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
455 Previously we abruptly closed the connection after reading a malware-
456 found indication; now we go on to read the "scan ok" response line,
459 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
460 ACL. Previously, a crash would result.
462 JH/19 Speed up macro lookups during configuration file read, by skipping non-
463 macro text after a replacement (previously it was only once per line) and
464 by skipping builtin macros when searching for an uppercase lead character.
466 JH/20 DANE support moved from Experimental to mainline. The Makefile control
467 for the build is renamed.
469 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
470 was allocated for every new TLS startup, meaning one per message. Fix
471 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
473 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
474 reported the original. Fix to report (as far as possible) the ACL
475 result replacing the original.
477 JH/23 Fix memory leak during multi-message connections using STARTTLS under
478 OpenSSL. Certificate information is loaded for every new TLS startup,
479 and the resources needed to be freed.
481 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
483 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
486 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
487 DATA response info to the (existing) per-recipient response info for
488 the "C=" log element. It can have useful tracking info from the
489 destination system. Patch from Simon Arlott.
491 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
492 length value. Previously this would segfault.
494 HS/02 Support Avast multiline protoocol, this allows passing flags to
495 newer versions of the scanner.
497 JH/28 Ensure that variables possibly set during message acceptance are marked
498 dead before release of memory in the daemon loop. This stops complaints
499 about them when the debug_store option is enabled. Discovered specifically
500 for sender_rate_period, but applies to a whole set of variables.
501 Do the same for the queue-runner and queue-list loops, for variables set
502 from spool message files. Do the same for the SMTP per-message loop, for
503 certain variables indirectly set in ACL operations.
505 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
506 as a multi-recipient message from a mailinglist manager). The coding had
507 an arbitrary cutoff number of characters while checking for more input;
508 enforced by writing a NUL into the buffer. This corrupted long / fast
509 input. The problem was exposed more widely when more pipelineing of SMTP
510 responses was introduced, and one Exim system was feeding another.
511 The symptom is log complaints of SMTP syntax error (NUL chars) on the
512 receiving system, and refused recipients seen by the sending system
513 (propating to people being dropped from mailing lists).
514 Discovered and pinpointed by David Carter.
516 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
517 replaced by the ${authresults } expansion.
519 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
521 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
522 allows proper process termination in container environments.
524 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
525 Previously the "final dot" had a newline after it; ensure it is CR,LF.
527 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
528 and "err_perm", deprecated since 4.83 when the RFC-defined words
529 "temperror" and "permerror" were introduced.
531 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
532 transport-filters or DKIM-signing. The restriction was lost in the
533 consolidation of verify-callout and delivery SMTP handling.
534 Extend the restriction to also cover ARC-signing.
536 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
537 in defer=pass mode supply a 450 to the initiator. Previously the message
540 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
541 tls_require_ciphers is used as before.
543 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
544 "pass_unscanned". Only tmpfails from the scanner are written to
545 the paniclog, as they may require admin intervention (permission
546 denied, license issues). Other scanner errors (like decompression
547 bombs) do not cause a paniclog entry.
549 JH/36 Fix reinitialisation of DKIM logging variable between messages.
550 Previously it was possible to log spurious information in receive log
553 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
554 triggered odd behaviour from Outlook Express clients.
556 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
559 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
560 since the IETF WG has not yet settled on that versus the original
561 "bare" representation.
563 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
564 Previously the millisecond value corrupted the output.
565 Fix also for syslog_pid=no and log_selector +pid, for which the pid
566 corrupted the output.
572 JH/01 Rework error string handling in TLS interface so that the caller in
573 more cases is responsible for logging. This permits library-sourced
574 string to be attached to addresses during delivery, and collapses
575 pairs of long lines into single ones.
577 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
578 during configuration. Wildcards are allowed and expanded.
580 JH/02 Rework error string handling in DKIM to pass more info back to callers.
581 This permits better logging.
583 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
584 do not close it down and have the child transport start it up again on
585 the passed-on TCP connection. Instead, proxy the child (and any
586 subsequent ones) for TLS via a unix-domain socket channel. Logging is
587 affected: the continued delivery log lines do not have any DNSSEC, TLS
588 Certificate or OCSP information. TLS cipher information is still logged.
590 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
591 identical IP addresses on different listening ports. Will also affect
594 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
595 add noisy ifdef guards to special-case this sillyness.
596 Patch from Bernd Kuhls.
598 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
599 than 255 are no longer allowed.
601 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
602 Disable the session-cache too, which might reduce our load. Since we
603 currrectly use a new context for every connection, both as server and
604 client, there is no benefit for these.
605 GnuTLS appears to not support tickets server-side by default (we don't
606 call gnutls_session_ticket_enable_server()) but client side is enabled
607 by default on recent versions (3.1.3 +) unless the PFS priority string
610 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
611 <https://reproducible-builds.org/specs/source-date-epoch/>.
613 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
614 the check for any unsuccessful recipients did not notice the limit, and
615 erroneously found still-pending ones.
617 JH/08 Pipeline CHUNKING command and data together, on kernels that support
618 MSG_MORE. Only in-clear (not on TLS connections).
620 JH/09 Avoid using a temporary file during transport using dkim. Unless a
621 transport-filter is involved we can buffer the headers in memory for
622 creating the signature, and read the spool data file once for the
623 signature and again for transmission.
625 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
626 4.77 as the kernel support then wasn't solid, having issues in 64bit
627 mode. Now, it's been long enough. Add support for FreeBSD also.
629 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
630 case where the routing stage had gathered several addresses to send to
631 a host before calling the transport for the first, we previously failed
632 to close down TLS in the old transport process before passing the TCP
633 connection to the new process. The new one sent a STARTTLS command
634 which naturally failed, giving a failed delivery and bloating the retry
635 database. Investigation and fix prototype from Wolfgang Breyha.
637 JH/12 Fix check on SMTP command input synchronisation. Previously there were
638 false-negatives in the check that the sender had not preempted a response
639 or prompt from Exim (running as a server), due to that code's lack of
640 awareness of the SMTP input buffering.
642 PP/04 Add commandline_checks_require_admin option.
643 Exim drops privileges sanely, various checks such as -be aren't a
644 security problem, as long as you trust local users with access to their
645 own account. When invoked by services which pass untrusted data to
646 Exim, this might be an issue. Set this option in main configuration
647 AND make fixes to the calling application, such as using `--` to stop
650 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
651 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
652 responses to those, into a single TLS record each way (this usually means
653 a single packet). As a side issue, smtp_enforce_sync now works on TLS
656 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
657 affects you only if you're dancing at the edge of the param size limits.
658 If you are, and this message makes sense to you, then: raise the
659 configured limit or use OpenSSL 1.1. Nothing we can do for older
662 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
663 cmdline element to get the documented default one. Previously it was
666 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
667 are unsuable for non-message-specific reasons - eg. connection timeout,
668 banner-time rejection.
670 JH/16 Fix logging of delivery remote port, when specified by router, under
673 PP/06 Repair manualroute's ability to take options in any order, even if one
674 is the name of a transport.
677 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
679 JH/17 Change the list-building routines interface to use the expanding-string
680 triplet model, for better allocation and copying behaviour.
682 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
683 Previously it was constructed the first time a possibly-matching string
684 was met in the configuration file input during startup; now it is done
687 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
688 compatible one, to avoid the (poorly documented) possibility of a config
689 file in the working directory redirecting the DB files, possibly correpting
690 some existing file. CVE-2017-10140 assigned for BDB.
692 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
693 cache-hot. Previously, although the result was properly cached, the
694 initial verify call returned a defer.
696 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
697 the main verify for receipient in uncached-mode.
699 JH/22 Retire historical build files to an "unsupported" subdir. These are
700 defined as "ones for which we have no current evidence of testing".
702 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
703 if present. Previously it was ignored.
705 JH/24 Start using specified-initialisers in C structure init coding. This is
706 a C99 feature (it's 2017, so now considered safe).
708 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
709 if was a fixed-sized field and bitmask ops via macros; it is now more
712 PP/07 GitHub PR 56: Apply MariaDB build fix.
713 Patch provided by Jaroslav Škarvada.
715 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
716 during Coverity cleanups [4.87 JH/47]
717 Diagnosis and fix provided by Michael Fischer v. Mollard.
719 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
720 the right size to place the terminating semicolon on its own folded
721 line, the header hash was calculated to an incorrect value thanks to
722 the (relaxed) space the fold became.
724 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
725 and confused the parent.
727 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
728 which could crash as a result. This could lead to undeliverable messages.
730 JH/28 Logging: "next input sent too soon" now shows where input was truncated
733 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
734 matters on fast-turnover and PID-randomising systems, which were getting
735 out-of-order delivery.
737 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
738 a possibly-overlapping copy. The symptom was that "Remote host closed
739 connection in response to HELO" was logged instead of the actual 4xx
742 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
743 Previously only that bufferd was discarded, resulting in SYMTP command
746 JH/32 DKIM: when a message has multiple signatures matching an identity given
747 in dkim_verify_signers, run the dkim acl once for each. Previously only
748 one run was done. Bug 2189.
750 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
751 "panic the current process" to "deliberately defer". The panic log is
752 still written with the problem list name; the mail and reject logs now
753 get a temp-reject line for the message that was being handled, saying
754 something like "domains check lookup or other defer". The SMTP 451
755 message is still "Temporary local problem".
757 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
758 A crafted sequence of BDAT commands could result in in-use memory beeing
759 freed. CVE-2017-16943.
761 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
762 from SMTP input. Previously it was always done; now only done for DATA
763 and not BDAT commands. CVE-2017-16944.
765 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
766 to the message (such as an overlong header line). Previously this was
767 not done and we did not exit BDAT mode. Followon from the previous item
768 though a different problem.
774 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
775 than -2003 did; needs libidn2 in addition to libidn.
777 JH/02 The path option on a pipe transport is now expanded before use.
779 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
780 Patch provided by "Björn", documentation fix added too.
782 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
783 missing a wire-to-host endian conversion.
785 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
786 close after a BDAT command line could be taken as a following command,
787 giving a synch failure. Fix by only checking for synch immediately
788 before acknowledging the chunk.
790 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
791 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
792 macro. Patches provided by Josh Soref.
794 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
795 Previously we did not; the RFC seems ambiguous and VRFY is not listed
796 by IANA as a service extension. However, John Klensin suggests that we
799 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
800 the dkim code may be unix-mode line endings rather than smtp wire-format
801 CRLF, so prepend a CR to any bare LF.
803 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
804 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
806 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
807 the first were themselves being wrongly included in the feed into dkim
808 processing; with most chunk sizes in use this resulted in an incorrect
809 body hash calculated value.
811 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
812 DKIM signature block, for verification. Although advised against by
813 standards it is specifically not ruled illegal.
815 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
817 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
818 missing a body hash (the bh= tag).
820 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
821 It seems that HAProxy sends the Proxy Protocol information in clear and
822 only then does a TLS startup, so do the same.
824 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
825 TCP connections (such as for Spamd) unless the daemon successfully set
826 Fast Open mode on its listening sockets. This fixes breakage seen on
827 too-old kernels or those not configured for Fast Open, at the cost of
828 requiring both directions being enabled for TFO, and TFO never being used
829 by non-daemon-related Exim processes.
831 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
832 endings, at least on the first header line. Try to canonify any that get
833 past that check, despite the cost.
835 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
836 now limited to an arbitrary five deep, while parsing addresses with the
837 strip_excess_angle_brackets option enabled.
839 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
840 instead leave the unprompted TLS handshake in socket buffer for the
841 TLS library to consume.
843 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
845 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
847 JH/16 Drop variables when they go out of scope. Memory management drops a whole
848 region in one operation, for speed, and this leaves assigned pointers
849 dangling. Add checks run only under the testsuite which checks all
850 variables at a store-reset and panics on a dangling pointer; add code
851 explicitly nulling out all the variables discovered. Fixes one known
852 bug: a transport crash, where a dangling pointer for $sending_ip_address
853 originally assigned in a verify callout, is re-used.
855 PP/06 Drop '.' from @INC in various Perl scripts.
857 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
859 PP/08 Reduce a number of compilation warnings under clang; building with
860 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
861 should be warning-free.
863 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
865 HS/01 Fix portability problems introduced by PP/08 for platforms where
866 realloc(NULL) is not equivalent to malloc() [SunOS et al].
868 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
869 chunk. This allows us to accept broken chunked messages. We need a more
870 general solution here.
872 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
873 already-broken messages in the queue.
875 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
877 JH/19 Fix reference counting bug in routing-generated-address tracking.
883 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
884 supports it and a size is available (ie. the sending peer gave us one).
886 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
887 years of being deprecated). The replacements are the ACLs
888 acl_smtp_mime and acl_not_smtp_mime.
890 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
891 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
892 or even an in-clear connection were permitted. Now, if the host lookup
893 was dnssec and dane was requested then the host is only used if the
894 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
895 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
896 if one fails this test.
897 This means that a poorly-configured remote DNS will make it incommunicado;
898 but it protects against a DNS-interception attack on it.
900 JH/04 Bug 1810: make continued-use of an open smtp transport connection
901 non-noisy when a race steals the message being considered.
903 JH/05 If main configuration option tls_certificate is unset, generate a
904 self-signed certificate for inbound TLS connections.
906 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
907 in rewrites and routers.
909 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
910 and logged a warning sing 4.83; now they are a configuration file error.
912 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
913 (lacking @domain). Apply the same qualification processing as RCPT.
915 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
917 JH/10 Support ${sha256:} applied to a string (as well as the previous
920 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
921 a cutthrough deliver is pending, as we always want to make a connection.
922 This also avoids re-routing the message when later placing the cutthrough
923 connection after a verify cache hit.
924 Do not update it with the verify result either.
926 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
927 when routing results in more than one destination address.
929 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
930 signing (which inhibits the cutthrough capability). Previously only
931 the presence of an option was tested; now an expansion evaluating as
932 empty is permissible (obviously it should depend only on data available
933 when the cutthrough connection is made).
935 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
936 the relevant preceding SMTP command did not note the pipelining mode.
938 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
939 Previously they were not counted.
941 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
942 as one having no matching records. Previously we deferred the message
943 that needed the lookup.
945 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
946 distinguished as "(=".
948 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
949 for missing MX records. Previously it only worked for missing A records.
951 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
953 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
954 after the data-go-ahead and data-ack. Patch from Jason Betts.
956 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
957 even for a "none" policy. Patch from Tony Meyer.
959 JH/22 Fix continued use of a connection for further deliveries. If a port was
960 specified by a router, it must also match for the delivery to be
963 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
964 When one of the recipients of a message was unsuitable for the connection
965 (has no matching addresses), we lost track of needing to mark it
966 deferred. As a result mail would be lost.
968 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
970 JH/25 Decoding ACL controls is now done using a binary search; the source code
971 takes up less space and should be simpler to maintain. Merge the ACL
972 condition decode tables also, with similar effect.
974 JH/26 Fix problem with one_time used on a redirect router which returned the
975 parent address unchanged. A retry would see the parent address marked as
976 delivered, so not attempt the (identical) child. As a result mail would
979 JH/27 Fix a possible security hole, wherein a process operating with the Exim
980 UID can gain a root shell. Credit to http://www.halfdog.net/ for
981 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
984 JH/28 Enable {spool,log} filesystem space and inode checks as default.
985 Main config options check_{log,spool}_{inodes,space} are now
986 100 inodes, 10MB unless set otherwise in the configuration.
988 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
989 Previously it only applied to the main-section connection policy
992 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
994 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
995 by me. Added RFC7919 DH primes as an alternative.
997 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
998 are not in the system include path.
1000 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1001 GnuTLS, when a session startup failed (eg because the client disconnected)
1002 Exim did stdio operations after fclose. This was exposed by a recent
1003 change which nulled out the file handle after the fclose.
1005 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1006 signed directly by the cert-signing cert, rather than an intermediate
1007 OCSP-signing cert. This is the model used by LetsEncrypt.
1009 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1011 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1012 an incoming connection.
1014 HS/02 Bug 1802: Do not half-close the connection after sending a request
1017 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1018 fallback to "prime256v1".
1020 JH/34 SECURITY: Use proper copy of DATA command in error message.
1021 Could leak key material. Remotely exploitable. CVE-2016-9963.
1027 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1028 and 3.4.4 - once the server is enabled to respond to an OCSP request
1029 it does even when not requested, resulting in a stapling non-aware
1030 client dropping the TLS connection.
1032 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1033 support variable-length bit vectors. No functional change.
1035 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1036 The I= interface field on outgoing lines is now after the H= remote
1037 host field, same as incoming lines. There is a separate
1038 outgoing_interface log selector which allows you to disable the
1041 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1042 If not running log_selector +smtp_connection the mainlog would be held
1043 open indefinitely after a "too many connections" event, including to a
1044 deleted file after a log rotate. Leave the per net connection logging
1045 leaving it open for efficiency as that will be quickly detected by the
1046 check on the next write.
1048 HS/01 Bug 1671: Fix post transport crash.
1049 Processing the wait-<transport> messages could crash the delivery
1050 process if the message IDs didn't exist for some reason. When
1051 using 'split_spool_directory=yes' the construction of the spool
1052 file name failed already, exposing the same netto behaviour.
1054 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1055 mime_regex ACL conditions.
1057 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1058 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1059 to HELO, local diagnostic string.
1061 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1062 log line to debug. Even when configured with a tls authenticator many
1063 client connections are expected to not authenticate in this way, so
1064 an authenticate fail is not an error.
1066 HS/02 Add the Exim version string to the process info. This way exiwhat
1067 gives some more detail about the running daemon.
1069 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1070 matter for fast-change records such as DNSBLs.
1072 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1073 retry record, even if constant. There may be multiple transports with
1074 different interface settings and the retry behaviour needs to be kept
1077 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1079 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1081 JH/10 Bug 840: fix log_defer_output option of pipe transport
1083 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1084 hosts_randomize. This matters a lot when combined with mua_wrapper.
1086 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1087 ${quote_pgsql:<string>} operator.
1089 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1091 JH/14 Change header file name for freeradius-client. Relevant if compiling
1092 with Radius support; from the Gentoo tree and checked under Fedora.
1094 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1096 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1097 option transport_filter, meaning no filtering.
1099 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1101 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1102 defaults to "*" (all hosts). The variable is now available when not built
1103 with TLS, default unset, mainly to enable keeping the testsuite sane.
1104 If a server certificate is not supplied (via tls_certificate) an error is
1105 logged, and clients will find TLS connections fail on startup. Presumably
1106 they will retry in-clear.
1107 Packagers of Exim are strongly encouraged to create a server certificate
1108 at installation time.
1110 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1111 with the $config_file variable.
1113 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1114 in transport context, after the attempt, and per-recipient. The latter type
1115 is per host attempted. The event data is the error message, and the errno
1116 information encodes the lookup type (A vs. MX) used for the (first) host,
1117 and the trailing two digits of the smtp 4xx response.
1119 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1120 to write to mainlog (or rejectlog, paniclog) in the window between file
1121 creation and permissions/ownership being changed. Particularly affects
1122 installations where exicyclog is run as root, rather than exim user;
1123 result is that the running daemon panics and dies.
1125 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1127 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1128 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1129 "pri" and "weight". Note that the previous implicit priority given by the
1130 list order is no longer honoured.
1132 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1133 for DKIM processing.
1135 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1136 by defining SUPPORT_SOCKS.
1138 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1139 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1140 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1141 variables are renamed to proxy_{local,external}_{address,port}.
1143 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1144 for a build by defining SUPPORT_I18N
1146 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1147 of the query string, and make ${quote_redis:} do that quoting.
1149 JH/29 Move Events support from Experimental to mainline, enabled by default
1150 and removable for a build by defining DISABLE_EVENT.
1152 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1154 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1155 cached by the daemon.
1157 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1158 by defining LOOKUP_REDIS. The libhiredis library is required.
1160 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1161 keys are given for lookup.
1163 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1164 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1165 only supported when built with TLS support. The PolarSSL SHA routines
1166 are still used when the TLS library is too old for convenient support.
1168 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1169 openssl_options), for security. OpenSSL forces this from version 1.1.0
1170 server-side so match that on older versions.
1172 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1173 allocation for $value could be released as the expansion processing
1174 concluded, but leaving the global pointer active for it.
1176 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1177 and to use the domains and local_parts ACL conditions.
1179 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1180 incorrectly not doubled on cutthrough transmission, hence seen as a
1181 body-termination at the receiving system - resulting in truncated mails.
1182 Commonly the sender saw a TCP-level error, and retransmitted the message
1183 via the normal store-and-forward channel. This could result in duplicates
1184 received - but deduplicating mailstores were liable to retain only the
1185 initial truncated version.
1187 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1189 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1191 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1192 we're in there, support oversigning also; bug 1309.
1194 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1196 HS/04 Add support for keep_environment and add_environment options.
1198 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1199 either intentional arithmetic overflow during PRNG, or testing config-
1202 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1203 delivery resulted in actual delivery. Cancel cutthrough before DATA
1206 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1207 rejects a recipient: pass the reject to the originator.
1209 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1210 Many were false-positives and ignorable, but it's worth fixing the
1213 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1214 for the new environment-manipulation done at startup. Move the routines
1215 from being local to tls.c to being global via the os.c file.
1217 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1218 an extract embedded as result-arg for a map, the first arg for extract
1219 is unavailable so we cannot tell if this is a numbered or keyed
1220 extraction. Accept either.
1226 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1229 JH/02 The smtp transport option "multi_domain" is now expanded.
1231 JH/03 The smtp transport now requests PRDR by default, if the server offers
1234 JH/04 Certificate name checking on server certificates, when exim is a client,
1235 is now done by default. The transport option tls_verify_cert_hostnames
1236 can be used to disable this per-host. The build option
1237 EXPERIMENTAL_CERTNAMES is withdrawn.
1239 JH/05 The value of the tls_verify_certificates smtp transport and main options
1240 default to the word "system" to access the system default CA bundle.
1241 For GnuTLS, only version 3.0.20 or later.
1243 JH/06 Verification of the server certificate for a TLS connection is now tried
1244 (but not required) by default. The verification status is now logged by
1245 default, for both outbound TLS and client-certificate supplying inbound
1248 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1251 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1252 Status Notification (bounce) messages are now MIME format per RFC 3464.
1253 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1254 under the control of the dsn_advertise_hosts option, and routers may
1255 have a dsn_lasthop option.
1257 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1258 default, modifiable by a malware= option. The list separator for
1259 the options can now be changed in the usual way. Bug 68.
1261 JH/10 The smtp_receive_timeout main option is now expanded before use.
1263 JH/11 The incoming_interface log option now also enables logging of the
1264 local interface on delivery outgoing connections.
1266 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1267 if the interface and destination host and port all match.
1269 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1272 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1273 Patch from Andrew Lewis.
1275 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1276 now supports optional time-restrictions, weighting, and priority
1277 modifiers per server. Patch originally by <rommer@active.by>.
1279 JH/16 The spamd_address main option now supports a mixed list of local
1280 and remote servers. Remote servers can be IPv6 addresses, and
1281 specify a port-range.
1283 JH/17 Bug 68: The spamd_address main option now supports an optional
1284 timeout value per server.
1286 JH/18 Bug 1581: Router and transport options headers_add/remove can
1287 now have the list separator specified.
1289 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1292 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1295 JH/21 Support for the A6 type of dns record is withdrawn.
1297 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1298 rather than the verbs used.
1300 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1301 from 255 to 1024 chars.
1303 JH/24 Verification callouts now attempt to use TLS by default.
1305 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1306 are generic router options now. The defaults didn't change.
1308 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1309 Original patch from Alexander Shikoff, worked over by JH.
1311 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1312 files if ZCAT_COMMAND is not executable.
1314 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1316 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1318 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1319 Normally benign, it bites when the pair was led to by a CNAME;
1320 modern usage is to not canonicalize the domain to a CNAME target
1321 (and we were inconsistent anyway for A-only vs AAAA+A).
1323 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1325 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1326 when evaluating $sender_host_dnssec.
1328 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1329 $sender_helo_dnssec variable.
1331 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1333 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1335 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1337 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1338 documented as working, but never had. Support all but $spam_report.
1340 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1341 added for tls authenticator.
1343 HS/03 Add perl_taintmode main config option
1349 TL/01 When running the test suite, the README says that variables such as
1350 no_msglog_check are global and can be placed anywhere in a specific
1351 test's script, however it was observed that placement needed to be near
1352 the beginning for it to behave that way. Changed the runtest perl
1353 script to read through the entire script once to detect and set these
1354 variables, reset to the beginning of the script, and then run through
1355 the script parsing/test process like normal.
1357 TL/02 The BSD's have an arc4random API. One of the functions to induce
1358 adding randomness was arc4random_stir(), but it has been removed in
1359 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1360 function when detected.
1362 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1363 cause callback expansion.
1365 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1366 syntax errors in an expansion can be treated as a string instead of
1367 logging or causing an error, due to the internal use of bool_lax
1368 instead of bool when processing it.
1370 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1371 server certificates when making smtp deliveries.
1373 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1375 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1377 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1379 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1380 Merged patch from Sebastian Wiedenroth.
1382 JH/05 Fix results-pipe from transport process. Several recipients, combined
1383 with certificate use, exposed issues where response data items split
1384 over buffer boundaries were not parsed properly. This eventually
1385 resulted in duplicates being sent. This issue only became common enough
1386 to notice due to the introduction of connection certificate information,
1387 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1389 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1390 size buffer was used, resulting in syntax errors when an expansion
1393 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1394 version 3.3.6 or later.
1396 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1397 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1398 both become "event_action", the variables become $event_name, $event_data
1399 and $event_defer_errno. There is a new variable $verify_mode, usable in
1400 routers, transports and related events. The tls:cert event is now also
1401 raised for inbound connections, if the main configuration event_action
1404 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1405 early OCSP support, but no stapling (appears to be less than 1.0.0).
1407 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1408 server certificate names available under the smtp transport option
1409 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1412 JH/10 Time-related extraction expansions from certificates now use the main
1413 option "timezone" setting for output formatting, and are consistent
1414 between OpenSSL and GnuTLS compilations. Bug 1541.
1416 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1417 encoded parameter in the incoming message. Bug 1558.
1419 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1420 include certificate info, eximon was claiming there were spoolfile
1423 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1425 JH/14 Log delivery-related information more consistently, using the sequence
1426 "H=<name> [<ip>]" wherever possible.
1428 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1429 are problematic for Debian distribution, omit them from the release
1432 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1434 JH/16 Fix string representation of time values on 64bit time_t architectures.
1437 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1438 output list separator was used.
1443 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1444 checkers that were complaining about end of non-void function with no
1447 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1448 This was a regression introduced in 4.83 by another bugfix.
1450 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1452 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1453 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1459 TF/01 Correctly close the server side of TLS when forking for delivery.
1461 When a message was received over SMTP with TLS, Exim failed to clear up
1462 the incoming connection properly after forking off the child process to
1463 deliver the message. In some situations the subsequent outgoing
1464 delivery connection happened to have the same fd number as the incoming
1465 connection previously had. Exim would try to use TLS and fail, logging
1466 a "Bad file descriptor" error.
1468 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1469 utilities have not been installed.
1471 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1472 temporary space as the ACL may create new global variables.
1474 TL/01 LDAP support uses per connection or global context settings, depending
1475 upon the detected version of the libraries at build time.
1477 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1478 to extract and use the src ip:port in logging and expansions as if it
1479 were a direct connection from the outside internet. PPv2 support was
1480 updated based on HAProxy spec change in May 2014.
1482 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1484 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1485 Properly escape header and check for NULL return.
1487 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1490 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1492 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1493 characters in header names, implemented as a verify condition.
1494 Contributed by Michael Fischer v. Mollard.
1496 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1497 results permerror and temperror. Previous values are deprecated but
1498 still accepted. In a future release, err_perm and err_temp will be
1499 completely removed, which will be a backward incompatibility if the
1500 ACL tests for either of these two old results. Patch contributed by
1501 user bes-internal on the mailing list.
1503 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1505 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1506 selectors, in both main and reject logs.
1508 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1511 JH/07 Add malware type "sock" for talking to simple daemon.
1513 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1515 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1516 routers/transports under cutthrough routing.
1518 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1519 numbers. Touch up "bool" conditional to keep the same definition.
1521 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1523 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1525 JH/12 Expand items in router/transport headers_add or headers_remove lists
1526 individually rather than the list as a whole. Bug 1452.
1528 Required for reasonable handling of multiple headers_ options when
1529 they may be empty; requires that headers_remove items with embedded
1530 colons must have them doubled (or the list-separator changed).
1532 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1533 view the policy declared in the DMARC record. Currently, $dmarc_status
1534 is a combined value of both the record presence and the result of the
1537 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1539 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1540 dnslookup router and the smtp transport (applying to the forward
1543 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1544 of ldap servers used for a specific lookup. Patch provided by Heiko
1547 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1548 New variable $lookup_dnssec_authenticated for observability.
1550 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1551 Patch submitted by Lars Timman.
1553 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1555 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1556 Requires trusted mode and valid format message id, aborts otherwise.
1557 Patch contributed by Heiko Schlichting.
1559 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1560 certextract with support for various fields. Bug 1358.
1562 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1563 is requested by default, modifiable by smtp transport option
1566 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1567 operate on certificate variables to give certificate fingerprints
1568 Also new ${sha256:cert_variable}.
1570 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1572 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1573 Christian Aistleitner.
1575 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1577 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1578 file. Patch from Wolfgang Breyha.
1580 JH/25 Expand the coverage of the delivery $host and $host_address to
1581 client authenticators run in verify callout. Bug 1476.
1583 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1584 align with daemon_smtp_ports. Bug 72.
1586 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1587 support and error reporting did not work properly.
1589 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1590 and is readable. Patch from Andrew Colin Kissa.
1592 TL/14 Enhance documentation of ${run expansion and how it parses the
1593 commandline after expansion, particularly in the case when an
1594 unquoted variable expansion results in an empty value.
1596 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1598 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1599 by renaming away from it. Addresses GH issue 15, reported by
1602 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1604 TL/15 SECURITY: prevent double expansion in math comparison functions
1605 (can expand unsanitized data). Not remotely exploitable.
1612 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1614 PP/02 Make -n do something, by making it not do something.
1615 When combined with -bP, the name of an option is not output.
1617 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1620 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1621 $sender_host_name and config options to manage this, and basic check
1624 PP/05 DSCP support for outbound connections and control modifier for inbound.
1626 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1627 (Only plugin which currently uses this is kerberos4, which nobody should
1628 be using, but we should make it available and other future plugins might
1629 conceivably use it, even though it would break NAT; stuff *should* be
1630 using channel bindings instead).
1632 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1633 name; added for Sendmail compatibility; requires admin caller.
1634 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1635 just ignore it); requires trusted caller.
1636 Also parse but ignore: -Ac -Am -X<logfile>
1639 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1641 TL/02 Add +smtp_confirmation as a default logging option.
1643 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1644 Patch by Magnus Holmgren from 2007-02-20.
1646 TL/04 Bugzilla 1281 - Spec typo.
1647 Bugzilla 1283 - Spec typo.
1648 Bugzilla 1290 - Spec grammar fixes.
1650 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1652 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1654 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1655 mailing list by by Dmitry Isaikin.
1657 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1659 JH/02 Support "G" suffix to numbers in ${if comparisons.
1661 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1663 NM/01 Bugzilla 1197 - Spec typo
1664 Bugzilla 1196 - Spec examples corrections
1666 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1668 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1669 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1672 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1673 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1675 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1676 "acl {{name}{arg}...}", and optional args on acl condition
1679 JH/05 Permit multiple router/transport headers_add/remove lines.
1681 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1683 JH/07 Avoid using a waiting database for a single-message-only transport.
1684 Performance patch from Paul Fisher. Bugzilla 1262.
1686 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1689 JH/09 Add $headers_added variable, with content from use of ACL modifier
1690 add_header (but not yet added to the message). Bugzilla 199.
1692 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1693 Pulled from Bugzilla 817 by Wolfgang Breyha.
1695 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1697 (nb: this is the same fix as in Exim 4.80.1)
1699 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1702 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1705 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1707 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1708 advertise SMTP AUTH mechanism to us, instead of a generic
1709 protocol violation error. Also, make Exim more robust to bad
1710 data from the Dovecot auth socket.
1712 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1714 When a queue runner is handling a message, Exim first routes the
1715 recipient addresses, during which it prunes them based on the retry
1716 hints database. After that it attempts to deliver the message to
1717 any remaining recipients. It then updates the hints database using
1720 So if a recipient address works intermittently, it can get repeatedly
1721 deferred at routing time. The retry hints record remains fresh so the
1722 address never reaches the final cutoff time.
1724 This is a fairly common occurrence when a user is bumping up against
1725 their storage quota. Exim had some logic in its local delivery code
1726 to deal with this. However it did not apply to per-recipient defers
1727 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1729 This change adds a proper retry rule check during routing so that the
1730 final cutoff time is checked against the message's age. We only do
1731 this check if there is an address retry record and there is not a
1732 domain retry record; this implies that previous attempts to handle
1733 the address had the retry_use_local_parts option turned on. We use
1734 this as an approximation for the destination being like a local
1735 delivery, as in LMTP.
1737 I suspect this new check makes the old local delivery cutoff check
1738 redundant, but I have not verified this so I left the code in place.
1740 TF/02 Correct gecos expansion when From: is a prefix of the username.
1742 Test 0254 submits a message to Exim with the header
1746 When I ran the test suite under the user fanf2, Exim expanded
1747 the header to contain my full name, whereas it should have added
1748 a Resent-Sender: header. It erroneously treats any prefix of the
1749 username as equal to the username.
1751 This change corrects that bug.
1753 GF/01 DCC debug and logging tidyup
1754 Error conditions log to paniclog rather than rejectlog.
1755 Debug lines prefixed by "DCC: " to remove any ambiguity.
1757 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1759 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1760 Bug spotted by Jeremy Harris; was flawed since initial commit.
1761 Would have resulted in OCSP responses post-SNI triggering an Exim
1762 NULL dereference and crash.
1764 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1766 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1767 Bug detection, analysis and fix by Samuel Thibault.
1768 Bugzilla 1331, Debian bug #698092.
1770 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1772 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1773 Server implementation by Todd Lyons, client by JH.
1774 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1775 config variable "prdr_enable" controls whether the server
1776 advertises the facility. If the client requests PRDR a new
1777 acl_data_smtp_prdr ACL is called once for each recipient, after
1778 the body content is received and before the acl_smtp_data ACL.
1779 The client is controlled by both of: a hosts_try_prdr option
1780 on the smtp transport, and the server advertisement.
1781 Default client logging of deliveries and rejections involving
1782 PRDR are flagged with the string "PRDR".
1784 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1785 fclose(). Diagnosis by Todd Lyons.
1787 PP/17 Update configure.default to handle IPv6 localhost better.
1788 Patch by Alain Williams (plus minor tweaks).
1791 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1792 This is now consistent with GnuTLS, and is now documented: the
1793 previous undocumented portable approach to treating the option as
1794 unset was to force an expansion failure. That still works, and
1795 an empty string is now equivalent.
1797 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1798 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1799 not performing validation itself.
1801 PP/20 Added force_command boolean option to pipe transport.
1802 Patch from Nick Koston, of cPanel Inc.
1804 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1807 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1809 PP/21 Fix eximon continuous updating with timestamped log-files.
1810 Broken in a format-string cleanup in 4.80, missed when I repaired the
1811 other false fix of the same issue.
1812 Report and fix from Heiko Schlichting.
1815 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1816 Report from Prashanth Katuri.
1818 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1819 It's SecureTransport, so affects any MacOS clients which use the
1820 system-integrated TLS libraries, including email clients.
1822 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1823 using a MIME ACL for non-SMTP local injection.
1824 Report and assistance in diagnosis by Warren Baker.
1826 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1828 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1830 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1831 last id that failed so it may be referenced in subsequent ACL's.
1833 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1836 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1837 ldap library initialization, allowing self-signed CA's to be
1838 used. Also properly sets require_cert option later in code by
1839 using NULL (global ldap config) instead of ldap handle (per
1840 session). Bug diagnosis and testing by alxgomz.
1842 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1843 the src/util/ subdirectory.
1845 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1846 renamed to Transport Post Delivery Action by Jeremy Harris, as
1849 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1850 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1851 redis_servers = needs to be configured which will be used by the redis
1852 lookup. Patch from Warren Baker, of The Packet Hub.
1854 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1856 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1857 hostname or reverse DNS when processing a host list. Used suggestions
1858 from multiple comments on this bug.
1860 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1862 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1863 Missed a few lines, added it to make the runtest require no keyboard
1866 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1867 contains upper case chars. Make router use caseful_local_part.
1869 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1870 support when GnuTLS has been built with p11-kit.
1876 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1878 This, or similar/improved, will also be change PP/11 of 4.82.
1884 PP/01 Handle short writes when writing local log-files.
1885 In practice, only affects FreeBSD (8 onwards).
1886 Bugzilla 1053, with thanks to Dmitry Isaikin.
1888 NM/01 Bugzilla 949 - Documentation tweak
1890 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1893 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1895 PP/02 Implemented gsasl authenticator.
1897 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1899 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1900 `pkg-config foo` for cflags/libs.
1902 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1903 with rest of GSASL and with heimdal_gssapi.
1905 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1906 `pkg-config foo` for cflags/libs for the TLS implementation.
1908 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1909 properties get this fed in as external SSF. A number of robustness
1910 and debugging improvements to the cyrus_sasl authenticator.
1912 PP/08 cyrus_sasl server now expands the server_realm option.
1914 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1915 Patch by Jeremy Harris.
1917 PP/10 Added dbmjz lookup type.
1919 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1921 PP/12 MAIL args handles TAB as well as SP, for better interop with
1922 non-compliant senders.
1923 Analysis and variant patch by Todd Lyons.
1925 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1926 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1927 Patch from Dirk Mueller <dmueller@suse.com>
1929 PP/13 tls_peerdn now print-escaped for spool files.
1930 Observed some $tls_peerdn in wild which contained \n, which resulted
1931 in spool file corruption.
1933 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1934 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1935 or write after TLS renegotiation, which otherwise led to messages
1938 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1939 as a tracking header (ie: a signed header comes before the signature).
1940 Patch from Wolfgang Breyha.
1942 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1943 comma-sep list; embedded commas doubled.
1945 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1947 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1949 Report and patch from Dmitry Banschikov.
1951 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1952 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1953 failures. We appear to now support TLS1.1+ with Exim.
1955 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1956 lets Exim select keys and certificates based upon TLS SNI from client.
1957 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1958 before an outbound SMTP session. New log_selector, +tls_sni.
1960 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1961 NULL dereference. Report and patch from Alun Jones.
1963 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1964 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1965 Not seeing resolver debug output on NetBSD, but suspect this is a
1966 resolver implementation change.
1968 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1969 Left warnings. Added "eximon gdb" invocation mode.
1971 PP/21 Defaulting "accept_8bitmime" to true, not false.
1973 PP/22 Added -bw for inetd wait mode support.
1975 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1976 locate the relevant includes and libraries. Made this the default.
1978 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1979 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1981 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1982 This may cause build issues on older platforms.
1984 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1985 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1986 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1987 Added SNI support via GnuTLS too.
1988 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1990 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1992 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1993 Applied second patch from Janne, implementing suggestion to default
1994 multiple-strings-in-record handling to match SPF spec.
1996 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1998 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1999 read-only, out of scope).
2000 Patch from Wolfgang Breyha, report from Stuart Northfield.
2002 PP/29 Fix three issues highlighted by clang analyser static analysis.
2003 Only crash-plausible issue would require the Cambridge-specific
2004 iplookup router and a misconfiguration.
2005 Report from Marcin Mirosław.
2007 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2009 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2010 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2011 As part of this, removing so much warning spew let me fix some minor
2012 real issues in debug logging.
2014 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2015 assignment on my part. Fixed.
2017 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2018 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2019 Janne Snabb (who went above and beyond: thank you).
2021 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2022 string otherwise requires a connection and a bunch more work and it's
2023 relatively easy to get wrong. Should also expose TLS library linkage
2026 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2027 64-bit ${eval} (JH/03).
2029 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2030 GNU libc to support some of the 64-bit stuff, should not lead to
2031 conflicts. Defined before os.h is pulled in, so if a given platform
2032 needs to override this, it can.
2034 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2035 protection layer was required, which is not implemented.
2036 Bugzilla 1254, patch from Wolfgang Breyha.
2038 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2039 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2040 tls_dhparam take prime identifiers. Also unbreak combination of
2041 OpenSSL+DH_params+TLSSNI.
2043 PP/39 Disable SSLv2 by default in OpenSSL support.
2049 PP/01 Solaris build fix for Oracle's LDAP libraries.
2050 Bugzilla 1109, patch from Stephen Usher.
2052 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2054 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2057 TF/02 Fix a couple more cases where we did not log the error message
2058 when unlink() failed. See also change 4.74-TF/03.
2060 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2061 lock up or crash if it happened to be inside a call to libc when it
2062 got a SIGUSR1 from exiwhat.
2064 The SIGUSR1 handler appends the current process status to the process
2065 log which is later printed by exiwhat. It used to use the general
2066 purpose logging code to do this, but several functions it calls are
2067 not safe for signals.
2069 The new output code in the SIGUSR1 handler is specific to the process
2070 log, and simple enough that it's easy to inspect for signal safety.
2071 Removing some special cases also simplifies the general logging code.
2072 Removing the spurious timestamps from the process log simplifies
2075 TF/04 Improved ratelimit ACL condition.
2077 The /noupdate option has been deprecated in favour of /readonly which
2078 has clearer semantics. The /leaky, /strict, and /readonly update modes
2079 are mutually exclusive. The update mode is no longer included in the
2080 database key; it just determines when the database is updated. (This
2081 means that when you upgrade Exim will forget old rate measurements.)
2083 Exim now checks that the per_* options are used with an update mode that
2084 makes sense for the current ACL. For example, when Exim is processing a
2085 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2086 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2087 must specify per_mail/readonly. If you omit the update mode it defaults to
2088 /leaky where that makes sense (as before) or /readonly where required.
2090 The /noupdate option is now undocumented but still supported for
2091 backwards compatibility. It is equivalent to /readonly except that in
2092 ACLs where /readonly is required you may specify /leaky/noupdate or
2093 /strict/noupdate which are treated the same as /readonly.
2095 A useful new feature is the /count= option. This is a generalization
2096 of the per_byte option, so that you can measure the throughput of other
2097 aggregate values. For example, the per_byte option is now equivalent
2098 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2100 The per_rcpt option has been generalized using the /count= mechanism
2101 (though it's more complicated than the per_byte equivalence). When it is
2102 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2103 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2104 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2105 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2106 Note that using per_rcpt with a non-readonly update mode in more than
2107 one ACL will cause the recipients to be double-counted. (The per_mail
2108 and per_byte options don't have this problem.)
2110 The handling of very low rates has changed slightly. If the computed rate
2111 is less than the event's count (usually one) then this event is the first
2112 after a long gap. In this case the rate is set to the same as this event's
2113 count, so that the first message of a spam run is counted properly.
2115 The major new feature is a mechanism for counting the rate of unique
2116 events. The new per_addr option counts the number of different
2117 recipients that someone has sent messages to in the last time period. It
2118 behaves like per_rcpt if all the recipient addresses are different, but
2119 duplicate recipient addresses do not increase the measured rate. Like
2120 the /count= option this is a general mechanism, so the per_addr option
2121 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2122 example, measure the rate that a client uses different sender addresses
2123 with the options per_mail/unique=$sender_address. There are further
2124 details in the main documentation.
2126 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2128 TF/06 Removed a few PCRE remnants.
2130 TF/07 Automatically extract Exim's version number from tags in the git
2131 repository when doing development or release builds.
2133 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2134 Bugzilla 879. Patch from Paul Fisher.
2136 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2137 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2140 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2142 PP/05 Variable $av_failed, true if the AV scanner deferred.
2143 Bugzilla 1078. Patch from John Horne.
2145 PP/06 Stop make process more reliably on build failure.
2146 Bugzilla 1087. Patch from Heiko Schlittermann.
2148 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2149 Bugzilla 1089. Patch from Heiko Schlittermann.
2151 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2152 Bugzilla 1131. Patch from Holger Weiß.
2154 PP/09 Handle IPv6 addresses with SPF.
2155 Bugzilla 860. Patch from Wolfgang Breyha.
2157 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2159 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2162 PP/11 match_* no longer expand right-hand-side by default.
2163 New compile-time build option, EXPAND_LISTMATCH_RHS.
2164 New expansion conditions, "inlist", "inlisti".
2166 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2168 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2170 PP/14 fix log_write() format string regression from TF/03.
2171 Bugzilla 1152. Patch from Dmitry Isaikin.
2177 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2179 PP/02 Harmonised TLS library version reporting; only show if debugging.
2180 Layout now matches that introduced for other libraries in 4.74 PP/03.
2182 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2184 PP/04 New "dns_use_edns0" global option.
2186 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2189 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2190 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2192 TK/01 Updated PolarSSL code to 0.14.2.
2193 Bugzilla 1097. Patch from Andreas Metzler.
2195 PP/07 Catch divide-by-zero in ${eval:...}.
2196 Fixes bugzilla 1102.
2198 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2201 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2202 format-string attack -- SECURITY: remote arbitrary code execution.
2204 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2205 time unintentionally subject to list matching rules, letting the header
2206 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2207 arbitrary string expansion). This allowed for information disclosure.
2209 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2210 INT_MIN/-1 -- value coerced to INT_MAX.
2216 NM/01 Workaround for PCRE version dependency in version reporting
2219 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2220 This fixes portability to compilers other than gcc, notably
2221 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2223 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2224 makefiles for portability to HP-UX and POSIX correctness.
2226 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2227 Also via indented variable definition in the Makefile.
2228 (Debugging by Oliver Heesakkers).
2230 PP/02 Restore caching of spamd results with expanded spamd_address.
2231 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2233 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2234 Improves build reliability. Fix from: Frank Elsner
2236 NM/02 Fix wide character breakage in the rfc2047 coding
2237 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2239 NM/03 Allow underscore in dnslist lookups
2240 Fixes bug 1026. Patch from Graeme Fowler
2242 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2243 Code patches from Adam Ciarcinski of NetBSD.
2245 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2248 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2249 is logged, to avoid truncation. Patch from John Horne.
2251 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2252 Patch from Jakob Hirsch.
2254 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2255 SQL string expansion failure details.
2256 Patch from Andrey Oktyabrski.
2258 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2259 Patch from Simon Arlott.
2261 PP/09 New lookups functionality failed to compile on old gcc which rejects
2262 extern declarations in function scope.
2263 Patch from Oliver Fleischmann
2265 PP/10 Use sig_atomic_t for flags set from signal handlers.
2266 Check getgroups() return and improve debugging.
2267 Fixed developed for diagnosis in bug 927 (which turned out to be
2270 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2271 Patch from Mark Zealey.
2273 PP/12 Bugzilla 1056: Improved spamd server selection.
2274 Patch from Mark Zealey.
2276 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2277 Based on patch from Heiko Schlittermann.
2279 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2280 Patch from Uwe Doering, sign-off by Michael Haardt.
2282 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2283 variable declaration deep within a block. Bug and patch from
2286 PP/15 lookups-Makefile IRIX compatibility coercion.
2288 PP/16 Make DISABLE_DKIM build knob functional.
2290 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2291 Patch by Simon Arlott
2293 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2294 variable argument macros. Our copy now differs from upstream.
2300 TF/01 Failure to get a lock on a hints database can have serious
2301 consequences so log it to the panic log.
2303 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2304 controlled using the smtp_confirmation log selector.
2306 TF/03 Include the error message when we fail to unlink a spool file.
2308 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2309 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2310 for maintaining out-of-tree patches for some time.
2312 PP/01 Bugzilla 139: Documentation and portability issues.
2313 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2314 Handle per-OS dynamic-module compilation flags.
2316 PP/02 Let /dev/null have normal permissions.
2317 The 4.73 fixes were a little too stringent and complained about the
2318 permissions on /dev/null. Exempt it from some checks.
2319 Reported by Andreas M. Kirchwitz.
2321 PP/03 Report version information for many libraries, including
2322 Exim version information for dynamically loaded libraries. Created
2323 version.h, now support a version extension string for distributors
2324 who patch heavily. Dynamic module ABI change.
2326 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2327 privilege escalation vulnerability whereby the Exim run-time user
2328 can cause root to append content of the attacker's choosing to
2331 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2334 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2335 If dropping privileges for untrusted macros, we disabled normal logging
2336 on the basis that it would fail; for the Exim run-time user, this is not
2337 the case, and it resulted in successful deliveries going unlogged.
2338 Fixed. Reported by Andreas Metzler.
2344 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2345 only prepend for the Resent-* case. Fixes regression introduced in
2346 Exim 4.70 by NM/22 for Bugzilla 607.
2348 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2349 increasing numbers of administrators be bitten by this.
2351 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2353 PP/03 Bugzilla 994: added openssl_options main configuration option.
2355 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2357 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2359 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2361 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2362 without a peer certificate, leading to a segfault because of an
2363 assumption that peers always have certificates. Be a little more
2364 paranoid. Problem reported by Martin Tscholak.
2366 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2367 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2368 NB: ClamAV planning to remove STREAM in "middle of 2010".
2369 CL also introduces -bmalware, various -d+acl logging additions and
2370 more caution in buffer sizes.
2372 PP/09 Implemented reverse_ip expansion operator.
2374 PP/10 Bugzilla 937: provide a "debug" ACL control.
2376 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2378 PP/12 Bugzilla 973: Implement --version.
2380 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2382 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2384 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2386 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2387 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2388 ignore trailing whitespace.
2390 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2392 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2395 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2396 Notification from Dr Andrew Aitchison.
2398 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2399 ExtendedDetectionInfo response format.
2400 Notification from John Horne.
2402 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2405 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2406 XSL and documented dependency on system catalogs, with examples of how
2409 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2412 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2413 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2414 configuration file which is writeable by the Exim user or group.
2416 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2417 of configuration files to cover files specified with the -C option if
2418 they are going to be used with root privileges, not just the default
2421 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2422 option (effectively making it always true).
2424 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2425 files to be used while preserving root privileges.
2427 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2428 that rogue child processes cannot use them.
2430 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2431 run-time user, instead of root.
2433 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2434 Exim run-time user without dropping privileges.
2436 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2437 result string, instead of calling string_vformat() twice with the same
2440 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2441 for other users. Others should always drop root privileges if they use
2442 -C on the command line, even for a whitelisted configure file.
2444 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2446 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2452 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2453 $data_path, and $header_path variables; fixed documentation bugs and
2456 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2457 exipick to access non-standard spools, including the "frozen" queue
2460 NM/01 Bugzilla 965: Support mysql stored procedures.
2461 Patch from Alain Williams
2463 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2465 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2466 Patch from Andreas Metzler
2468 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2469 Patch from Kirill Miazine
2471 NM/05 Bugzilla 671: Added umask to procmail example.
2473 JJ/03 installed exipick 20100323.0, fixing doc bug
2475 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2476 directory. Notification and patch from Dan Rosenberg.
2478 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2480 TK/02 Improve log output when DKIM signing operation fails.
2482 MH/01 Treat the transport option dkim_domain as a colon separated
2483 list, not as a single string, and sign the message with each element,
2484 omitting multiple occurences of the same signer.
2486 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2487 Bugzilla 985, 986. Patch by Simon Arlott
2489 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2490 Patch by Simon Arlott
2492 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2493 MBX locking. Notification from Dan Rosenberg.
2499 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2501 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2503 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2505 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2507 NM/04 Bugzilla 915: Flag broken perl installation during build.
2513 TK/01 Added patch by Johannes Berg that expands the main option
2514 "spamd_address" if it starts with a dollar sign.
2516 TK/02 Write list of recipients to X-Envelope-Sender header when building
2517 the mbox-format spool file for content scanning (suggested by Jakob
2520 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2521 (http://www.dcc-servers.net/) support via dccifd. Activated by
2522 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2524 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2525 by Mark Daniel Reidel <mr@df.eu>.
2527 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2528 When building exim an external PCRE library is now needed -
2529 PCRE is a system library on the majority of modern systems.
2530 See entry on PCRE_LIBS in EDITME file.
2532 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2533 conversation. Added nologin parameter to request.
2534 Patch contributed by Kirill Miazine.
2536 TF/01 Do not log submission mode rewrites if they do not change the address.
2538 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2540 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2541 log files in place. Contributed by Roberto Lima.
2543 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2545 TF/03 Bugzilla 615: When checking the local_parts router precondition
2546 after a local_part_suffix or local_part_prefix option, Exim now
2547 does not use the address's named list lookup cache, since this
2548 contains cached lookups for the whole local part.
2550 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2551 Robert Millan. Documentation is in experimental-spec.txt.
2553 TF/04 Bugzilla 668: Fix parallel build (make -j).
2555 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2557 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2558 Patch provided by Jan Srzednicki.
2560 TF/05 Leading white space used to be stripped from $spam_report which
2561 wrecked the formatting. Now it is preserved.
2563 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2564 that they are available at delivery time.
2566 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2568 TF/08 TLS error reporting now respects the incoming_interface and
2569 incoming_port log selectors.
2571 TF/09 Produce a more useful error message if an SMTP transport's hosts
2572 setting expands to an empty string.
2574 NM/06 Bugzilla 744: EXPN did not work under TLS.
2575 Patch provided by Phil Pennock.
2577 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2578 Patch provided by Richard Godbee.
2580 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2581 acl_smtp_notquit, added index entry.
2583 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2584 Patch provided by Eugene Bujak.
2586 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2587 accept(). Patch provided by Maxim Dounin.
2589 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2590 Patch provided by Phil Pennock.
2592 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2594 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2595 Patch provided by Brad "anomie" Jorsch.
2597 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2598 Patch provided by Dean Brooks.
2600 TK/05 Add native DKIM support (does not depend on external libraries).
2602 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2603 Patch provided by Graeme Fowler.
2605 NM/16 Bugzilla 851: Documentation example syntax fix.
2607 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2609 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2612 NM/19 Bugzilla 745: TLS version reporting.
2613 Patch provided by Phil Pennock.
2615 NM/20 Bugzilla 167: bool: condition support.
2616 Patch provided by Phil Pennock.
2618 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2619 clients. Patch provided by Phil Pennock.
2621 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2622 Patch provided by Brad "anomie" Jorsch.
2624 NM/23 Bugzilla 687: Fix misparses in eximstats.
2625 Patch provided by Heiko Schlittermann.
2627 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2628 Patch provided by Heiko Schlittermann.
2630 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2631 plus update to original patch.
2633 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2635 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2636 Patch provided by David Brownlee.
2638 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2640 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2642 NM/30 Bugzilla 888: TLS documentation bugfixes.
2644 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2646 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2647 Unlike the original bugzilla I have changed all shell scripts in src tree.
2649 NM/33 Bugzilla 898: Transport filter timeout fix.
2650 Patch by Todd Rinaldo.
2652 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2653 Patch by Serge Demonchaux.
2655 NM/35 Bugzilla 39: Base64 decode bug fixes.
2656 Patch by Jakob Hirsch.
2658 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2660 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2662 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2664 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2670 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2671 ALT-N's libdkim that I have put here:
2672 http://duncanthrax.net/exim-experimental/
2674 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2675 were called 'true' and it seems that C99 defines that as a reserved
2676 keyword to be used with 'bool' variable types. That means you could
2677 not include C99-style headers which use bools without triggering
2678 build errors in sieve.c.
2680 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2681 as mailq or other aliases. Changed the --help handling significantly
2682 to do whats expected. exim_usage() emits usage/help information.
2684 SC/01 Added the -bylocaldomain option to eximstats.
2686 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2688 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2690 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2696 PH/01 Another patch from the Sieve maintainer.
2698 PH/02 When an IPv6 address is converted to a string for single-key lookup
2699 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2700 dots are used instead of colons so that keys in lsearch files need not
2701 contain colons. This was done some time before quoting was made available
2702 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2703 (notated using the quote facility) so as to distinguish them from IPv4
2704 keys. This meant that lookups for IP addresses in host lists did not work
2705 for iplsearch lookups.
2707 This has been fixed by arranging for IPv6 addresses to be expressed with
2708 colons if the lookup type is iplsearch. This is not incompatible, because
2709 previously such lookups could never work.
2711 The situation is now rather anomalous, since one *can* have colons in
2712 ordinary lsearch keys. However, making the change in all cases is
2713 incompatible and would probably break a number of configurations.
2715 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2718 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2719 conversion specification without a maximum field width, thereby enabling
2720 a rogue spamd server to cause a buffer overflow. While nobody in their
2721 right mind would setup Exim to query an untrusted spamd server, an
2722 attacker that gains access to a server running spamd could potentially
2723 exploit this vulnerability to run arbitrary code as the Exim user.
2725 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2726 $primary_hostname instead of what libspf2 thinks the hosts name is.
2728 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2729 a directory entry by the name of the lookup key. Previously, if a
2730 symlink pointed to a non-existing file or a file in a directory that
2731 Exim lacked permissions to read, a lookup for a key matching that
2732 symlink would fail. Now it is enough that a matching directory entry
2733 exists, symlink or not. (Bugzilla 503.)
2735 PH/03 The body_linecount and body_zerocount variables are now exported in the
2738 PH/04 Added the $dnslist_matched variable.
2740 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2741 This means they are set thereafter only if the connection becomes
2744 PH/06 Added the client_condition to authenticators so that some can be skipped
2745 by clients under certain conditions.
2747 PH/07 The error message for a badly-placed control=no_multiline_responses left
2748 "_responses" off the end of the name.
2750 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2752 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2753 (without spaces) instead of just copying the configuration text.
2755 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2757 PH/11 Added $max_received_linelength.
2759 PH/12 Added +ignore_defer and +include_defer to host lists.
2761 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2762 way in which PCRE > 7.0 is built.
2764 PH/14 Implemented queue_only_load_latch.
2766 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2767 MAIL command. The effect was to mangle the value on 64-bit systems.
2769 PH/16 Another patch from the Sieve maintainer.
2771 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2773 PH/18 If a system quota error occurred while trying to create the file for
2774 a maildir delivery, the message "Mailbox is full" was not appended to the
2775 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2776 only to a quota excession during the actual writing of the file.
2778 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2779 characters?) which causes problems in log lines. The DN values are now
2780 passed through string_printing() before being added to log lines.
2782 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2783 and InterBase are left for another time.)
2785 PH/21 Added message_body_newlines option.
2787 PH/22 Guard against possible overflow in moan_check_errorcopy().
2789 PH/23 POSIX allows open() to be a macro; guard against that.
2791 PH/24 If the recipient of an error message contained an @ in the local part
2792 (suitably quoted, of course), incorrect values were put in $domain and
2793 $local_part during the evaluation of errors_copy.
2799 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2800 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2803 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2804 issue a MAIL command.
2806 PH/02 In an ACL statement such as
2808 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2810 if a client was not listed at all, or was listed with a value other than
2811 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2812 the condition was not true (as it should be), so access was not denied.
2813 The bug was that the ! inversion was incorrectly passed on to the second
2814 item. This has been fixed.
2816 PH/03 Added additional dnslists conditions == and =& which are different from
2817 = and & when the dns lookup returns more than one IP address.
2819 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2820 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2822 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2823 FSYNC, which compiles an option called disable_fsync that allows for
2824 bypassing fsync(). The documentation is heavily laced with warnings.
2826 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2828 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2829 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2830 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2831 including adding "make clean"; (3) Added -fPIC when compiling the test
2832 dynamically loaded module, to get rid of a warning.
2834 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2835 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2836 = 0s. The bug is otherwise harmless.
2838 PH/07 There was a bug in the dovecot authenticator such that the value of
2839 $auth1 could be overwritten, and so not correctly preserved, after a
2840 successful authentication. This usually meant that the value preserved by
2841 the server_setid option was incorrect.
2843 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2845 PH/09 Installed PCRE release 7.0.
2847 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2848 run for batched SMTP input. It is now run at the start of every message
2849 in the batch. While fixing this I discovered that the process information
2850 (output by running exiwhat) was not always getting set for -bs and -bS
2851 input. This is fixed, and it now also says "batched" for BSMTP.
2853 PH/11 Added control=no_pipelining.
2855 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2856 patch, slightly modified), and move the expansion of helo_data till after
2857 the connection is made in the smtp transport (so it can use these
2860 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2862 PH/14 Added log_selector = +pid.
2864 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2866 PH/16 Add ${if forany and ${if forall.
2868 PH/17 Added dsn_from option to vary the From: line in DSNs.
2870 PH/18 Flush SMTP output before performing a callout, unless control =
2871 no_callout_flush is set.
2873 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2874 was true (the default) a successful delivery failed to delete the retry
2875 item, thus causing premature timeout of the address. The bug is now
2878 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2880 PH/21 Long custom messages for fakedefer and fakereject are now split up
2881 into multiline responses in the same way that messages for "deny" and
2882 other ACL rejections are.
2884 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2885 with slight modification.
2887 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2888 draft, changing the syntax and factoring some duplicate code".
2890 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2891 for deliveries of the second and subsequent messages over the same SMTP
2894 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2895 ${reduce, with only minor "tidies".
2897 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2899 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2900 expansion side effects.
2902 PH/27 When a message times out after an over-quota error from an Exim-imposed
2903 quota, the bounce message says "mailbox is full". This message was not
2904 being given when it was a system quota that was exceeded. It now should
2907 MH/03 Made $recipients available in local_scan(). local_scan() already has
2908 better access to the recipient list through recipients_list[], but
2909 $recipients can be useful in postmaster-provided expansion strings.
2911 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2912 in the case of a MAIL command with additional options following the
2913 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2914 were accidentally chopped off.
2916 PH/29 SMTP synchronization checks are implemented when a command is read -
2917 there is a check that no more input is waiting when there shouldn't be
2918 any. However, for some commands, a delay in an ACL can mean that it is
2919 some time before the response is written. In this time, more input might
2920 arrive, invalidly. So now there are extra checks after an ACL has run for
2921 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2922 pipelining has not been advertised.
2924 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2926 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2927 returned in the SMTP response when smtp_return_error_details was false.
2928 This has been fixed.
2930 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2931 instead of the C I/O that was originally supplied, because problems were
2932 reported on Solaris.
2934 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2935 Exim which did not show up earlier: it was assuming that a call to
2936 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2937 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2938 was a macro that became an assignment, so it seemed to work. This has
2939 changed to a proper function call with a void return, hence the compile
2940 error. Exim's code has been fixed.
2942 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2945 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2947 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2949 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2950 because it turns out to be too big or there is a timeout) there is no
2951 "Completed" line in the log. When some messages of this type were
2952 selected by exigrep, they were listed as "not completed". Others were
2953 picked up by some special patterns. I have improved the selection
2954 criteria to be more general.
2956 PH/38 The host_find_failed option in the manualroute router can now be set
2957 to "ignore", to completely ignore a host whose IP address cannot be
2958 found. If all hosts are ignored, the behaviour is controlled by the new
2959 host_all_ignored option.
2961 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2962 homing or because of multiple MX records with /mx) generated more than
2963 one IP address, and the following item turned out to be the local host,
2964 all the secondary addresses of the first item were incorrectly removed
2965 from the list, along with the local host and any following hosts (which
2966 is what is supposed to happen).
2968 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2969 whoever called Exim into the -H file. In the case of the daemon it was
2970 behaving confusingly. When first started, it used values for whoever
2971 started the daemon, but after a SIGHUP it used the Exim user (because it
2972 calls itself on a restart). I have changed the code so that it now always
2975 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2976 message are rejected with the same error (e.g. no authentication or bad
2977 sender address), and a DATA command is nevertheless sent (as can happen
2978 with PIPELINING or a stupid MUA), the error message that was given to the
2979 RCPT commands is included in the rejection of the DATA command. This is
2980 intended to be helpful for MUAs that show only the final error to their
2983 PH/42 Another patch from the Sieve maintainer.
2985 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2986 Eximstats - Fixed some broken HTML links and added missing column headers
2988 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2989 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2991 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2993 PH/43 Yet another patch from the Sieve maintainer.
2995 PH/44 I found a way to check for a TCP/IP connection going away before sending
2996 the response to the final '.' that terminates a message, but only in the
2997 case where the client has not sent further data following the '.'
2998 (unfortunately, this is allowed). However, in many cases there won't be
2999 any further data because there won't be any more messages to send. A call
3000 to select() can be used: if it shows that the input is "ready", there is
3001 either input waiting, or the socket has been closed. An attempt to read
3002 the next input character can distinguish the two cases. Previously, Exim
3003 would have sent an OK response which the client would never have see.
3004 This could lead to message repetition. This fix should cure that, at
3005 least in a lot of common cases.
3007 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3008 advertised in response to EHLO.
3014 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3015 fixed by 4.65/MH/01 (is this a record?) are fixed:
3017 (i) An empty string was always treated as zero by the numeric comparison
3018 operators. This behaviour has been restored.
3020 (ii) It is documented that the numeric comparison operators always treat
3021 their arguments as decimal numbers. This was broken in that numbers
3022 starting with 0 were being interpreted as octal.
3024 While fixing these problems I realized that there was another issue that
3025 hadn't been noticed. Values of message_size_limit (both the global option
3026 and the transport option) were treated as octal if they started with 0.
3027 The documentation was vague. These values are now always treated as
3028 decimal, and I will make that clear in the documentation.
3034 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3035 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3038 MH/01 Don't check that the operands of numeric comparison operators are
3039 integers when their expansion is in "skipping" mode (fixes bug
3040 introduced by 4.64-PH/07).
3042 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3043 child addresses, Exim now panics and dies. Previously, because the count
3044 is held in a short int, deliveries were likely to be lost. As such a
3045 large number of recipients for a single message is ridiculous
3046 (performance will be very, very poor), I have chosen to impose a limit
3047 rather than extend the field.
3053 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3054 leftover -K file (the existence of which was triggered by #402).
3055 While we were at it, introduced process PID as part of the -K
3056 filename. This should rule out race conditions when creating
3059 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3060 processing considerably. Previous code took too long for large mails,
3061 triggering a timeout which in turn triggers #401.
3063 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3064 in the DK code in transports.c. sendfile() is not really portable,
3065 hence the _LINUX specificness.
3067 TF/01 In the add_headers option to the mail command in an Exim filter,
3068 there was a bug that Exim would claim a syntax error in any
3069 header after the first one which had an odd number of characters
3072 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3073 callout verification, Exim cached a "reject" for the entire domain. This
3074 is correct for most verifications, but it is not correct for a recipient
3075 verification with use_sender or use_postmaster set, because in that case
3076 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3077 case of MAIL FROM:<> rejection from other early rejections (e.g.
3078 rejection of HELO). When verifying a recipient using a non-null MAIL
3079 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3080 Whatever the result of the callout, the value of the domain cache is
3081 left unchanged (for any other kind of callout, getting as far as trying
3082 RCPT means that the domain itself is ok).
3084 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3087 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3088 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3089 session, the daemon ignores ECONNECT errors and logs others; it now
3090 ignores EPIPE as well.
3092 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3093 (quoted-printable decoding).
3095 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3096 later the small subsequent patch to fix an introduced bug.
3098 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3100 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3102 PH/08 An error is now given if message_size_limit is specified negative.
3104 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3105 to be given (somewhat) arbitrary names.
3107 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3110 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3111 miscellaneous code fixes
3113 PH/10 Added the log_reject_target ACL modifier to specify where to log
3116 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3117 hostname. This is wrong, because it relates to the incoming message (and
3118 probably the interface on which it is arriving) and not to the outgoing
3119 callout (which could be using a different interface). This has been
3120 changed to use the value of the helo_data option from the smtp transport
3121 instead - this is what is used when a message is actually being sent. If
3122 there is no remote transport (possible with a router that sets up host
3123 addresses), $smtp_active_hostname is used.
3125 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3126 tweaks were necessary in order to get it to work (see also 21 below):
3127 (a) The code assumed that strncpy() returns a negative number on buffer
3128 overflow, which isn't the case. Replaced with Exim's string_format()
3130 (b) There were several signed/unsigned issues. I just did the minimum
3131 hacking in of casts. There is scope for a larger refactoring.
3132 (c) The code used strcasecmp() which is not a standard C function.
3133 Replaced with Exim's strcmpic() function.
3134 (d) The code set only $1; it now sets $auth1 as well.
3135 (e) A simple test gave the error "authentication client didn't specify
3136 service in request". It would seem that Dovecot has changed its
3137 interface. Fortunately there's a specification; I followed it and
3138 changed what the client sends and it appears to be working now.
3140 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3143 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3144 address A is aliased to B and C, where B exists and C does not. Without
3145 -v the output is "A verified" because verification stops after a
3146 successful redirection if more than one address is generated. However,
3147 with -v the child addresses are also verified. Exim was outputting "A
3148 failed to verify" and then showing the successful verification for C,
3149 with its parentage. It now outputs "B failed to verify", showing B's
3150 parentage before showing the successful verification of C.
3152 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3153 look up a TXT record in a specific list after matching in a combined
3156 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3157 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3158 they consult the DNS. I had assumed they would set it the way they
3159 wanted; and indeed my experiments on Linux seem to show that in some
3160 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3161 To be on the safe side, however, I have now made the interface to
3162 host_find_byname() similar to host_find_bydns(), with an argument
3163 containing the DNS resolver options. The host_find_byname() function now
3164 sets these options at its start, just as host_find_bydns() does. The smtp
3165 transport options dns_qualify_single and dns_search_parents are passed to
3166 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3167 of host_find_byname() use the default settings of RES_DEFNAMES
3168 (qualify_single) but not RES_DNSRCH (search_parents).
3170 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3171 spool_read_header() do less string testing, by means of a preliminary
3172 switch on the second character of optional "-foo" lines. (This is
3173 overdue, caused by the large number of possibilities that now exist.
3174 Originally there were few.) While I was there, I also converted the
3175 str(n)cmp tests so they don't re-test the leading "-" and the first
3176 character, in the hope this might squeeze out yet more improvement.
3178 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3179 flag allowing group syntax was set by the header_syntax check but not
3180 turned off, possible causing trouble later; (2) The flag was not being
3181 set at all for the header_verify test, causing "group"-style headers to
3182 be rejected. I have now set it in this case, and also caused header_
3183 verify to ignore an empty address taken from a group. While doing this, I
3184 came across some other cases where the code for allowing group syntax
3185 while scanning a header line wasn't quite right (mostly, not resetting
3186 the flag correctly in the right place). These bugs could have caused
3187 trouble for malformed header lines. I hope it is now all correct.
3189 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3190 with the "reply" argument non-NULL. The code, however (which originally
3191 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3192 but it didn't always do it. This confused somebody who was copying the
3193 code for some other use. I have removed all the tests.
3195 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3196 feature that was used to support insecure browsers during the U.S. crypto
3197 embargo. It requires special client support, and Exim is probably the
3198 only MTA that supported it -- and would never use it because real RSA is
3199 always available. This code has been removed, because it had the bad
3200 effect of slowing Exim down by computing (never used) parameters for the
3201 RSA_EXPORT functionality.
3203 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3204 authenticator to fail if there's a tab character in the incoming data
3205 (there should never be unless someone is messing about, as it's supposed
3206 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3207 if the connection is using TLS or if the remote IP is the same as the
3208 local IP, and the "valid-client-cert option" if a client certificate has
3211 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3212 authenticators. This can be used for authorization after authentication
3213 succeeds. (In the case of plaintext, it servers for both authentication
3216 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3217 if any retry times were supplied.
3219 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3220 connection, where there is no client IP address to check. In this
3221 situation, the verify now always succeeds.
3223 PH/25 Applied John Jetmore's -Mset patch.
3225 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3227 PH/27 In a string expansion for a processed (not raw) header when multiple
3228 headers of the same name were present, leading whitespace was being
3229 removed from all of them, but trailing whitespace was being removed only
3230 from the last one. Now trailing whitespace is removed from each header
3231 before concatenation. Completely empty headers in a concatenation (as
3232 before) are ignored.
3234 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3235 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3237 PH/29 [Removed. This was a change that I later backed out, and forgot to
3238 correct the ChangeLog entry (that I had efficiently created) before
3239 committing the later change.]
3241 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3242 address errors (4xx response to RCPT) over the same connection as other
3243 messages routed to the same hosts. Such deliveries are always "forced",
3244 so retry times are not inspected. This resulted in far too many retries
3245 for the affected addresses. The effect occurred only when there were more
3246 hosts than the hosts_max_try setting in the smtp transport when it had
3247 the 4xx errors. Those hosts that it had tried were not added to the list
3248 of hosts for which the message was waiting, so if all were tried, there
3249 was no problem. Two fixes have been applied:
3251 (i) If there are any address or message errors in an SMTP delivery, none
3252 of the hosts (tried or untried) are now added to the list of hosts
3253 for which the message is waiting, so the message should not be a
3254 candidate for sending over the same connection that was used for a
3255 successful delivery of some other message. This seems entirely
3256 reasonable: after all the message is NOT "waiting for some host".
3257 This is so "obvious" that I'm not sure why it wasn't done
3258 previously. Hope I haven't missed anything, but it can't do any
3259 harm, as the worst effect is to miss an optimization.
3261 (ii) If, despite (i), such a delivery is accidentally attempted, the
3262 routing retry time is respected, so at least it doesn't keep
3263 hammering the server.
3265 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3266 in ${readsocket because some servers need this prod.
3268 PH/32 Added some extra debug output when updating a wait-xxx database.
3270 PH/33 The hint "could be header name not terminated by colon", which has been
3271 given for certain expansion errors for a long time, was not being given
3272 for the ${if def:h_colon_omitted{... case.
3274 PH/34 The spec says: "With one important exception, whenever a domain list is
3275 being scanned, $domain contains the subject domain." There was at least
3276 one case where this was not true.
3278 PH/35 The error "getsockname() failed: connection reset by peer" was being
3279 written to the panic log as well as the main log, but it isn't really
3280 panic-worthy as it just means the connection died rather early on. I have
3281 removed the panic log writing for the ECONNRESET error when getsockname()
3284 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3285 runs only) independently of the message's sender address. This meant
3286 that, if the 4xx error was in fact related to the sender, a different
3287 message to the same recipient with a different sender could confuse
3288 things. In particular, this can happen when sending to a greylisting
3289 server, but other circumstances could also provoke similar problems.
3290 I have changed the default so that the retry time for these errors is now
3291 based a combination of the sender and recipient addresses. This change
3292 can be overridden by setting address_retry_include_sender=false in the
3295 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3296 remote server are returned as part of bounce messages. This was not
3297 happening for LMTP over a pipe (the lmtp transport), but now it is the
3298 same for both kinds of LMTP.
3300 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3301 in header lines that were in fact CNAMEs. This is no longer the case.
3303 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3304 and queue runs started by the daemon processed all messages. This has
3305 been fixed so that -R and -S can now usefully be given with -q<time>.
3307 PH/40 Import PCRE release 6.7 (fixes some bugs).
3309 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3311 PH/42 Give an error if -q is specified more than once.
3313 PH/43 Renamed the variables $interface_address and $interface_port as
3314 $received_ip_address and $received_port, to make it clear that these
3315 values apply to message reception, and not to the outgoing interface when
3316 a message is delivered. (The old names remain recognized, of course.)
3318 PH/44 There was no timeout on the connect() call when using a Unix domain
3319 socket in the ${readsocket expansion. There now is.
3321 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3322 be meaningful with "accept".
3324 SC/01 Eximstats V1.43
3325 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3327 SC/02 Eximstats V1.44
3328 Use a glob alias rather than an array ref in the generated
3329 parser. This improves both readability and performance.
3331 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3332 Collect SpamAssassin and rejection statistics.
3333 Don't display local sender or destination tables unless
3334 there is data to show.
3335 Added average volumes into the top table text output.
3337 SC/04 Eximstats V1.46
3338 Collect data on the number of addresses (recipients)
3339 as well as the number of messages.
3341 SC/05 Eximstats V1.47
3342 Added 'Message too big' to the list of mail rejection
3343 reasons (thanks to Marco Gaiarin).
3345 SC/06 Eximstats V1.48
3346 Mainlog lines which have GMT offsets and are too short to
3347 have a flag are now skipped.
3349 SC/07 Eximstats V1.49 (Alain Williams)
3350 Added the -emptyok flag.
3352 SC/08 Eximstats V1.50
3353 Fixes for obtaining the IP address from reject messages.
3355 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3356 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3357 whitespace changes from 4.64-PH/27
3359 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3362 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3363 are found, allow negative numbers in numeric criteria)
3365 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3367 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3368 to match changes made in 4.64-PH/43
3370 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3372 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3373 "domain" is actually an IP address, and force a failure. This locks out
3374 those revolvers/nameservers that support "A-for-A" lookups, in
3375 contravention of the specifications.
3377 PH/48 When a host name was looked up from an IP address, and the subsequent
3378 forward lookup of the name timed out, the host name was left in
3379 $sender_host_name, contrary to the specification.
3381 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3382 restricted to single-key lookups, Exim was not diagnosing an error if
3383 * or *@ was used with a query-style lookup.
3385 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3387 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3388 long ago, but noone interested enough thought of it. Let's just say that
3389 the "1.1" means that there are some new functions that weren't there at
3390 some point in the past.
3392 PH/51 Error processing for expansion failure of helo_data from an smtp
3393 transport during callout processing was broken.
3395 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3396 tested/used via the -bh/-bhc/-bs options.
3398 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3399 bug, fixed in subsequent PCRE releases).
3401 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3402 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3404 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3410 SC/01 Use a glob alias rather than an array ref in eximstats generated
3411 parser. This improves both readability and performance.
3413 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3414 Don't display local sender or destination tables in eximstats unless
3415 there is data to show.
3416 Added average volumes into the eximstats top table text output.
3418 SC/03 Collect data on the number of addresses (recipients) as well
3419 as the number of messages in eximstats.
3421 TF/01 Correct an error in the documentation for the redirect router. Exim
3422 does (usually) call initgroups() when daemonizing.
3424 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3425 with consistent privilege compared to when running as a daemon.
3427 TF/03 Note in the spec that $authenticated_id is not set for local
3428 submissions from trusted users.
3430 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3431 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3433 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3434 by adding some example configuration directives to the default
3435 configuration file. A little bit of work is required to uncomment the
3436 directives and define how usernames and passwords are checked, but
3437 there is now a framework to start from.
3439 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3440 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3441 without this. I don't know how relevant this is to other LDAP libraries.
3443 PH/02 Add the verb name to the "unknown ACL verb" error.
3445 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3447 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3449 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3450 directory not expanded when it should be if an expanded home directory
3451 was set for the address (which is overridden by the transport).
3453 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3456 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3457 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3458 because it is too late at that time, and has no effect.
3460 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3461 security issue with \' (bugzilla #107). I could not use the
3462 PQescapeStringConn() function, because it needs a PGconn value as one of
3465 PH/08 When testing addresses using -bt, indicate those final addresses that
3466 are duplicates that would not cause an additional delivery. At least one
3467 person was confused, thinking that -bt output corresponded to deliveries.
3468 (Suppressing duplicates isn't a good idea as you lose the information
3469 about possibly different redirections that led to the duplicates.)
3471 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3472 systems where poll() doesn't work, in particular OS X.
3474 PH/10 Added more information to debugging output for retry time not reached.
3476 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3477 operations in malware.c.
3479 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3482 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3483 syslog_duplication set false, log lines that would normally be written
3484 both the the main log and to the reject log were not written to syslog at
3487 PH/14 In the default configuration, change the use of "message" in ACL warn
3488 statements to "add_header".
3490 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3491 not followed by a command (e.g. "seen endif").
3493 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3494 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3497 PH/17 Added extra conditions to the default value of delay_warning_condition
3501 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3502 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3503 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3506 The Auto-Submitted: and various List- headers are standardised, whereas I
3507 don't think Precedence: ever was.
3509 PH/18 Refactored debugging code in route_finduser() to show more information,
3510 in particular, the error code if getpwnam() issues one.
3512 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3513 This is apparently needed in addition to the PH/07 change above to avoid
3514 any possible encoding problems.
3516 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3517 but not after initializing Perl.
3519 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3520 output them only if debugging. By default they are written stderr,
3521 apparently, which is not desirable.
3523 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3526 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3529 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3531 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3532 authenticated or an ident call has been made. Suppress the default
3533 values for $authenticated_id and $authenticated_sender (but permit -oMai
3534 and -oMas) when testing with -bh.
3536 PH/24 Re-jigged the order of the tests in the default configuration so that the
3537 tests for valid domains and recipients precede the DNS black list and CSA
3538 tests, on the grounds that those ones are more expensive.
3540 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3541 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3542 This bug exists in every version of Exim that I still have, right back to
3545 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3546 However, an attempt to turn on case-sensitivity in a regex key by
3547 including (?-i) didn't work because the subject string was already
3548 lowercased, and the effects were non-intuitive. It turns out that a
3549 one-line patch can be used to allow (?-i) to work as expected.
3555 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3556 other effects) broke the use of negated acl sub-conditions.
3558 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3561 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3562 "Deny" causes Exim to reject the incoming connection with a 554 error.
3563 Unfortunately, if there is a major crisis, such as a disk failure,
3564 tcp-wrappers gives "deny", whereas what one would like would be some
3565 kind of temporary error. A kludge has been added to help with this.
3566 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3567 554 error is used if errno is still zero or contains ENOENT (which occurs
3568 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3571 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3573 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3574 errors. Otherwise a message that provokes a temporary error (when other
3575 messages do not) can cause a whole host to time out.
3577 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3578 addresses were routed directly to files or pipes from a redirect router.
3579 File deliveries just didn't batch; pipe deliveries might have suffered
3582 PH/06 A failure to get a lock for a hints database would erroneously always say
3583 "Failed to get write lock", even when it was really a read lock.
3585 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3586 of 0600. This has been changed to use the value of the lockfile_mode
3587 option (which defaults to 0600).
3589 PH/08 Applied small patch from the Sieve maintainer.
3591 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3592 folder from quota calculations, a direct delivery into this folder messed
3593 up the contents of the maildirsize file. This was because the regex was
3594 used only to exclude .Trash (or whatever) when the size of the mailbox
3595 was calculated. There was no check that a delivery was happening into an
3596 excluded directory. This bug has been fixed by ignoring all quota
3597 processing for deliveries into excluded directories.
3599 PH/10 Added the maildirfolder_create_regex option to appendfile.
3605 PH/01 The code for finding all the local interface addresses on a FreeBSD
3606 system running IPv6 was broken. This may well have applied to all BSD
3607 systems, as well as to others that have similar system calls. The broken
3608 code found IPv4 interfaces correctly, but gave incorrect values for the
3609 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3610 that it would not match correctly against @[] and not recognize the IPv6
3613 PH/02 The ipliteral router was not recognizing addresses of the form user@
3614 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3616 PH/03 Added disable_ipv6.
3618 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3619 decoded form, because it is most often used to construct To: headers
3620 lines in autoreplies, and the decoded form may well be syntactically
3621 invalid. However, $reply_address has leading white space removed, and all
3622 newlines turned into spaces so that the autoreply transport does not
3625 PH/05 If group was specified without a user on a router, and no group or user
3626 was specified on a transport, the group from the router was ignored.
3628 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3629 for visible compile-time settings that can be used to change these
3630 numbers, for those that want even more. Backwards compatibility with old
3631 spool files has been maintained. However, going back to a previous Exim
3632 release will lost any variables that are in spool files.
3634 PH/07 Two small changes when running in the test harness: increase delay when
3635 passing a TCP/IP connection to a new process, in case the original
3636 process has to generate a bounce, and remove special handling of
3637 127.0.0.2 (sic), which is no longer necessary.
3639 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3640 be the same on different OS.
3642 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3645 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3646 whether --show-vars was specified or not
3648 JJ/02 exipick: Added support for new ACL variable spool format introduced
3651 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3652 syntactically invalid From: or Reply-to: line, and a filter used this to
3653 generate an autoreply, and therefore failed to obtain an address for the
3654 autoreply, Exim could try to deliver to a non-existent relative file
3655 name, causing unrelated and misleading errors. What now happens is that
3656 it logs this as a hard delivery error, but does not attempt to create a
3659 PH/11 The exinext utility has a -C option for testing purposes, but although
3660 the given file was scanned by exinext itself; it wasn't being passed on
3661 when Exim was called.
3663 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3664 an end-of-file indication when reading a command response.
3666 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3667 compiled. In many other places in Exim, IPv6 addresses are always
3668 recognized, so I have changed this. It also means that IPv4 domain
3669 literals of the form [IPV4:n.n.n.n] are now always recognized.
3671 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3672 used if the router is not running as root, for example, when verifying at
3673 ACL time, or when using -bh. The debugging output from this situation was
3674 non-existent - all you got was a failure to exec. I have made two
3677 (a) Failures to set uid/gid, the current directory, or a process leader
3678 in a subprocess such as that created by queryprogram now generate
3679 suitable debugging output when -d is set.
3681 (b) The queryprogram router detects when it is not running as root,
3682 outputs suitable debugging information if -d is set, and then runs
3683 the subprocess without attempting to change uid/gid.
3685 PH/15 Minor change to Makefile for building test_host (undocumented testing
3688 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3689 additional section of a DNS packet that returns MX or SRV records.
3690 Instead, it always explicitly searches for A/AAAA records. This avoids
3691 major problems that occur when a DNS server includes only records of one
3692 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3693 fixed another bug: if SRV records were looked up and the corresponding
3694 address records were *not* found in the additional section, the port
3695 values from the SRV records were lost.
3697 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3698 using the correct key (the original address) when searching the retry
3699 rules in order to find which one to use for generating the retry hint.
3701 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3702 adding the default one. Similarly, if it contains a Reply-To: header, the
3703 errors_reply_to option, if set, is not used.
3705 PH/19 When calculating a retry time, Exim used to measure the "time since
3706 failure" by looking at the "first failed" field in the retry record. Now
3707 it does not use this if it is later than than the arrival time of the
3708 message. Instead it uses the arrival time. This makes for better
3709 behaviour in cases where some deliveries succeed, thus re-setting the
3710 "first failed" field. An example is a quota failure for a huge message
3711 when small messages continue to be delivered. Without this change, the
3712 "time since failure" will always be short, possible causing more frequent
3713 delivery attempts for the huge message than are intended.
3714 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3716 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3717 $1, $2, $3) because the numerical variables can be reset during some
3718 expansion items (e.g. "match"), thereby losing the authentication data.
3720 PH/21 Make -bV show the size of off_t variables so that the test suite can
3721 decide whether to run tests for quotas > 2G.
3723 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3724 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3725 filecount value is greater than 2G or if a quota value is greater than 2G
3726 on a system where the size of off_t is not greater than 4, a panic error
3729 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3730 never match. The debug and -bh output now contains an explicit error
3731 message indicating a malformed IPv4 address or mask.
3733 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3734 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3735 PH/23 above applies.
3737 PH/25 Do not write to syslog when running in the test harness. The only
3738 occasion when this arises is a failure to open the main or panic logs
3739 (for which there is an explicit test).
3741 PH/26 Added the /no_tell option to "control=freeze".
3743 PH/27 If a host name lookup failed very early in a connection, for example, if
3744 the IP address matched host_lookup and the reverse lookup yielded a name
3745 that did not have a forward lookup, an error message of the form "no IP
3746 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3747 could be logged. Now it outputs the IP address instead of "NULL".
3749 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3750 allows the sender and the authenticated sender to be set when
3751 submitting a message from within Exim. Since child_open_exim() is
3752 documented for local_scan(), the new function should be too.
3754 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3755 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3756 results in an empty string is now treated as unset.
3758 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3760 PH/31 Added sender_verify_fail logging option.
3762 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3763 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3764 tidied the source and removed it altogether.
3766 PH/33 When a queue run was abandoned because the load average was too high, a
3767 log line was always written; now it is written only if the queue_run log
3768 selector is set. In addition, the log line for abandonment now contains
3769 information about the queue run such as the pid. This is always present
3770 in "start" and "stop" lines but was omitted from the "abandon" line.
3772 PH/34 Omit spaces between a header name and the colon in the error message that
3773 is given when verify = headers_syntax fails (if there are lots of them,
3774 the message gets confusing).
3776 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3777 names, as there are now some PTR records that contain slashes. This check
3778 is only to protect against broken name servers that fall over on strange
3779 characters, so the fact that it applies to all lookups doesn't matter.
3781 PH/36 Now that the new test suite is complete, we can remove some of the
3782 special code in Exim that was needed for the old test suite. For example,
3783 sorting DNS records because real resolvers return them in an arbitrary
3784 order. The new test suite's fake resolver always returns records in the
3787 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3788 bounces) except when queue_only is set, to avoid logging races between
3789 the different processes.
3791 PH/38 Panic-die if .include specifies a non-absolute path.
3793 PH/39 A tweak to the "H" retry rule from its user.
3795 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3796 a label. They prevented compilation on older perls.
3798 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3799 a warning to be raised on newish perls.
3801 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3802 on queue. Changes to match documented behaviour of showing count of
3803 messages matching specified criteria.
3805 PH/40 Changed the default ident timeout from 30s to 5s.
3807 PH/41 Added support for the use of login_cap features, on those BSD systems
3808 that have them, for controlling the resources used by pipe deliveries.
3810 PH/42 The content-scanning code uses fopen() to create files in which to put
3811 message data. Previously it was not paying any attention to the mode of
3812 the files. Exim runs with umask(0) because the rest of the code creates
3813 files with open(), and sets the required mode explicitly. Thus, these
3814 files were ending up world-writeable. This was not a big issue, because,
3815 being within the spool directory, they were not world-accessible. I have
3816 created a function called modefopen, which takes an additional mode
3817 argument. It sets umask(777), creates the file, chmods it to the required
3818 mode, then resets the umask. All the relevant calls to fopen() in the
3819 content scanning code have been changed to use this function.
3821 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3822 to 24 hours. This avoids potential overflow problems when processing G
3823 and H retry rules. I suspect nobody ever tinkers with this value.
3825 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3827 PH/45 When the plaintext authenticator is running as a client, the server's
3828 challenges are checked to ensure they are valid base64 strings. By
3829 default, the authentication attempt is cancelled if an invalid string is
3830 received. Setting client_ignore_invalid_base64 true ignores these errors.
3831 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3832 they are received. Thus, the responses can be made to depend on the
3833 challenges. If an invalid string is ignored, an empty string is placed in
3836 PH/46 Messages that are created by the autoreply transport now contains a
3837 References: header, in accordance with RFCs 2822 and 3834.
3839 PH/47 Added authenticated_sender_force to the smtp transport.
3841 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3843 PH/49 Installed latest patch from the Sieve maintainer.
3845 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3846 was also set, the appendfile transport was unnecessarily scanning a
3847 directory of message files (e.g. for maildir delivery) to find the count
3848 of files (along with the size), even though it did not need this
3849 information. It now does the scan only if it needs to find either the
3850 size of the count of files.
3852 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3854 PH/52 Two bugs concerned with error handling when the smtp transport is
3857 (i) Exim was not creating retry information for temporary errors given
3858 for individual recipients after the DATA command when the smtp transport
3859 was used in LMTP mode. This meant that they could be retried too
3860 frequently, and not timed out correctly.
3862 (ii) Exim was setting the flag that allows error details to be returned
3863 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3864 recipients that were returned after the DATA command.
3866 PH/53 This is related to PH/52, but is more general: for any failing address,
3867 when detailed error information was permitted to be returned to the
3868 sender, but the error was temporary, then after the final timeout, only
3869 "retry timeout exceeded" was returned. Now it returns the full error as
3870 well as "retry timeout exceeded".
3872 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3873 do this, and (what is worse) MTAs that accept it.
3875 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3876 will now be deprecated.
3878 PH/56 New os.c-cygwin from the Cygwin maintainer.
3880 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3881 formats (previously only available in exim formats via -bpr, -bpru,
3882 and -bpra. Now also available in native and exiqgrep formats)
3884 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3885 with very large, slow to parse queues
3887 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3889 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3891 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3892 responses to authentication challenges, though it was showing the
3893 challenges; (ii) I've removed the CR characters from the debug output for
3896 PH/58 Allow for the insertion of a newline as well as a space when a string
3897 is turned into more than one encoded-word during RFC 2047 encoding. The
3898 Sieve code now uses this.
3900 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3901 data_4xx, lost_connection, tls_required.
3903 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3904 message was being sent as an SMTP response.
3906 PH/61 Add -l and -k options to exicyclog.
3908 PH/62 When verifying, if an address was redirected to one new address, so that
3909 verification continued, and the new address failed or deferred after
3910 having set something in $address_data, the value of $address_data was not
3911 passed back to the ACL. This was different to the case when no
3912 redirection occurred. The value is now passed back in both cases.
3914 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3915 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3916 use login_cap.h, so on its own it isn't the distinguishing feature. The
3917 new name refers directly to the setclassresources() function.
3919 PH/65 Added configuration files for NetBSD3.
3921 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3923 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3924 is preferred over IPv4.
3926 PH/68 The bounce_return_message and bounce_return_body options were not being
3927 honoured for bounces generated during the reception of non-SMTP messages.
3928 In particular, this applied to messages rejected by the ACL. This bug has
3929 been fixed. However, if bounce_return_message is true and bounce_return_
3930 body is false, the headers that are returned for a non-SMTP message
3931 include only those that have been read before the error was detected.
3932 (In the case of an ACL rejection, they have all been read.)
3934 PH/69 The HTML version of the specification is now built in a directory called
3935 spec_html instead of spec.html, because the latter looks like a path with
3936 a MIME-type, and this confuses some software.
3938 PH/70 Catch two compiler warnings in sieve.c.
3940 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3941 function verify_get_ident() calls ip_connect() to connect a socket, but
3942 if the "connect()" function timed out, ip_connect() used to close the
3943 socket. However, verify_get_ident() also closes the socket later, and in
3944 between Exim writes to the log, which may get opened at this point. When
3945 the socket was closed in ip_connect(), the log could get the same file
3946 descriptor number as the socket. This naturally causes chaos. The fix is
3947 not to close the socket in ip_connect(); the socket should be closed by
3948 the function that creates it. There was only one place in the code where
3949 this was missing, in the iplookup router, which I don't think anybody now
3950 uses, but I've fixed it anyway.
3952 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3953 well as to direct DNS lookups. Otherwise the handling of names in host
3954 lists is inconsistent and therefore confusing.
3960 PH/01 Two changes to the default runtime configuration:
3962 (1) Move the checks for relay_from_hosts and authenticated clients from
3963 after to before the (commented out) DNS black list checks.
3965 (2) Add control=submission to the relay_from_hosts and authenticated
3966 clients checks, on the grounds that messages accepted by these
3967 statements are most likely to be submissions.
3969 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3971 (1) Generate an error if the third argument for the ${prvs expansion is
3974 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3977 (3) Reset the variables that are obtained from the first argument of
3978 ${prvscheck and used in the second argument before leaving the code,
3979 because their memory is reclaimed, so using them afterwards may do
3982 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3983 one (it's much easier than Tom thought :-).
3985 (5) Because of (4), we can now allow for the use of $prvscheck_result
3986 inside the third argument.
3988 PH/03 For some reason, the default setting of PATH when running a command from
3989 a pipe transport was just "/usr/bin". I have changed it to
3992 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3993 anything to be listed in the output from -bV.
3995 PH/05 When a filter generated an autoreply, the entire To: header line was
3996 quoted in the delivery log line, like this:
3998 => >A.N.Other <ano@some.domain> <original@ddress> ...
4000 This has been changed so that it extracts the operative address. There
4001 may be more than one such address. If so, they are comma-separated, like
4004 => >ano@some.domain,ona@other.domain <original@ddress> ...
4006 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4007 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4008 not being looked up in the rDNS to get a host name, Exim was showing the
4009 IP address twice in Received: lines, even though the IP addresses were
4010 identical. For example:
4012 Received: from [1.2.3.4] (helo=[1.2.3.4])
4014 However, if the real host name was known, it was omitting the HELO data
4015 if it matched the actual IP address. This has been tidied up so that it
4016 doesn't show the same IP address twice.
4018 PH/07 When both +timestamp and +memory debugging was on, the value given by
4019 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4020 called by the debug printing, thereby overwriting the timestamp buffer.
4021 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4023 PH/08 When the original message was included in an autoreply transport, it
4024 always said "this is a copy of the message, including all the headers",
4025 even if body_only or headers_only was set. It now gives an appropriate
4028 PH/09 Applied a patch from the Sieve maintainer which:
4030 o fixes some comments
4031 o adds the (disabled) notify extension core
4032 o adds some debug output for the result of if/elsif tests
4033 o points to the current vacation draft in the documentation
4034 and documents the missing references header update
4038 o fixes a bug in processing the envelope test (when testing
4039 multiple envelope elements, the last element determined the
4042 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4043 Electronic Mail") by including:
4045 Auto-submitted: auto-generated
4047 in the messages that it generates (bounce messages and others, such as
4048 warnings). In the case of bounce messages for non-SMTP messages, there was
4049 also a typo: it was using "Auto_submitted" (underscore instead of
4050 hyphen). Since every message generated by Exim is necessarily in response
4051 to another message, thes have all been changed to:
4053 Auto-Submitted: auto-replied
4055 in accordance with these statements in the RFC:
4057 The auto-replied keyword:
4059 - SHOULD be used on messages sent in direct response to another
4060 message by an automatic process,
4062 - MUST NOT be used on manually-generated messages,
4064 - MAY be used on Delivery Status Notifications (DSNs) and Message
4065 Disposition Notifications (MDNs),
4067 - MUST NOT be used on messages generated by automatic or periodic
4068 processes, except for messages which are automatic responses to
4071 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4072 to the default Received: header definition.
4074 PH/12 Added log selector acl_warn_skipped (default on).
4076 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4077 variables because (a) they are in the wrong storage pool and (b) even if
4078 they were copied, it wouldn't work properly because of the caching.
4080 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4081 checking when decoding. Apparently there are clients that generate
4082 overlong encoded strings. Why am I not surprised?
4084 PH/15 If the first argument of "${if match_address" was not empty, but did not
4085 contain an "@" character, Exim crashed. Now it writes a panic log message
4086 and treats the condition as false.
4088 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4090 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4091 extension "envelope-auth". The code is finished and in agreement with
4092 other implementations, but there is no documentation so far and in fact,
4093 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4094 not changing the active code.
4096 Print executed "if" and "elsif" statements when debugging is used. This
4097 helps a great deal to understand what a filter does.
4099 Document more things not specified clearly in RFC3028. I had all this
4100 sorted out, when out of a sudden new issues came to my mind. Oops."
4102 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4105 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4106 canonical form (as documented). However, after a host name lookup from
4107 the IP address, check_host() was doing a simple string comparison with
4108 addresses acquired from the DNS when checking that the found name did
4109 have the original IP as one of its addresses. Since any found IPv6
4110 addresses are likely to be in abbreviated form, the comparison could
4111 fail. Luckily, there already exists a function for doing the comparison
4112 by converting both addresses to binary, so now that is used instead of
4113 the text comparison.
4115 PH/20 There was another similar case to PH/19, when a complete host name was
4116 given in a host list; looking up its IP address could give an abbreviated
4117 form, whereas the current host's name might or might not be abbreviated.
4118 The same fix has been applied.
4124 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4125 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4128 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4129 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4131 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4133 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4134 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4135 Also, for "Auto-Submitted", treat anything other than "no" as
4136 non-personal, in accordance with RFC 3834. (Previously it treated
4137 anything starting "auto-" as non-personal.)
4139 TF/01 The control=submission/name=... option had a problem with syntax
4140 errors if the name included a slash character. The /name= option
4141 now slurps the rest of the string, so it can include any characters
4142 but it must come last in the list of options (after /sender_retain
4145 PH/05 Some modifications to the interface to the fake nameserver for the new
4153 TK/01 Added the "success_on_redirect" address verification option. See
4154 NewStuff for rationale and an example.
4156 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4158 PH/02 Patch to exigrep to allow it to work on syslog lines.
4160 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4161 fread() to skip over the body file's header line, because in Cygwin the
4162 header line is locked and is inaccessible.
4164 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4165 co-exist for some time) to make it clear that it is the Exim ID that is
4166 referenced, not the Message-ID: header line.
4168 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4169 string_format() function, because snprintf() does not exist on all
4172 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4173 ${stat: expansion item.
4175 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4176 protocol synchronization error", to keep the pedants happy.
4178 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4179 well as for IRIX systems, when gcc is being used. See the host.c source
4182 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4184 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4187 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4188 transport and to the smtp transport in LMTP mode.
4190 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4192 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4193 run a filter in a subprocess. This could lead to confusion in subsequent
4194 lookups in the parent process. There should also be a search_tidyup() at
4195 the end of the subprocess.
4197 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4198 only if the host matched helo_try_verify_hosts, which caused the
4199 verification to occur when the EHLO/HELO command was issued. The ACL just
4200 tested the remembered result. Now, if a previous verification attempt has
4201 not happened, "verify = helo" does it there and then.
4203 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4205 TK/03 Fix log output including CR from clamd.
4207 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4208 exist provoked a memory error which could cause a segfault.
4210 PH/15 Installed PCRE 6.2
4212 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4214 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4215 of the problem. Specifically, suggested +O2 rather than +O1 for the
4218 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4220 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4221 record, the port was not correctly logged when the outgoing_port log
4222 selector was set (it logged the transort's default port).
4224 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4225 fallback_hosts, and "hosts" in the smtp transport.
4227 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4228 host errors such as "Connection refused".
4230 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4231 authentication with radiusclient 0.4.9:
4233 - Error returned from rc_read_config was caught wrongly
4234 - Username/password not passed on to radius server due to wrong length.
4236 The presumption is that some radiusclient API changes for 4.51/PH/17
4237 were not taken care of correctly. The code is still untested by me (my
4238 Linux distribution still has 0.3.2 of radiusclient), but it was
4239 contributed by a Radius user.
4241 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4242 expanding the "port" option of the smtp transport.
4244 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4245 while reading a MIME header. Thanks to Tom Hughes for a patch.
4247 PH/24 Include config.h inside local_scan.h so that configuration settings are
4250 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4251 that in an ACL for RCPT (for example), you can examine exactly what was
4254 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4255 commands, but it was not correctly comparing the address with the actual
4256 client host address. Thus, it would show the EHLO address in Received:
4257 header lines when this was not necessary.
4259 PH/27 Added the % operator to ${eval:}.
4261 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4262 it should be ignoring failures (because with -C, for example, it has lost
4263 privilege). It wasn't ignoring creation failures other than "already
4266 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4269 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4270 because an input line was too long, either on its own, or by virtue of
4271 too many continuations, the temporary file was not being removed, and the
4272 return code was incorrect.
4274 PH/31 Missing "BOOL" in function definition in filtertest.c.
4276 PH/32 Applied Sieve patches from the maintainer.
4278 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4280 PH/33 Added "verify = not_blind".
4282 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4283 Local/Makefile (with some defaults set). These are used in built scripts
4284 such as exicyclog, but they have never been used in the exim_install
4285 script (though there are many overriding facilities there). I have
4286 arranged that the exim_install script now takes note of these two
4289 PH/35 Installed configuration files for Dragonfly.
4291 PH/36 When a locally submitted message by a trusted user did not contain a
4292 From: header, and the sender address was obtained from -f or from an SMTP
4293 MAIL command, and the trusted user did not use -F to supply a sender
4294 name, $originator_name was incorrectly used when constructing a From:
4295 header. Furthermore, $originator_name was used for submission mode
4296 messages from external hosts without From: headers in a similar way,
4297 which is clearly wrong.
4299 PH/37 Added control=suppress_local_fixups.
4301 PH/38 When log_selector = +received_sender was set, and the addition of the
4302 sender made the log line's construction buffer exactly full, or one byte
4303 less than full, an overflow happened when the terminating "\n" was
4306 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4307 when the result of a list match is failure because a DNS lookup failed.
4309 PH/40 RM_COMMAND is now used in the building process.
4311 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4312 the "build-* directories that it finds.
4314 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4315 address in a domain literal was a prefix of an interface address.
4317 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4318 when verifying a sender address, unless rewrite_headers is false.
4320 PH/44 Wrote a long comment about why errors_to addresses are verified as
4321 recipients, not senders.
4323 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4324 the ratelimit ACL was added.
4326 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4328 PH/46 Added extra information about PostgreSQL errors to the error string.
4330 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4331 avoiding the need to install special zones in a real server. This is
4332 backwards compatible; if it can't find the fake resolver, it drops back.
4333 Thus, both old and new test suites can be run.
4335 TF/02 Added util/ratelimit.pl
4337 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4338 clock is set back in time.
4340 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4341 Candler <B.Candler@pobox.com>.
4343 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4344 OK for addresses that are the result of redirections.
4346 PH/48 A number of further additions for the benefit of the new test suite,
4347 including a fake gethostbyname() that interfaces to the fake DNS resolver
4350 TF/06 The fix for widen_domains has also been applied to qualify_single and
4351 search_parents which are the other dnslookup options that can cause
4354 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4357 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4359 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4360 service" if it encountered an MX record with an empty target hostname.
4361 The message is now "an MX or SRV record indicated no SMTP service".
4363 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4364 if the DNS of the sending site is misconfigured. This is quite a
4365 common situation. This change restores the behaviour of treating a
4366 helo verification defer as a failure.
4368 PH/51 If self=fail was set on a router, the bounce message did not include the
4369 actual error message.
4375 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4377 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4378 command itself ended in error, the underlying message about the transport
4379 filter timeout was being overwritten with the pipe command error. Now the
4380 underlying error message should be appended to the second error message.
4382 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4384 PH/02 Reduce the amount of output that "make" produces by default. Full output
4385 can still be requested.
4387 PH/03 The warning log line about a condition test deferring for a "warn" verb
4388 was being output only once per connection, rather than after each
4389 occurrence (because it was using the same function as for successful
4390 "warn" verbs). This seems wrong, so I have changed it.
4392 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4393 it should not have, which might have caused a crash in the right
4394 circumstances, but probably never did.
4396 PH/04 Installed a modified version of Tony Finch's patch to make submission
4397 mode fix the return path as well as the Sender: header line, and to
4398 add a /name= option so that you can make the user's friendly name appear
4401 TF/03 Added the control = fakedefer ACL modifier.
4403 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4404 Mark Lowes for thorough testing.
4406 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4408 TK/03 Merged latest SRS patch from Miles Wilton.
4410 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4411 with the definition in sysexits.h (which is #included earlier).
4412 Fortunately, Exim does not actually use EX_OK. The code used to try to
4413 preserve the sysexits.h value, by assuming that macro definitions were
4414 scanned for macro replacements. I have been disabused of this notion,
4415 so now the code just undefines EX_OK before #including unistd.h.
4417 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4418 in the smtp transport. When a block could not be written in a single
4419 write() function, the timeout was being re-applied to each part-write.
4420 This seems wrong - if the receiver was accepting one byte at a time it
4421 would take for ever. The timeout is now adjusted when this happens. It
4422 doesn't have to be particularly precise.
4424 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4425 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4427 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4428 without a domain if the check to <postmaster@domain> fails.
4430 SC/01 Eximstats: added -xls and the ability to specify output files
4431 (patch written by Frank Heydlauf).
4433 SC/02 Eximstats: use FileHandles for outputting results.
4435 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4437 SC/04 Eximstats: fixed display of large numbers with -nvr option
4439 SC/05 Eximstats: fixed merging of reports with empty tables.
4441 SC/06 Eximstats: added the -include_original_destination flag
4443 SC/07 Eximstats: removed tabs and trailing whitespace.
4445 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4447 TK/06 MBOX spool code: Add real "From " MBOX separator line
4448 so the .eml file is really in mbox format (even though
4449 most programs do not really care). Patch from Alex Miller.
4451 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4452 The latter is generated from $received_to and is only set if the
4453 message has one envelope recipient. SA can use these headers,
4454 obviously out-of-the-box. Patch from Alex Miller.
4456 PH/08 The ${def test on a variable was returning false if the variable's
4457 value was "0", contrary to what the specification has always said!
4458 The result should be true unless the variable is empty.
4460 PH/09 The syntax error of a character other than { following "${if
4461 def:variable_name" (after optional whitespace) was not being diagnosed.
4462 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4463 accidental colon was present, for example, could give incorrect results.
4465 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4466 result is used (not including appendfile, where other changes are about
4469 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4470 This involved changing a lot of size variables from int to off_t. It
4471 should work with maildirs and everything.
4473 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4474 spamd dying while we are connected to it.
4476 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4479 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4480 function for BDB 4.3.
4482 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4484 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4485 the sources for PCRE have been re-arranged and more files are now
4488 PH/15 The code I had for printing potentially long long variables in PH/11
4489 above was not the best (it lost precision). The length of off_t variables
4490 is now inspected at build time, and an appropriate printing format (%ld
4491 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4492 to be "long long int" or "long int". This is needed for the internal
4493 formatting function string_vformat().
4495 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4496 the configuration file to be ":syslog", then the script "guesses" where
4497 the logs files are, rather than using the compiled in default. In our
4498 case the guess is not the same as the compiled default, so the script
4499 suddenly stopped working when I started to use syslog. The patch checks
4500 to see if log_file_path is "". If so, it attempts to read it from exim
4501 with no configuration file to get the compiled in version, before it
4502 falls back to the previous guessing code."
4504 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4505 implementing BATV in an Exim configuration. See NewStuff for the gory
4508 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4509 Makefile that are specific to HP-UX.
4511 PH/18 If the "use_postmaster" option was set for a recipient callout together
4512 with the "random" option, the postmaster address was used as the MAIL
4513 FROM address for the random test, but not for the subsequent recipient
4514 test. It is now used for both.
4516 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4517 patch removes a few documentation additions to RFC 3028, because the
4518 latest draft now contains them. It adds the new en;ascii-case comparator
4519 and a new error check for 8bit text in MIME parts. Comparator and
4520 require names are now matched exactly. I enabled the subaddress
4521 extension, but it is not well tested yet (read: it works for me)."
4523 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4524 rework some of the code of TK/09 above to avoid the hardwired use of
4525 "%lld" and "long long". Replaced the call to snprintf() with a call to
4528 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4529 records point to non-existent hosts", "retry timeout exceeded", and
4530 "retry time not reached for any host after a long failure period".
4532 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4533 experimental DomainKeys support:
4535 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4536 (2) On an error such as an illegally used "control", the wrong name for
4537 the control was given.
4539 These problems did NOT occur unless DomainKeys support was compiled.
4541 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4543 PH/24 Added ${if match_ip condition.
4545 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4546 databases so that it will be absolutely obvious if a crash occurs in the
4547 DB library. This is a regular occurrence (often caused by mis-matched
4550 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4551 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4552 picked up on a user's system that detects such things. There doesn't seem
4553 to be a gcc warning option for this - only an attribute that has to be
4554 put on the function's prototype. It seems that in Fedora Core 4 they have
4555 set this on a number of new functions. No doubt there will be more in due
4558 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4559 specify a transport. However, if an address that was verified by such a
4560 router was the subject of a callout, Exim crashed because it tried to
4561 read the rcpt_include_affixes from the non-existent transport. Now it
4562 just assumes that the setting of that option is false. This bug was
4563 introduced by 4.51/PH/31.
4565 PH/28 Changed -d+all to exclude +memory, because that information is very
4566 rarely of interest, but it makes the output a lot bigger. People tend to
4567 do -d+all out of habit.
4569 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4570 code in os-type was giving problems when libc.so lives in lib64, like on
4573 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4574 aren't the modern standard, and it seems that some systems' include files
4575 don't always have them. Exim was already checking for some of the newer
4576 ones like T_AAAA, and defining it itself. I've added checks for all the
4577 record types that Exim uses.
4579 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4580 not automatically generating a new one, as it is supposed to. This
4581 prevented TLS from working. If the file did exist, but contained invalid
4582 data, a new version was generated, as expected. It was only the case of a
4583 non-existent file that was broken.
4585 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4586 with a change in libdomainkeys > 0.64.
4588 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4589 from DNS. If the selector record carries the flag, it now has
4590 precedence over the domain-wide flag.
4592 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4594 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4595 the use of a port name that isn't defined in /etc/services) occurred, the
4596 message was deferred as in a normal delivery, and thus remained on the
4597 spool, instead of being failed because of the mua_wrapper setting. This
4598 is now fixed, and I tidied up some of the mua_wrapper messages at the
4601 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4602 the messages in a hash of arrays rather than using individual hashes.
4603 This is a bit cleaner and results in dramatic memory savings, albeit
4604 at a slight CPU cost.
4606 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4607 as requested by Marc Sherman.
4609 SC/10 Eximstats: added histograms for user specified patterns as requested
4612 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4614 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4615 fopen() in the content-scanning modules that did not already have it.
4621 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4622 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4624 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4626 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4628 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4629 to test Sieve filters that use "vacation".
4631 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4632 that changes the way the GnuTLS parameters are stored in the cache file.
4633 The new format can be generated externally. For backward compatibility,
4634 if the data in the cache doesn't make sense, Exim assumes it has read an
4635 old-format file, and it generates new data and writes a new file. This
4636 means that you can't go back to an older release without removing the
4639 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4640 work if there are any delivery delays because "one_time" forces the
4641 parent to be marked "delivered", so its unseen clone is never tried
4642 again. For this reason, Exim now forbids the simultaneous setting of
4645 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4646 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4647 entry if you want to know the details. The fix, however, neglected to
4648 consider the case when local delivery batching is involved. The test for
4649 "previously delivered" was not happening when checking to see if an
4650 address could be batched with a previous (undelivered) one; under
4651 certain circumstances this could lead to multiple deliveries to the same
4654 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4655 in its include files, and this causes problems building Exim.
4657 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4658 header_syntax) but Exim was just ignoring anything given after a slash.
4659 In particular, this caused confusion with an attempt to use "verify =
4660 reverse_host_lookup/defer_ok". An error is now given when options are
4661 supplied for verify items that do not have them. (Maybe reverse_host_
4662 lookup should have a defer_ok option, but that's a different point.)
4664 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4665 defined by RFC 821) to 2048, because there were problems with some AUTH
4666 commands, and RFC 1869 says the size should be increased for extended
4667 SMTP commands that take arguments.
4669 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4672 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4673 "unknown" error; now it says that the functionality isn't in the binary.
4675 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4676 an address' error message when a string expansion fails (syntax or
4677 whatever). Otherwise the password may appear in the log. Following change
4678 PH/42 below, there is no longer a chance of it appearing in a bounce
4681 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4683 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4684 of its addresses was ever tried. (Bugzilla bug #2.)
4686 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4687 the result incorrectly in the debug output. (It correctly added a newline
4688 to what was transported.)
4690 TF/01 Added $received_time.
4692 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4693 commented out examples of how to interface to a virus scanner and to
4694 SpamAssassin. Also added commented examples of av_scanner and
4695 spamd_address settings.
4697 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4698 and controls are allowed in which ACLs. There were a couple of minor
4699 errors. Some of the entries in the conditions table (which is a table of
4700 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4701 negation of where the condition IS allowed.
4703 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4705 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4706 header file does not have a version number, so I've had to invent a new
4707 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4708 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4709 radiusclient), but it was contributed by a Radius user.
4711 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4712 files or directories, for OpenSSL.
4714 PH/19 When an Exim process that is running as root has to create an Exim log
4715 file, it does so in a subprocess that runs as exim:exim so as to get the
4716 ownership right at creation (otherwise, other Exim processes might see
4717 the file with the wrong ownership). There was no test for failure of this
4718 fork() call, which would lead to the process getting stuck as it waited
4719 for a non-existent subprocess. Forks do occasionally fail when resources
4720 run out. I reviewed all the other calls to fork(); they all seem to check
4723 PH/20 When checking for unexpected SMTP input at connect time (before writing
4724 the banner), Exim was not dealing correctly with a non-positive return
4725 from the read() function. If the client had disconnected by this time,
4726 the result was a log entry for a synchronization error with an empty
4727 string after "input=" when read() returned zero. If read() returned -1
4728 (an event I could not check), uninitialized data bytes were printed.
4729 There were reports of junk text (parts of files, etc) appearing after
4732 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4734 PH/22 Added support for macro redefinition, and (re)definition in between
4735 driver and ACL definitions.
4737 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4738 forgetting to use the resulting value; it was using the unexpanded value.
4740 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4741 hadn't been configured. The fix is from Juergen Kreileder, who
4742 understands it better than I do:
4744 "Here's what I see happening with three configured cyrus_sasl
4745 authenticators configured (plain, login, cram-md5):
4747 On startup auth_cyrus_sasl_init() gets called for each of these.
4748 This means three calls to sasl_listmech() without a specified mech_list.
4749 => SASL tests which mechs of all available mechs actually work
4750 => three warnings about OTP not working
4751 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4753 With the patch, sasl_listmech() also gets called three times. But now
4754 SASL's mech_list option is set to the server_mech specified in the the
4755 authenticator. Or in other words, the answer from sasl_listmech()
4756 gets limited to just the mech you're testing for (which is different
4758 => the return list contains just 'plain' or 'login', 'cram-md5' or
4759 nothing depending on the value of ob->server_mech.
4761 I've just tested the patch: Authentication still works fine,
4762 unavailable mechs specified in the exim configuration are still
4763 caught, and the auth.log warnings about OTP are gone."
4765 PH/25 When debugging is enabled, the contents of the command line are added
4766 to the debugging output, even when log_selector=+arguments is not
4769 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4770 answer is "GNU", and only if the return is "GNU/something" is the answer
4773 PH/27 $acl_verify_message is now set immediately after the failure of a
4774 verification in an ACL, and so is available in subsequent modifiers. In
4775 particular, the message can be preserved by coding like this:
4777 warn !verify = sender
4778 set acl_m0 = $acl_verify_message
4780 Previously, $acl_verify_message was set only while expanding "message"
4781 and "log_message" when a very denied access.
4783 PH/28 Modified OS/os.c-Linux with
4785 -#ifndef OS_LOAD_AVERAGE
4786 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4788 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4789 nomenclature these days.)
4791 PH/29 Installed patch from the Sieve maintainer that adds the options
4792 sieve_useraddress and sieve_subaddress to the redirect router.
4794 PH/30 In these circumstances:
4795 . Two addresses routed to the same list of hosts;
4796 . First host does not offer TLS;
4797 . First host accepts first address;
4798 . First host gives temporary error to second address;
4799 . Second host offers TLS and a TLS session is established;
4800 . Second host accepts second address.
4801 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4802 and peerdn, if requested) that were in fact used only for the second
4805 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4806 attention to any local part prefix or suffix that was matched by the
4807 router that accepted the address. It now behaves in the same way as it
4808 does for delivery: the affixes are removed from the local part unless
4809 rcpt_include_affixes is set on the transport.
4811 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4812 timeout during the DATA phase of an incoming message.
4814 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4815 applied a patch sanctioned by the Sieve maintainer.
4817 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4818 the uid or gid is negative. A case of a negative gid caused this to be
4819 noticed. The fix allows for either to be negative.
4821 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4822 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4825 PH/36 The change PH/12 above was broken. Fixed it.
4827 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4828 the grounds that routing the same address twice would always produce the
4829 same answer. This might have been true once, but it is certainly no
4830 longer true now. Routing a child address may depend on the previous
4831 routing that produced that child. Some complicated redirection strategies
4832 went wrong when messages had multiple recipients, and made Exim's
4833 behaviour dependent on the order in which the addresses were given.
4835 I have moved the duplicate checking until after the routing is complete.
4836 Exim scans the addresses that are assigned to local and remote
4837 transports, and removes any duplicates. This means that more work will be
4838 done, as duplicates will always all be routed, but duplicates are
4839 presumably rare, so I don't expect this is of any significance.
4841 For deliveries to pipes, files, and autoreplies, the duplicate checking
4842 still happens during the routing process, since they are not going to be
4845 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4846 It corrects a timeout issue with spamd. This is Ian's comment: "The
4847 background is that sometimes spamd either never reads data from a
4848 connection it has accepted, or it never writes response data. The exiscan
4849 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4850 blindly assumes that writes won't block so it may never time out."
4852 PH/39 Allow G after quota size as well as K and M.
4854 PH/40 The value set for $authenticated_id in an authenticator may not contain
4855 binary zeroes or newlines because the value is written to log lines and
4856 to spool files. There was no check on this. Now the value is run through
4857 the string_printing() function so that such characters are converted to
4858 printable escape sequences.
4860 PH/41 $message_linecount is a new variable that contains the total number of
4861 lines in the message. Compare $body_linecount, which is the count for the
4864 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4865 bounce and delay warning messages, except in certain special cases, which
4868 (a) An SMTP error message from a remote host;
4869 (b) A message specified in a :fail: redirection;
4870 (c) A message specified in a "fail" command in a system filter;
4871 (d) A message specified in a FAIL return from the queryprogram router;
4872 (e) A message specified by the cannot_route_message router option.
4874 In these cases only, Exim does include the error details in bounce and
4875 warning messages. There are also a few cases where bland messages such
4876 as "unrouteable address" or "local delivery error" are given.
4878 PH/43 $value is now also set for the "else" part of a ${run expansion.
4880 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4881 being worked on, but at least Exim now implements the latest version to
4884 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4885 process to complete was treated as a delivery failure, a timeout while
4886 writing the message to the pipe was logged, but erroneously treated as a
4887 successful delivery. Such timeouts include transport filter timeouts. For
4888 consistency with the overall process timeout, these timeouts are now
4889 treated as errors, giving rise to delivery failures by default. However,
4890 there is now a new Boolean option for the pipe transport called
4891 timeout_defer, which, if set TRUE, converts the failures into defers for
4892 both kinds of timeout. A transport filter timeout is now identified in
4895 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4896 systems where "make" and "gmake" are different, calling "gmake" at top
4897 level broke things. I've arranged for the value of $(MAKE) to be passed
4898 from the Makefile to this script so that it can call the same version of
4902 A note about Exim versions 4.44 and 4.50
4903 ----------------------------------------
4905 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4906 changes of various kinds. As a consequence, a big documentation update was
4907 needed. This delayed the release for rather longer than seemed good, especially
4908 in the light of a couple of (minor) security issues. Therefore, the changes
4909 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4910 release. So 4.44 and 4.50 are in effect two different branches that both start
4913 I have left the 4.50 change log unchanged; it contains all the changes since
4914 4.43. The change log for 4.44 is below; many of its items are identical to
4915 those for 4.50. This seems to be the most sensible way to preserve the
4916 historical information.
4922 1. Minor wording change to the doc/README.SIEVE file.
4924 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4925 computation of the current number of files was incorrect.
4927 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4928 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4931 4. Give more explanation in the error message when the command for a transport
4932 filter fails to execute.
4934 5. There are several places where Exim runs a non-Exim command in a
4935 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4936 was being done only for the command run by the queryprogram router. It is
4937 now done for all such subprocesses. The other cases are: ${run, transport
4938 filters, and the commands run by the lmtp and pipe transports.
4940 6. Added CONFIGURE_GROUP build-time option.
4942 7. Some older OS have a limit of 256 on the maximum number of file
4943 descriptors. Exim was using setrlimit() to set 1000 as a large value
4944 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4945 systems. I've change it so that if it can't get 1000, it tries for 256.
4947 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4948 was an oversight, and furthermore, ever since the addition of extra
4949 controls (e.g. 4.43/32), the checks on when to allow different forms of
4950 "control" were broken. There should now be diagnostics for all cases when a
4951 control that does not make sense is encountered.
4953 9. Added the /retain_sender option to "control=submission".
4955 10. $recipients is now available in the predata ACL (oversight).
4957 11. Tidy the search cache before the fork to do a delivery from a message
4958 received from the command line. Otherwise the child will trigger a lookup
4959 failure and thereby defer the delivery if it tries to use (for example) a
4960 cached ldap connection that the parent has called unbind on.
4962 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4963 of $address_data from the recipient verification was clobbered by the
4964 sender verification.
4966 13. The value of address_data from a sender verification is now available in
4967 $sender_address_data in subsequent conditions in the ACL statement.
4969 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4971 15. Added a new option "connect=<time>" to callout options, to set a different
4974 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4975 was its contents. (It was OK if the option was not defined at all.)
4977 17. A "Completed" log line is now written for messages that are removed from
4978 the spool by the -Mrm option.
4980 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4981 information about exactly what failed.
4983 19. Added -dd to debug only the daemon process.
4985 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4986 handles timeouts, both on the server side and network timeouts. Renamed the
4987 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4989 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4990 It is now set to "smtps".
4992 22. $host_address is now set to the target address during the checking of
4993 ignore_target_hosts.
4995 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4996 being passed; this would have caused $sender_host_name to have been used if
4997 matching the list had actually called for a host name (not very likely,
4998 since this list is usually IP addresses). A host name is now passed as
5001 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5002 code that specifies a non-restarting handler (typically sigaction() in
5003 modern systems) in an attempt to fix a rare and obscure crash bug.
5005 25. Narrowed the window for a race in the daemon that could cause it to ignore
5006 SIGCHLD signals. This is not a major problem, because they are used only to
5007 wake it up if nothing else does.
5009 26. A malformed maildirsize file could cause Exim to calculate negative values
5010 for the mailbox size or file count. Odd effects could occur as a result.
5011 The maildirsize information is now recalculated if the size or filecount
5014 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5015 support for a long time. Removed HAVE_SYS_VFS_H.
5017 28. Installed the latest version of exipick from John Jetmore.
5019 29. In an address list, if the pattern was not a regular expression, an empty
5020 subject address (from a bounce message) matched only if the pattern was an
5021 empty string. Non-empty patterns were not even tested. This was the wrong
5022 because it is perfectly reasonable to use an empty address as part of a
5023 database query. An empty address is now tested by patterns that are
5024 lookups. However, all the other forms of pattern expect the subject to
5025 contain a local part and a domain, and therefore, for them, an empty
5026 address still always fails if the pattern is not itself empty.
5028 30. Exim went into a mad DNS loop when attempting to do a callout where the
5029 host was specified on an smtp transport, and looking it up yielded more
5030 than one IP address.
5032 31. Re-factored the code for checking spool and log partition space into a
5033 function that finds that data and another that does the check. The former
5034 is then used to implement four new variables: $spool_space, $log_space,
5035 $spool_inodes, and $log_inodes.
5037 32. The RFC2047 encoding function was originally intended for short strings
5038 such as real names; it was not keeping to the 75-character limit for
5039 encoded words that the RFC imposes. It now respects the limit, and
5040 generates multiple encoded words if necessary. To be on the safe side, I
5041 have increased the buffer size for the ${rfc2047: expansion operator from
5044 33. It is now permitted to omit both strings after an "if" condition; if the
5045 condition is true, the result is "true". As before, when the second string
5046 is omitted, a false condition yields an empty string. This makes it less
5047 cumbersome to write custom ACL and router conditions.
5049 34. Failure to deliver a bounce message always caused it to be frozen, even if
5050 there was an errors_to setting on the router. The errors_to setting is now
5053 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5054 canonical form (fully expanded) before being placed in
5055 $sender_host_address.
5057 36. The table in the code that translates DNS record types into text (T_A to
5058 "A" for instance) was missing entries for NS and CNAME. It is just possible
5059 that this could have caused confusion if both these types were looked up
5060 for the same domain, because the text type is used as part of Exim's
5061 per-process caching. But the chance of anyone hitting this buglet seems
5064 37. The dnsdb lookup has been extended in a number of ways.
5066 (1) There is a new type, "zns", which walks up the domain tree until it
5067 finds some nameserver records. It should be used with care.
5069 (2) There is a new type, "mxh", which is like "mx" except that it returns
5070 just the host names, not the priorities.
5072 (3) It is now possible to give a list of domains (or IP addresses) to be
5073 looked up. The behaviour when one of the lookups defers can be
5074 controlled by a keyword.
5076 (4) It is now possible to specify the separator character for use when
5077 multiple records are returned.
5079 38. The dnslists ACL condition has been extended: it is now possible to supply
5080 a list of IP addresses and/or domains to be looked up in a particular DNS
5083 39. Added log_selector=+queue_time_overall.
5085 40. When running the queue in the test harness, wait just a tad after forking a
5086 delivery process, to get repeatability of debugging output.
5088 41. Include certificate and key file names in error message when GnuTLS fails
5089 to set them up, because the GnuTLS error message doesn't include the name
5090 of the failing file when there is a problem reading it.
5092 42. Allow both -bf and -bF in the same test run.
5094 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5096 44. The "Exiscan patch" is now merged into the mainline Exim source.
5098 45. Sometimes the final signoff response after QUIT could fail to get
5099 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5100 < 0 before doing a fflush(). This bug looks as though it goes back to the
5101 introduction of TLS in release 3.20, but "sometimes" must have been rare
5102 because the tests only now provoked it.
5104 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5105 (this can affect the format of dates).
5107 47. exim_tidydb, when checking for the continued existence of a message for
5108 which it has found a message-specific retry record, was not finding
5109 messages that were in split spool directories. Consequently, it was
5110 deleting retry records that should have stayed in existence.
5112 48. Steve fixed some bugs in eximstats.
5114 49. The SPA authentication driver was not abandoning authentication and moving
5115 on to the next authenticator when an expansion was forced to fail,
5116 contradicting the general specification for all authenticators. Instead it
5117 was generating a temporary error. It now behaves as specified.
5119 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5120 (the order specifies the preference for clients). The order is now AES256,
5121 AES128, 3DES, ARCFOUR128.
5123 51. Small patch to Sieve code - explicitly set From: when generating an
5126 52. Exim crashed if a remote delivery caused a very long error message to be
5127 recorded - for instance if somebody sent an entire SpamAssassin report back
5128 as a large number of 550 error lines. This bug was coincidentally fixed by
5129 increasing the size of one of Exim's internal buffers (big_buffer) that
5130 happened as part of the Exiscan merge. However, to be on the safe side, I
5131 have made the code more robust (and fixed the comments that describe what
5134 53. Now that there can be additional text after "Completed" in log lines (if
5135 the queue_time_overall log selector is set), a one-byte patch to exigrep
5136 was needed to allow it to recognize "Completed" as not the last thing in
5139 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5140 patch that reportedly fixes this has been added. I am not expert enough to
5141 create a test for it. This is what the patch creator wrote:
5143 "I found a little strange behaviour of ldap code when working with
5144 Windows 2003 AD Domain, where users was placed in more than one
5145 Organization Units. When I tried to give exim partial DN, the exit code
5146 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5147 But simultaneously result of request was absolutely normal ldap result,
5148 so I produce this patch..."
5150 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5151 REFERENCE, so I have modified the code to exclude the patch when that macro
5154 55. Some experimental protocols are using DNS PTR records for new purposes. The
5155 keys for these records are domain names, not reversed IP addresses. The
5156 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5157 leaves it alone. Component reversal etc. now happens only for IP addresses.
5160 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5162 57. Double the size of the debug message buffer (to 2048) so that more of very
5163 long debug lines gets shown.
5165 58. The exicyclog utility now does better if the number of log files to keep
5166 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5168 59. Two changes related to the smtp_active_hostname option:
5170 (1) $smtp_active_hostname is now available as a variable.
5171 (2) The default for smtp_banner uses $smtp_active_hostname instead
5172 of $primary_hostname.
5174 60. The host_aton() function is supposed to be passed a string that is known
5175 to be a valid IP address. However, in the case of IPv6 addresses, it was
5176 not checking this. This is a hostage to fortune. Exim now panics and dies
5177 if the condition is not met. A case was found where this could be provoked
5178 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5179 components; fortuitously, this particular loophole had already been fixed
5180 by change 4.50/55 above.
5182 If there are any other similar loopholes, the new check in host_aton()
5183 itself should stop them being exploited. The report I received stated that
5184 data on the command line could provoke the exploit when Exim was running as
5185 exim, but did not say which command line option was involved. All I could
5186 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5187 running as the user.
5190 61. There was a buffer overflow vulnerability in the SPA authentication code
5191 (which came originally from the Samba project). I have added a test to the
5192 spa_base64_to_bits() function which I hope fixes it.
5195 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5196 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5198 63. The daemon start-up calls getloadavg() while still root for those OS that
5199 need the first call to be done as root, but it missed one case: when
5200 deliver_queue_load_max is set with deliver_drop_privilege. This is
5201 necessary for the benefit of the queue runner, because there is no re-exec
5202 when deliver_drop_privilege is set.
5204 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5205 This has been fixed.
5207 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5208 was in use, was not putting the data itself into the right store pool;
5209 consequently, it could be overwritten for a subsequent message in the same
5210 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5213 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5215 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5216 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5217 the function were treating the return as a boolean value, which happened to
5218 work because 0=false and not-0=true, but is not correct code.
5220 68. The host_aton() function was not handling scoped IPv6 addresses (those
5221 with, for example, "%eth0" on the end) correctly.
5223 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5224 negated items (that is, ~something) in unsigned ints. Some compilers
5225 apparently mutter when there is no cast.
5227 70. If an address verification called from an ACL failed, and did not produce a
5228 user-specific message (i.e. there was only a "system" message), nothing was
5229 put in $acl_verify_message. In this situation, it now puts the system
5232 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5233 session; change 4.31/43 added the unwanted input to the log line - except
5234 that it did not do this in the start of session case. It now does.
5236 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5237 This is wrong and can cause the other end to generate a synchronization
5238 error if it is another Exim or anything else that does the synchronization
5239 check. A QUIT command is no longer sent after a timeout.
5241 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5242 during host lookups.
5244 74. The defer_ok option of callout verification was not working if it was used
5245 when verifying addresses in header lines, that is, for this case:
5247 verify = header_sender/callout=defer_ok
5249 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5250 those file descriptors could be used for SMTP connections. If anything
5251 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5252 could be sent to the SMTP client, causing chaos. The daemon now opens
5253 stdin, stdout, and stderr to /dev/null when it puts itself into the
5256 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5257 log by default. The user can override this with suitable Perl magic.
5259 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5260 the log message when discard triggers, was not working for the DATA ACL or
5261 for the non-SMTP ACL.
5263 78. Error message wording change in sieve.c.
5265 79. If smtp_accept_max_per_host was set, the number of connections could be
5266 restricted to fewer than expected, because the daemon was trying to set up
5267 a new connection before checking whether the processes handling previous
5268 connections had finished. The check for completed processes is now done
5269 earlier. On busy systems, this bug wouldn't be noticed because something
5270 else would have woken the daemon, and it would have reaped the completed
5273 80. If a message was submitted locally by a user whose login name contained one
5274 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5275 It caused a spool format error. I have fixed the spool reading code. A
5276 related problem was that the "from" clause in the Received: line became
5277 illegal because of the space(s). It is now covered by ${quote_local_part.
5279 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5282 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5284 83. Patch from Sieve maintainer to fix unterminated string problem in
5285 "vacation" handling.
5287 84. Some minor changes to the Linux configuration files to help with other
5288 OS variants using glibc.
5290 85. One more patch for Sieve to update vacation handling to latest spec.
5293 ----------------------------------------------------
5294 See the note above about the 4.44 and 4.50 releases.
5295 ----------------------------------------------------
5301 1. Change 4.43/35 introduced a bug that caused file counts to be
5302 incorrectly computed when quota_filecount was set in an appendfile
5305 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5306 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5309 3. Give more explanation in the error message when the command for a transport
5310 filter fails to execute.
5312 4. There are several places where Exim runs a non-Exim command in a
5313 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5314 was being done only for the command run by the queryprogram router. It is
5315 now done for all such subprocesses. The other cases are: ${run, transport
5316 filters, and the commands run by the lmtp and pipe transports.
5318 5. Some older OS have a limit of 256 on the maximum number of file
5319 descriptors. Exim was using setrlimit() to set 1000 as a large value
5320 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5321 systems. I've change it so that if it can't get 1000, it tries for 256.
5323 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5324 was an oversight, and furthermore, ever since the addition of extra
5325 controls (e.g. 4.43/32), the checks on when to allow different forms of
5326 "control" were broken. There should now be diagnostics for all cases when a
5327 control that does not make sense is encountered.
5329 7. $recipients is now available in the predata ACL (oversight).
5331 8. Tidy the search cache before the fork to do a delivery from a message
5332 received from the command line. Otherwise the child will trigger a lookup
5333 failure and thereby defer the delivery if it tries to use (for example) a
5334 cached ldap connection that the parent has called unbind on.
5336 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5337 of $address_data from the recipient verification was clobbered by the
5338 sender verification.
5340 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5341 was its contents. (It was OK if the option was not defined at all.)
5343 11. A "Completed" log line is now written for messages that are removed from
5344 the spool by the -Mrm option.
5346 12. $host_address is now set to the target address during the checking of
5347 ignore_target_hosts.
5349 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5350 being passed; this would have caused $sender_host_name to have been used if
5351 matching the list had actually called for a host name (not very likely,
5352 since this list is usually IP addresses). A host name is now passed as
5355 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5356 code that specifies a non-restarting handler (typically sigaction() in
5357 modern systems) in an attempt to fix a rare and obscure crash bug.
5359 15. Narrowed the window for a race in the daemon that could cause it to ignore
5360 SIGCHLD signals. This is not a major problem, because they are used only to
5361 wake it up if nothing else does.
5363 16. A malformed maildirsize file could cause Exim to calculate negative values
5364 for the mailbox size or file count. Odd effects could occur as a result.
5365 The maildirsize information is now recalculated if the size or filecount
5368 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5369 support for a long time. Removed HAVE_SYS_VFS_H.
5371 18. Updated exipick to current release from John Jetmore.
5373 19. Allow an empty sender to be matched against a lookup in an address list.
5374 Previously the only cases considered were a regular expression, or an
5377 20. Exim went into a mad DNS lookup loop when doing a callout where the
5378 host was specified on the transport, if the DNS lookup yielded more than
5381 21. The RFC2047 encoding function was originally intended for short strings
5382 such as real names; it was not keeping to the 75-character limit for
5383 encoded words that the RFC imposes. It now respects the limit, and
5384 generates multiple encoded words if necessary. To be on the safe side, I
5385 have increased the buffer size for the ${rfc2047: expansion operator from
5388 22. Failure to deliver a bounce message always caused it to be frozen, even if
5389 there was an errors_to setting on the router. The errors_to setting is now
5392 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5393 canonical form (fully expanded) before being placed in
5394 $sender_host_address.
5396 24. Updated eximstats to version 1.33
5398 25. Include certificate and key file names in error message when GnuTLS fails
5399 to set them up, because the GnuTLS error message doesn't include the name
5400 of the failing file when there is a problem reading it.
5402 26. Expand error message when OpenSSL has problems setting up cert/key files.
5405 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5406 (this can affect the format of dates).
5408 28. exim_tidydb, when checking for the continued existence of a message for
5409 which it has found a message-specific retry record, was not finding
5410 messages that were in split spool directories. Consequently, it was
5411 deleting retry records that should have stayed in existence.
5413 29. eximstats updated to version 1.35
5414 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5415 1.35 - bugfix such that pie charts by volume are generated correctly
5417 30. The SPA authentication driver was not abandoning authentication and moving
5418 on to the next authenticator when an expansion was forced to fail,
5419 contradicting the general specification for all authenticators. Instead it
5420 was generating a temporary error. It now behaves as specified.
5422 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5423 (the order specifies the preference for clients). The order is now AES256,
5424 AES128, 3DES, ARCFOUR128.
5426 31. Small patch to Sieve code - explicitly set From: when generating an
5429 32. Exim crashed if a remote delivery caused a very long error message to be
5430 recorded - for instance if somebody sent an entire SpamAssassin report back
5431 as a large number of 550 error lines. This bug was coincidentally fixed by
5432 increasing the size of one of Exim's internal buffers (big_buffer) that
5433 happened as part of the Exiscan merge. However, to be on the safe side, I
5434 have made the code more robust (and fixed the comments that describe what
5437 33. Some experimental protocols are using DNS PTR records for new purposes. The
5438 keys for these records are domain names, not reversed IP addresses. The
5439 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5440 leaves it alone. Component reversal etc. now happens only for IP addresses.
5443 34. The host_aton() function is supposed to be passed a string that is known
5444 to be a valid IP address. However, in the case of IPv6 addresses, it was
5445 not checking this. This is a hostage to fortune. Exim now panics and dies
5446 if the condition is not met. A case was found where this could be provoked
5447 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5448 components; fortuitously, this particular loophole had already been fixed
5449 by change 4.50/55 or 4.44/33 above.
5451 If there are any other similar loopholes, the new check in host_aton()
5452 itself should stop them being exploited. The report I received stated that
5453 data on the command line could provoke the exploit when Exim was running as
5454 exim, but did not say which command line option was involved. All I could
5455 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5456 running as the user.
5459 35. There was a buffer overflow vulnerability in the SPA authentication code
5460 (which came originally from the Samba project). I have added a test to the
5461 spa_base64_to_bits() function which I hope fixes it.
5464 36. The daemon start-up calls getloadavg() while still root for those OS that
5465 need the first call to be done as root, but it missed one case: when
5466 deliver_queue_load_max is set with deliver_drop_privilege. This is
5467 necessary for the benefit of the queue runner, because there is no re-exec
5468 when deliver_drop_privilege is set.
5470 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5471 was in use, was not putting the data itself into the right store pool;
5472 consequently, it could be overwritten for a subsequent message in the same
5473 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5476 38. Sometimes the final signoff response after QUIT could fail to get
5477 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5478 < 0 before doing a fflush(). This bug looks as though it goes back to the
5479 introduction of TLS in release 3.20, but "sometimes" must have been rare
5480 because the tests only now provoked it.
5486 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5487 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5488 Now it returns an integer. A number of calls were still expecting a T/F
5489 return. Fortuitously, in all cases, the tests worked in OK situations,
5490 which is the norm. However, things would have gone wrong on any write
5491 failures on the smtp file descriptor. This function is used when sending
5492 messages over SMTP and also when doing verify callouts.
5494 2. When Exim is called to do synchronous delivery of a locally submitted
5495 message (the -odf or -odi options), it no longer closes stderr before doing
5498 3. Implemented the mua_wrapper option.
5500 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5502 5. Implemented the functions header_remove(), header_testname(),
5503 header_add_at_position(), and receive_remove_recipient(), and exported them
5506 6. If an ACL "warn" statement specified the addition of headers, Exim already
5507 inserted X-ACL-Warn: at the start if there was no header name. However, it
5508 was not making this test for the second and subsequent header lines if
5509 there were newlines in the string. This meant that an invalid header could
5510 be inserted if Exim was badly configured.
5512 7. Allow an ACL "warn" statement to add header lines at the start or after all
5513 the Received: headers, as well as at the end.
5515 8. Added the rcpt_4xx retry error code.
5517 9. Added postmaster_mailfrom=xxx to callout verification option.
5519 10. Added mailfrom=xxxx to the callout verification option, for verify=
5522 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5523 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5525 12. Inserted some casts to stop certain compilers warning when using pointer
5526 differences as field lengths or precisions in printf-type calls (mostly
5527 affecting debugging statements).
5529 13. Added optional readline() support for -be (dynamically loaded).
5531 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5532 same clock tick as a message's arrival, so that its received time was the
5533 same as the "first fail" time on the retry record, and that message
5534 remained on the queue past the ultimate address timeout, every queue runner
5535 would try a delivery (because it was past the ultimate address timeout) but
5536 after another failure, the ultimate address timeout, which should have then
5537 bounced the address, did not kick in. This was a "< instead of <=" error;
5538 in most cases the first failure would have been in the next clock tick
5539 after the received time, and all would be well.
5541 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5542 being recognized when the domain list was tested by the match_domain
5543 condition in an expansion string.
5545 16. Added the ${str2b64: operator.
5547 17. Exim was always calling setrlimit() to set a large limit for the number of
5548 processes, without checking whether the existing limit was already
5549 adequate. (It did check for the limit on file descriptors.) Furthermore,
5550 errors from getrlimit() and setrlimit() were being ignored. Now they are
5551 logged to the main and panic logs, but Exim does carry on, to try to do its
5552 job under whatever limits there are.
5554 18. Imported PCRE 5.0.
5556 19. Trivial typo in log message " temporarily refused connection" (the leading
5559 20. If the log selector return_path_on_delivery was set and an address was
5560 redirected to /dev/null, the delivery process crashed because it assumed
5561 that a return path would always be set for a "successful" delivery. In this
5562 case, the whole delivery is bypassed as an optimization, and therefore no
5565 21. Internal re-arrangement: the function for sending a challenge and reading
5566 a response while authentication was assuming a zero-terminated challenge
5567 string. It's now changed to take a pointer and a length, to allow for
5568 binary data in such strings.
5570 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5572 23. Exim was not respecting finduser_retries when seeking the login of the
5573 uid under which it was called; it was always trying 10 times. (The default
5574 setting of finduser_retries is zero.) Also, it was sleeping after the final
5575 failure, which is pointless.
5577 24. Implemented tls_on_connect_ports.
5579 25. Implemented acl_smtp_predata.
5581 26. If the domain in control=submission is set empty, Exim assumes that the
5582 authenticated id is a complete email address when it generates From: or
5583 Sender: header lines.
5585 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5586 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5587 chown and chgrp in /bin and hostname in /usr/bin.
5589 28. Exim was keeping the "process log" file open after each use, just as it
5590 does for the main log. This opens the possibility of it remaining open for
5591 long periods when the USR1 signal hits a daemon. Occasional processlog
5592 errors were reported, that could have been caused by this. Anyway, it seems
5593 much more sensible not to leave this file open at all, so that is what now
5596 29. The long-running daemon process does not normally write to the log once it
5597 has entered its main loop, and it closes the log before doing so. This is
5598 so that log files can straightforwardly be renamed and moved. However,
5599 there are a couple of unusual error situations where the daemon does write
5600 log entries, and I had neglected to close the log afterwards.
5602 30. The text of an SMTP error response that was received during a remote
5603 delivery was being truncated at 512 bytes. This is too short for some of
5604 the long messages that one sometimes sees. I've increased the limit to
5607 31. It is now possible to make retry rules that apply only when a message has a
5608 specific sender, in particular, an empty sender.
5610 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5611 it possible to be selective about when SMTP synchronization is enforced.
5613 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5615 32. Implemented hosts_connection_nolog.
5617 33. Added an ACL for QUIT.
5619 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5622 35. Added mailbox_size and mailbox_filecount to appendfile.
5624 36. Added control = no_multiline_responses to ACLs.
5626 37. There was a bug in the logic of the code that waits for the clock to tick
5627 in the case where the clock went backwards by a substantial amount such
5628 that the microsecond fraction of "now" was more than the microsecond
5629 fraction of "then" (but the whole seconds number was less).
5631 38. Added support for the libradius Radius client library this is found on
5632 FreeBSD (previously only the radiusclient library was supported).
5638 1. When certain lookups returned multiple values in the form name=value, the
5639 quoting of the values was not always being done properly. Specifically:
5640 (a) If the value started with a double quote, but contained no whitespace,
5642 (b) If the value contained whitespace other than a space character (i.e.
5643 tabs or newlines or carriage returns) it was not quoted.
5644 This fix has been applied to the mysql and pgsql lookups by writing a
5645 separate quoting function and calling it from the lookup code. The fix
5646 should probably also be applied to nisplus, ibase and oracle lookups, but
5647 since I cannot test any of those, I have not disturbed their existing code.
5649 2. A hit in the callout cache for a specific address caused a log line with no
5650 reason for rejecting RCPT. Now it says "Previous (cached) callout
5651 verification failure".
5653 3. There was an off-by-one bug in the queryprogram router. An over-long
5654 return line was truncated at 256 instead of 255 characters, thereby
5655 overflowing its buffer with the terminating zero. As well as fixing this, I
5656 have increased the buffer size to 1024 (and made a note to document this).
5658 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5659 when Exim is waiting for an SMTP response from a remote server, Exim
5660 restarts its select() call on the socket, thereby resetting its timeout.
5661 This is not a problem when such interrupts are rare. Somebody set up a cron
5662 job to run exiwhat every 2 minutes, which is less than the normal select()
5663 timeout (5 or 10 minutes). This meant that the select() timeout never
5664 kicked in because it was always reset. I have fixed this by comparing the
5665 time when an interrupt arrives with the time at the start of the first call
5666 to select(). If more time than the timeout has elapsed, the interrupt is
5667 treated as a timeout.
5669 5. Some internal re-factoring in preparation for the addition of Sieve
5670 extensions (by MH). In particular, the "personal" test is moved to a
5671 separate function, and given an option for scanning Cc: and Bcc: (which is
5672 not set for Exim filters).
5674 6. When Exim created an email address using the login of the caller as the
5675 local part (e.g. when creating a From: or Sender: header line), it was not
5676 quoting the local part when it contained special characters such as @.
5678 7. Installed new OpenBSD configuration files.
5680 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5681 try to make them clearer.
5683 9. Callout options, other than the timeout value, were being ignored when
5684 verifying sender addresses in header lines. For example, when using
5686 verify = header_sender/callout=no_cache
5688 the cache was (incorrectly) being used.
5690 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5691 only the Cygwin environment.
5693 11. When return_path_on_delivery was set as a log selector, if different remote
5694 addresses in the same message used different return paths and parallel
5695 remote delivery occurred, the wrong values would sometimes be logged.
5696 (Whenever a remote delivery process finished, the return path value from
5697 the most recently started remote delivery process was logged.)
5699 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5700 lines when AUTH and/or TLS are in use. This is the "received protocol"
5701 field. Exim used to use "asmtp" for authenticated SMTP, without any
5702 indication (in the protocol name) for TLS use. Now it follows the RFC and
5703 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5704 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5705 names appear in log lines as well as in Received: header lines.
5707 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5708 extensions, and comparison tests, and to fix some bugs.
5710 14. Changes to the "personal" filter test:
5712 (1) The test was buggy in that it was just doing the equivalent of
5713 "contains" tests on header lines. For example, if a user's address was
5714 anne@some.where, the "personal" test would incorrectly be true for
5716 To: susanne@some.where
5718 This test is now done by extracting each address from the header in turn,
5719 and checking the entire address. Other tests that are part of "personal"
5720 are now done using regular expressions (for example, to check local parts
5721 of addresses in From: header lines).
5723 (2) The list of non-personal local parts in From: addresses has been
5724 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5725 taken from the Sieve specification recommendations.
5727 (3) If the message contains any header line starting with "List-" it is
5728 treated as non-personal.
5730 (4) The test for "circular" in the Subject: header line has been removed
5731 because it now seems ill-conceived.
5733 15. Minor typos in src/EDITME comments corrected.
5735 16. Installed latest exipick from John Jetmore.
5737 17. If headers_add on a router specified a text string that was too long for
5738 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5739 of string_sprintf() is now avoided.
5741 18. $message_body_size was not set (it was always zero) when running the DATA
5742 ACL and the local_scan() function.
5744 19. For the "mail" command in an Exim filter, no default was being set for
5745 the once_repeat time, causing a random time value to be used if "once" was
5746 specified. (If the value happened to be <= 0, no repeat happened.) The
5747 default is now 0s, meaning "never repeat". The "vacation" command was OK
5748 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5749 (I found it when inspecting the code).
5751 20. There is now an overall timeout for performing a callout verification. It
5752 defaults to 4 times the callout timeout, which applies to individual SMTP
5753 commands during the callout. The overall timeout applies when there is more
5754 than one host that can be tried. The timeout is checked before trying the
5755 next host. This prevents very long delays if there are a large number of
5756 hosts and all are timing out (e.g. when the network connections are timing
5757 out). The value of the overall timeout can be changed by specifying an
5758 additional sub-option for "callout", called "maxwait". For example:
5760 verify = sender/callout=5s,maxwait=20s
5762 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5763 to the end before writing, but this should make it even safer).
5765 22. Exim was forgetting that it had advertised PIPELINING for the second and
5766 subsequent messages on an SMTP connection. It was also not resetting its
5767 memory on STARTTLS and an internal HELO.
5769 23. When Exim logs an SMTP synchronization error within a session, it now
5770 records whether PIPELINING has been advertised or not.
5772 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5773 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5774 rather than long int.
5776 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5778 26. Added the never_mail option to autoreply.
5784 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5785 crash if the getsockname() call failed; this can happen if a connection is
5786 closed very soon after it is established. The problem was simply in the
5787 order in which certain operations were done, causing Exim to try to write
5788 to the SMTP stream before it had set up the file descriptor. The bug has
5789 been fixed by making things happen in the correct order.
5795 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5796 before the connection was closed, thus losing the rejection response.
5798 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5799 some early Solaris releases, but causes trouble in current releases where
5800 socklen_t is defined.
5802 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5805 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5808 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5809 that releasing the top of it at the end releases what was used for sub-
5810 expansions (unless the block got too big). However, discard this block if
5811 the first thing is a variable or header, so that we can use its block when
5812 it is dynamic (useful for very large $message_headers, for example).
5814 6. Lookups now cache *every* query, not just the most recent. A new, separate
5815 store pool is used for this. It can be recovered when all lookup caches are
5816 flushed. Lookups now release memory at the end of their result strings.
5817 This has involved some general refactoring of the lookup sources.
5819 7. Some code has been added to the store_xxx() functions to reduce the amount
5820 of flapping under certain conditions.
5822 8. log_incoming_interface used to affect only the <= reception log lines. Now
5823 it causes the local interface and port to be added to several more SMTP log
5824 lines, for example "SMTP connection from", and rejection lines.
5826 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5828 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5830 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5831 could be overwritten at the end of the current message (or the start of a
5832 new message if it was set in a HELO ACL). The value is now preserved for
5833 the duration of the SMTP connection.
5835 12. If a transport had a headers_rewrite setting, and a matching header line
5836 contained an unqualified address, that address was qualified, even if it
5837 did not match any rewriting rules. The underlying bug was that the values
5838 of the flags that permit the existence of unqualified sender and recipient
5839 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5840 non-local messages, and by -bnq for local messages) were not being
5841 preserved with the message after it was received.
5843 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5844 "next input=" as part of the text comprising the host identity instead of
5845 the correct text. The code was using the same buffer for two different
5846 strings. However, depending on which order the printing function evaluated
5847 its arguments, the bug did not always show up. Under Linux, for example, my
5848 test suite worked just fine.
5850 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5851 doesn't work with some older versions of Perl. It has been changed to "my",
5852 which in any case is probably the better facility to use.
5854 15. A really picky compiler found some instances of statements for creating
5855 error messages that either had too many or two few arguments for the format
5858 16. The size of the buffer for calls to the DNS resolver has been increased
5859 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5860 for addresses that have a lot of PTR records. This alleviates a problem; it
5861 does not fully solve it.
5863 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5864 buffer now truncates the list and logs the incident, which is the same
5865 action as happens when Exim is looking up a host name and its aliases.
5866 Previously in this situation something unpredictable would happen;
5867 sometimes it was "internal error: store_reset failed".
5869 18. If a server dropped the connection unexpectedly when an Exim client was
5870 using GnuTLS and trying to read a response, the client delivery process
5871 crashed while trying to generate an error log message.
5873 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5874 string, for example:
5876 warn message = H1: something\nH2: something
5878 the text was added as a single header line from Exim's point of view
5879 though it ended up OK in the delivered message. However, searching for the
5880 second and subsequent header lines using $h_h2: did not work. This has been
5881 fixed. Similarly, if a system filter added multiple headers in this way,
5882 the routers could not see them.
5884 20. Expanded the error message when iplsearch is called with an invalid key to
5885 suggest using net-iplsearch in a host list.
5887 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5888 ACLs are no longer actually imposed (and a message to that effect is
5891 22. If a "gecos" field in a passwd entry contained escaped characters, in
5892 particular, if it contained a \" sequence, Exim got it wrong when building
5893 a From: or a Sender: header from that name. A second bug also caused
5894 incorrect handling when an unquoted " was present following a character
5895 that needed quoting.
5897 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5898 was not being matched caselessly.
5900 24. Arranged for all hyphens in the exim.8 source to be escaped with
5903 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5904 properly. Recipient callout cache records were still being keyed to include
5905 the sender, even when use_sender was set false. This led to far more
5906 callouts that were necessary. The sender is no longer included in the key
5907 when use_sender is false.
5909 26. Added "control = submission" modifier to ACLs.
5911 27. Added the ${base62d: operator to decode base 62 numbers.
5913 28. dnsdb lookups can now access SRV records.
5915 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5916 the configuration file.
5918 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5919 (-v) mode. This makes the output for a verbose queue run more intelligible.
5921 31. Added a use_postmaster feature to recipient callouts.
5923 32. Added the $body_zerocount variable, containing the number of binary zero
5924 bytes in the message body.
5926 33. The time of last modification of the "new" subdirectory is now used as the
5927 "mailbox time last read" when there is a quota error for a maildir
5930 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5932 35. Added +ignore_unknown as a special item in host lists.
5934 36. Code for decoding IPv6 addresses in host lists is now included, even if
5935 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5936 address was recognized as an IP address, but was then not correctly decoded
5937 into binary, causing unexpected and incorrect effects when compared with
5944 1. Very minor rewording of debugging text in manualroute to say "list of
5945 hosts" instead of "hostlist".
5947 2. If verify=header_syntax was set, and a header line with an unqualified
5948 address (no domain) and a large number of spaces between the end of the
5949 name and the colon was received, the reception process suffered a buffer
5950 overflow, and (when I tested it) crashed. This was caused by some obsolete
5951 code that should have been removed. The fix is to remove it!
5953 3. When running in the test harness, delay a bit after writing a bounce
5954 message to get a bit more predictability in the log output.
5956 4. Added a call to search_tidyup() just before forking a reception process. In
5957 theory, someone could use a lookup in the expansion of smtp_accept_max_
5958 per_host which, without the tidyup, could leave open a database connection.
5960 5. Added the variables $recipient_data and $sender_data which get set from a
5961 lookup success in an ACL "recipients" or "senders" condition, or a router
5962 "senders" option, similar to $domain_data and $local_part_data.
5964 6. Moved the writing of debug_print from before to after the "senders" test
5967 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5968 problems for message scanning, either using a data ACL, or using
5969 local_scan() because the Received: header was not generated till after they
5970 were called (in order to set the time as the time of reception completion).
5971 I have revised the way this works. The header is now generated after the
5972 body is received, but before the ACL or local_scan() are called. After they
5973 are run, the timestamp in the header is updated.
5979 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5980 before starting a queue runner without re-exec. This happened only when
5981 deliver_drop_privilege was set or when the Exim user was set to root. The
5982 effect of the bug was that timeouts during subsequent deliveries caused
5983 crashes instead of being properly handled. The handler is now left at its
5984 default (and expected) setting.
5986 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5987 message, again when deliver_drop_privilege is set or Exim is run as root.
5988 The bug described in (1) was not present in this case, but the tidying up
5989 of the other signals was missing. I have made the two cases consistent.
5991 3. The ignore_target_hosts setting on a manualroute router was being ignored
5992 for hosts that were looked up using the /MX notation.
5994 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5997 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5998 operated on the sender address. After changing the $sender_address to <>
5999 for the sender address verify, Exim was re-instated it as the original
6000 (before rewriting) address, but remembering that it had rewritten it, so it
6001 wasn't rewriting it again. This bug also had the effect of breaking the
6002 sender address verification caching when the sender address was rewritten.
6004 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6005 This has been changed so that if the ip literal address matches
6006 ignore_target_hosts, the router declines.
6008 7. Added expansion conditions match_domain, match_address, and match_local_
6009 part (NOT match_host).
6011 8. The placeholder for the Received: header didn't have a length field set.
6013 9. Added code to Exim itself and to exim_lock to test for a specific race
6014 condition that could lead to file corruption when using MBX delivery. The
6015 issue is with the lockfile that is created in /tmp. If this file is removed
6016 after a process has opened it but before that process has acquired a lock,
6017 there is the potential for a second process to recreate the file and also
6018 acquire a lock. This could lead to two Exim processes writing to the file
6019 at the same time. The added code performs the same test as UW imapd; it
6020 checks after acquiring the lock that its file descriptor still refers to
6021 the same named file.
6023 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6024 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6027 11. Added the smtp_active_hostname option. If used, this will typically be made
6028 to depend on the incoming interface address. Because $interface_address is
6029 not set up until the daemon has forked a reception process, error responses
6030 that can happen earlier (such as "too many connections") no longer contain
6033 12. If an expansion in a condition on a "warn" statement fails because a lookup
6034 defers, the "warn" statement is abandoned, and the next ACL statement is
6035 processed. Previously this caused the whole ACL to be aborted.
6037 13. Added the iplsearch lookup type.
6039 14. Added ident_timeout as a log selector.
6041 15. Added tls_certificate_verified as a log selector.
6043 16. Added a global option tls_require_ciphers (compare the smtp transport
6044 option of the same name). This controls incoming TLS connections.
6046 17. I finally figured out how to make tls_require_ciphers do a similar thing
6047 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6048 before starting the TLS session.
6050 18. Tabs are now shown as \t in -bP output.
6052 19. If the log selector return_path_on_delivery was set, Exim crashed when
6053 bouncing a message because it had too many Received: header lines.
6055 20. If two routers both had headers_remove settings, and the first one included
6056 a superfluous trailing colon, the final name in the first list and the
6057 first name in the second list were incorrectly joined into one item (with a
6058 colon in the middle).
6064 1. Added -C and -D options to the exinext utility, mainly to make it easier
6065 to include in the automated testing, but these could be helpful when
6066 multiple configurations are in use.
6068 2. The exinext utility was not formatting the output nicely when there was
6069 an alternate port involved in the retry record key, nor when there was a
6070 message id as well (for retries that were specific to a specific message
6071 and a specific host). It was also confused by IPv6 addresses, because of
6072 the additional colons they contain. I have fixed the IPv4 problem, and
6073 patched it up to do a reasonable job for IPv6.
6075 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6076 delivery, the log line now contains "pipelined" if PIPELINING was used.
6078 4. An SMTP transport process used to panic and die if the bind() call to set
6079 an explicit outgoing interface failed. This has been changed; it is now
6080 treated in the same way as a connect() failure.
6082 5. A reference to $sender_host_name in the part of a conditional expansion
6083 that was being skipped was still causing a DNS lookup. This no longer
6086 6. The def: expansion condition was not recognizing references to header lines
6087 that used bh_ and bheader_.
6089 7. Added the _cache feature to named lists.
6091 8. The code for checking quota_filecount in the appendfile transport was
6092 allowing one more file than it should have been.
6094 9. For compatibility with Sendmail, the command line option
6102 and sets the incoming protocol and host name (for trusted callers). The
6103 host name and its colon can be omitted when only the protocol is to be set.
6104 Note the Exim already has two private options, -pd and -ps, that refer to
6105 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6106 "s", but I don't think that's a major issue.
6108 10. A number of refactoring changes to the code, none of which should affect
6111 (a) The number of logging options was getting close to filling up the
6112 32-bit word that was used as a bit map. I have split them into two classes:
6113 those that are passed in the argument to log_write(), and those that are
6114 only ever tested independently outside of that function. These are now in
6115 separate 32-bit words, so there is plenty of room for expansion again.
6116 There is no change in the user interface or the logging behaviour.
6118 (b) When building, for example, log lines, the code previously used a
6119 macro that called string_cat() twice, in order to add two strings. This is
6120 not really sufficiently general. Furthermore, there was one instance where
6121 it was actually wrong because one of the argument was used twice, and in
6122 one call a function was used. (As it happened, calling the function twice
6123 did not affect the overall behaviour.) The macro has been replaced by a
6124 function that can join an arbitrary number of extra strings onto a growing
6127 (c) The code for expansion conditions now uses a table and a binary chop
6128 instead of a serial search (which was left over from when there were very
6129 few conditions). Also, it now recognizes conditions like "pam" even when
6130 the relevant support is not compiled in: a suitably worded error message is
6131 given if an attempt is made to use such a condition.
6133 11. Added ${time_interval:xxxxx}.
6135 12. A bug was causing one of the ddress fields not to be passed back correctly
6136 from remote delivery subprocesses. The field in question was not being
6137 subsequently used, so this caused to problems in practice.
6139 13. Added new log selectors queue_time and deliver_time.
6141 14. Might have fixed a bug in maildirsizefile handling that threw up
6142 "unexpected character" debug warnings, and recalculated the data
6143 unnecessarily. In any case, I expanded the warning message to give more
6146 15. Added the message "Restricted characters in address" to the statements in
6147 the default ACL that block characters like @ and % in local parts.
6149 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6150 Three changes have been made:
6152 (a) There was a serious bug; a negative response to MAIL caused the whole
6153 recipient domain to be cached as invalid, thereby blocking all messages
6154 to all local parts at the same domain, from all senders. This bug has
6155 been fixed. The domain is no longer cached after a negative response to
6156 MAIL if the sender used is not empty.
6158 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6161 (c) A new callout option, "use_sender" has been added for people who want
6162 the modified behaviour.
6168 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6171 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6172 indeed breaks things for older releases.
6174 3. Added additional logging to the case where there is a problem reading data
6175 from a filter that is running in a subprocess using a pipe, in order to
6176 try to track down a specific problem.
6178 4. Testing facility fudge: when running in the test harness and attempting
6179 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6180 getting "No route to host". Convert this to a timeout.
6182 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6185 6. Some OS don't have socklen_t but use size_t instead. This affects the
6186 fifth argument of getsockopt() amongst other things. This is now
6187 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6188 can be set for individual OS. I have set it for SunOS5, OSF1, and
6189 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6190 some earlier ones do not.
6192 7. Change 4.30/15 was not doing the test caselessly.
6194 8. The standard form for an IPv6 address literal was being rejected by address
6195 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6196 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6197 this, as well as the form without the "IPv6" on the front (but only when
6198 address literals are enabled, of course).
6200 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6202 10. Exim crashed if a message with an empty sender address specified by -f
6203 encountered a router with an errors_to setting. This could be provoked only
6204 by a command such as
6208 where an empty string was supplied; "<>" did not hit this bug.
6210 11. Installed PCRE release 4.5.
6212 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6213 remained set. It is now erased.
6215 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6216 times from message ids (which are base 36 rather than the normal 62).
6218 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6219 were being counted as actual protocol errors, and logged if the log
6220 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6221 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6222 and DATA following a set of rejected RCPTs do not count as protocol errors.
6223 In other words, Exim assumes they were pipelined, though this may not
6224 actually be the case. Of course, in all cases the client gets an
6225 appropriate error code.
6227 15. If a lookup fails in an ACL condition, a message about the failure may
6228 be available; it is used if testing the ACL cannot continue, because most
6229 such messages specify what the cause of the deferral is. However, some
6230 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6231 that caused an old message to be retained and used if a later statement
6232 caused a defer, replacing the real cause of the deferral.
6234 16. If an IP address had so many PTR records that the DNS lookup buffer
6235 was not large enough to hold them, Exim could crash while trying to process
6236 the truncated data. It now detects and logs this case.
6238 17. Further to 4.21/58, another change has been made: if (and only if) the
6239 first line of a message (the first header line) ends with CRLF, a bare LF
6240 in a subsequent header line has a space inserted after it, so as not to
6241 terminate the header.
6243 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6244 unnecessarily, used atoi() instead of strtol(), and didn't check the
6245 termination when getting file sizes from file names by regex.
6247 19. Completely re-implemented the support for maildirsize files, in the light
6248 of a number of problems with the previous contributed implementation
6249 (4.30/29). In particular:
6251 . If the quota is zero, the maildirsize file is maintained, but no quota is
6254 . If the maildir directory does not exist, it is created before any attempt
6255 to write a maildirsize file.
6257 . The quota value in the file is just a cache; if the quota is changed in
6258 the transport, the new value overrides.
6260 . A regular expression is available for excluding directories from the
6263 20. The autoreply transport checks the characters in options that define the
6264 message's headers; it allows continued headers, but it was checking with
6265 isspace() after an embedded newline instead of explicitly looking for a
6268 21. If all the "regular" hosts to which an address was routed had passed their
6269 expiry times, and had not reached their retry times, the address was
6270 bounced, even if fallback hosts were defined. Now Exim should go on to try
6273 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6274 equivalent code in the SMTP transport. Some hosts send humungous responses
6275 to HELO/EHLO, more than 1024 it seems.
6277 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6278 gives compiler warnings in some environments. I've now done it "properly",
6281 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6282 (because of problems with the built-in one) was declared to return uschar *
6283 instead of char *, causing compiler failure.
6285 25. Fixed a file descriptor leak when processing alias/forward files.
6287 26. Fixed a minor format string issue in dbfn.c.
6289 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6291 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6292 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6295 29. When a sender address is verified, it is cached, to save repeating the test
6296 when there is more than one recipient in a message. However, when the
6297 verification involves a callout, it is possible for different callout
6298 options to be set for different recipients. It is too complicated to keep
6299 track of this in the cache, so now Exim always runs a verification when a
6300 callout is required, relying on the callout cache for the optimization.
6301 The overhead is duplication of the address routing, but this should not be
6304 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6305 to be verified with callout=postmaster, and the main callout worked but the
6306 postmaster check failed, the verification correctly failed. However, if a
6307 subsequent RCPT command asked for sender verification *without* the
6308 postmaster check, incorrect caching caused this verification also to fail,
6311 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6312 it was not caching the DNS options (qualify_single, search_parents) that
6313 were used when the lookup failed. A subsequent lookup with different
6314 options therefore always gave the same answer, though there were cases
6315 where it should not have. (Example: a "domains = !$mx_any" option on a
6316 dnslookup router: the "domains" option is always processed without any
6317 widening, but the router might have qualify_single set.) Now Exim uses the
6318 cached value only when the same options are set.
6320 32. Added John Jetmore's "exipick" utility to the distribution.
6322 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6323 than a timeout (e.g. a certificate is required, and is not provided), an
6324 Exim server now closes the connection immediately. Previously it waited for
6325 the client to close - but if the client is SSL, it seems that they each
6326 wait for each other, leading to a delay before one of them times out.
6328 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6329 maintained 0.8.x compatibility because I don't think many are using it, and
6330 it is clearly obsolete.
6332 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6335 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6336 client certificate was expired. A simple patch fixes this, though I don't
6337 understand the full logic of why the verify callback is called multiple
6340 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6341 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6342 which causes problems with some clients (such as the Certicom SSL Plus
6343 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6344 disables the coutermeasure allowing Eudora to connect."
6346 38. Exim was not checking that a write() to a log file succeeded. This could
6347 lead to Bad Things if a log got too big, in particular if it hit a file
6348 size limit. Exim now panics and dies if it cannot write to a log file, just
6349 as it does if it cannot open a log file.
6351 39. Modified OS/Makefile-Linux so that it now contains
6353 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6355 The two -D definitions ensure that Exim is compiled with large file
6356 support, which makes it possible to handle log files that are bigger than
6359 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6360 instance) a domain was checked against a named list that involved a lookup,
6361 causing $domain_data to be set, then another domain was checked against the
6362 same list, then the first domain was re-checked, the value of $domain_data
6363 after the final check could be wrong. In particular, if the second check
6364 failed, it could be set empty. This bug probably also applied to
6367 41. The strip_trailing_dot option was not being applied to the address given
6368 with the -f command-line option.
6370 42. The code for reading a message's header from the spool was incrementing
6371 $received_count, but never initializing it. This meant that the value was
6372 incorrect (doubled) while delivering a message in the same process in which
6373 it was received. In the most common configuration of Exim, this never
6374 happens - a fresh exec is done - but it can happen when
6375 deliver_drop_privilege is set.
6377 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6378 it now includes up to 150 characters of the unexpected data in the log
6381 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6382 and building data strings. The size of both of these buffers was 10 000
6383 bytes - far larger than anybody would *ever* want, thought I. Needless to
6384 say, somebody hit the limit. I have increased the maximum line length to
6385 20 000 and the maximum data length of concatenated lines to 100 000. I have
6386 also fixed two bugs, because there was no checking on these buffers. Tsk,
6387 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6388 buffer is too small.
6390 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6391 lsearch lookups. Now it does.
6393 46. When parsing a route_list item in a manualroute router, a fixed-length
6394 buffer was used for the list of hosts. I made this 1024 bytes long,
6395 thinking that nobody would ever have a list of hosts that long. Wrong.
6396 Somebody had a whole pile of complicated expansion conditions, and the
6397 string was silently truncated, leading to an expansion error. It turns out
6398 that it is easier to change to an unlimited length (owing to other changes
6399 that have happened since this code was originally written) than to build
6400 structure for giving a limitation error. The length of the item that
6401 expands into the list of hosts is now unlimited.
6403 47. The lsearch lookup could not handle data where the length of text line was
6404 more than 4095 characters. Such lines were truncated, leading to shortened
6405 data being returned. It should now handle lines of any length.
6407 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6408 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6411 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6412 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6413 no longer happens in comment lines. A list of replacements is now placed
6414 at the head of all of the source files, except those whose only change is
6415 to replace PERL_COMMAND in the very first #! line.
6417 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6418 messages on the queue, with a bottom-up merge sort, using code contributed
6419 by Michael Haardt. This should make operations like -bp somewhat faster on
6420 large queues. It won't affect queue runners, except when queue_run_in_order
6423 51. Installed eximstats 1.31 in the distribution.
6425 52. Added support for SRV lookups to the dnslookup router.
6427 53. If an ACL referred to $message_body or $message_body_end, the value was not
6428 reset for any messages that followed in the same SMTP session.
6430 54. The store-handling optimization for building very long strings was not
6431 differentiating between the different store pools. I don't think this
6432 actually made any difference in practice, but I've tidied it.
6434 55. While running the routers to verify a sender address, $sender_address
6435 was still set to the sender address. This is wrong, because when routing to
6436 send a bounce to the sender, it would be empty. Therefore, I have changed
6437 it so that, while verifying a sender address, $sender_address is set to <>.
6438 (There is no change to what happens when verifying a recipient address.)
6440 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6441 target A or AAAA records (if not already returned) without resetting the
6442 qualify_single or search_parents options of the DNS resolver. These are
6443 inappropriate in this case because the targets of MX and SRV records must
6444 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6445 target that, when qualified, matched something in the local domain. These
6446 two options are now turned off when doing these lookups.
6448 57. It seems that at least some releases of Reiserfs (which does not have the
6449 concept of a fixed number of inodes) returns zero and not -1 for the
6450 number of available inodes. This interacted badly with check_spool_inodes,
6451 which assumed that -1 was the "no such thing" setting. What I have done is
6452 to check that the total number of inodes is greater than zero before doing
6453 the test of how many are available.
6455 58. When a "warn" ACL statement has a log_message modifier, the message is
6456 remembered, and not repeated. This is to avoid a lot of repetition when a
6457 message has many recipients that cause the same warning to be written.
6458 However, Exim was preserving the list of already written lines for an
6459 entire SMTP session, which doesn't seem right. The memory is now reset if a
6460 new message is started.
6462 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6463 debugging output unless log_rewrite was also set.
6465 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6466 of (int)(handle) when we know that handle contains (void *)(-1).
6468 61. The Exim daemon panic-logs an error return when it closes the incoming
6469 connection. However "connection reset by peer" seems to be common, and
6470 isn't really an error worthy of noting specially, so that particular error
6473 62. When Exim is trying to find all the local interfaces, it used to panic and
6474 die if the ioctl to get the interface flags failed. However, it seems that
6475 on at least one OS (Solaris 9) it is possible to have an interface that is
6476 included in the list of interfaces, but for which you get a failure error
6477 for this call. This happens when the interface is not "plumbed" into a
6478 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6479 failure of the "get flags" call assumes that the interface is down.
6481 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6482 makes life easier for people who are doing arithmetic on fields extracted
6483 from dates, where you often get leading zeros that should not be
6484 interpreted as octal.
6486 64. Added qualify_domain to the redirect router, to override the global
6489 65. If a pathologically long header line contained very many addresses (the
6490 report of this problem mentioned 10 000) and each of them was rewritten,
6491 Exim could use up a very large amount of memory. (It kept on making new
6492 copies of the header line as it rewrote, and never released the old ones.)
6493 At the expense of a bit more processing, the header rewriting function has
6494 been changed so that it no longer eats memory in this way.
6496 66. The generation of the Received: header has been moved from the time that a
6497 message starts to be received, to the time that it finishes. The timestamp
6498 in the Received: header should now be very close to that of the <= log
6499 line. There are two side-effects of this change:
6501 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6502 logged header lines no longer include the local Received: line, because
6503 it has not yet been created. The same applies to a copy of the message
6504 that is returned to a non-SMTP sender when a message is rejected.
6506 (b) When a filter file is tested using -bf, no additional Received: header
6507 is added to the test message. After some thought, I decided that this
6510 This change does not affect the value of $received_for. It is still set
6511 after address rewriting, but before local_scan() is called.
6513 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6515 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6516 gave an unhelpful panic error message, and a defer error. I have managed to
6517 change this behaviour so that it now rejects any supplied certificate,
6518 which seems right, as the list of acceptable certificates is empty.
6520 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6521 gave an unhelpful defer error. I have not managed to make this reject any
6522 supplied certificates, but the error message it gives is "no certificate
6523 supplied", which is not helpful.
6525 70. exigrep's output now also includes lines that are not associated with any
6526 message, but which match the given pattern. Implemented by a patch from
6527 Martin Sluka, which also tidied up the Perl a bit.
6529 71. Recipient callout verification, like sender verification, was using <> in
6530 the MAIL FROM command. This isn't really the right thing, since the actual
6531 sender may affect whether the remote host accepts the recipient or not. I
6532 have changed it to use the actual sender in the callout; this means that
6533 the cache record is now keyed on a recipient/sender pair, not just the
6534 recipient address. There doesn't seem to be a real danger of callout loops,
6535 since a callout by the remote host to check the sender would use <>.
6536 [SEE ABOVE: changed after hitting problems.]
6538 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6539 temporary errors. However, in the case of such a code being given after
6540 the end of a data transmission (i.e. after ".") Exim was failing to write
6541 a retry record for the message. (Yes, there was some broken host that was
6542 actually sending 8xx at this point.)
6544 73. An unknown lookup type in a host list could cause Exim to panic-die when
6545 the list was checked. (An example that provoked this was putting <; in the
6546 middle of a list instead of at the start.) If this happened during a DATA
6547 ACL check, a -D file could be left lying around. This kind of configuration
6548 error no longer causes Exim to die; instead it causes a defer error. The
6549 incident is still logged to the main and panic logs.
6551 74. Buglet left over from Exim 3 conversion. The message "too many messages
6552 in one connection" was written to the rejectlog but not the mainlog, except
6553 when address rewriting (yes!) was being logged.
6555 75. Added write_rejectlog option.
6557 76. When a system filter was run not as root (that is, when system_filter_user
6558 was set), the values of the $n variables were not being returned to the
6559 main process; thus, they were not subsequently available in the $sn
6562 77. Added +return_path_on_delivery log selector.
6564 78. A connection timeout was being treated differently from recipients deferred
6565 when testing hosts_max_try with a message that was older than the host's
6566 retry timeout. (The host should not be counted, thus allowing all hosts to
6567 be tried at least once before bouncing.) This may have been the cause of an
6568 occasionally reported bug whereby a message would remain on the queue
6569 longer than the retry timeout, but would be bounced if a delivery was
6570 forced. I say "may" because I never totally pinned down the problem;
6571 setting up timeout/retry tests is difficult. See also the next item.
6573 79. The ultimate address timeout was not being applied to errors that involved
6574 a combination of host plus message (for example, a timeout on a MAIL
6575 command). When an address resolved to a number of possible hosts, and they
6576 were not all tried for each delivery (e.g. because of hosts_max_try), a
6577 message could remain on the queue longer than the retry timeout.
6579 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6582 81. Fixed an obscure SMTP outgoing bug which required at least the following
6583 conditions: (a) there was another message waiting for the same server;
6584 (b) the server returned 5xx to all RCPT commands in the first message so
6585 that the message was not completed; (c) the server dropped the connection
6586 or gave a negative response to the RSET that Exim sends to abort the
6587 transaction. The observed case was a dropped connection after DATA that had
6588 been sent in pipelining mode. That is, the server had advertised PIPELINING
6589 but was not implementing it correctly. The effect of the bug was incorrect
6590 behaviour, such as trying another host, and this could lead to a crash.
6596 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6597 and daemon.c were passed as pointers to ints; they should have been
6598 pointers to socklen_t variables (which are typically unsigned ints).
6600 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6603 3. Fixed a really odd bug that affected only the testing scheme; patching a
6604 certain fixed string in the binary changed the value of another string that
6605 happened to be identical to the end of the original first string.
6607 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6608 name", it returns that address as the IP address. On some operating
6609 systems (e.g. Solaris), it also passes back the IP address string as the
6610 "host name". However, on others (e.g. Linux), it passes back an empty
6611 string. Exim wasn't checking for this, and was changing the host name to an
6612 empty string, assuming it had been canonicalized.
6614 5. Although rare, it is permitted to have more than one PTR record for a given
6615 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6616 all the names associated with an address, because they do in Solaris.
6617 However, it seems that they do not in Linux for data that comes from the
6618 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6619 I found this out when I moved to a new Linux workstation and tried to run
6620 the Exim test suite.
6622 To get round this problem I have changed the code so that it now does its
6623 own call to the DNS to look up PTR records when searching for a host name.
6624 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6625 addresses that are only in /etc/hosts are still found.
6627 This behaviour is, however, controlled by an option called host_lookup_
6628 order, which defaults to "bydns:byaddr". If people want to use the other
6629 order, or indeed, just use one or the other means of lookup, they can
6630 specify it in this variable.
6632 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6633 some operating systems, this comes back from gethostbyaddr() as an empty
6634 string, and this is what Exim used to test for. However, it seems that in
6635 other systems, "." is yielded. Exim now tests for this case too.
6637 7. The values of check_spool_space and check_log_space are now held internally
6638 as a number of kilobytes instead of an absolute number of bytes. If a
6639 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6640 kilobyte. This means that much larger values can be stored.
6642 8. Exim monitor: an attempt to get the action menu when not actually pointing
6643 at a message produces an empty menu entitled "No message selected". This
6644 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6645 no entries in it ("Shell widget menu has zero width and/or height"). So I
6646 have added a single, blank menu entry in this case.
6648 9. Added ${quote_local_part.
6650 10. MIME decoding is now applied to the contents of Subject: header lines when
6653 11. Now that a reference to $sender_host_address automatically causes a reverse
6654 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6655 host lookup before query-style lookups in lists that might use this
6656 variable. This has therefore been abolished, and the "net-" prefix is no
6657 longer necessary for query-style lookups.
6659 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6660 have been a typo for LFLAGS, so it has been changed.
6662 13. The install script calls Exim with "-C /dev/null" in order to find the
6663 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6664 to be output. However, since Exim outputs its version number before the
6665 error, it didn't break the script. It just looked ugly. I fixed this by
6666 always allowing "-C /dev/null" if the caller is root.
6668 14. Ignore overlarge ACL variable number when reading spool file - insurance
6669 against a later release with more variables having written the file.
6671 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6672 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6673 this, as well as the form without the "IPv6" on the front.
6675 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6676 OS/Makefile-Darwin file.
6678 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6679 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6681 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6682 variable. However, some OS use ints for inodes. I've added cast to long int
6683 to get rid of the compiler warning.
6685 19. I had forgotten to lock out "/../" in configuration file names when
6686 ALT_CONFIG_PREFIX was set.
6688 20. Routers used for verification do not need to specify transports. However,
6689 if such a router generated a host list, and callout was configured, Exim
6690 crashed, because it could not find a port number from the (non-existent)
6691 transport. It now assumes port 25 in this circumstance.
6693 21. Added the -t option to exigrep.
6695 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6696 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6697 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6698 them is set, LOOKUP_LSEARCH is forced.
6700 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6702 24. Added sender and host information to the "rejected by local_scan()" log
6703 line; previously there was no indication of these.
6705 25. Added .include_if_exists.
6707 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6708 Linux. It turns out that not all file systems support this. Apparently some
6709 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6710 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6711 not supported on the file descriptor, is now ignored when Exim is trying to
6712 sync a directory. This applies only to Linux.
6714 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6716 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6717 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6720 29. Added support for maildirsize files from supplied patch (modified a bit).
6722 30. The use of :fail: followed by an empty string could lead Exim to respond to
6723 sender verification failures with (e.g.):
6725 550 Verification failed for <xxx>
6726 550 Sender verify failed
6728 where the first response line was missing the '-' that indicates it is not
6729 the final line of the response.
6731 31. The loop for finding the name of the user that called Exim had a hardwired
6732 limit of 10; it now uses the value of finduser_retries, which is used for
6733 all other user lookups.
6735 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6738 33. Exim was neglecting to zero errno before one call of strtol() when
6739 expanding a string and expecting an integer value. On some systems this
6740 resulted in spurious "integer overflow" errors. Also, it was casting the
6741 result into an int without checking.
6743 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6744 did not work. The code looks as if it has *never* worked, though it appears
6745 to have been documented since at least release 1.62. I have made it work.
6747 35. The "timeout_DNS" error in retry rules, also documented since at least
6748 1.62, also never worked. As it isn't clear exactly what this means, and
6749 clearly it isn't a major issue, I have abolished the feature by treating it
6750 as "timeout", and writing a warning to the main and panic logs.
6752 36. The display of retry rules for -brt wasn't always showing the error code
6755 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6756 timeout_connect_A, timeout_connect_MX.
6758 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6759 to the empty sender.
6761 39. The daemon was not analysing the content of -oX till after it had closed
6762 stderr and disconnected from the controlling terminal. This meant that any
6763 syntax errors were only noted on the panic log, and the return code from
6764 the command was 0. By re-arranging the code a little, I've made the
6765 decoding happen first, so such errors now appear on stderr, and the return
6766 code is 1. However, the actual setting up of the sockets still happens in
6767 the disconnected process, so errors there are still only recorded on the
6770 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6771 connections (as happens on some IP stacks) was logged at start up time as
6772 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6773 it from "IPv6 and IPv4", which means that two separate sockets are being
6776 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6777 says whether AF_INET or AF_INET6 was passed as an argument.
6779 42. Exiwhat output was messed up when time zones were included in log
6782 43. Exiwhat now gives more information about the daemon's listening ports,
6783 and whether -tls-on-connect was used.
6785 44. The "port" option of the smtp transport is now expanded.
6787 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6788 silently ignored. Now an error message is written to the main and panic
6791 46. There's a new ACL modifier called "logwrite" which writes to a log file
6792 as soon as it is encountered.
6794 47. Added $local_user_uid and $local_user_gid at routing time.
6796 48. Exim crashed when trying to verify a sender address that was being
6799 49. Exim was recognizing only a space character after ".include". It now also
6800 recognizes a tab character.
6802 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6803 extracting the relevant information from the specification. The man page no
6804 longer contains scrambled data for the -d option, and I've added a section
6805 at the front about calling Exim under different names.
6807 51. Added "extra_headers" argument to the "mail" command in filter files.
6809 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6812 53. Installed eximstats 1.29.
6814 54. Added transport_filter_timeout as a generic transport option.
6816 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6817 Cc: header lines. This was required by RFC 822, but it not required by RFC
6820 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6821 incoming messages that did not have them. Now it does so only if the
6822 message originates locally, that is, if there is no associated remote host
6823 address. When Resent- header lines are present, this applies to the Resent-
6824 lines rather than the non-Resent- lines.
6826 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6827 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6829 58. Messages for configuration errors now include the name of the main
6830 configuration file - useful now that there may be more than one file in a
6831 list (.included file names were always shown).
6833 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6834 for those rare installations that do not start the daemon as root or run it
6835 setuid root. I've cut out the call to initgroups() if the daemon is not
6838 60. The Exim user and group can now be bound into the binary as text strings
6839 that are looked up at the start of Exim's processing.
6841 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6843 62. Added $mailstore_basename variable.
6845 63. Installed patch to sieve.c from Michael Haardt.
6847 64. When Exim failed to open the panic log after failing to open the main log,
6848 the original message it was trying to log was written to stderr and debug
6849 output, but if they were not available (the usual case in production), it
6850 was lost. Now it is written to syslog before the two lines that record the
6851 failures to open the logs.
6853 65. Users' Exim filters run in subprocesses under the user's uid. It is
6854 possible for a "deliver" command or an alias in a "personal" command to
6855 provoke an address rewrite. If logging of address rewriting is configured,
6856 this fails because the process is not running as root or exim. There may be
6857 a better way of dealing with this, but for the moment (because 4.30 needs
6858 to be released), I have disabled address rewrite logging when running a
6859 filter in a non-root, non-exim process.
6865 1. The buildconfig auxiliary program wasn't quoting the value set for
6866 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6867 not defined. This bug was masked in 4.22 by the effect that was fixed in
6870 2. Some messages that were rejected after a message id was allocated were
6871 shown as "incomplete" by exigrep. It no longer does this for messages that
6872 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6874 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6875 have allow_domain_literals set, the ID did not get logged in the <= line.
6876 Domain literals are now always recognized in Message-ID: header lines.
6878 4. The first argument for a ${extract expansion item is the key name or field
6879 number. Leading and trailing spaces in this item were not being ignored,
6880 causing some misleading effects.
6882 5. When deliver_drop_privilege was set, single queue runner processes started
6883 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6884 same command in the process it spins off) were not dropping privilege.
6886 6. When the daemon running as "exim" started a queue runner, it always
6887 re-executed Exim in the spun-off process. This is a waste of effort when
6888 deliver_drop_privilege is set. The new process now just calls the
6889 queue-runner function directly.
6895 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6898 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6899 ignored. Though the use of -oP was forcing the writing of a pid file, it
6900 was always written to the default place.
6902 3. If the message "no IP address found for host xxxx" is generated during
6903 incoming verification, it is now followed by identification of the incoming
6904 connection (so you can more easily find what provoked it).
6906 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6908 5. Added some features to "harden" Exim a bit more against certain attacks:
6910 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6911 be put in Local/Makefile. This is like the never_users runtime option,
6912 but it cannot be overridden. The default setting is "root".
6914 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6915 prefix string with which any file named in a -C command line option
6918 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6919 is retained for -C and -D only if the caller of Exim is root. Without
6920 it, the exim user may also use -C and -D and retain privilege.
6922 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6923 command line option is disabled.
6925 6. Macro names set by the -D option must start with an upper case letter, just
6926 like macro names defined in the configuration file.
6928 7. Added "dereference=" facility to LDAP.
6930 8. Two instances of the typo "uknown" in the source files are fixed.
6932 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6933 the Configure-Makefile script screwed up while processing it.
6935 10. Incorporated PCRE 4.4.
6937 11. The SMTP synchronization check was not operating right at the start of an
6938 SMTP session. For example, it could not catch a HELO sent before the client
6939 waited for the greeting. There is now a check for outstanding input at the
6940 point when the greeting is written. Because of the duplex, asynchronous
6941 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6942 way, but not yet received, when the check is performed.
6944 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6945 on TCP/IP sockets, because this apparently causes some broken clients to
6948 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6949 unchanged) from the Cygwin maintainer.
6951 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6952 was supported instead of testing for mbx. Effectively a typo.
6954 15. The spa authenticator server code was not checking that the input it
6955 received was valid base64.
6957 16. The debug output line for the "set" modifier in ACLs was not showing the
6958 name of the variable that was being set.
6960 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6962 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6963 lookup on its own. Something else was needed to trigger the lookup. For
6964 example, a match in host_lookup or the need for a host name in a host list.
6965 Now, if $sender_host_name is referenced and the host name has not yet been
6966 looked up, a lookup is performed. If the lookup fails, the variable remains
6967 empty, and $host_lookup_failed is set to "1".
6969 19. Added "eqi" as a case-independent comparison operator.
6971 20. The saslauthd authentication condition could segfault if neither service
6972 nor realm was specified.
6974 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6975 error message that was logged was misleading, and incoming SMTP
6976 connections were dropped. The message is now more accurate, and temporary
6977 errors are given to SMTP connections.
6979 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6980 (except RSET) with a 421 error, until QUIT is received. However, it was
6981 failing to send a response to QUIT.
6983 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6984 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6985 ACL is now run after the helo_try_verify_hosts code.
6987 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6988 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6989 case-independent, so other case variants are also recognized). Apparently
6990 some systems use these upper case variants.
6992 25. If more than two messages were waiting for the same host, and a transport
6993 filter was specified for the transport, Exim sent two messages over the
6994 same TCP/IP connection, and then failed with "socket operation on non-
6995 socket" when it tried to send the third.
6997 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6999 27. The extern definition of crypt16() in expand.c was not being excluded when
7000 the OS had its own crypt16() function.
7002 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7003 as a preferred synonym for return_size_limit, both as an option and as an
7006 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7008 30. Changed the default configuration ACL to relax the local part checking rule
7009 for addresses that are not in any local domains. For these addresses,
7010 slashes and pipe symbols are allowed within local parts, but the sequence
7011 /../ is explicitly forbidden.
7013 31. SPA server authentication was not clearing the challenge buffer before
7016 32. log_message in a "warn" ACL statement was writing to the reject log as
7017 well as to the main log, which contradicts the documentation and doesn't
7018 seem right (because no rejection is happening). So I have stopped it.
7020 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7021 However, I am unable to do any testing of this.
7023 34. Fixed an infelicity in the appendfile transport. When checking directories
7024 for a mailbox, to see if any needed to be created, it was accidentally
7025 using path names with one or more superfluous leading slashes; tracing
7026 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7028 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7029 contents are added to the log line that is written for every discarded
7030 recipient. (Previously a log_message setting was ignored.)
7032 36. The ${quote: operator now quotes the string if it is empty.
7034 37. The install script runs exim in order to find its version number. If for
7035 some reason other than non-existence or emptiness, which it checks, it
7036 could not run './exim', it was installing it with an empty version number,
7037 i.e. as "exim-". This error state is now caught, and the installation is
7040 38. An argument was missing from the function that creates an error message
7041 when Exim fails to connect to the socket for saslauthd authentication.
7042 This could cause Exim to crash, or give a corrupted message.
7044 39. Added isip, isip4, and isip6 to ${if conditions.
7046 40. The ACL variables $acl_xx are now saved with the message, and can be
7047 accessed later in routers, transports, and filters.
7049 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7050 strings in the file are not string-expanded.
7052 42. If a MAIL command specified a SIZE value that was too large to fit into an
7053 int variable, the check against message_size_limit failed. Such values are
7054 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7055 day this will have to be increased, but I don't think I want to be around
7056 when emails are that large.
7063 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7064 iconv() is not standard in FreeBSD.
7066 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7067 IPv6 enabled. The observed symptom was a segmentation fault on return from
7068 the function os_common_find_running_interfaces() in src/os.c.
7070 3. In the check_special_case() function in daemon.c I had used "errno" as an
7071 argument name, which causes warnings on some systems. This was basically a
7072 typo, since it was named "eno" in the comments!
7074 4. The code that waits for the clock to tick (at a resolution of some fraction
7075 of a second) so as to ensure message-id uniqueness was always waiting for
7076 at least one whole tick, when it could have waited for less. [This is
7077 almost certainly not relevant at current processor speeds, where it is
7078 unlikely to ever wait at all. But we try to future-proof.]
7080 5. The function that sleeps for a time interval that includes fractions of a
7081 second contained a race. It did not block SIGALRM between setting the
7082 timer, and suspending (a couple of lines later). If the interval was short
7083 and the sigsuspend() was delayed until after it had expired, the suspension
7084 never ended. On busy systems this could lead to processes getting stuck for
7087 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7088 process, before it forks any delivery processes. The open lookup caching
7089 mechanism meant that the open file or database connection was passed into
7090 the delivery process. The problem was that delivery processes always tidy
7091 up cached lookup data. This could cause a problem for the next delivery
7092 process started by the queue runner, because the external queue runner
7093 process does not know about the closure. So the next delivery process
7094 still has data in the lookup cache. In the case of a file lookup, there was
7095 no problem because closing a file descriptor in a subprocess doesn't affect
7096 the parent. However, if the lookup was caching a connection to a database,
7097 the connection was closed, and the second delivery process was likely to
7098 see errors such as "PGSQL: query failed: server closed the connection
7099 unexpectedly". The problem has been fixed by closing all cached lookups
7100 in a queue runner before running a delivery process.
7102 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7103 seem to have the "const" qualifier which it has on other OS. I've
7106 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7107 *of the same type* (client or server) with the same public name that an
7108 error should be diagnosed.
7110 9. When Exim looked up a host name for an IP address, but failed to find the
7111 original IP address when looking up the host name (a safety check), it
7112 output the message "<ip address> does not match any IP for NULL", which was
7113 confusing, to say the least. The bug was that the host name should have
7114 appeared instead of "NULL".
7116 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7117 user that is built into the binary, and the -C or -D options is used, root
7118 privilege is dropped before the configuration file is read. In addition,
7119 logging is switched to stderr instead of the normal log files. If the
7120 configuration then re-defines the Exim user, the unprivileged environment
7121 is probably not what is expected, so Exim logs a panic warning message (but
7124 However, if deliver_drop_privilege is set, the unprivileged state may well
7125 be exactly what is intended, so the warning has been cut out in that case,
7126 and Exim is allowed to try to write to its normal log files.
7132 1. smtp_return_error_details was not giving details for temporary sender
7133 or receiver verification errors.
7135 2. Diagnose a configuration error if two authenticators have the same public
7138 3. Exim used not to create the message log file for a message until the first
7139 delivery attempt. This could be confusing when incoming messages were held
7140 for policy or load reasons. The message log file is now created at the time
7141 the message is received, and an initial "Received" line is written to it.
7143 4. The automatically generated man page for command line options had a minor
7144 bug that caused no ill effects; however, a more serious problem was that
7145 the procedure for building the man page automatically didn't always
7146 operate. Consequently, release 4.20 contains an out-of-date version. This
7147 shouldn't happen again.
7149 5. When building Exim with embedded Perl support, the script that builds the
7150 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7151 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7153 6. The freeze_tell option was not being used for messages that were frozen on
7154 arrival, either by an ACL or by local_scan().
7156 7. Added the smtp_incomplete_transaction log selector.
7158 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7159 was accepting AUTH without a new EHLO.
7161 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7162 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7165 10. Logging of TCP/IP connections (when configured) now happens in the main
7166 daemon process instead of the child process, so that the TCP/IP connection
7167 count is more accurate (but it can never be perfect).
7169 11. The use of "drop" in a nested ACL was not being handled correctly in the
7170 outer ACL. Now, if condition failure induced by the nested "drop" causes
7171 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7172 or "require"), the connection is dropped.
7174 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7175 that yield "discard" can now be used with an "accept" or a "discard" verb,
7176 but an error is generated for any others (because I can't see a useful way
7177 to define what should happen).
7179 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7180 are now processed in the same way as lines in the Exim configuration file.
7181 In particular, continuation lines are supported.
7183 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7185 15. Added -ti meaning -t -i.
7187 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7188 domains, and warn by logging if others are found.
7190 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7191 returned from GIFCONF when Exim is trying to find the list of interfaces on
7192 a host. The code in os.c has been modified to copy each ifreq to an aligned
7193 structure in all cases.
7195 Also, in some cases, the returned ifreq's were being copied to a 'struct
7196 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7197 means the last couple of bytes of an IPv6 address could be chopped if the
7198 ifreq contained only a normal sockaddr (14 bytes storage).
7200 18. Named domain lists were not supported in the hosts_treat_as_local option.
7201 An entry such as +xxxx was not recognized, and was treated as a literal
7204 19. Ensure that header lines added by a DATA ACL are included in the reject log
7205 if the ACL subsequently rejects the message.
7207 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7208 MD5 (which is deprecated).
7210 21. When testing a filter file using -bf, Exim was writing a message when it
7211 took the sender from a "From " line in the message, but it was not doing so
7212 when it took $return_path from a Return-Path: header line. It now does.
7214 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7215 with a valid header line field name (a series of printing characters
7216 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7218 23. Changed "disc" in the source to "disk" to conform to the documentation and
7219 the book and for uniformity.
7221 24. Ignore Sendmail's -Ooption=value command line item.
7223 25. When execve() failed while trying to run a command in a pipe transport,
7224 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7225 could be confused with a return value of 69 from the command itself. This
7226 has been changed to 127, the value the shell returns if it is asked to run
7227 a non-existent command. The wording for the related log line suggests a
7228 non-existent command as the problem.
7230 26. If received_header_text expands to an empty string, do not add a Received:
7231 header line to the message. (Well, it adds a token one on the spool, but
7232 marks it "old" so that it doesn't get used or transmitted.)
7234 27. Installed eximstats 1.28 (addition of -nt option).
7236 28. There was no check for failure on the call to getsockname() in the daemon
7237 code. This can fail if there is a shortage of resources on the system, with
7238 ENOMEM, for example. A temporary error is now given on failure.
7240 29. Contrary to the C standard, it seems that in some environments, the
7241 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7242 program. Exim now does this explicitly; it affects the formatting of
7243 timestamps using strftime().
7245 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7246 complaints. I've now initialized all the variables, to avoid this.
7248 32. Header lines added by a system filter were not being "seen" during
7249 transport-time rewrites.
7251 33. The info_callback() function passed to OpenSSL is set up with type void
7252 (*)(SSL *, int, int), as described somewhere. However, when calling the
7253 function (actually a macro) that sets it up, the type void(*)() is
7254 expected. I've put in a cast to prevent warnings from picky compilers.
7256 34. If a DNS black list lookup found a CNAME record, but there were no A
7257 records associated with the domain it pointed at, Exim crashed.
7259 35. If a DNS black list lookup returned more than one A record, Exim ignored
7260 all but the first. It now scans all returned addresses if a particular IP
7261 value is being sought. In this situation, the contents of the
7262 $dnslist_value variable are a list of all the addresses, separated by a
7265 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7266 to accept a host name and all its aliases if the forward lookup for any of
7267 them yielded the IP address of the incoming connection. Now it accepts only
7268 those names whose forward lookup yields the correct IP address. Any other
7269 names are discarded. This closes a loophole whereby a rogue DNS
7270 administrator could create reverse DNS records to break through a
7271 wildcarded host restriction in an ACL.
7273 37. If a user filter or a system filter that ran in a subprocess used any of
7274 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7275 the wrong values were passed to the pipe command ($thisaddress had the
7276 value of $0, $0 had the value of $1, etc). This bug was introduced by
7277 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7279 38. Improved the line breaking for long SMTP error messages from ACLs.
7280 Previously, if there was no break point between 40 and 75 characters, Exim
7281 left the rest of the message alone. Two changes have been made: (a) I've
7282 reduced the minimum length to 35 characters; (b) if it can't find a break
7283 point between 35 and 75 characters, it looks ahead and uses the first one
7284 that it finds. This may give the occasional overlong line, but at least the
7285 remaining text gets split now.
7287 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7288 file descriptors might be low, and that setting 1000 would always raise it.
7289 It turns out that in some environments, the limit is already over 1000 and
7290 that lowering it causes trouble. So now Exim takes care not to decrease it.
7292 40. When delivering a message, the value of $return_path is set to $sender_
7293 address at the start of routing (routers may change the value). By an
7294 oversight, this default was not being set up when an address was tested by
7295 -bt or -bv, which affected the outcome if any router or filter referred to
7298 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7299 log only when "message" or "log_message" are set. However, if one of the
7300 conditions was an address verification, or a call to a nested ACL, the
7301 messages generated by the underlying test were being passed through. This
7302 no longer happens. The underlying message is available in $acl_verify_
7303 message for both "message" and "log_message" expansions, so it can be
7304 passed through if needed.
7306 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7307 new expansion $bh_ to give the encoded byte string without charset
7308 translation. Translation happens only if iconv() is available; HAVE_ICONV
7309 indicates this at build time. HEADERS_CHARSET gives the charset to
7310 translate to; headers_charset can change it in the configuration, and
7311 "headers charset" can change it in an individual filter file.
7313 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7314 that creates RFC 2047 encoded "words" labels them as that charset instead
7315 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7316 expansion operator; (ii) when Exim creates a From: line for a local
7317 message; (iii) when a header line is rewritten to include a "phrase" part.
7319 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7320 buggy, causing it to skip the first lines of messages whose message ID
7321 ended in 'D'. This would not have bitten before Exim release 4.14, because
7322 message IDs were unlikely to end in 'D' before then. The effect was to have
7323 incorrect size information for certain domains.
7325 45. #include "config.h" was missing at the start of the crypt16.c module. This
7326 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7329 46. If there was a timeout during a "random" callout check, Exim treated it as
7330 a failure of the random address, and carried on sending RSET and the real
7331 address. If the delay was just some slowness somewhere, the response to the
7332 original RCPT would be taken as a response to RSET and so on, causing
7333 mayhem of various kinds.
7335 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7336 when I implemented it. It didn't allow for the fact that some option values
7337 may legitimately be negative (e.g. size_addition), and it didn't even do
7338 the right test for positive values.
7340 48. Domain names in DNS records are case-independent. Exim always looks them up
7341 in lower case. Some resolvers return domain names in exactly the case they
7342 appear in the zone file, that is, they may contain uppercase letters. Not
7343 all resolvers do this - some return always lower case. Exim was treating a
7344 change of case by a resolver as a change of domain, similar to a widening
7345 of a domain abbreviation. This triggered its re-routing code and so it was
7346 trying to route what was effectively the same domain again. This normally
7347 caused routing to fail (because the router wouldn't handle the domain
7348 twice). Now Exim checks for this case specially, and just changes the
7349 casing of the domain that it ultimately uses when it transmits the message
7352 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7355 50. If a filter generated a file delivery with a non-absolute name (possible if
7356 no home directory exists for the router), the forbid_file option was not
7359 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7360 the existing equality matching.
7362 52. Exim was using ints instead of ino_t variables in some places where it was
7363 dealing with inode numbers.
7365 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7366 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7367 called TMPDIR, and if it finds it is different, it changes its value.
7369 54. The smtp_printf() function is now made available to local_scan() so
7370 additional output lines can be written before returning. There is also an
7371 smtp_fflush() function to enable the detection of a dropped connection.
7372 The variables smtp_input and smtp_batched_input are exported to
7375 55. Changed the default runtime configuration: the message "Unknown user"
7376 has been removed from the ACL, and instead placed on the localuser router,
7377 using the cannot_route_message feature. This means that any verification
7378 failures that generate their own messages won't get overridden. Similarly,
7379 the "Unrouteable address" message that was in the ACL for unverifiable
7380 relay addresses has also been removed.
7382 56. Added hosts_avoid_esmtp to the smtp transport.
7384 57. The exicyclog script was not checking for the esoteric option
7385 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7386 will work only if exicyclog is run under the appropriate euid.
7388 58. Following a discussion on the list, the rules by which Exim recognises line
7389 endings on incoming messages have been changed. The -dropcr and drop_cr
7390 options are now no-ops, retained only for backwards compatibility. The
7391 following line terminators are recognized: LF CRLF CR. However, special
7392 processing applies to CR:
7394 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7395 nor a local message in the state where . is a terminator.
7397 (ii) If a bare CR is encountered in a header line, an extra space is added
7398 after the line terminator so as not to end the header. The reasoning
7399 behind this is that bare CRs in header lines are most likely either
7400 to be mistakes, or people trying to play silly games.
7402 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7403 was being incorrectly given as 18 bytes larger than it should have been.
7404 This is a VOB (very old bug).
7406 60. This may never have affected anything current, but just in case it has:
7407 When the local host is found other than at the start of a list of hosts,
7408 the local host, those with the same MX, and any that follow, are discarded.
7409 When the list in question was part of a longer list of hosts, the following
7410 hosts (not currently being processed) were also being discarded. This no
7411 longer happens. I'm not sure if this situation could ever has previously
7414 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7417 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7418 header. This is something that is recommended in a new Internet Draft, and
7419 is something that is documented as being done by Sendmail. There are two
7420 possible values. For messages generated by the autoreply transport, Exim
7423 Auto-Submitted: auto-replied
7425 whereas for all other generated messages (e.g. bounces) it adds
7427 Auto-Submitted: auto-generated
7429 63. The "personal" condition in filters now includes a test for the
7430 Auto-Submitted: header. If it contains the string "auto-" the message it
7431 not considered personal.
7433 64. Added rcpt_include_affixes as a generic transport option.
7435 65. Added queue_only_override (default true).
7437 66. Added the syslog_duplication option.
7439 67. If what should have been the first header line of a message consisted of
7440 a space followed by a colon, Exim was mis-interpreting it as a header line.
7441 It isn't of course - it is syntactically invalid and should therefore be
7442 treated as the start of the message body. The misbehaviour could have
7443 caused a number of strange effects, including loss of data in subsequent
7444 header lines, and spool format errors.
7446 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7447 client host had authenticated. This control can now be exercised by an ACL
7448 for more flexibility.
7450 69. By default, callouts do not happen when testing with -bh. There is now a
7451 variant, -bhc, which does actually run the callout code, including
7452 consulting and updating the callout cache.
7454 70. Added support for saslauthd authentication, courtesy of Alexander
7457 71. If statvfs() failed on the spool or log directories while checking their
7458 size for availability, Exim confusingly gave the error "space shortage".
7459 Furthermore, in debugging mode it crashed with a floating point exception.
7460 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7461 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7462 serious problem, Exim now writes to the main and panic logs when this
7463 happens, with details of the failure. It then refuses to accept the
7464 incoming message, giving the message "spool directory problem" or "log
7465 directory problem" with a 421 code for SMTP messages.
7467 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7468 0, 1, and 2 exist, because some OS complain for execs without them (see
7469 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7470 descriptors. However, the code omitted to check that the open succeeded,
7471 causing mysterious errors if for some reason the permissions on /dev/null
7472 got screwed. Now Exim writes a message to the main and panic logs, and
7473 bombs out if it can't open /dev/null.
7475 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7476 interact so that it is all more flexible. It is supposed to remain
7477 backwards compatible. Also added extra_local_interfaces.
7479 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7480 to bomb out with an assertion failure - to the client this appears as a
7481 connection drop. This problem occurs in the part of the code that was taken
7482 from the Samba project. Fortunately, the assertion is in a very simple
7483 function, so I have fixed this by reproducing the function inline in the
7484 one place where it is called, and arranging for authentication to fail
7485 instead of killing the process with assert().
7487 75. The SPA client code was not working when the server requested OEM rather
7488 than Unicode encoding.
7490 76. Added code to make require_files with a specific uid setting more usable in
7491 the case where statting the file as root fails - usually a non-root-mounted
7492 NFS file system. When this happens and the failure is EACCES, Exim now
7493 forks a subprocess and does the per-uid checking as the relevant uid.
7495 77. Added process_log_path.
7497 78. If log_file_path was not explicitly set, a setting of check_log_space or
7498 check_log_inodes was ignored.
7500 79. If a space check for the spool or log partitions fails, the incident is now
7501 logged. Of course, in the latter case the data may get lost...
7503 80. Added the %p formatting code to string_format() so that it can be used to
7504 print addresses in debug_print(). Adjusted all the address printing in the
7505 debugging in store.c to use %p rather than %d.
7507 81. There was a concern that a line of code in smtp_in.c could overflow a
7508 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7509 initially expressed, the concern was not well-founded, because trailing
7510 spaces are removed early. However, if the trailing spaces were followed by
7511 a NULL, they did not get removed, so the overflow was possible. Two fixes
7514 (a) I re-wrote the offending code in a cleaner fashion.
7515 (b) If an incoming SMTP command contains a NULL character, it is rejected
7518 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7519 runs initgroups(), so that if the Exim user is in any additional groups,
7520 they will be used during message reception.
7526 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob